Download MBA IB and Marketing 1st and 2nd Semester Quantitative Techniques for Managers Notes

Download MBA (Master of Business Administration) IB and Marketing 1st and 2nd Semester Quantitative Techniques for Managers Notes


MBA-H2040 Quantitative Techniques for Managers



UNIT I





1 INTRODUCTION TO OPERATIONS RESEARCH






ON
S



S
E



L

LESSON STRUCTURE





1.1 Introduction



1.2 History of Operations Research



1.3 Stages of Development of Operations



Research



1.4 Relationship Between Manager and



OR Specialist



1.5 OR Tools and Techniques



1.6 Applications of Operations Research



1.7 Limitations of Operations Research



1.8 Summary

Objectives

1.9 Key Terms

After Studying this lesson, you should be able

1.10 Self Assessment Questions

to:

1.11 Further References

Understand the meaning, purpose, and



tools of Operations Research



Describe the history of Operations



Research



Describe the Stages of O.R



Explain the Applications of Operations



Research



Describe the Limitations of Operation



Research



Understand the OR specialist and



Manager relationship





































MBA-H2040 Quantitative Techniques for Managers
1.1 Introduction

The British/Europeans refer to "operational research", the Americans to "operations research" - but both

are often shortened to just "OR" - which is the term we will use.

Another term which is used for this field is "management science" ("MS"). The Americans

sometimes combine the terms OR and MS together and say "OR/MS" or "ORMS". Yet other terms

sometimes used are "industrial engineering" ("IE") and "decision science" ("DS"). In recent years there

has been a move towards a standardization upon a single term for the field, namely the term "OR".

Operation Research is a relatively new discipline. The contents and the boundaries of the OR are

not yet fixed. Therefore, to give a formal definition of the term Operations Research is a difficult task.

The OR starts when mathematical and quantitative techniques are used to substantiate the decision being

taken. The main activity of a manager is the decision making. In our daily life we make the decisions

even without noticing them. The decisions are taken simply by common sense, judgment and expertise

without using any mathematical or any other model in simple situations. But the decision we are

concerned here with are complex and heavily responsible. Examples are public transportation network

planning in a city having its own layout of factories, residential blocks or finding the appropriate product

mix when there exists a large number of products with different profit contributions and production

requirement etc.

Operations Research tools are not from any one discipline. Operations Research takes tools from

different discipline such as mathematics, statistics, economics, psychology, engineering etc. and

combines these tools to make a new set of knowledge for decision making. Today, O.R. became a

professional discipline which deals with the application of scientific methods for making decision, and

especially to the allocation of scarce resources. The main purpose of O.R. is to provide a rational basis

for decisions making in the absence of complete information, because the systems composed of human,

machine, and procedures may do not have complete information.

Operations Research can also be treated as science in the sense it describing, understanding and

predicting the systems behaviour, especially man-machine system. Thus O.R. specialists are involved in

three classical aspect of science, they are as follows:

i) Determining the systems behaviour
ii) Analyzing the systems behaviour by developing appropriate models
iii) Predict the future behaviour using these models




The emphasis on analysis of operations as a whole distinguishes the O.R. from other research

and engineering. O.R. is an interdisciplinary discipline which provided solutions to problems of military

operations during World War II, and also successful in other operations. Today business applications are


MBA-H2040 Quantitative Techniques for Managers
primarily concerned with O.R. analysis for the possible alternative actions. The business and industry

befitted from O.R. in the areas of inventory, reorder policies, optimum location and size of warehouses,

advertising policies, etc.

As stated earlier defining O.R. is a difficult task. The definitions stressed by various experts and

Societies on the subject together enable us to know what O.R. is, and what it does. They are as follows:

1. According to the Operational Research Society of Great Britain (OPERATIONAL RESEARCH

QUARTERLY, l3(3):282, l962), Operational Research is the attack of modern science on
complex problems arising in the direction and management of large systems of men, machines,
materials and money in industry, business, government and defense. Its distinctive approach is to
develop a scientific model of the system, incorporating measurements of factors such as change
and risk, with which to predict and compare the outcomes of alternative decisions, strategies or
controls. The purpose is to help management determine its policy and actions scientifically.


2. Randy Robinson stresses that Operations Research is the application of scientific methods to

improve the effectiveness of operations, decisions and management. By means such as analyzing
data, creating mathematical models and proposing innovative approaches, Operations Research
professionals develop scientifically based information that gives insight and guides decision-
making. They also develop related software, systems, services and products.



3. Morse and Kimball have stressed O.R. is a quantitative approach and described it as " a scientific

method of providing executive departments with a quantitative basis for decisions regarding the
operations under their control".



4. Saaty considers O.R. as tool of improving quality of answers. He says, "O.R. is the art of giving

bad answers to problems which otherwise have worse answers".



5. Miller and Starr state, "O.R. is applied decision theory, which uses any scientific, mathematical

or logical means to attempt to cope with the problems that confront the executive, when he tries
to achieve a thorough-going rationality in dealing with his decision problem".



6. Pocock stresses that O.R. is an applied Science. He states "O.R. is scientific methodology

(analytical, mathematical, and quantitative) which by assessing the overall implication of various
alternative courses of action in a management system provides an improved basis for
management decisions".


1.2 History of Operations Research

Operation Research is a relatively new discipline. Whereas 70 years ago it would have been possible to

study mathematics, physics or engineering (for example) at university it would not have been possible to

study Operation Research, indeed the term O.R. did not exist then. It was really only in the late 1930's

that operational research began in a systematic fashion, and it started in the UK. As such it would be

interesting to give a short history of O.R.

1936



MBA-H2040 Quantitative Techniques for Managers
Early in 1936 the British Air Ministry established Bawdsey Research Station, on the east coast, near

Felixstowe, Suffolk, as the centre where all pre-war radar experiments for both the Air Force and the

Army would be carried out. Experimental radar equipment was brought up to a high state of reliabilit y

and ranges of over 100 miles on aircraft were obtained.

It was also in 1936 that Royal Air Force (RAF) Fighter Command, charged specifically with the

air defense of Britain, was first created. It lacked however any effective fighter aircraft - no Hurricanes

or Spitfires had come into service - and no radar data was yet fed into its very elementary warning and

control system.

It had become clear that radar would create a whole new series of problems in fighter direction

and control so in late 1936 some experiments started at Biggin Hill in Kent into the effective use of such

data. This early work, attempting to integrate radar data with ground based observer data for fighter

interception, was the start of OR.

1937

The first of three major pre-war air-defence exercises was carried out in the summer of 1937. The

experimental radar station at Bawdsey Research Station was brought into operation and the information

derived from it was fed into the general air-defense warning and control system. From the early warning

point of view this exercise was encouraging, but the tracking information obtained from radar, after

filtering and transmission through the control and display network, was not very satisfactory.

1938

In July 1938 a second major air-defense exercise was carried out. Four additional radar stations had been

installed along the coast and it was hoped that Britain now had an aircraft location and control system

greatly improved both in coverage and effectiveness. Not so! The exercise revealed, rather, that a new

and serious problem had arisen. This was the need to coordinate and correlate the additional, and often

conflicting, information received from the additional radar stations. With the outbreak of war apparently

imminent, it was obvious that something new - drastic if necessary - had to be attempted. Some new

approach was needed.

Accordingly, on the termination of the exercise, the Superintendent of Bawdsey Research Station, A.P.

Rowe, announced that although the exercise had again demonstrated the technical feasibility of the radar

system for detecting aircraft, its operational achievements still fell far short of requirements. He

therefore proposed that a crash program of research into the operational - as opposed to the technical -

aspects of the system should begin immediately. The term "operational research" [RESEARCH into


MBA-H2040 Quantitative Techniques for Managers
(military) OPERATIONS] was coined as a suitable description of this new branch of applied science.

The first team was selected from amongst the scientists of the radar research group the same day.

1939

In the summer of 1939 Britain held what was to be its last pre-war air defence exercise. It involved some

33,000 men, 1,300 aircraft, 110 antiaircraft guns, 700 searchlights, and 100 barrage balloons. This

exercise showed a great improvement in the operation of the air defence warning and control system.

The contribution made by the OR team was so apparent that the Air Officer Commander-in-Chief RAF

Fighter Command (Air Chief Marshal Sir Hugh Dowding) requested that, on the outbreak of war, they

should be attached to his headquarters at Stanmore in north London.



Initially, they were designated the "Stanmore Research Section". In 1941 they were redesignated

the "Operational Research Section" when the term was formalised and officially accepted, and similar

sections set up at other RAF commands.



1940

On May 15th 1940, with German forces advancing rapidly in France, Stanmore Research Section was

asked to analyses a French request for ten additional fighter squadrons (12 aircraft a squadron - so 120

aircraft in all) when losses were running at some three squadrons every two days (i.e. 36 aircraft every 2

days). They prepared graphs for Winston Churchill (the British Prime Minister of the time), based upon

a study of current daily losses and replacement rates, indicating how rapidly such a move would deplete

fighter strength. No aircraft were sent and most of those currently in France were recalled.



This is held by some to be the most strategic contribution to the course of the war made by OR

(as the aircraft and pilots saved were consequently available for the successful air defense of Britain, the

Battle of Britain).

1941 onward

In 1941, an Operational Research Section (ORS) was established in Coastal Command which was to

carry out some of the most well-known OR work in World War II.

The responsibility of Coastal Command was, to a large extent, the flying of long-range sorties by

single aircraft with the object of sighting and attacking surfaced U-boats (German submarines). The

technology of the time meant that (unlike modern day submarines) surfacing was necessary to recharge

batteries, vent the boat of fumes and recharge air tanks. Moreover U-boats were much faster on the

surface than underwater as well as being less easily detected by sonar.



MBA-H2040 Quantitative Techniques for Managers

Thus the Operation Research started just before World War II in Britain with the establishment

of teams of scientists to study the strategic and tactical problems involved in military operations. The

objective was to find the most effective utilization of limited military resources by the use of quantitative

techniques. Following the end of the war OR spread, although it spread in different ways in the UK and

USA.

In 1951 a committee on Operations Research formed by the National Research Council of USA,

and the first book on "Methods of Operations Research", by Morse and Kimball, was published. In 1952

the Operations Research Society of America came into being.

Success of Operations Research in army attracted the attention of the industrial mangers who

were seeking solutions to their complex business problems. Now a days, almost every organization in all

countries has staff applying operations research, and the use of operations research in government has

spread from military to wide variety of departments at all levels. The growth of operations research has

not limited to the U.S.A. and U.K., it has reached many countries of the world.

India was one the few first countries who started using operations research. In India, Regional

Research Laboratory located at Hyderabad was the first Operations Research unit established during

1949. At the same time another unit was set up in Defense Science Laboratory to solve the Stores,

Purchase and Planning Problems. In 1953, Operations Research unit was established in Indian Statistical

Institute, Calcutta, with the objective of using Operations Research methods in National Planning and

Survey. In 1955, Operations Research Society of India was formed, which is one of the first members of

International Federation of Operations Research societies. Today Operations Research is a popular

subject in management institutes and schools of mathematics.

1.3 Stages of Development of Operations Research

The stages of development of O.R. are also known as phases and process of O.R, which has six

important steps. These six steps are arranged in the following order:

Step I: Observe the problem environment

Step II: Analyze and define the problem

Step III: Develop a model

Step IV: Select appropriate data input

Step V: Provide a solution and test its reasonableness

Step VI: Implement the solution


MBA-H2040 Quantitative Techniques for Managers






Step I: Observe the problem environment

The first step in the process of O.R. development is the problem environment observation. This step

includes different activities; they are conferences, site visit, research, observations etc. These activities

provide sufficient information to the O.R. specialists to formulate the problem.

Step II: Analyze and define the problem

This step is analyzing and defining the problem. In this step in addition to the problem definition the

objectives, uses and limitations of O.R. study of the problem also defined. The outputs of this step are

clear grasp of need for a solution and its nature understanding.

Step III: Develop a model

This step develops a model; a model is a representation of some abstract or real situation. The models

are basically mathematical models, which describes systems, processes in the form of equations,

formula/relationships. The different activities in this step are variables definition, formulating equations

etc. The model is tested in the field under different environmental constraints and modified in order to

work. Some times the model is modified to satisfy the management with the results.

Step IV: Select appropriate data input

A model works appropriately when there is appropriate data input. Hence, selecting appropriate input

data is important step in the O.R. development stage or process. The activities in this step include

internal/external data analysis, fact analysis, and collection of opinions and use of computer data banks.

The objective of this step is to provide sufficient data input to operate and test the model developed in

Step_III.

Step V: Provide a solution and test its reasonableness

This step is to get a solution with the help of model and input data. This solution is not implemented

immediately, instead the solution is used to test the model and to find there is any limitations. Suppose if

the solution is not reasonable or the behaviour of the model is not proper, the model is updated and

modified at this stage. The output of this stage is the solution(s) that supports the current organizational

objectives.

Step VI: Implement the solution



MBA-H2040 Quantitative Techniques for Managers
At this step the solution obtained from the previous step is implemented. The implementation of the

solution involves mo many behavioural issues. Therefore, before implementation the implementation

authority has to resolve the issues. A properly implemented solution results in quality of work and gains

the support from the management.

The process, process activities, and process output are summarized in the following Table 1-1.

Process Activities

Process

Process Output

Site visits, Conferences,



Sufficient information and

Step 1:

Observations, Research

support to proceed

Observe the problem
environment



Define: Use, Objectives,

Clear grasp of need for and

limitations

Step 2:

nature of solution requested

Analyze and define



the problem



Define interrelationships,

Models that works under stated

Formulate equations,

environmental constraints

Step 3:
Develop a Model

Use known O.R. Model ,

Search alternate Model

Analyze: internal-external data,

Sufficient inputs to operate and

facts

Step 4:

test model

Select appropriate data

Collect options,

input

Use computer data banks

Test the model

Solution(s) that support current

Step 5:

organizational goals

find limitations

Provide a solution and
test its reasonableness


update the model



Step 6:
Implement the
solution
MBA-H2040 Quantitative Techniques for Managers

Resolve behavioural issues

Improved working and

Management support for longer

Sell the idea

run operation of model

Give explanations

Management involvement

Table 1-1: Process, Process activities, Process output of O.R. development stages

1.4 Relationship between the Manager and O.R. Specialist

The key responsibility of manager is decision making. The role of the O.R. specialist is to help the

manager make better decisions. Figure 1-1 explains the relationship between the O.R. specialist and the

manager/decision maker.

STEPS IN PROBLEM RECOGNITION,





INVOLVEMENT: O.R. SPECIALIST or

FORMULATION AND SOLUTION





MANAGER



Recognize

from

organizational

Manager

symptoms that a problem exists.









Decide what variables are involved; state
the problem in quantitative relationships

Manager and O.R. Specialist

among the variables.





















I



nvestigat

e met

hods f

or sol

ving t

he





problems as stated above; determine

O.R. Specialist



appropriate quantitative tools to be used.



















Attempt solutions to the problems; find



various solutions; state assumptions

O.R. Specialist

underlying

these

solutions;

test



alternative solutions.



















Determine which solution is most
effective because of practical constraints

Manager and O.R. Specialist



wit hin the organi zation; decide what t

he



solution means for the organization.



MBA-H2040 Quantitative Techniques for Managers




















Choose the solution to be used.

Manager





`

S

e l l ' t

h

e

de

c i s

i

o

n

t

o

o

pe

r

a

t i

n

g

m

a

na

ge

r

s ;

Manager and O.R. Specialist

get their understanding and cooperation.

Figure 1-1 Relationship between Manager/Decision Maker and O.R. Specialists

1.5 O.R. Tools and Techniques

Operations Research uses any suitable tools or techniques available. The common frequently used

tools/techniques are mathematical procedures, cost analysis, electronic computation. However,

operations researchers given special importance to the development and the use of techniques like linear

programming, game theory, decision theory, queuing theory, inventory models and simulation. In

addition to the above techniques, some other common tools are non-linear programming, integer

programming, dynamic programming, sequencing theory, Markov process, network scheduling

(PERT/CPM), symbolic Model, information theory, and value theory. There is many other Operations

Research tools/techniques also exists. The brief explanations of some of the above techniques/tools are

as follows:

Linear Programming:

This is a constrained optimization technique, which optimize some criterion within some constraints. In

Linear programming the objective function (profit, loss or return on investment) and constraints are

linear. There are different methods available to solve linear programming.

Game Theory:

This is used for making decisions under conflicting situations where there are one or more

players/opponents. In this the motive of the players are dichotomized. The success of one player tends to

be at the cost of other players and hence they are in conflict.

Decision Theory:

Decision theory is concerned with making decisions under conditions of complete certainty about the

future outcomes and under conditions such that we can make some probability about what will happen

in future.

Queuing Theory:


MBA-H2040 Quantitative Techniques for Managers
This is used in situations where the queue is formed (for example customers waiting for service, aircrafts

waiting for landing, jobs waiting for processing in the computer system, etc). The objective here is

minimizing the cost of waiting without increasing the cost of servicing.

Inventory Models:

Inventory model make a decisions that minimize total inventory cost. This model successfully reduces

the total cost of purchasing, carrying, and out of stock inventory.



Simulation:

Simulation is a procedure that studies a problem by creating a model of the process involved in the

problem and then through a series of organized trials and error solutions attempt to determine the best

solution. Some times this is a difficult/time consuming procedure. Simulation is used when actual

experimentation is not feasible or solution of model is not possible.

Non-linear Programming:

This is used when the objective function and the constraints are not linear in nature. Linear relationships

may be applied to approximate non-linear constraints but limited to some range, because approximation

becomes poorer as the range is extended. Thus, the non-linear programming is used to determine the

approximation in which a solution lies and then the solution is obtained using linear methods.

Dynamic Programming:

Dynamic programming is a method of analyzing multistage decision processes. In this each elementary

decision depends on those preceding decisions and as well as external factors.

Integer Programming:

If one or more variables of the problem take integral values only then dynamic programming method is

used. For example number or motor in an organization, number of passenger in an aircraft, number of

generators in a power generating plant, etc.

Markov Process:

Markov process permits to predict changes over time information about the behavior of a system is

known. This is used in decision making in situations where the various states are defined. The

probability from one state to another state is known and depends on the current state and is independent

of how we have arrived at that particular state.



MBA-H2040 Quantitative Techniques for Managers
Network Scheduling:

This technique is used extensively to plan, schedule, and monitor large projects (for example computer

system installation, R & D design, construction, maintenance, etc.). The aim of this technique is

minimize trouble spots (such as delays, interruption, production bottlenecks, etc.) by identifying the

critical factors. The different activities and their relationships of the entire project are represented

diagrammatically with the help of networks and arrows, which is used for identifying critical activities

and path. There are two main types of technique in network scheduling, they are:

Program Evaluation and Review Technique (PERT) ? is used when activities time is not known
accurately/ only probabilistic estimate of time is available.

Critical Path Method (CPM) ? is used when activities time is know accurately.

Information Theory:

This analytical process is transferred from the electrical communication field to O.R. field. The

objective of this theory is to evaluate the effectiveness of flow of information with a given system. This

is used mainly in communication networks but also has indirect influence in simulating the examination

of business organizational structure with a view of enhancing flow of information.

1.6 Applications of Operations Research

Today, almost all fields of business and government utilizing the benefits of Operations Research. There

are voluminous of applications of Operations Research. Although it is not feasible to cover all

applications of O.R. in brief. The following are the abbreviated set of typical operations research

applications to show how widely these techniques are used today:

Accounting:





Assigning audit teams effectively
Credit policy analysis
Cash flow planning
Developing standard costs
Establishing costs for byproducts
Planning of delinquent account strategy



Construction:





Project scheduling, monitoring and control
Determination of proper work force
Deployment of work force
Allocation of resources to projects



Facilities Planning:





Factory location and size decision
Estimation of number of facilities required
Hospital planning


MBA-H2040 Quantitative Techniques for Managers

International logistic system design
Transportation loading and unloading
Warehouse location decision

Finance:





Building cash management models
Allocating capital among various alternatives
Building financial planning models
Investment analysis
Portfolio analysis
Dividend policy making



Manufacturing:





Inventory control
Marketing balance projection
Production scheduling
Production smoothing



Marketing:





Advertising budget allocation
Product introduction timing
Selection of Product mix
Deciding most effective packaging alternative



Organizational Behavior / Human Resources:





Personnel planning
Recruitment of employees
Skill balancing
Training program scheduling
Designing organizational structure more effectively



Purchasing:





Optimal buying
Optimal reordering
Materials transfer



Research and Development:





R & D Projects control
R & D Budget allocation
Planning of Product introduction

1.7 Limitations of Operations Research

Operations Research has number of applications; similarly it also has certain limitations. These

limitations are mostly related to the model building and money and time factors problems involved in its

application. Some of them are as given below:

i)

Distance between O.R. specialist and Manager

Operations Researchers job needs a mathematician or statistician, who might not be aware of
the business problems. Similarly, a manager is unable to understand the complex nature of
Operations Research. Thus there is a big gap between the two personnel.

ii)

Magnitude of Calculations



MBA-H2040 Quantitative Techniques for Managers

The aim of the O.R. is to find out optimal solution taking into consideration all the factors. In
this modern world these factors are enormous and expressing them in quantitative model and
establishing relationships among these require voluminous calculations, which can be
handled only by machines.

iii)

Money and Time Costs

The basic data are subjected to frequent changes, incorporating these changes into the
operations research models is very expensive. However, a fairly good solution at present may
be more desirable than a perfect operations research solution available in future or after some
time.

iv)

Non-quantifiable Factors

When all the factors related to a problem can be quantifiable only then operations research
provides solution otherwise not. The non-quantifiable factors are not incorporated in O.R.
models. Importantly O.R. models do not take into account emotional factors or qualitative
factors.

v)

Implementation

Once the decision has been taken it should be implemented. The implementation of decisions
is a delicate task. This task must take into account the complexities of human relations and
behavior and in some times only the psychological factors.

1.8 Summary

Operations Research is relatively a new discipline, which originated in World War II, and became very

popular throughout the world. India is one of the few first countries in the world who started using

operations research. Operations Research is used successfully not only in military/army operations but

also in business, government and industry. Now a day's operations research is almost used in all the

fields.



Proposing a definition to the operations research is a difficult one, because its boundary and

content are not fixed. The tools for operations search is provided from the subject's viz. economics,

engineering, mathematics, statistics, psychology, etc., which helps to choose possible alternative courses

of action. The operations research tool/techniques include linear programming, non-linear programming,

dynamic programming, integer programming, Markov process, queuing theory, etc.



Operations Research has a number of applications. Similarly it has a number of limitations,

which is basically related to the time, money, and the problem involves in the model building. Day-by-

day operations research gaining acceptance because it improve decision making effectiveness of the

managers. Almost all the areas of business use the operations research for decision making.

1.9 Key Terms


MBA-H2040 Quantitative Techniques for Managers
OR: Operations Research.

MS: Management Science.

Symbolic Model: An abstract model, generally using mathematical symbols.

Criterion: is measurement, which is used to evaluation of the results.

Integer Programming: is a technique, which ensures only integral values of variables in the problem.

Dynamic Programming: is a technique, which is used to analyze multistage decision process.

Linear Programming: is a technique, which optimizes linear objective function under limited

constraints.

Inventory Model: these are the models used to minimize total inventory costs.

Optimization: Means maximization or minimization.





1.10 Self Assessment Questions

Q1. Define Operations Research.
Q2. Describe the relationship between the manager and O.R. specialist.
Q3. Explain the various steps in the O.R. development process.
Q4. Discuss the applications of O.R.
Q5. Discuss the limitation of O.R.
Q6. Describe different techniques of O.R.
Q7. Discuss few areas of O.R. applications in your organization or organization you are familiar with.


1.11 Further References

Hamdy A Taha, 1999. Introduction to Operations Research, PHI Limited, New Delhi.

Sharma, J.K., 1989. Mathematical Models in Operations Research, Tata McGraw Hill Publishing
Company Ltd., New Delhi.
Beer, Stafford, 1966. Decision and Control, John Wiley & Sons, Inc., New York.
Levin, Rubin, Stinson, Gardner, 1992. Quantitative Approaches to Management, Tata McGraw Hill
Publishing Company Ltd. New Delhi.
Wagner, Harvery M., 1975. Principles of Operations Research, PHI, Egnlewood Cliffs, N.J.











MBA-H2040 Quantitative Techniques for Managers
















UNIT I












ON
S

2 LINEAR PROGRAMMING ?GRAPHICAL METHOD

S
E



L





LESSON STRUCTURE




2.1 Introduction to Linear Programming



2.2 Linear Programming Problem



Formulation



2.3 Formulation with Different Types of



Constraints



2.4 Graphical Analysis of Linear



Programming



2.5 Graphical Linear Programming Solution



2.6 Multiple Optimal Solutions



2.7 Unbounded Solution



2.8 Infeasible Solution



2.9 Summary



2.10 Key Terms



2.11 Self Assessment Questions



2.12 Key Solutions



2.13 Further References








MBA-H2040 Quantitative Techniques for Managers














Objectives
After studying this lesson, you should be able
to:


Formulate Linear Programming Problem
Identify the characteristics of linear

programming problem

Make a graphical analysis of the linear

programming problem

Solve the problem graphically
Identify the various types of solutions










MBA-H2040 Quantitative Techniques for Managers

2.1 Introduction to Linear Programming




Linear Programming is a special and versatile technique which can be applied to a variety of

management problems viz. Advertising, Distribution, Investment, Production, Refinery Operations, and

Transportation analysis. The linear programming is useful not only in industry and business but also in

non-profit sectors such as Education, Government, Hospital, and Libraries. The linear programming

method is applicable in problems characterized by the presence of decision variables. The objective

function and the constraints can be expressed as linear functions of the decision variables. The

decision variables represent quantities that are, in some sense, controllable inputs to the system being

modeled. An objective function represents some principal objective criterion or goal that measures the

effectiveness of the system such as maximizing profits or productivity, or minimizing cost or

consumption. There is always some practical limitation on the availability of resources viz. man,

material, machine, or time for the system. These constraints are expressed as linear equations involving

the decision variables. Solving a linear programming problem means determining actual values of the

decision variables that optimize the objective function subject to the limitation imposed by the

constraints.



The main important feature of linear programming model is the presence of linearity in the

problem. The use of linear programming model arises in a wide variety of applications. Some model

may not be strictly linear, but can be made linear by applying appropriate mathematical transformations.

Still some applications are not at all linear, but can be effectively approximated by linear models. The

ease with which linear programming models can usually be solved makes an attractive means of dealing

with otherwise intractable nonlinear models.


2.2 Linear Programming Problem Formulation

The linear programming problem formulation is illustrated through a product mix problem. The product

mix problem occurs in an industry where it is possible to manufacture a variety of products. A product

has a certain margin of profit per unit, and uses a common pool of limited resources. In this case the

linear programming technique identifies the products combination which will maximize the profit

subject to the availability of limited resource constraints.


Example 2.1:

Suppose an industry is manufacturing tow types of products P1 and P2. The profits per Kg of the two

products are Rs.30 and Rs.40 respectively. These two products require processing in three types of

machines. The following table shows the available machine hours per day and the time required on each


18
MBA-H2040 Quantitative Techniques for Managers
machine to produce one Kg of P1 and P2. Formulate the problem in the form of linear programming

model.



Profit/Kg

P1

P2

Total available Machine

Rs.30

Rs.40

hours/day

Machine 1

3

2

600

Machine 2

3

5

800

Machine 3

5

6

1100


Solution:






The procedure for linear programming problem formulation is as follows:




Introduce the decision variable as follows:



Let x1 = amount of P1



x2 = amount of P2




In order to maximize profits, we establish the objective function as






30x1 + 40x2




Since one Kg of P1 requires 3 hours of processing time in machine 1 while the corresponding

requirement of P2 is 2 hours. So, the first constraint can be expressed as





3x1 + 2x2 600



Similarly, corresponding to machine 2 and 3 the constraints are











3x1 + 5x2 800





5x1 + 6x2 1100

In addition to the above there is no negative production, which may be represented algebraically as





x1 0

;

x2 0


Thus, the product mix problem in the linear programming model is as follows:


Maximize





30x1 + 40x2



Subject to:





3x1 + 2x2 600





3x1 + 5x2 800





5x1 + 6x2 1100





x1 0, x2 0


2.3 Formulation with Different Types of Constraints




19

MBA-H2040 Quantitative Techniques for Managers
The constraints in the previous example 2.1 are of "less than or equal to" type. In this section we are

going to discuss the linear programming problem with different constraints, which is illustrated in the

following Example 2.2.

Example 2.2:

A company owns two flour mills viz. A and B, which have different production capacities for high,

medium and low quality flour. The company has entered a contract to supply flour to a firm every month

with at least 8, 12 and 24 quintals of high, medium and low quality respectively. It costs the company

Rs.2000 and Rs.1500 per day to run mill A and B respectively. On a day, Mill A produces 6, 2 and 4

quintals of high, medium and low quality flour, Mill B produces 2, 4 and 12 quintals of high, medium

and low quality flour respectively. How many days per month should each mill be operated in order to

meet the contract order most economically.

Solution:

Let us define x1 and x2 are the mills A and B. Here the objective is to minimize the cost of the machine

runs and to satisfy the contract order. The linear programming problem is given by

Minimize





2000x1 + 1500x2



Subject to:





6x1 + 2x2 8





2x1 + 4x2 12





4x1 + 12x2 24






x1 0, x2 0


2.4 Graphical Analysis of Linear Programming

This section shows how a two-variable linear programming problem is solved graphically, which is
illustrated as follows:

Example 2.3:
Consider the product mix problem discussed in section 2.2



Maximize





30x1 + 40x2



Subject to:





3x1 + 2x2 600





3x1 + 5x2 800





5x1 + 6x2 1100





x1 0, x2 0





20
MBA-H2040 Quantitative Techniques for Managers
From the first constraints 3x1 + 2x2 600, draw the line 3x1 + 2x2 = 600 which passes through the point
(200, 0) and (0, 300). This is shown in the following graph as line 1.














300



3x1 + 2x2 = 600(line 1)











200







B



X2






C



100






3x1 + 5x2 = 800(line 2)







5x1 + 6x2 = 1100(line 3)



A





D

0

50

100

150

X1

200

275












Graph 1: Three closed half planes and Feasible Region

Half Plane

- A linear inequality in two variables is called as a half plane.

Boundary

- The corresponding equality (line) is called as the boundary of the half plane.

Close Half Plane ? Half plane with its boundary is called as a closed half plane.



21

MBA-H2040 Quantitative Techniques for Managers
In this case we must decide in which side of the line 3x1 + 2x2 = 600 the half plane is located. The

easiest way to solve the inequality for x2 is









3x1 600 ? 2x2

And for the fixed x1, the coordinates satisfy this inequality are smaller than the corresponding ordinate

on the line and thus the inequality is satisfied for all the points below the line 1.

Similarly, we have to determine the closed half planes for the inequalities 3x1 + 5x2 800 and
5x1 + 6x2 1100 (line 2 and line 3 in the graph). Since all the three constraints must be satisfied

simultaneously we have consider the intersection of these three closed half planes. The complete

intersection of these three closed half planes is shown in the above graph as ABCD. The region ABCD

is called the feasible region, which is shaded in the graph.

Feasible Solution:

Any non-negative value of x1, x2 that is x1 0 and x2 0 is known as feasible solution of the linear

programming problem if it satisfies all the existing constraints.

Feasible Region:


The collection of all the feasible solution is called as the feasible region.


Example 2.4:

In the previous example we discussed about the less than or equal to type of linear programming

problem, i.e. maximization problem. Now consider a minimization (i.e. greater than or equal to type)

linear programming problem formulated in Example 2.2.





Minimize




2000x1 + 1500x2



Subject to:





6x1 + 2x2 8





2x1 + 4x2 12





4x1 + 12x2 24






x1 0, x2 0




The three lines 6x1 + 2x2 = 8, 2x1 + 4x2 = 12, and 4x1 + 12x2 = 24 passes through the point

(1.3,0) (0,4), (6,0) (0,3) and (6,0) (0,2). The feasible region for this problem is shown in the following

Graph 2. In this problem the constraints are of greater than or equal to type of feasible region, which is

bounded on one side only.







22
MBA-H2040 Quantitative Techniques for Managers








8






6






X2

A



4

6x



1 + 2x2 8



B




2

2x1 + 4x2 12






C



4x1 + 12x2 24

0



X1



2

4

6

8










Graph 2: Feasible Region


2.5 Graphical Liner Programming Solution

A two variable linear programming problem can be easily solved graphically. The method is simple but

the principle of solution is depends on certain analytical concepts, they are:

Convex Region:

A region R is convex if and only if for any two points on the region R the line connecting those points

lies entirely in the region R.

Extreme Point:

The extreme point E of a convex region R is a point such that it is not possible to locate two distinct

points in R, so that the line joining them will include E. The extreme points are also called as corner

points or vertices.



Thus, the following result provides the solution to the linear programming model:

"If the minimum or maximum value of a linear function defined over a convex region exists,

then it must be on one of the extreme points".



23

MBA-H2040 Quantitative Techniques for Managers


In this section we are going to describe linear programming graphical solution for both the

maximization and minimization problems, discussed in Example 2.3 and Example 2.4.




Example 2. 5:


Consider the maximization problem described in Example 2.3.




Maximize





30x1 + 40x2

Subject to:





3x1 + 2x2 600





3x1 + 5x2 800





5x1 + 6x2 1100

M = 30x





x

1 +40x2

1 0, x2 0


The feasible region identified in the Example 2.3 is a convex polygon, which is illustrated in the
following Graph 3. The extreme point of this convex region are A, B, C, D and E.



300












200





B






X2

C








100






D



A



E











0

50

100

150

X1

200

275



24
MBA-H2040 Quantitative Techniques for Managers








Graph 3: Graphical Linear Programming Solution




In this problem the objective function is 30x1 + 40x2. Let be M is a parameter, the graph 30x1 +

40x2 = M is a group of parallel lines with slope ? 30/40. Some of these lines intersects the feasible region

and contains many feasible solutions, whereas the other lines miss and contain no feasible solution. In

order to maximize the objective function, we find the line of this family that intersects the feasible

region and is farthest out from the origin. Note that the farthest is the line from the origin the greater will

be the value of M.



Observe that the line 30x1 + 40x2 = M passes through the point D, which is the intersection of the

lines 3x1 + 5x2 = 800 and 5x1 + 6x2 = 1100 and has the coordinates x1 = 170 and x2 = 40. Since D is the

only feasible solution on this line the solution is unique.



The value of M is 6700, which is the objective function maximum value. The optimum value

variables are x1 = 170 and X2 = 40.



The following Table 1 shows the calculation of maximum value of the objective function.



Extreme Point

Coordinates

Objective Function

X1 X2

30x1 + 40x2

A

X1 = 0

X2 = 0

0

B

X1 = 0

X2 = 160

6400

C

X1 = 110

X2 = 70

6100

D

X1 = 170

X2 = 40

6700

E

X1 = 200

X2 = 0

6000



Table 1: Shows the objective function Maximum value calculation


Example 2.6:


Consider the minimization problem described in Example 2.4.




Minimize





2000x1 + 1500x2



Subject to:





6x1 + 2x2 8





2x1 + 4x2 12





4x1 + 12x2 24





x1 0, x2 0




The feasible region for this problem is illustrated in the following Graph 4. Here each of the half

planes lies above its boundary. In this case the feasible region is infinite. In this case, we are concerned

with the minimization; also it is not possible to determine the maximum value. As in the previous



25

MBA-H2040 Quantitative Techniques for Managers
example let us introduce a parameter M in the objective function i.e. 2000x1 + 1500x2 = M and draw the

lines for different values of M, which is shown in the following Table 2.










8







6





X2

A



4





B




2






C

0





2000x1+ 1500x

2=M

X1



2

4

6

8










Graph 4: Graphical Linear Programming Solution



Extreme Point

Coordinates

Objective Function

X1 X2

2000x1 + 1500x2

A

X1 = 0

X2 = 4

6000

B

X1 = 0.5

X2 = 2.75

5125

C

X1 = 6

X2 = 0

12000







Table 2: Shows the objective function Minimum value computation





The minimum value is 5125 at the extreme point B, which is the value of the M (objective

function). The optimum values variables are X1 = 0.5 and X2 = 2.75.


2.6 Multiple Optimal Solutions



26
MBA-H2040 Quantitative Techniques for Managers
When the objective function passed through only the extreme point located at the intersection of two

half planes, then the linear programming problem possess unique solutions. The previous examples i.e.

Example 2.5 and Example 2.6 are of this types (which possessed unique solutions).



When the objective function coincides with one of the half planes generated by the constraints in

the problem, will possess multiple optimal solutions. In this section we are going to discuss about the

multiple optimal solutions of linear programming problem with the help of the following Example 2.7.



Example 2.7:

A company purchasing scrap material has two types of scarp materials available. The first type has 30%

of material X, 20% of material Y and 50% of material Z by weight. The second type has 40% of

material X, 10% of material Y and 30% of material Z. The costs of the two scraps are Rs.120 and

Rs.160 per kg respectively. The company requires at least 240 kg of material X, 100 kg of material Y

and 290 kg of material Z. Find the optimum quantities of the two scraps to be purchased so that the

company requirements of the three materials are satisfied at a minimum cost.

Solution

First we have to formulate the linear programming model. Let us introduce the decision variables x1 and

x2 denoting the amount of scrap material to be purchased. Here the objective is to minimize the

purchasing cost. So, the objective function here is



Minimize





120x1 + 160x2




Subject to:





0.3x1 + 0.4x2 240





0.2x1 + 0.1x2 100





0.5x1 + 0.3x2 290






x1 0; x2 0


Multiply by 10 both sides of the inequalities, then the problem becomes

Minimize





120x1 + 160x2



Subject to:





3x1 + 4x2 2400





2x1 + x2 1000





5x1 + 3x2 2900






x1 0; x2 0





27

MBA-H2040 Quantitative Techniques for Managers
Let us introduce parameter M in the objective function i.e. 120x1 + 160x2 = M. Then we have to

determine the different values for M, which is shown in the following Table 3.

Extreme Point

Coordinates

Objective Function

X1 X2

120x1 + 160x2

A

X1 = 0

X2 = 1000

160000

B

X1 = 150

X2 = 740

136400

C

X1 = 400

X2 = 300

96000

D

X1=800

X2=0

96000





Table 3: Shows the calculation of Minimum objective function value



Note that there are two minimum value for the objective function (M=96000). The feasible

region and the multiple solutions are indicated in the following Graph 5.



1100



A



1000



900





800

B



700





5x1 + 3x2 2900



600




X2

500




2x



1 + x2 1000

400





300

C





200



3x1 + 4x2 2400




100



28

0

D
MBA-H2040 Quantitative Techniques for Managers










X1

Graph 5: Feasible Region, Multiple Optimal Solutions





The extreme points are A, B, C, and D. One of the objective functions 120x1 + 160x2 = M family

coincides with the line CD at the point C with value M=96000, and the optimum value variables are x1 =

400, and x2 = 300. And at the point D with value M=96000, and the optimum value variables are x1 =

800, and x2 = 0.



Thus, every point on the line CD minimizes objective function value and the problem contains

multiple optimal solutions.

2.7 Unbounded Solution

When the feasible region is unbounded, a maximization problem may don't have optimal solution, since

the values of the decision variables may be increased arbitrarily. This is illustrated with the help of the

following problem.



Maximize





3x1 + x2



Subject to:





x1 + x2 6





-x1 + x2 6





-x1 + 2x2 -6

and





x1, x2 0



Graph 6 shows the unbounded feasible region and demonstrates that the objective function can

be made arbitrarily large by increasing the values of x1 and x2 within the unbounded feasible region. In

this case, there is no point (x1, x2) is optimal because there are always other feasible points for which

objective function is larger. Note that it is not the unbounded feasible region alone that precludes an

optimal solution. The minimization of the function subject to the constraints shown in the Graph 6

would be solved at one the extreme point (A or B).

The unbounded solutions typically arise because some real constraints, which represent a

practical resource limitation, have been missed from the linear programming formulation. In such

situation the problem needs to be reformulated and re-solved.







29

MBA-H2040 Quantitative Techniques for Managers























-x1 + x2 = 6




















A



6

x1 + x2 = 6




5




4

-x1 + 2x2 = 6



X2




3










2



Graph 6: Unbounded Feasible Region



1









1

2

3

4

5

6

B

X1

2.8 Infeasible Solution

A linear programming problem is said to be infeasible if no feasible solution of the problem exists. This

section describes infeasible solution of the linear programming problem with the help of the following

Example 2.8.

Example 2.8:


30
MBA-H2040 Quantitative Techniques for Managers

Minimize


200x1 + 300x2


Subject to:


0.4x1 + 0.6x2 240



0.2x1 + 0.2x2 80



0.4x1 + 0.3x2 180



x1, x2 0


On multiplying both sides of the inequalities by 10, we get



4x1 + 6x2 2400



2x1 + 2x2 800



4x1 + 3x2 1800



700





A



600



4x1 + 3x2 = 1800




500





B



400

X2





300





200

F



2x1 + 2x2 = 800




100



4x1 + 6x2 = 2400




0



D



C

E

500

600



100

200

300

400

X1










Graph 7: Infeasible Solution




The region right of the boundary AFE includes all the solutions which satisfy the first (4x1 + 6x2

2400) and the third (4x1 + 3x2 1800) constraints. The region left of the BC contains all solutions

which satisfy the second constraint (2x1 + 2x2 800).



31

MBA-H2040 Quantitative Techniques for Managers


Hence, there is no solution satisfying all the three constraints (first, second, and third). Thus, the

linear problem is infeasible. This is illustrated in the above Graph 7.

2.9 Summary

In Operations Research linear programming is a versatile technique with wide applications in various

management problems. Linear Programming problem has a number of characteristics. That is first we

have to identify the decision variable. The problem must have a well defined objective function, which

are expressed in terms of the decision variables.



The objective function may have to be maximized when it indicates the profit or production or

contribution. If the objective function represents cost, in this case the objective function has to be

minimized.



The management problem is expressed in terms of the decision variables with the objective

function and constraints. A linear programming problem is solved graphically if it contains only two

variables.


2.10 Key Terms

Objective Function: is a linear function of the decision variables representing the objective of the

manager/decision maker.


Constraints: are the linear equations or inequalities arising out of practical limitations.

Decision Variables: are some physical quantities whose values indicate the solution.

Feasible Solution: is a solution which satisfies all the constraints (including the non-negative) presents

in the problem.


Feasible Region: is the collection of feasible solutions.

Multiple Solutions: are solutions each of which maximize or minimize the objective function.



Unbounded Solution: is a solution whose objective function is infinite.

Infeasible Solution: means no feasible solution.

2.11 Self Assessment Questions

Q1. A juice company has its products viz. canned apple and bottled juice with profit margin Rs.4 and
Rs.2 respectively pre unit. The following table shows the labour, equipment, and ingredients to produce
each product per unit.



Canned Apple

Bottled Juice

Total

Labour

2.0

3.0

12.0

Equipment

3.2

1.0

8.0



32
MBA-H2040 Quantitative Techniques for Managers
Ingredients

2.4

2.0

9.0


Formulate the linear programming problem (model) specifying the product mix which will maximize the
profit without exceeding the levels of resources.


Q2. An organization is interested in the analysis of two products which can be produces from the idle
time of labour, machine and investment. It was notified on investigation that the labour requirement of
the first and the second products was 4 and 5 units respectively and the total available man hours was
48. Only first product required machine hour utilization of one hour per unit and at present only 10 spare
machine hours are available. Second product needs one unit of byproduct per unit and the daily
availability of the byproduct is 12 units. According to the marketing department the sales potential of
first product cannot exceed 7 units. In a competitive market, first product can be sold at a profit of Rs.6
and the second product at a profit of Rs.10 per unit.



Formulate the problem as a linear programming model. Also determine graphically the feasible

region. Identify the redundant constraints if any.

Q3. Find graphically the feasible region of the linear programming problem given in Q1.

Q4. A bed mart company is in the business of manufacturing beds and pillows. The company has 40
hours for assembly and 32 hours for finishing work per day. Manufacturing of a bed requires 4 hours for
assembly and 2 hours in finishing. Similarly a pillow requires 2 hours for assembly and 4 hours for
finishing. Profitability analysis indicates that every bed would contribute Rs.80, while a pillow
contribution is Rs.55 respectively. Find out the daily production of the company to maximize the
contribution (profit).

Q5. Maximize




1170x1 + 1110x2

Subject to:




9x1 + 5x2 500





7x1 + 9x2 300





5x1 + 3x2 1500





7x1 + 9x2 1900





2x1 + 4x2 1000





x1, x2 0

Find graphically the feasible region and the optimal solution.

Q6. Solve the following LP problem graphically

Minimize





2x1 +1.7x2

Subject to:




0.15x1 + 0.10x2 1.0





0.75x1 + 1.70x2 7.5





1.30x1 + 1.10x2 10.0






x1, x2 0


Q7. Solve the following LP problem graphically




33

MBA-H2040 Quantitative Techniques for Managers


Maximize





2x1 + 3x2



Subject to:





x1 ? x2 1





x1 + x2 3





x1, x2 0


Q8. Graphically solve the following problem of LP


Maximize





3x1 + 2x2



Subject to:





2x1 ? 3x2 0





3x1 + 4x2 -12





x1, x2 0



Q9. Solve the following problem graphically



Maximize





4x1 + 4x2



Subject to:





-2x1 + x2 1





x1 2





x1 + x2 3





x1, x2 0


2.12 Key Solutions

Q1.




Canned Apple x1





Bottled Juice x2






Maximize







4x1 + 2x2






Subject to:







2x1 + 3x2 12







3.2x1 + x2 8







2.4x1 + 2x2 9







x1, x2 0


Q2.




First Product x1





Second Product x2









Maximize







6x1 + 10x2





Subject to:







4x1 + 5x2 48







x1 10



34
MBA-H2040 Quantitative Techniques for Managers






x2 12







x1 7








x1, x2 0





The constraints x1 10 is redundant.

Q4.


Beds = 8



Pillows = 4



Maximum Profits is: Rs.860


Q5.





Optimum variables values are: x1=271.4, x2=0



The maximum value is: 317573


Q6.


Optimum variables values are: x1=6.32, x2=1.63



The minimum values is: 15.4


Q7.


The solution is unbounded


Q8.


The problem has no feasible solution


Q9.


The problem has multiple solutions with the following optimum variable values:



x1=2, x2 =1 or x1=2/3, x2=7/3




The Maximum objective function value is: 12










2.13 Further References

Mittal, K.V. 1976. Optimization Methods in Operations Research and Systems Analysis, Wiley Eastern
Ltd, New Delhi.

Taha, H.A1999. Operations Research An Introduction, PHI Ltd., New Delhi.

Richard I.Levin, David S. Rubin, Joel P. Stinson, Everette S.Gardner, Jr.1992. Quantitative Approaches
to Management, McGraw Hill, NJ.















35

MBA-H2040 Quantitative Techniques for Managers





















UNIT I












ON
S

3 LINEAR PROGRAMMING ? SIMPLEX METHOD

S
E



L





LESSON STRUCTURE




3.1 Introduction



3.2 Basics of Simplex Method



3.3 Simplex Method Computation



3.4 Simplex Method with More Than Two



Variables



3.5 Two Phase and M Method



3.5.1 Two Phase Method



3.5.2 M Method



3.6 Multiple Solutions



3.7 Unbounded Solution



3.8 Infeasible Solution



3.9 Summary



3.10

Key Terms



3.11

Self Assessment Questions



3.12

Key Solutions



3.13

3.13 Further References













36
MBA-H2040 Quantitative Techniques for Managers













Objectives
After Studying this lesson, you should be able
to:

Understand the basics of simplex method
Explain the simplex calculations
Describe various solutions of Simplex

Method

Understand two phase and M method













37

MBA-H2040 Quantitative Techniques for Managers
3.1 Introduction

The Linear Programming with two variables can be solved graphically. The graphical method of solving

linear programming problem is of limited application in the business problems as the number of

variables is substantially large. If the linear programming problem has larger number of variables, the

suitable method for solving is Simplex Method. The simplex method is an iterative process, through

which it reaches ultimately to the minimum or maximum value of the objective function.



The simplex method also helps the decision maker/manager to identify the following:

Redundant Constraints

Multiple Solutions

Unbounded Solution

Infeasible Problem


3.2 Basics of Simplex Method

The basic of simplex method is explained with the following linear programming problem.

Example 3.1:


Maximize





60x1 + 70x2



Subject to:





2x1 + x2 300





3x1 + 4x2 509





4x1 + 7x2 812






x1, x2 0


Solution


First we introduce the variables







s3, s4, s5 0



So that the constraints becomes equations, thus






2x1 + x2 + s3 = 300





3x1 + 4x2 + s4 = 509





4x1 + 7x2 + s5 = 812

Corresponding to the three constraints, the variables s3, s4, s5 are called as slack variables. Now, the

system of equation has three equations and five variables.


There are two types of solutions they are basic and basic feasible, which are discussed as follows:

Basic Solution

We may equate any two variables to zero in the above system of equations, and then the system will

have three variables. Thus, if this system of three equations with three variables is solvable such a

solution is called as basic solution.



38
MBA-H2040 Quantitative Techniques for Managers



For example suppose we take x1=0 and x2=0, the solution of the system with remaining three

variables is s3=300, s4=509 and s5=812, this is a basic solution and the variables s3, s4, and s5 are known

as basic variables where as the variables x1, x2 are known as non-basic variables.



The number of basic solution of a linear programming problem is depends on the presence of the

number of constraints and variables. For example if the number of constraints is m and the number of

variables including the slack variables is n then there are at most nCn-m = nCm basic solutions.

Basic Feasible Solution

A basic solution of a linear programming problem is called as basic feasible solutions if it is feasible it

means all the variables are non-negative. The solution s3=300, s4=509 and s5=812 is a basic feasible

solution.



The number of basic feasible solution of a linear programming problem is depends on the

presence of the number of constraints and variables. For example if the number of constraints is m and

the number of variables including the slack variables is n then there are at most nCn-m = nCm basic

feasible solutions.



Every basic feasible solution is an extreme point of the convex set of feasible solutions and every

extreme point is a basic feasible solution of the set of given constraints. It is impossible to identify the

extreme points geometrically if the problem has several variables but the extreme points can be

identified using basic feasible solutions. Since one the basic feasible solution will maximize or minimize

the objective function, the searching of extreme points can be carry out starting from one basic feasible

solution to another.



The Simplex Method provides a systematic search so that the objective function increases in the

cases of maximization progressively until the basic feasible solution has been identified where the

objective function is maximized.

3.3 Simplex Method Computation

This section describes the computational aspect of simplex method. Consider the following linear

programming problem






Maximize







60x1 + 70x2





Subject to:







2x1 + x2 + s3 = 300







3x1 + 4x2 + s4 = 509







4x1 + 7x2 + s5 = 812







x1, x2, s3, s4 ,s5 0





39

MBA-H2040 Quantitative Techniques for Managers




The profit Z=60x1 + 70x2 i.e. Maximize 60x1 + 70x2



The standard form can be summarized in a compact table form as




In this problem the slack variables s3, s4, and s5 provide a basic feasible solution from which the

simplex computation starts. That is s3==300, s4=509 and s5=812. This result follows because of the

special structure of the columns associated with the slacks.



If z represents profit then z=0 corresponding to this basic feasible solution. We represent by CB

the coefficient of the basic variables in the objective function and by XB the numerical values of the

basic variable.



So that the numerical values of the basic variables are: XB1=300, XB2=509, XB3=812. The profit

z=60x1+70x2 can also expressed as z-60x1-70x2=0. The simplex computation starts with the first

compact standard simplex table as given below:







CB

Basic

Cj

60 70 0 0 0





Variables XB

x1 x2 s3 s4 s5



















0

s3

300

2 1 1 0 0



0

s4

509

3 4 0 1 0



0

s5

812

4 7 0 0 1









z

-60 -70 0 0 0













Table 1




In the objective function the coefficients of the variables are CB1=CB2=CB3=0. The topmost row

of the Table 1 denotes the coefficient of the variables x1, x2, s3, s4, s5 of the objective function

respectively. The column under x1 indicates the coefficient of x1 in the three equations respectively.

Similarly the remaining column also formed.



On seeing the equation z=60x1+70x2 we may observe that if either x1 or x2, which is currently

non-basic is included as a basic variable so that the profit will increase. Since the coefficient of x2 is

higher we choose x2 to be included as a basic variable in the next iteration. An equivalent criterion of

choosing a new basic variable can be obtained the last row of Table 1 i.e. corresponding to z.



Since the entry corresponding to x2 is smaller between the two negative values, x2 will be

included as a basic variable in the next iteration. However with three constraints there can be only three

basic variables.



Thus, by bringing x2 a basic variable one of the existing basic variables becomes non-basic. The

question here is How to identify this variable? The following statements give the solution to this

question.



Consider the first equation i.e. 2x1 + x2 + s3 = 300



40
MBA-H2040 Quantitative Techniques for Managers

From this equation




2x1+s3=300-x2

But x1=0. Hence, in order that s30




300-x20





i.e. x2300

Similarly consider the second equation i.e. 3x1 + 4x2 + s4 = 509
From this equation




3x1+s4=509-4x2

But, x1=0. Hence, in order that s40




509-4x20





i.e. x2509/9


Similarly consider the third equation i.e. 4x1 + 7x2 + s5 = 812
From this equation




4x1+s5=812-7x2

But x1=0. Hence, in order that s50




812-7x20





i.e. x2812/7


Therefore the three equation lead to

x2300,

x2509/9,

x2812/7


Thus x2=Min (x2300, x2509/9, x2812/7) it means



x2=Min (x2300/1, x2509/9, x2812/7)=116


Therefore x2=116

If x2=116, you may be note from the third equation

7x2+s5=812



i.e. s5=0

Thus, the variable s5 becomes non-basic in the next iteration.

So that the revised values of the other two basic variables are



S3=300-x2=184



S4=509-4*116=45


Refer to Table 1, we obtain the elements of the next Table i.e. Table 2 using the following rules:


1. We allocate the quantities which are negative in the z-row. Suppose if all the quantities are

positive, the inclusion of any non-basic variable will not increase the value of the objective
function. Hence the present solution maximizes the objective function. If there are more than one
negative values we choose the variable as a basic variable corresponding to which the z value is
least as this is likely to increase the more profit.





41

MBA-H2040 Quantitative Techniques for Managers

2. Let xj be the incoming basic variable and the corresponding elements of the jth row column be

denoted by Y1j, Y2j and Y3j respectively. If the present values of the basic variables are XB1,
XB2 and XB3 respectively, then we can compute.



Min [XB1/Y1j, XB2/Y2j, XB3/Y3j] for Y1j, Y2j, Y3j>0.
Note that if any Yij0, this need not be included in the comparison. If the minimum occurs
corresponding to XBr/Yrj then the rth basic variable will become non-basic in the next iteration.


3. Using the following rules the Table 2 is computed from the Table 1.


i.

The revised basic variables are s3, s4 and x2. Accordingly, we make CB1=0, CB2=0
and CB3=70.


ii.

As x2 is the incoming basic variable we make the coefficient of x2 one by dividing
each element of row-3 by 7. Thus the numerical value of the element corresponding
to x1 is 4/7, corresponding to s5 is 1/7 in Table 2.


iii.

The incoming basic variable should appear only in the third row. So we multiply the
third-row of Table 2 by 1 and subtract it from the first-row of Table 1 element by
element. Thus the element corresponding to x2 in the first-row of Table 2 is 0.



Therefore the element corresponding to x1 is

2-1*4/7=10/7 and the element corresponding to s5 is
0-1*1/7=-1/7




In this way we obtain the elements of the first and the second row in Table 2. In Table 2 the

numerical values can also be calculated in a similar way.




CB

Basic

Cj

60 70 0 0 0





Variables XB

x1 x2 s3 s4 s5













0

s3

184

10/7 0 1 0 -1/7



0

s4

45

5/7 0 0 1 -4/7



70

x2

116

4/7 1 0 0 1/7













zj-cj

-140/7 0 0 0 70/7













Table 2




Let CB1, CB2, Cb3 be the coefficients of the basic variables in the objective function. For

example in Table 2 CB1=0, CB2=0 and CB3=70. Suppose corresponding to a variable J, the quantity zj is

defined as zj=CB1, Y1+CB2, Y2j+CB3Y3j. Then the z-row can also be represented as Zj-Cj.

For example:


z1 - c1 = 10/7*0+5/7*0+70*4/7-60 = -140/7




z5 ? c5 = -1/7*0-4/7*0+1/7*70-0 = 70/7


1. Now we apply rule (1) to Table 2. Here the only negative zj-cj is z1-c1 = -140/7

Hence x1 should become a basic variable at the next iteration.



42
MBA-H2040 Quantitative Techniques for Managers






2. We compute the minimum of the ratio


184 , 45, 116



644 , 63 , 203

Min 10 5 4

= Min 5



= 63

7 7 7




This minimum occurs corresponding to s4, it becomes a non basic variable in next iteration.

3. Like Table 2, the Table 3 is computed sing the rules (i), (ii), (iii) as described above.




CB

Basic

Cj

60 70 0 0 0





Variables XB

x1 x2 s3 s4 s5













0

s3

94

0 0 1 -2 1



60

x1

63

1 0 0 7/5 -4/5



70

x2

80

0 1 0 -4/5 3/5













zj-cj

0 0 0 28 -6













Table 3



1. z5 ? c5 < 0 should be made a basic variable in the next iteration.



2. Now compute the minimum ratios



94, 80
Min 1 3

= 94

5



Note: Since y25 = -4/5 < 0, the corresponding ratio is not taken for comparison.



The variable s3 becomes non basic in the next iteration.



3. From the Table 3, Table 4 is calculated following the usual steps.




CB

Basic

Cj

60 70 0 0 0





Variables XB

x1 x2 s3 s4 s5



















0

s5

94

0 0 1 -2 1



60

x1

691/5

1 0 4/5 -1/5 0



70

x2

118/5

0 1 -3/5 2/5 0













zj-cj

0 0 6 16 0



43

MBA-H2040 Quantitative Techniques for Managers


Note that zj ? cj 0 for al j, so that the objective function can't be improved any further.

Thus, the objective function is maximized for x1 = 691/5 and x2=118/5 and

The maximum value of the objective function is 9944.


3.4

Simplex Method with More Than Two Variables



In previous section we discussed the simplex method of linear programming problem with two decision
variables. The simplex method computational procedure can be readily extended to linear programming
problems with more than two variables. This is illustrated in this section with the help of the product
mix problem given in the following Example 3.2.

Example 3.2

An organization has three machine shops viz. A, B and C and it produces three product viz. X, Y and Z
using these three machine shops. Each product involves the operation of the machine shops. The time
available at the machine shops A, B and C are 100, 72 and 80 hours respectively. The profit per unit of
product X, Y and Z is $22, $6

and $2 respectively. The







following table shows the time

required for each operation for

10

7

2

unit amount of each product.

Determine

an

appropriate







product mix so as to maximize

the profit.

2

3

4













Machine

1

2

1

Shops





A B C



Products



Profit/unit








X $22








Y $6








Z $2






Available Hours 100 72 80


Solution

First we have to develop linear programming formulation. The linear programming formulation of the
product mix problem is:







Maximize









22x1 + 6x2 + 2x3








Subject to:









10x1 + 2x2 + x3 100









7x1 + 3x2 + 2x3 72









2x1 + 4x2 + x3 80










x1, x2, x3 0




44
MBA-H2040 Quantitative Techniques for Managers
We introduce slack variables s4, s5 and s6 to make the inequalities equation.


Thus, the problem can be stated as







Maximize









22x1 + 6x2 + 2x3








Subject to:









10x1 + 2x2 + x3 + s4 = 100









7x1 + 3x2 + 2x3 + s5 = 72









2x1 + 4x2 + x3 + s6 = 80









x1, x2, x3, s4, s5, s6 0



From the above equation the simplex Table 1 can be obtained in a straight forward manner. Here

the basic variables are s4, s5 and s6. Therefore CB1 = CB2 = CB3 = 0.


CB

Basic

Cj

22

6

2

0

0

0



Variable XB

x1

x2

x3

s4

s5

s6



















0

s4

100

10

2

1

1

0

0

0

s5

72

7

3

2

0

1

0

0

s6

80

2

4

1

0

0

1




zj-cj



-22

-6

-2

0

0

0













Table 1



1. z1-c1 = -22 is the smallest negative value. Hence x1 should be taken as a basic variable in the next

iteration.

2. Calculate the minimum of the ratios





Min 100 , 72 , 80 = 10







10 7 2





The variable s4 corresponding to which minimum occurs is made a non basic variable.
3. From the Table 1, the Table 2 is calculated using the following rules:


i.

The revised basic variables are x1, s5, s6. Accordingly we make CB1=22, CB2=0 and
CB3=0.

ii.

Since x1 is the incoming variable we make x1 coefficient one by dividing each
element of row 1 by 10. Thus the numerical value of the element corresponding to x2
is 2/10, corresponding to x3 is 1/10, corresponding to s4 is 1/10, corresponding to s5 is
0/10 and corresponding to s6 is 0/10 in Table 2.

iii.

The incoming basic variable should only appear in the first row. So we multiply first
row of Table 2 by 7 and subtract if from the second row of Table 1 element by
element.
Thus,
The element corresponding to x1 in the second row of Table 2 is zero

The element corresponding to x2 is 3 ? 7 * 2 = 16












10 10











By using this way we get the elements of the second and the third row in Table 2.



45

MBA-H2040 Quantitative Techniques for Managers


Similarly, the calculation of numerical values of basic variables in Table 2 is done.



CB

Basic

Cj

22

6

2

0

0

0



Variable XB

x1

x2

x3

s4

s5

s6



















22

x1

10

1

2/10

1/10

1/10

0

0

0

s5

2

0

16/10 13/10 -7/10

1

0

0

s6

60

0

18/5

4/5

-1/5

0

1




zj-cj



0

-8/5

1/5

12/5

0

0













Table 2



1. z2-c2 = -8/5. So x2 becomes a basic variable in the next iteration.
2. Calculate the minimum of the ratios









10, 7 , 60

Min 2 16 18 = Min 50, 70, 300 = 70







10 10 5 16 18 16





Hence the variable s5 will be a non basic variable in the next iteration.



3. From Table 2, the Table 3 is calculated using the rules (i), (ii) and (iii) mentioned above.


CB

Basic

Cj

22

6

2

0

0

0



Variable XB

x1

x2

x3

s4

s5

s6



















22

x1

73/8

1

0

-1/16

3/16

-1/8

0

6

x2

30/8

0

1

13/16 -7/16

5/8

0

0

s6

177/4 0

0

-17/8

11/8

-9/4

1




zj-cj



0

0

24/16 24/16 1

0













Table 3


Note that all zj ? cj 0, so that the solution is x1 = 73/8, x2 = 30/8 and s6 = 177/4 maximizes the
objective function.

The Maximum Profit is: 22*73/8 + 6*30/8 = 1606/8 + 180/8













= 1786/8 = $223.25









3.5 Tow Phase and M-Method

In the last two section we discussed the simplex method was applied to linear programming problems

with less than or equal to () type constraints. Thus, there we could introduce slack variables which

provide an initial basic feasible solution of the problem.



46
MBA-H2040 Quantitative Techniques for Managers


Generally, the linear programming problem can also be characterized by the presence of both

`less than or equal to' type or `greater than or equal to ()' type constraints. In such case it is not always

possible to obtain an initial basic feasible solution using slack variables.



The greater than or equal to type of linear programming problem can be solved by using the

following methods:



1. Two Phase Method
2. M Method

In this section we will discuss these two methods.

3.5.1 Two Phase Method

We discuss the Two Phase Method with the help of the following Example 3.3.

Example 3.3




Minimize







12.5x1 + 14.5x2






Subject to:







x1 + x2 2000







0.4x1 + 0.75x2 1000







0.075x1 + 0.1x2 200








x1, x2 0

Solution

Here the objective function is to be minimized; the values of x1 and x2 which minimized this objective
function are also the values which maximize the revised objective function i.e.





Maximize







-12.5x1 ? 14.5x2




We can multiply the second and the third constraints by 100 and 1000 respectively for the

convenience of calculation.



Thus, the revised linear programming problem is:





Maximize







-12.5x1 ? 14.5x2





Subject to:







x1 + x2 2000







40x1 + 75x2 100000







75x1 + 100x2 200000







x1, x2 0




Now we convert the two inequalities by introducing surplus variables s3 and s4 respectively.

The third constraint is changed into an equation by introducing a slack variable s5.



Thus, the linear programming problem becomes as



47

MBA-H2040 Quantitative Techniques for Managers


Maximize





-12.5x1 ? 14.5x2 = -25/2x1 ? 29/2x2




Subject to:





x1 + x2 -s3

= 2000





40x1 + 75x2 -s4 = 100000





75x1 + 100x2 +s5 = 200000






x1, x2,s3,s4,s5 0



Even though the surplus variables can convert greater than or equal to type constraints into

equations they are unable to provide initial basic variables to start the simplex method calculation. So

we may have to introduce two more additional variables a6 and a7 called as artificial variable to

facilitate the calculation of an initial basic feasible solution.




In this method the calculation is carried out in tow phases hence tow phase method.


Phase I


In this phase we will consider the following linear programming problem






Maximize







-a6 ?a7






Subject to:







x1 + x2 -s3 +a6 = 2000







40x1 + 75x2 -s4 + a7 = 100000







75x1 + 100x2 +s5 = 200000








x1, x2.s3,s4,s5,a6,a7 0







The initial basic feasible solution of the problem is







a6 = 2000, a7=100000 and s5 = 200000.




As the minimum value of the objective function of the Phase I is zero at the end of the Phase I

calculation both a6 and a7 become zero.


CB

Basic

C

0

0

0

0

0

-1

-1



j



variables XB

x

x

s

s

s

a

a



1

2

3

4

5

6

7























-1

a

2000

1

1

-1

0

0

1

0



6

-1

a

100000 40

75

0

-1

0

0

1



7

0

s

200000 75

100

0

0

1

0

0



5



z



j-cj

-41

-76

1

1

0

0

0













Table 1



Here x2 becomes a basic variable and a7 becomes non basic variable in the next iteration. It is no

longer considered for re-entry in the table.




48
MBA-H2040 Quantitative Techniques for Managers


CB

Basic

Cj

0

0

0

0

0

-1





variables XB

x1

x2

s3

s4

s5

a6























-1

a6

2000/3

7/15

0

-1

1/75

0

1











































0

x2

4000/3

8/15

1

0

-1/75

0

0

























0

s5

200000/3 65/3

0

0

4/3

1

0







zj-cj

-1/15

0

1

-1/75

0

0




Table 2




Then x1 becomes a basic variable and a6 becomes a non basic variable in the next iteration.




CB

Basic

Cj

0

0

0

0

0





variables XB

x1

x2

s3

s4

s5





















0

x1

10000/7

1

0

-15/7

1/35

0







































0

x2

4000/7

0

1

8/7

-1/35

0







































0

s5

250000/7 0

0

325/7

16/21

1





zj-cj

0

0

0

0

0














Table 3



The calculation of Phase I end at this stage. Note that, both the artificial variable have been

removed and also found a basic feasible solution of the problem.



The basic feasible solution is:








x1 = 10000/7, x2 = 4000/2, s5 = 250000/7.


Phase II

The initial basic feasible solution obtained at the end of the Phase I calculation is used as the initial basic
feasible solution of the problem. In this Phase II calculation the original objective function is introduced
and the usual simplex procedure is applied to solve the linear programming problem.




49

MBA-H2040 Quantitative Techniques for Managers


CB

Basic

Cj

-25/2

-29/2

0

0

0





variables XB

x1

x2

s3

s4

s5





















-25/2

x1

10000/7

1

0

-15/7

1/35

0







































-29/2

x2

4000/7

0

1

8/7

-1/35

0







































0

s5

250000/7 0

0

325/7

5/7

1





zj-cj

0

0

143/14 2/35

0














Table 1




In this Table 1 all zj-cj 0 the current solution maximizes the revised objective function.




Thus, the solution of the problem is:








x1 = 10000/7 = 1428 and x2 = 4000/7 = 571.4 and




The Minimum Value of the objective function is: 26135.3


3.5.2 M Method

In this method also we need artificial variables for determining the initial basic feasible solution. The M

method is explained in the next Example 3.4 with the help of the previous Example 3.3.


Example 3.4




Maximize







-12.5x1 ? 14.5x2






Subject to:







x1 + x2 ?s3 = 2000







40x1 + 75x2 -s4 = 100000







75x1 + 100x2 +s5 = 200000








x1, x2, s3, s4, s5 0.





Introduce the artificial variables a6 and a7 in order to provide basic feasible solution in the second and

third constraints. The objective function is revised using a large positive number say M.



Thus, instead of the original problem, consider the following problem i.e.




Maximize







-12.5x1 ? 14.5x2 ? M (a6 + a7)



Subject to:







x1 + x2 ?s3 + a6 = 2000







40x1 + 75x2 -s4 +a7 = 100000







75x1 + 100x2 +s5 = 200000







x1, x2, s3, s4, s5, a6, a7 0.



50
MBA-H2040 Quantitative Techniques for Managers
The coefficient of a6 and a7 are large negative number in the objective function. Since the objective

function is to be maximized in the optimum solution, the artificial variables will be zero. Therefore, the

basic variable of the optimum solution are variable other than the artificial variables and hence is a basic

feasible solution of the original problem.



The successive calculation of simplex tables is as follows:



CB

Basic

Cj

-12.5

-14.5 0

0

0

-M

-M



variables XB

x1

x2

s3

s4

s5

a6

a7





















-M

a6

2000

1

1

-1

0

0

1

0









































-M

a7

100000 40

75

0

-1

0

0

1









































0

s5

200000 75

100

0

0

1

0

0



zj-cj

-41M

-76M M

M

0

0

0

+12.5

+14.5













Table 1



Since M is a large positive number, the coefficient of M in the zj ? cj row would decide the

entering basic variable. As -76M < -41M, x2 becomes a basic variable in the next iteration replacing a7.

The artificial variable a7 can't be re-entering as basic variable.



CB

Basic

Cj

-12.5

-14.5 0

0

0

-M



variables XB

x1

x2

s3

s4

s5

a6



















-M

a6

2000/3

7/15

0

-1

1/75

0

1





































-14.5

x2

4000/3

8/15

1

0

-1/75

0

0





































0

s5

200000/3 65/3

0

0

4/3

1

0



zj-cj

-7/15M

0

M

-M/75

0

0

+143/30



+29/150













Table 2




Now x1 becomes a basic variable replacing a6. Like a7 the variable a6 also artificial variable so it

can't be re-entering in the table.




51

MBA-H2040 Quantitative Techniques for Managers


CB

Basic

Cj

-12.5

-14.5 0

0

0





variables XB

x1

x2

s3

s4

s5























-12.5

x1

10000/7

1

0

-15/7

1/35

0







































-14.5

x2

4000/7

0

1

8/7

-1/35

0







































0

s5

250000/7 0

0

325/7

16/21

1





zj-cj

0

0

143/14 2/35

0


















Table 3




Hence

The optimum solution of the problem is x1 = 10000/7, x2 = 4000/7 and

The Minimum Value of the Objective Function is: 26135.3


3.6 Multiple Solutions

The simplex method also helps in identifying multiple solutions of a linear programming problem. This
is explained with the help of the following Example 3.5.



Example 3.5



Consider the following linear programming problem.




Maximize





2000x1 + 3000x2




Subject to:





6x1 + 9x2 100





2x1 + x2 20






x1, x2 0.


Solution


Introduce the slack variables s3 and s4, so that the inequalities can be converted in to equation as

follows:




6x1 + 9x2 + s3 = 100





2x1 + x2 + s4 = 20






x1, x2, s3, s4 0.



The computation of simple procedure and tables are as follows:





52
MBA-H2040 Quantitative Techniques for Managers


CB

Basic

Cj

2000

3000

0

0





variables XB

x1

x2

s3

s4



































0

s3

100

6

9

1

0



































0

s4

20

2

1

0

1













































zj-cj

-2000 -3000 0

0














Table 1




CB

Basic

Cj

2000

3000

0

0





variables XB

x1

x2

s3

s4



































0

x2

100/9

2/3

1

1/9

0



































0

s4

80/9

4/3

0

-1/9

1



















































zj-cj

0

0

3000/9 0














Table 2




Here zj-cj 0 for al the variables so that we can't improve the simplex table any more. Hence it

is optimum.



The optimum solution is x1 = 0, x2 = 100/9 and




The maximum value of the objective function is: 100000/3 = 33333.33.




However, the zj-cj value corresponding to the non basic variable x1 is also zero. This indicates

that there is more than one optimum solution for the problem exists.



In order to calculate the value of the alternate optimum solution we have to introduce x1 as a

basic variable replacing s4. The next Table 3 shows the computation of this.



53

MBA-H2040 Quantitative Techniques for Managers


CB

Basic

C

2000

3000

0

0



j



variables XB

x

x

s

s



1

2

3

4

































3000

x

20/3

0

1

1/6

1/2



2

































2000

x

20/3

1

0

-1/12

3/4



1

















































z





j-cj

0

0

1000/3 3000

















Table 3




Thus,





x1 = 20/3, x2 = 20/3 also maximize the objective function and






The Maximum value of the objective function is: 100000/3 = 33333.33




Thus, the problem has multiple solutions.


3.7 Unbounded Solution

In this section we will discuss how the simplex method is used to identify the unbounded solution. This
is explained with the help of the following Example 3.6.

Example 3.6


Consider the following linear programming problem.





Maximize







5x1 + 4x2





Subject to:









x1 ? x2 8







x1 7








x1, x2 0.


Solution :

Introduce the slack variables s3 and s4, so that the inequalities becomes as equation as follows:






x1 + s3 = 7
x1 ? x2 + s4 = 8








x1, x2, s3, s4 0.




The calculation of simplex procedures and tables are as follows:





54
MBA-H2040 Quantitative Techniques for Managers


CB

Basic

C

5

4

0

0



j



variables XB

x

x

s

s



1

2

3

4

































0

s

7

1

0

1

0



3

















0

s

8

1

-1

0

1



4































z



j-cj

-5

-4

0

0













Table 1




CB

Basic

C

5

4

0

0



j



variables XB

x

x

s

s



1

2

3

4

































5

x

7

1

0

1

0



1

















0

s

1

0

-1

-1

1



4































z



j-cj

0

-4

5

0













Table 2




Note that z2-c2 < 0 which indicates x2 should be introduced as a basic variable in the next

iteration. However, both y120, y220.



Thus, it is not possible to proceed with the simplex method of calculation any further as we

cannot decide which variable will be non basic at the next iteration. This is the criterion for unbounded

solution.


NOTE: If in the course of simplex computation zj-cj < 0 but yij 0 for al i then the problem has no
finite solution.



But in this case we may observe that the variable x2 is unconstrained and can be increased

arbitrarily. This is why the solution is unbounded.

3.8 Infeasible Solution




This section illustrates how to identify the infeasible solution using simplex method. This is explained

with the help of the following Example 3.7.


Example 3.7



55

MBA-H2040 Quantitative Techniques for Managers
Consider the following problem.




Minimize







200x1 + 300x2











Subject to:







2x1 + 3x2 1200







x1 + x2 400







2x1 + 3/2x2 900








x1, x2 0

Solution

Since it is a minimization problem we have to convert it into maximization problem and introduce the

slack, surplus and artificial variables. The problem appears in the following manner after doing all these

procedure.





Maximize









-200x1 - 300x2



Subject to:







2x1 + 3x2 -s3 +a6 = 1200







x1 + x2 +s4 = 400







2x1 + 3/2x2 - s5 + a7 = 900








x1, x2, s3, s4, s5, a6, a7 0


Here the a6 and a7 are artificial variables. We use two phase method to solve this problem.

Phase I


Maximize





-a6 ?a7



Subject to:





2x1 + 3x2 -s3 +a6 = 1200





x1 + x2 +s4 = 400





2x1 + 3/2x2 - s5 + a7 = 900






x1, x2, s3, s4, s5, a6, a7 0






The calculation of simplex procedures and tables are as follows:



CB

Basic

C

0

0

0

0

0

-1

-1



j



variables XB

x

x

s

s

s

a

a



1

2

3

4

5

6

7























-1

a

1200

2

3

-1

0

0

1

0



6

0

s

400

1

1

0

1

0

0

1



4

-1

a

900

2

3/2

0

0

-1

0

0



7



z



j-cj

-4

-9/2

1

0

1

0

0













Table 1




56
MBA-H2040 Quantitative Techniques for Managers


CB

Basic

C

0

0

0

0

0

-1



j



variables XB

x

x

s

s

s

a



1

2

3

4

5

7





















0

x

400

2/3

1

-1/3

0

0

0



2

0

s

0

1/3

0

1/3

1

0

0



4

-1

a

300

1

0

1/2

0

-1

1



7



z



j-cj

-1

0

-1/2

0

1

0













Table 2




CB

Basic

C

0

0

0

0

0

-1



j



variables XB

x

x

s

s

s

a



1

2

3

4

5

7





















0

x

400

0

1

-1

-2

0

0



2

0

x

0

1

0

1

3

0

0



1

-1

a

300

0

0

-1/2

-3

-1

1



7



z



j-cj

0

0

1/2

3

1

0













Table 3




Note that zj-cj 0 for al the variables but the artificial variable a7 is still a basic variable. This

situation indicates that the problem has no feasible solution.

3.9 Summary

The simplex method is very useful and appropriate method for solving linear programming problem

having more than tow variables. The slack variables are introduced for less that or equal to type, surplus

variables are introduce for greater than or equal to type of linear programming problem. The basic

feasible solution is important in order to solve the problem using the simplex method.



A basic feasible solution of a system with m-equations and n-variables has m non-negative

variables called as basic variables and n-m variables with value zero known as non-basic variables. The

objective function is maximized or minimized at one of the basic feasible solutions.



Surplus variables can't provide the basic feasible solution instead artificial variables are used to

get the basic feasible solutions and it initiate the simplex procedure. Two phase and M-Method are

available to solve linear programming problem in these case.




The simplex method also used to identify the multiple, unbounded and infeasible solutions.


3.10 Key Terms




Basic Variable: Variable of a basic feasible solution has n non-negative value.

Non Basic Variable: Variable of a feasible solution has a value equal to zero.




57

MBA-H2040 Quantitative Techniques for Managers
Artificial Variable: A non-negative variable introduced to provide basic feasible solution and initiate
the simplex procedures.

Slack Variable: A variable corresponding to a type constraint is a non-negative variable introduced to
convert the inequalities into equations.

Surplus Variable: A variable corresponding to a type constraint is a non-negative variable introduced
to convert the constraint into equations.

Basic Solution: System of m-equation and n-variables i.e. m<n is a solution where at least n-m
variables are zero.

Basic Feasible Solution: System of m-equation and n-variables i.e. m<n is a solution where m variables
are non-negative and n-m variables are zero.

Optimum Solution: A solution where the objective function is minimized or maximized.

3.11 Self Assessment Questions

Q1. A soft drinks company has a two products viz. Coco-cola and Pepsi with profit of $2 an $1 per unit.
The following table illustrates the labour, equipment and materials to produce per unit of each product.
Determine suitable product mix which maximizes the profit using simplex method.







Pepsi Coco-cola Total Resources














Labour

12

3

2



1

2.3





Equipment

6.9





1

1.4





Material

4.9





Q2. A factory produces three using three types of ingredients viz. A, B and C in different proportions.
The following table shows the requirements o various ingredients as inputs per kg of the products.













Ingredients



Products



A



B



C


1









4

8

8



2

4

6

4







3

8

4

0






The three profits coefficients are 20, 20 and 30 respectively. The factory has 800 kg of ingredients A,
1800 kg of ingredients B and 500 kg of ingredient C.

Determine the product mix which will maximize the profit and also find out maximum profit.



58
MBA-H2040 Quantitative Techniques for Managers

Q3. Solve the following linear programming problem using two phase and M method.



Maximize





12x1 + 15x2 + 9x3



Subject to:





8x1 + 16x2 + 12x3 250





4x1 + 8x2 + 10x3 80





7x1 + 9x2 + 8x3 =105





x1, x2, x3 0


Q4. Solve the following linear programming problem using simplex method.



Maximize





3x1 + 2x2



Subject to:





x1 ?x2 1





x1 + x2 3





x1, x2 0


Q5. Solve the following linear programming problem using simplex method.


Maximize





x1 + x2



Subject to:





-2x1 + x2 1





x1 2





x1 + x2 3





x1, x2, x3 0





Q6. Maximize




P = 3x1 + 4x2 + x3

Subject to:




x1 + 2x2 + x3 6





2x1 +2x3 4





3x1 + x2 + x3 9





x1, x2, x3 0

3.12 Key Solutions

Q1. Coco-Cola = 20/9, Pepsi = 161/90
Maximum Profit = $6.23

Q2. x1 = 0, x2 = 125, x3 = 75/2
Maximum Profit = 5375

Q3. x1 = 6, x2 = 7, x3 = 0
Maximum Profit = 177

Q4. Unbounded Solution

Q5. x1 = 2, x2 = 1 or x1 = 2/3, x2 = 7/3



59

MBA-H2040 Quantitative Techniques for Managers
Maximum Profit = 3.

Q6. x1 = 2, x2 = 2, x3 = 0
Maximum P = 14


3.13. Further References

Hamdy A Taha, 1999. Introduction to Operations Research, PHI Limited, New Delhi.

Mustafi, C.K. 1988. Operations Research Methods and Practices, Wiley Eastern Limited, New Delhi.

Levin, R and Kirkpatrick, C.A. 1978. Quantitative Approached to Management, Tata McGraw Hill,
Kogakusha Ltd., International Student Edition.

Peterson R and Silver, E. A. 1979. Decision Systems for Inventory Management and Production
Planning.

Handley, G and T.N. Whitin. 1983. Analysis of Inventory Systems, PHI.

Starr, M.K. and D.W. Miller. 1977. Inventory Control Theory and Practice, PHI.




















UNIT I







60
MBA-H2040 Quantitative Techniques for Managers









ON
S

4 DUAL LINEAR PROGRAMMING PROBLEMS

S
E



L







LESSON STRUCTURE




4.1 Introduction



4.2 Dual Problem Formulation



4.3 Dual Problem Properties



4.4 Simple Way of Solving Dual Problem



4.5 Summary



4.6 Key Terms



4.7 Self Assessment Questions



4.8 Key Solutions



4.9 Further References

Objectives



After Studying this lesson, you should be able



to:



Understand the Dual Linear programming



Problem



Formulate a Dual Problem



Solve the Dual Linear Programming



Problem



Understand the Properties of a Dual



Problem

































61

MBA-H2040 Quantitative Techniques for Managers
4.1 Introduction




For every linear programming problem there is a corresponding linear programming problem called the

dual. If the original problem is a maximization problem then the dual problem is minimization problem

and if the original problem is a minimization problem then the dual problem is maximization problem.

In either case the final table of the dual problem will contain both the solution to the dual problem and

the solution to the original problem.



The solution of the dual problem is readily obtained from the original problem solution if the

simplex method is used.



The formulation of the dual problem also sometimes referred as the concept of duality is helpful

for the understanding of the linear programming. The variable of the dual problem is known as the dual

variables or shadow price of the various resources. The dual problem is easier to solve than the original

problem. The dual problem solution leads to the solution of the original problem and thus efficient

computational techniques can be developed through the concept of duality. Finally, in the competitive

strategy problem solution of both the original and dual problem is necessary to understand the complete

problem.

4.2 Dual Problem Formulation




If the original problem is in the standard form then the dual problem can be formulated using the

following rules:

The number of constraints in the original problem is equal to the number of dual variables. The

number of constraints in the dual problem is equal to the number of variables in the original
problem.

The original problem profit coefficients appear on the right hand side of the dual problem

constraints.

If the original problem is a maximization problem then the dual problem is a minimization

problem. Similarly, if the original problem is a minimization problem then the dual problem is a
maximization problem.

The original problem has less than or equal to () type of constraints while the dual problem has

greater than or equal to () type constraints.

The coefficients of the constraints of the original problem which appear from left to right are

placed from top to bottom in the constraints of the dual problem and vice versa.


The Dual Linear Programming Problem is explained with the help of the following Example 4.1.


Example 4.1



Consider the following product mix problem:





62
MBA-H2040 Quantitative Techniques for Managers
Three machine shops A, B, C produces three types of products X, Y, Z respectively. Each product

involves operation of each of the machine shops. The time required for each operation on various

products is given as follows:



Machine Shops





Products

A

B



C

Profit per unit















X

$12

10

7

2















Y

$3

2

3

4















Z

$1

1

2

1



Available Hours

100

77 80


The available hours at the machine shops A, B, C are 100, 77, and 80 only. The profit per unit of

products X, Y, and Z is $12, $3, and $1 respectively.

Solution:



The formulation of Linear Programming (original problem) is as follows:

Maximize




12x1 + 3x2 + x3

Subject to:




10x1 + 2x2 + x3 100





7x1 + 3x2 + 2x3 77





2x1 + 4x2 + x3 80











x1, x2, x3 0


We introduce the slack variables s4, s5 and s6 then the equalities becomes as:





Maximize







12x1 + 3x2 + x3









Subject to:







10x1 + 2x2 + x3

+ s4 = 100







7x1 + 3x2 + 2x3

+ s5 = 77







2x1 + 4x2 + x3

+s6 = 80













x1, x2, x3, s4, s5, s6 0


Form the above equations, the first simplex table is obtained is as follows:


CB

Basic

Cj

12

3

1

0

0

0



Variable XB

x1

x2

x3

s4

s5

s6



















0

s4

100

10

2

1

1

0

0



63

MBA-H2040 Quantitative Techniques for Managers

0

s5

77

7

3

2

0

1

0

0

s6

80

2

4

1

0

0

1




zj-cj



-12

-3

-1

0

0

0











Table 1

Note that the basic variables are s4, s5 and s6. Therefore CB1 = 0, CB2 = 0, CB3 = 0.








1. The smallest negative element in the above table of z1 ? c1 is -12. Hence, x1 becomes a basic

variable in the next iteration.



2. Determine the minimum ratios



Min 100, 72, 80 = 10







10 7 2





Here the minimum value is s4, which is made as a non-basic variable.



3. The next Table 2 is calculated using the following rules:


(i)

The revised basic variables are x1, s5, s6. Accordingly we make CB1=22, CB2=0 and
CB3=0.


(ii)

Since x1 is the incoming variable we make x1 coefficient one by dividing each
element of row 1 by 10. Thus the numerical value of the element corresponding to x2
is 2/10, corresponding to x3 is 1/10, corresponding to s4 is 1/10, corresponding to s5 is
0/10 and corresponding to s6 is 0/10 in Table 2.



(iii)

The incoming basic variable should only appear in the first row. So we multiply first
row of Table 2 by 7 and subtract if from the second row of Table 1 element by
element.
Thus,
The element corresponding to x1 in the second row of Table 2 is zero

The element corresponding to x2 is 3 ? 7 * 2 = 16












10 10











By using this way we get the elements of the second and the third row in Table 2.

Similarly, the calculation of numerical values of basic variables in Table 2 is done.



CB

Basic

Cj

22

6

2

0

0

0



Variable XB

x1

x2

x3

s4

s5

s6



















12

x1

10

1

2/10

1/10

1/10

0

0

0

s5

7

0

16/10 13/10 -7/10

1

0

0

s6

60

0

18/5

4/5

-1/5

0

1




zj-cj



0

-3/5

1/5

6/5

0

0














Table 2



64
MBA-H2040 Quantitative Techniques for Managers


4. z2-c2 = -3/5. So x2 becomes a basic variable in the next iteration.







5. Determine the minimum of the ratios









10, 7 , 60

Min 2 16 18 = Min 50, 70, 300 = 70







10 10 5 16 18 16





So that the variable s5 will be a non basic variable in the next iteration.



6. From Table 2, the Table 3 is calculated using the rules (i), (ii) and (iii) mentioned above.


CB

Basic

Cj

12

3

1

0

0

0



Variable XB

x1

x2

x3

s4

s5

s6



















12

x1

73/8

1

0

-1/16

3/16

-1/8

0

3

s5

35/8

0

1

13/16 -7/16

5/8

0

0

s6

177/4 0

0

-17/8

11/8

-9/4

1




zj-cj



0

0

11/16 15/16 3/8

0











Table 3


Since all the zi ? cj 0, the optimum solution is as:







x1 = 73/8 and x2 = 35/8 and


The Maximum Profit is: $981/8 = $122.625

Suppose an investor is deciding to purchase the resources A, B, C. What offers is he going to produce?

Let, assume that W1, W2 and W3 are the offers made per hour of machine time A, B and C respectively.
Then these prices W1, W2 and W3 must satisfy the conditions given below:


1. W1, W2, W3 0

2. Assume that the investor is behaving in a rational manner; he would try to bargain as much as

possible so that the total annual payable to the produces would be as little as possible. This leads
to the following condition:



Minimize


100W1 + 77W2 + 80W3





65

MBA-H2040 Quantitative Techniques for Managers

3. The total amount offer by the investor to the three resources viz. A, B and C required to produce

one unit of each product must be at least as high as the profit gained by the producer per unit.



Since, these resources enable the producer to earn the specified profit corresponding to the
product he would not like to sell it for anything less assuming he is behaving rationally. This
leads to the following conditions:




10w1 + 7w2 + 2w3 12





2w1 + 3w2 + 4w3 3





w1 + 2w2 + w3 1


Thus, in this case we have a linear problem to ascertain the values of the variable w1, w2, w3. The
variables w1, w2 and w3 are called as dual variables.


Note:


The original (primal) problem illustrated in this example



a. considers the objective function maximization
b. contains type constraints
c. has non-negative constraints

This original problem is called as primal problem in the standard form.


4.3 Dual Problem Properties






The following are the different properties of dual programming problem:



i.

If the original problem is in the standard form, then the dual problem solution is obtained
from the zj ? cj values of slack variables.

For example: In the Example 4.1, the variables s4, s4 and s6 are the slack variables. Hence
the dual problem solution is w1 = z4 ? c4 = 15/16, w2 = z5 ? c5 = 3/8 and w3 = z6 ? c6 = 0.


ii.

The original problem objective function maximum value is the minimum value of the dual
problem objective function.



For example:




From the above Example 4.1 we know that the original problem maximum values

is 981/8 = 122.625. So that the minimum value of the dual problem objective function is







100*15/16 + 77*3/8 + 80*0 = 981/8

Here the result has an important practical implication. If both producer and investor analyzed
the problem then neither of the two can outmaneuver the other.




iii.

Shadow Price: A resource shadow price is its unit cost, which is equal to the increase in
profit to be realized by one additional unit of the resource.



For example:




Let the minimum objective function value is expressed as:





66
MBA-H2040 Quantitative Techniques for Managers





100*15/16 + 77*3/8 + 80*0

If the first resource is increased by one unit the maximum profit also increases by 15/16,
which is the first dual variable of the optimum solution. Therefore, the dual variables are also
referred as the resource shadow price or imputed price. Note that in the previous example the
shadow price of the third resource is zero because there is already an unutilized amount, so
that profit is not increased by more of it until the current supply is totally exhausted.


iv.

In the originals problem, if the number of constraints and variables is m and n then the
constraint and variables in the dual problem is n and m respectively. Suppose the slack
variables in the original problem is represented by y1, y1, ....., yn and the surplus variables are
represented by z1, z2, ..., zn in the dual problem.


v.

Suppose, the original problem is not in a standard form, then the dual problem structure is
unchanged. However, if a constraint is greater than or equal to type, the corresponding dual
variable is negative or zero. Similarly, if a constraint in the original problem is equal to type,
then the corresponding dual variables is unrestricted in sign.


Example 4.2



Consider the following linear programming problem




Maximize





22x1 + 25x2 +19x3




Subject to:





18x1 + 26x2 + 22x3 350





14x1 + 18x2 + 20x3 180





17x1 + 19x2 + 18x3 = 205





x1, x2, x3 0


Note that this is a primal or original problem.

The corresponding dual problem for this problem is as follows:



Minimize





250w1 + 80w2 +105w3







Subject to:





18w1 + 4w2 + 7w3 22





26w1 + 18w2 + 19w3 25





22w1 + 20w2 + 18w3 19






w1 0, w2 , and w3 is unrestricted in sign (+ or -).


Now, we can solve this using simplex method as usual.

4.4 Simple Way of Solving Dual Problem

Solving of dual problem is simple; this is illustrated with the help of the following Example 4.3.




67

MBA-H2040 Quantitative Techniques for Managers
Example 4.3:





Minimize







P = x1 + 2x2







Subject to:







x1 + x2 8







2x1 + y 12







x1 1

Solution:

Step 1: Set up the P-matrix and its transpose

1 1 8
2 1 12


P = 1 0 1

1 2 0






1 2 1 1


PT = 1 1 0 2

8

12 1 0



w1 w2 w3 s1 s2 g z



1 2 1 1 0 0 1


Step 2: Form the

constraints and the

1 1 0 0 1 0 2

objective function

for the dual



-8 -12 -1 0 0 1 0





w1

+ 2w2 + w3 1





w1 + w2 2





z = 8w1 = 12w2 + 2

Step 3: Construct the initial simplex tableau for the dual











Since there are no negative entries in the last column above the third row, we have a standard

simplex problem. The most negative number in the bottom row to the left of the last column is

-12. This establishes the pivot column. The smallest nonnegative ratio is 1/2. The pivot element

is 2 in the w2-column.



68
MBA-H2040 Quantitative Techniques for Managers
Step 4: Pivoting





w1 w2 w3 s1 s2 g z

Pivoting about the

2 we get:

1 2 1 1 0 0 1



1/2 0 -1/2 -1/2 1 0 3/2



-2 0 5 6 0 1 6











w1 w2 w3 s1 s2 g z



1/2 1 1/2 1/2 0 0 1/2



1 1 0 0 1 0 2



-8 -12 -1 0 0 1 0





w1 w2 w3 s1 s2 g z



1/2 1 1/2 1/2 0 0 1/2



1/2 0 -1/2 -1/2 1 0 3/2



-2 0 5 6 0 1 6

The most negative entry in the bottom row to the left of the last column is -2. The smallest non-

negative ratio is the 1/2 in the first row. This is the next pivot element.









w1 w2 w3 s1 s2 g z



1/2 1 1/2 1/2 0 0 1/2



1/2 0 -1/2 -1/2 1 0 3/2

Pivoting about the

-2 0 5 6 0 1 6

1/2:






69

MBA-H2040 Quantitative Techniques for Managers










Since there are no negative entries in the bottom row and to the left of the last column, the

process is complete. The solutions are at the feet of the slack variable columns.

Therefore,

w1 w2 w3 s1 s2 g z



The

optimum solution

1 2 1 1 0 0 1

provided by x1 = 8

and x2 = 0

0 -1 -1 -1 1 0 1



The

Minimum Value

is: 8

0 4 7 8 0 1 8

4.5 Summary

For every linear programming problem there is a dual problem. The variables of the dual problem are

called as dual variables. The variables have economic value, which can be used for planning its

resources. The dual problem solution is achieved by the simplex method calculation of the original

(primal) problem. The dual problem solution has certain properties, which may be very useful for

calculation purposes.

4.6 Key Terms

Original Problem: This is the original linear programming problem, also called as primal problem.

Dual Problem: A dual problem is a linear programming problem is another linear programming
problem formulated from the parameters of the primal problem.

Dual Variables: Dual programming problem variables.

Optimum Solution: The solution where the objective function is minimized or maximized.

Shadow Price: Price of a resource is the change in the optimum value of the objective function per unit
increase of the resource.

4.7 Self Assessment Questions

Q1. An organization manufactures three products viz. A, B and c. The required raw material per piece of

product A, B and C is 2kg, 1kg, and 2kg. Assume that the total weekly availability is 50 kg. In order to

produce the products the raw materials are processed on a machine by the labour force and on a weekly



70
MBA-H2040 Quantitative Techniques for Managers
the availability of machine hours is 30. Assume that the available total labour hour is 26. The following

table illustrates time required per unit of the three products.











The profit per unit from the products A, B and C are #25, #30 and #40.

Formulate the dual linear programming problem and determine the optimum values of the dual
variables.

Q2. Consider the following dual problem





Minimize







3w1 + 4w2






Subject to:







3w1 + 4w2 24







2w1 + w2 10







5w1 + 3w2 29







w1, w2 0

4.8 Key Solutions

Q1. Minimize




50w1 + 30w2 + 26w3




Subject to:





2w1 + 0.5w2 + w3 25





w1 + 3w2 + 2w3 30





2w1 + w2 + w3 40






w1 = 50/3 = 16.6, w2 = 0 and w3 = 20/3 = 6.6


Q2.

Maximize





24x1 + 10x2 + 29x3




Subject

to:



Product

Labour

Machine



3x1 + 2x2 + 5x3 3



Hour

Hour



4x1 + x2 + 3x3 4







A

0.5

1





x1, x2, x3 0 and w1 = 4, w2 = 3



B

3

2



Objective Function Maximum

Value is: 24

C

1

1



71

MBA-H2040 Quantitative Techniques for Managers

4.9 Further References

Mustafi, C.K. 1988. Operations Research Methods and Practices, Wiley Eastern Limited, New Delhi.

Hamdy A Taha, 1999. Introduction to Operations Research, PHI Limited, New Delhi.

Peterson R and Silver, E. A. 1979. Decision Systems for Inventory Management and Production
Planning.

Levin, R and Kirkpatrick, C.A. 1978. Quantitative Approached to Management, Tata McGraw Hill,
Kogakusha Ltd., International Student Edition.





72
MBA-H2040 Quantitative Techniques for Managers



UNIT II











ON



S
S

1 TRANSPORTATION PROBLEM

E
L






LESSON STRUCTURE




1.5 Introduction



1.6 Transportation Algorithm



1.7 Basic Feasible Solution of a



Transportation Problem



1.8 Modified Distribution Method



1.9 Unbalanced Transportation Problem



1.10

Degenerate

Transportation



Problem



1.11

Transshipment Problem



1.12

Transportation

Problem



Maximization



1.13

Summary



1.14

Key Terms



1.15

Self Assessment Questions



1.16

Key Solutions



1.17

Further References

Objectives



After Studying this lesson, you should be able



to:







Formulation of a Transportation Problem



Determine basic feasible solution using



various methods



Understand the MODI, Stepping Stone



Methods for cost minimization



Make unbalanced Transportation Problem



into balanced one using appropriate



method



Solve Degenerate Problem



Formulate and Solve Transshipment



Problem



Describe suitable method for maximizing



the objective function instead of



minimizing










73

MBA-H2040 Quantitative Techniques for Managers
1.1 Introduction

A special class of linear programming problem is Transportation Problem, where the objective is to

minimize the cost of distributing a product from a number of sources (e.g. factories) to a number of

destinations (e.g. warehouses) while satisfying both the supply limits and the demand requirement.

Because of the special structure of the Transportation Problem the Simplex Method of solving is

unsuitable for the Transportation Problem. The model assumes that the distributing cost on a given rout

is directly proportional to the number of units distributed on that route. Generally, the transportation

model can be extended to areas other than the direct transportation of a commodity, including among

others, inventory control, employment scheduling, and personnel assignment.



The transportation problem special feature is illustrated here with the help of following Example 1.1.

Example 1.1:

Suppose a manufacturing company owns three factories (sources) and distribute his products to five

different retail agencies (destinations). The following table shows the capacities of the three factories,

the quantity of products required by the various retail agencies and the cost of shipping one unit of the

product from each of the three factories to each of the five retail agencies.





Retail Agency



Factories

1

2

3

4

5

Capacity

1

1

9

13

36

51

50

2

24

12

16

20

1

100

3

14

33

1

23

26

150



Requirement

100

60

50

50

40

300



Usually the above table is referred as Transportation Table, which provides the basic information

regarding the transportation problem. The quantities inside the table are known as transportation cost per

unit of product. The capacity of the factories 1, 2, 3 is 50, 100 and 150 respectively. The requirement of

the retail agency 1, 2, 3, 4, 5 is 100,60,50,50, and 40 respectively.




In this case, the transportation cost of one unit

from factory 1 to retail agency 1 is 1,
from factory 1 to retail agency 2 is 9,
from factory 1 to retail agency 3 is 13, and so on.




A transportation problem can be formulated as linear programming problem using variables with

two subscripts.


Let



74
MBA-H2040 Quantitative Techniques for Managers




x11=Amount to be transported from factory 1 to retail agency 1





x12= Amount to be transported from factory 1 to retail agency 2





........





........





........





........





x35= Amount to be transported from factory 3 to retail agency 5.



Let the transportation cost per unit be represented by C11, C12, .....C35 that is C11=1, C12=9, and so on.


Let the capacities of the three factories be represented by a1=50, a2=100, a3=150.



Let the requirement of the retail agencies are b1=100, b2=60, b3=50, b4=50, and b5=40.




Thus, the problem can be formulated as







Minimize





C11x11+C12x12+...............+C35x35




Subject to:





x11 + x12 + x13 + x14 + x15 = a1





x21 + x22 + x23 + x24 + x25 = a2





x31 + x32 + x33 + x34 + x35 = a3











x11 + x21 + x31 = b1





x12 + x22 + x32 = b2





x13 + x23 + x33 = b3





x14 + x24 + x34 = b4





x15 + x25 + x35 = b5






x11, x12, ......, x35 0.




Thus, the problem has 8 constraints and 15 variables. So, it is not possible to solve such a

problem using simplex method. This is the reason for the need of special computational procedure to

solve transportation problem. There are varieties of procedures, which are described in the next section.

1.2 Transportation Algorithm

The steps of the transportation algorithm are exact parallels of the simplex algorithm, they are:

Step 1: Determine a starting basic feasible solution, using any one of the following three methods

1. North West Corner Method
2. Least Cost Method
3. Vogel Approximation Method


Step 2: Determine the optimal solution using the following method


1. MODI (Modified Distribution Method) or UV Method.


1.3 Basic Feasible Solution of a Transportation Problem




75

MBA-H2040 Quantitative Techniques for Managers
The computation of an initial feasible solution is illustrated in this section with the help of the

example1.1 discussed in the previous section. The problem in the example 1.1 has 8 constraints and 15

variables we can eliminate one of the constraints since a1 + a2 + a3 = b1 + b2 + b3 + b4 +b5. Thus now the

problem contains 7 constraints and 15 variables. Note that any initial (basic) feasible solution has at

most 7 non-zero Xij. Generally, any basic feasible solution with m sources (such as factories) and n

destination (such as retail agency) has at most m + n -1 non-zero Xij.



The special structure of the transportation problem allows securing a non artificial basic feasible

solution using one the following three methods.



4. North West Corner Method
5. Least Cost Method
6. Vogel Approximation Method



The difference among these three methods is the quality of the initial basic feasible solution they

produce, in the sense that a better that a better initial solution yields a smaller objective value. Generally

the Vogel Approximation Method produces the best initial basic feasible solution, and the North West

Corner Method produces the worst, but the North West Corner Method involves least computations.



North West Corner Method:




The method starts at the North West (upper left) corner cell of the tableau (variable x11).

Step -1: Allocate as much as possible to the selected cell, and adjust the associated amounts of capacity
(supply) and requirement (demand) by subtracting the allocated amount.

Step -2: Cross out the row (column) with zero supply or demand to indicate that no further assignments
can be made in that row (column). If both the row and column becomes zero simultaneously, cross out
one of them only, and leave a zero supply or demand in the uncrossed out row (column).

Step -3: If exactly one row (column) is left uncrossed out, then stop. Otherwise, move to the cell to the
right if a column has just been crossed or the one below if a row has been crossed out. Go to step -1.

Example 1.2:

Consider the problem discussed in Example 1.1 to illustrate the North West Corner Method of
determining basic feasible solution.



Retail Agency



Factories

1

2

3

4

5

Capacity

1

1

9

13

36

51

50

2

24

12

16

20

1

100

3

14

33

1

23

26

150



Requirement

100

60

50

50

40

300






76
MBA-H2040 Quantitative Techniques for Managers




The allocation is shown in the following tableau:






















Capacity















1 9 13 36 51

50



50







24 12 16 20 1

100 50



50 50



14 33 1 23 26









10

50

50

40

150 140 90 40



Requirement 100

60

50

50

40

50 10


The arrows show the order in which the allocated (bolded) amounts are generated. The starting

basic solution is given as
x11 = 50,








x21 = 50, x22 = 50









x32 = 10, x33 = 50, x34 = 50, x35 = 40










The corresponding transportation cost is






50 * 1 + 50 * 24 + 50 * 12 + 10 * 33 + 50 * 1 + 50 * 23 + 40 * 26 = 4420




It is clear that as soon as a value of Xij is determined, a row (column) is eliminated from further

consideration. The last value of Xij eliminates both a row and column. Hence a feasible solution

computed by North West Corner Method can have at most m + n ? 1 positive Xij if the transportation

problem has m sources and n destinations.


Least Cost Method

The least cost method is also known as matrix minimum method in the sense we look for the row and

the column corresponding to which Cij is minimum. This method finds a better initial basic feasible

solution by concentrating on the cheapest routes. Instead of starting the allocation with the northwest

cell as in the North West Corner Method, we start by allocating as much as possible to the cell with the

smallest unit cost. If there are two or more minimum costs then we should select the row and the column

corresponding to the lower numbered row. If they appear in the same row we should select the lower

numbered column. We then cross out the satisfied row or column, and adjust the amounts of capacity

and requirement accordingly. If both a row and a column is satisfied simultaneously, only one is crossed

out. Next, we look for the uncrossed-out cell with the smallest unit cost and repeat the process until we

are left at the end with exactly one uncrossed-out row or column.



77

MBA-H2040 Quantitative Techniques for Managers

Example 1.3:

The least cost method of determining initial basic feasible solution is illustrated with the help of problem

presented in the section 1.1.
























Capacity















1 9 13 36 51

50



50







24 12 16 20 1

100 60



60

40



14 33 1 23 26







50

50

50

150 100 50



Requirement 100

60

50

50

40

50





The Least Cost method is applied in the following manner:


We observe that C11=1 is the minimum unit cost in the table. Hence X11=50 and the first row is

crossed out since the row has no more capacity. Then the minimum unit cost in the uncrossed-out row

and column is C25=1, hence X25=40 and the fifth column is crossed out. Next C33=1is the minimum unit

cost, hence X33=50 and the third column is crossed out. Next C22=12 is the minimum unit cost, hence

X22=60 and the second column is crossed out. Next we look for the uncrossed-out row and column now

C31=14 is the minimum unit cost, hence X31=50 and crossed out the first column since it was satisfied.

Finally C34=23 is the minimum unit cost, hence X34=50 and the fourth column is crossed out.


So that the basic feasible solution developed by the Least Cost Method has transportation cost is

1 * 50 + 12 * 60 + 1 * 40 + 14 * 50 + 1 * 50 + 23 * 50 = 2710

Note that the minimum transportation cost obtained by the least cost method is much lower than

the corresponding cost of the solution developed by using the north-west corner method.

Vogel Approximation Method (VAM):

VAM is an improved version of the least cost method that generally produces better solutions. The steps
involved in this method are:

Step 1: For each row (column) with strictly positive capacity (requirement), determine a penalty by
subtracting the smallest unit cost element in the row (column) from the next smallest unit cost element
in the same row (column).




78
MBA-H2040 Quantitative Techniques for Managers
Step 2: Identify the row or column with the largest penalty among all the rows and columns. If the
penalties corresponding to two or more rows or columns are equal we select the topmost row and the
extreme left column.

Step 3: We select Xij as a basic variable if Cij is the minimum cost in the row or column with largest
penalty. We choose the numerical value of Xij as high as possible subject to the row and the column
constraints. Depending upon whether ai or bj is the smaller of the two ith row or jth column is crossed out.

Step 4: The Step 2 is now performed on the uncrossed-out rows and columns until all the basic variables
have been satisfied.

Example 1.4:

Consider the following transportation problem




Destination



Origin

1

2

3

4

ai

1

20

22

17

4

120

2

24

37

9

7

70

3

32

37

20

15

50

bj

60

40

30

110

240


Note: ai=capacity (supply)
bj=requirement (demand)

Now, compute the penalty for various rows and columns which is shown in the following table:




Destination



Origin

1

2

3

4

ai Column
Penalty

1

20

22

17

4

120 13

2

24

37

9

7

70 2

3

32

37

20

15

50 5

bj

60

40

30

110

240













Row Penalty

4

15

8

3





Look for the highest penalty in the row or column, the highest penalty occurs in the second column and

the minimum unit cost i.e. cij in this column is c12=22. Hence assign 40 to this cell i.e. x12=40 and cross

out the second column (since second column was satisfied. This is shown in the following table:



Destination



Origin

1

2

3

4

ai Column
Penalty

1

20

22 40

17

4

80 13

2

24

37

9

7

70 2

3

32

37

20

15

50 5

bj

60

40

30

110

240



79

MBA-H2040 Quantitative Techniques for Managers













Row Penalty

4

15

8

3




The next highest penalty in the uncrossed-out rows and columns is 13 which occur in the first row and
the minimum unit cost in this row is c14=4, hence x14=80 and cross out the first row. The modified table
is as follows:





Destination



Origin

1

2

3

4

ai Column
Penalty

1

20

22

17

4 0 13

40

80

2

24

37

9

7

70 2

3

32

37

20

15

50 5

bj

60

40

30

110

240













Row Penalty

4

15

8

3





The next highest penalty in the uncrossed-out rows and columns is 8 which occurs in the third column
and the minimum cost in this column is c23=9, hence x23=30 and cross out the third column with
adjusted capacity, requirement and penalty values. The modified table is as follows:




Destination



Origin

1

2

3

4

ai Column
Penalty

1

20

22

17

4 0 13

40

80

2

24

37

9

7

40 17

30

3

32

37

20

15

50 17

bj

60

40

30

110

240













Row Penalty

8

15

8

8




The next highest penalty in the uncrossed-out rows and columns is 17 which occurs in the second row
and the smallest cost in this row is c24=15, hence x24=30 and cross out the fourth column with the
adjusted capacity, requirement and penalty values. The modified table is as follows:




Destination



Origin

1

2

3

4

ai Column
Penalty

1

20

22

17

4 0 13

40

80

2

24

37

9

7

10 17

30

30

3

32

37

20

15

50 17

bj

60

40

30

110

240



80
MBA-H2040 Quantitative Techniques for Managers













Row Penalty

8

15

8

8





The next highest penalty in the uncrossed-out rows and columns is 17 which occurs in the second row
and the smallest cost in this row is c21=24, hence xi21=10 and cross out the second row with the adjusted
capacity, requirement and penalty values. The modified table is as follows:





Destination



Origin

1

2

3

4

ai Column
Penalty

1

20

22

17

4 0 13

40

80

2

24

37

9

7

0 17

10

30

30

3

32

37

20

15

50 17

bj

60

40

30

110

240













Row Penalty

8

15

8

8




The next highest penalty in the uncrossed-out rows and columns is 17 which occurs in the third row and
the smallest cost in this row is c31=32, hence xi31=50 and cross out the third row or first column. The
modified table is as follows:




Destination



Origin

1

2

3

4

ai Column
Penalty

1

20

22

17

4 0 13

40

80

2

24

37

9

7

0 17

10

30

30

3

32

37

20

15

0 17

50

bj

60

40

30

110

240













Row Penalty

8

15

8

8




The transportation cost corresponding to this choice of basic variables is

22 * 40 + 4 * 80 + 9 * 30 + 7 * 30 + 24 * 10 + 32 * 50 = 3520

1.4 Modified Distribution Method

The Modified Distribution Method, also known as MODI method or u-v method, which provides a

minimum cost solution (optimal solution) to the transportation problem. The following are the steps

involved in this method.





81

MBA-H2040 Quantitative Techniques for Managers
Step 1: Find out the basic feasible solution of the transportation problem using any one of the three
methods discussed in the previous section.

Step 2: Introduce dual variables corresponding to the row constraints and the column constraints. If
there are m origins and n destinations then there will be m+n dual variables. The dual variables
corresponding to the row constraints are represented by ui, i=1,2,.....m where as the dual variables
corresponding to the column constraints are represented by vj, j=1,2,.....n. The values of the dual
variables are calculated from the equation given below



ui + vj = cij if xij > 0


Step 3: Any basic feasible solution has m + n -1 xij > 0. Thus, there will be m + n -1 equation to
determine m + n dual variables. One of the dual variables can be chosen arbitrarily. It is also to be noted
that as the primal constraints are equations, the dual variables are unrestricted in sign.

Step 4: If xij=0, the dual variables calculated in Step 3 are compared with the cij values of this allocation
as cij ? ui ? vj. If al cij ? ui ? vj 0, then by the theorem of complementary slackness it can be shown that
the corresponding solution of the transportation problem is optimum. If one or more cij ? ui ? vj < 0, we
select the cell with the least value of cij ? ui ? vj and allocate as much as possible subject to the row and
column constraints. The allocations of the number of adjacent cell are adjusted so that a basic variable
becomes non-basic.

Step 5: A fresh set of dual variables are calculated and repeat the entire procedure from Step 1 to Step 5.

Example 1.5:
For example consider the transportation problem given below:






















Supply















1 9 13 36 51

50











24 12 16 20 1

100









14 33 1 23 26









150

Demand 100

70



50

40

40

300



Step 1: First we have to determine the basic feasible solution. The basic feasible solution using least
cost method is



x11=50, x22=60, x25=40, x31=50, x32=10, x33=50 and x34=40


Step 2: The dual variables u1, u2, u3 and v1, v2, v3, v4, v5 can be calculated from the corresponding cij
values, that is



u1+v1=1

u2+v2=12

u2+v5=1

u3+v1=14



u3+v2=33

u3+v3=1

u3+v4=23


Step 3: Choose one of the dual variables arbitrarily is zero that is u3=0 as it occurs most often in the
above equations. The values of the variables calculated are



82
MBA-H2040 Quantitative Techniques for Managers


u1= -13, u2= -21, u3=0
v1=14, v2=33, v3=1, v4=23, v5=22


Step 4: Now we calculate cij ? ui ? vj values for all the cells where xij=0 (.e. unallocated cell by the basic
feasible solution)
That is

Cell(1,2)= c12-u1-v2 = 9+13-33 = -11
Cell(1,3)= c13-u1-v3 = 13+13-1 = 25
Cell(1,4)= c14-u1-v4 = 36+13-23 = 26
Cell(1,5)= c15-u1-v5 = 51+13-22 = 42
Cell(2,1)= c21-u2-v1 = 24+21-14 = 31
Cell(2,3)= c23-u2-v3 = 16+21-1 = 36
Cell(2,4)= c24-u2-v4 = 20+21-23 = 18
Cell(3,5)= c35-u3-v5 = 26-0-22 = 4


Note that in the above calculation all the cij ? ui ? vj 0 except for cel (1, 2) where c12 ? u1 ? v2 = 9+13-
33 = -11.


Thus in the next iteration x12 will be a basic variable changing one of the present basic variables

non-basic. We also observe that for allocating one unit in cell (1, 2) we have to reduce one unit in cells

(3, 2) and (1, 1) and increase one unit in cell (3, 1). The net transportation cost for each unit of such

reallocation is







-33 -1 + 9 +14 = -11



The maximum that can be allocated to cell (1, 2) is 10 otherwise the allocation in the cell (3, 2)

will be negative. Thus, the revised basic feasible solution is



x11=40, x12=10, x22=60, x25=40, x31=60, x33=50, x34=40


1.5 Unbalanced Transportation Problem

In the previous section we discussed about the balanced transportation problem i.e. the total supply

(capacity) at the origins is equal to the total demand (requirement) at the destination. In this section we

are going to discuss about the unbalanced transportation problems i.e. when the total supply is not equal

to the total demand, which are called as unbalanced transportation problem.



In the unbalanced transportation problem if the total supply is more than the total demand then

we introduce an additional column which will indicate the surplus supply with transportation cost zero.

Similarly, if the total demand is more than the total supply an additional row is introduced in the

transportation table which indicates unsatisfied demand with zero transportation cost.

Example 1.6:



Consider the following unbalanced transportation problem



Warehouses



83

MBA-H2040 Quantitative Techniques for Managers
Plant w1 w2 w3 Supply


20

17

25

X

400







10

10

20

Y

500







Demand 400





400





500


In this problem the demand is 1300 whereas the total supply is 900. Thus, we now introduce an

additional row with zero transportation cost denoting the unsatisfied demand. So that the modified

transportation problem table is as follows:



Warehouses

Plant w1 w2 w3 Supply


20

17

25

X

400







10

10

20

Y

500



Unsatisfied

0

0

0

Demand



400




Demand



400





400



500

1300



Now we can solve as balanced problem discussed as in the previous sections.

1.6. Degenerate Transportation Problem

In a transportation problem, if a basic feasible solution with m origins and n destinations has less than m

+ n -1 positive Xij i.e. occupied cells, then the problem is said to be a degenerate transportation

problem. The degeneracy problem does not cause any serious difficulty, but it can cause computational

problem wile determining the optimal minimum solution.



There fore it is important to identify a degenerate problem as early as beginning and take the

necessary action to avoid any computational difficulty. The degeneracy can be identified through the

following results:

"In a transportation problem, a degenerate basic feasible solution exists if and only if some

partial sum of supply (row) is equal to a partial sum of demand (column). For example the following

transportation problem is degenerate. Because in this problem



a1 = 400 = b1

a2 + a3 = 900 = b2 + b3



Warehouses

Plant w1 w2 w3 Supply (ai)


84
MBA-H2040 Quantitative Techniques for Managers


20

17

25

X

400







10

10

20

Y

500



Unsatisfied

0

0

0

demand

400




Demand (bj)



400





400



500

1300




There is a technique called perturbation, which helps to solve the degenerate problems.


Perturbation Technique:



The degeneracy of the transportation problem can be avoided if we ensure that no partial sum of

ai (supply) and bj (demand) is equal. We set up a new problem where




ai = ai +d



i = 1, 2, ......, m



bj = bj



j = 1, 2, ......, n -1



bn = bn + md

d > 0




This modified problem is constructed in such a way that no partial sum of ai is equal to the bj.

Once the problem is solved, we substitute d = 0 leading to optimum solution of the original problem.

Example: 1.7


Consider the above problem



Warehouses

Plant w1 w2 w3 Supply (ai)


20

17

25

X

400 + d







10

10

20

Y

500 + d



Unsatisfied

0

0

0

demand

400 + d




Demand (bj)



400





400



500 + 3d 1300 + 3d


Now this modified problem can be solved by using any of the three methods viz. North-west Corner,
Least Cost, or VAM.

1.7 Transshipment Problem

There could be a situation where it might be more economical to transport consignments in several sages

that is initially within certain origins and destinations and finally to the ultimate receipt points, instead of

transporting the consignments from an origin to a destination as in the transportation problem.



85

MBA-H2040 Quantitative Techniques for Managers


The movement of consignment involves tow different modes of transport viz. road and railways

or between stations connected by metre gauge and broad gauge lines. Similarly it is not uncommon to

maintain dumps for central storage of certain bulk material. These require transshipment.



Thus for the purpose of transshipment the distinction between an origin and destination is

dropped so that from a transportation problem with m origins and n destinations we obtain a

transshipment problem with m + n origins and m + n destinations.



The formulation and solution of a transshipment problem is illustrated with the following

Example 1.8.





Example 1.8:

Consider the following transportation problem where the origins are plants and destinations are depots.

Table 1

Depot



Plant



X





Y



Z



Supply









A







150





$1

$3

$15





B







300



$3

$5

$25



Demand



150



150



150 450




When each plant is also considered as a destination and each depot is also considered as an

origin, there are altogether five origins and five destinations. So that some additional cost data are

necessary, they are as follows:

Table 2

Unit transportation cost From Plant To Plant

To











Plant A



Plant B







From











Plant A

0

55











Plant B

2

0







Table 3

Unit transportation cost From Depot To Depot

To









Depot X

Depot Y

Depot Z





86
MBA-H2040 Quantitative Techniques for Managers



From













Depot X

0

25

2









Depot Y

2

0

3









Depot Z

55

3

0

Table 4

Unit transportation cost From Depot to Plant



To











Plant A



Plant B







From











Depot X

3

15











Depot Y

25

3







Depot Z

45

55



Now, from the Table

1, Table 2, Table 3, Table 4

we obtain the transportation formulation of the transshipment problem, which is shown in the Table 5.


Table 5

Transshipment Table

Supply







A



B



X



Y



Z















A

150+450=600

0

55

1

3

15















B

300+450=750

2

0

3

5

25















X

450

3

15

0

25

2















Y

450

25

3

2

0

3















Z

450

45

55

55

3

0




Demand

450



450

150+450= 150+450= 150+450=















600

600



600




A buffer stock of 450 which is the total supply and total demand in the original transportation

problem is added to each row and column of the transshipment problem. The resulting transportation

problem has m + n = 5 origins and m + n = 5 destinations.



By solving the transportation problem presented in the Table 5, we obtain



x11=150

x13=300

x14=150

x21=3001

x22=450

x33=300

x35=150

x44=450

x55=450




The transshipment problem explanation is as follows:



87

MBA-H2040 Quantitative Techniques for Managers


1. Transport x21=300 from plant B to plant A. This increase the availability at plant A to 450 units

including the 150 originally available from A.

2. From plant A transport x13=300 to depot X and x14=150 to depot Y.
3. From depot X transport x35=150 to depot Z.



Thus, the total cost of transshipment is:

2*300 + 3 * 150 + 1*300 + 2*150 = $1650

Note: The consignments are transported from pants A, B to depots X, Y, Z only according to the
transportation Table 1, the minimum transportation cost schedule is x13=150 x21=150 x22=150 with a
minimum cost of 3450.



Thus, transshipment reduces the cost of consignment movement.






1.8 Transportation Problem Maximization

There are certain types of transportation problem where the objective function is to be maximized

instead of minimized. These kinds of problems can be solved by converting the maximization problem

into minimization problem. The conversion of maximization into minimization is done by subtracting

the unit costs from the highest unit cost of the table.



The maximization of transportation problem is illustrated with the following Example 1.9.


Example 1.9:

A company has three factories located in three cities viz. X, Y, Z. These factories supplies consignments

to four dealers viz. A, B, C and D. The dealers are spread all over the country. The production capacity

of these factories is 1000, 700 and 900 units per month respectively. The net return per unit product is

given in the following table.

Dealers





Factory

A



B



C

D



capacity























X

6

6

6

4

1000

















Y

4

2

4

5

700

















Z

900

5

6

7

8






Requirement 900

800 500 400 2600




Determine a suitable allocation to maximize the total return.





88
MBA-H2040 Quantitative Techniques for Managers


This is a maximization problem. Hence first we have to convert this in to minimization problem.

The conversion of maximization into minimization is done by subtracting the unit cost of the table from

the highest unit cost.



Look the table, here 8 is the highest unit cost. So, subtract all the unit cost from the 8, and then

we get the revised minimization transportation table, which is given below.




Dealers





Factory

A



B



C

D



capacity























X

2

2

2

4

1000 = a1

















Y

4

6

4

3

700 =a2

















Z

900 =a3

3

2

1

0






Requirement 900=b1

800=b2 500=b3 400=b4 2600


Now we can solve the problem as a minimization problem.



The problem here is degenerate, since the partial sum of a1=b2+b3 or a3=b3. So consider the

corresponding perturbed problem, which is shown below.


Dealers





Factory

A



B



C

D



capacity























X

2

2

2

4

1000+d

















Y

4

6

4

3

700+d

















Z

900+d

3

2

1

0






Requirement 900

800 500 400+3d 2600+3d





First we have to find out the basic feasible solution. The basic feasible solution by lest cost

method is x11=100+d, x22=700-d, x23=2d, x33=500-2d and x34=400+3d.



Once if the basic feasible solution is found, next we have to determine the optimum solution

using MODI (Modified Distribution Method) method. By using this method we obtain





u1+v1=2



u1+v2=2

u2+v2=6





u2+v3=4



u3+v3=1

u3+v4=0


Taking u1=0 arbitrarily we obtain




u1=0, u2=4, u3=1 and





v1=2, v2=3, v3=0


On verifying the condition of optimality, we know that



89

MBA-H2040 Quantitative Techniques for Managers


C12-u1-v2 < 0

and

C32-u3-v2 <0


So, we allocate x12=700-d and make readjustment in some of the other basic variables.

The revised values are:

x11=200+d, x12=800, x21=700-d, x23=2d, x33=500-3d, and x34=400+3d


u1+v1=2

u1+v2=2

u2+v1=4

u2+v3=4

u3+v3=1

u3+v4=0

Taking u1=0 arbitrarily we obtain




u1=0, u2=2, u3=-1





v1=2, v2=2, v3=2, v4=1

Now, the optimality condition is satisfied.

Finally, taking d=0 the optimum solution of the transportation problem is


X11=200, x12=800, x21=700, x33=500 and x34=400


Thus, the maximum return is:


6*200 + 6*800 + 4*700 + 7*500 + 8*400 = 15500


1.9 Summary

Transportation Problem is a special kind of linear programming problem. Because of the transportation

problem special structure the simplex method is not suitable. But which may be utilized to make

efficient computational techniques for its solution.



Generally transportation problem has a number of origins and destination. A certain amount of

consignment is available in each origin. Similarly, each destination has a certain demand/requirements.

The transportation problem represents amount of consignment to be transported from different origins to

destinations so that the transportation cost is minimized with out violating the supply and demand

constraints.



There are two phases in the transportation problem. First is the determination of basic feasible

solution and second is the determination of optimum solution.

There are three methods available to determine the basic feasible solution, they are

1. North West Corner Method
2. Least Cost Method or Matrix Minimum Method
3. Vogel's Approximation Method (VAM)

In order to determine optimum solution we can use either one of the following method


1. Modified Distribution (MODI) Method

Or

2. Stepping Stone Method



Transportation problem can be generalized into a Transshipment Problem where transportation

of consignment is possible from origin to origin or destination as well as destination to origin or


90
MBA-H2040 Quantitative Techniques for Managers
destination. The transshipment problem may be result in an economy way of shipping in some

situations.

1.10 Key Terms



Origin:

is the location from which the shipments are dispatched.

Destination: is the location to which the shipments are transported.
Unit Transportation Cost: is the transportation cost per unit from an origin to destination.
Perturbation Technique: is a method of modifying a degenerate transportation problem in order to
solve the degeneracy.

1.11 Self Assessment Questions

Q1. Four companies viz. W, X, Y and Z supply the requirements of three warehouses viz. A, B and C
respectively. The companies' availability, warehouses requirements and the unit cost of transportation
are given in the following table. Find an initial basic feasible solution using


a. North West Corner Method
b. Least Cost Method
c. Vogel Approximation Method (VAM)




Warehouses





Company



A



B

C

Supply





















W

10

8

9

15















X

5

2

3

20















Y

30

6

7

4









Z

35

7

6

9






Requirement

25



26

49 100



Q2. Find the optimum Solution of the following Problem using MODI method.













Destination







Source

1

2

3

Capacity







A 42









8

9

10







B







30



9

11

11







C

28









10

12

9








Demand 35 40 25 100






91

MBA-H2040 Quantitative Techniques for Managers
Q3. The ABT transport company ships truckloads of food grains from three sources viz. X, Y, Z to four
mills viz. A, B, C, D respectively. The supply and the demand together with the unit transportation cost
per truckload on the different routes are described in the following transportation table. Assume that the
unit transportation costs are in hundreds of dollars. Determine the optimum minimum shipment cost of
transportation using MODI method.



Mill





Source

A



B



C

D



Supply























X

10

2

20

11

15

















Y

12

7

9

20

25

















Z

10

4

14

16

18






Demand 5 15 15 15



Q4. An organization has three plants at X, Y, Z which supply to warehouses located at A, B, C, D, and E
respectively. The capacity of the plants is 800, 500 and 900 per month and the requirement of the
warehouses is 400, 400, 500, 400 and 800 units respectively. The following table shows the unit
transportation cost.







A

B



C

D E







X













$5

$8

$6

$6

$3

Y













$4

$7

$7

$6

$6

Z

























$8

$4

$6

$6

$3









Determine an optimum distribution for the organization in order to minimize the total cost of
transportation.




Q5. Solve the following transshipment problem

Consider a transportation problem has tow sources and three depots. The availability, requirements and
unit cost are as follows:

Depot





Source

D1

D2

D3 Availability







S1







30



9

8

1















S2

1

7

8

30



Requirement 20 20 20 60


In addition to the above, suppose that the unit cost of transportation from source to source and from
depot to depot are as:

Source



92
MBA-H2040 Quantitative Techniques for Managers










S1

S2











S1





Source

0

1













S2 2

0





Depot







D1

D2

D3







D1









0

2

1





Depot







D2

2

0

9









D3 1

9

0



Find out minimum transshipment cost of the problem and also compare this cost with the corresponding
minimum transportation cost.


Q6. Saravana Store, T.Nagar, Chennai interested to purchase the following type and quantities of dresses


Dress

V

W

X

Y

Z

Type

Quantity

150

100

75

250

200


Four different dress makers are submitted the tenders, who undertake to supply not more than the
quantities indicated below:


Dress

A

B

C

D

Maker

Dress

300

250

150

200

Quantity


Saravana Store estimates that its profit per dress will vary according to the dress maker as indicates in
the following table:







V W

X

Y Z













A

2.75

3.5

4.25

2.25

1.5













B

3

3.25

4.5

1.75

1













C

2.5

3.5

4.75

2

1.25













D

3.25

2.75

4

2.5

1.75



Determine how should the orders to be places for the dresses so as to maximize the profit.


93

MBA-H2040 Quantitative Techniques for Managers

1.12 Key Solutions


Q1. a. x11 = 15, x21 = 10, x22 = 10, x32 = 16, x33 = 14, x43 = 35

Minimum Cost is: 753

b. x13 = 15, x22 = 20, x33 = 30, x41 = 25, x42 = 6, x43 = 4
Minimum Cost is: 542

c. x13=15, x22=20, x33=30, x41=25, x42=6, x43=4



Minimum Cost is: 542


Q2. x11=2, x12=40, x21=30, x31=3, x33=25


Minimum Transportation Optimal cost is: 901.


Q3. x12=5, x14=10, x22=10, x23=15, x31=5, x34=5


Minimum Optimal Cost is: $435


Q4. x15=800, x21=400, x24=100, x32=400, x33=200, x34=300, x43=300 (supply shortage)


Minimum Cost of Transportation is: $9200


Q5. Transportation Problem


S1-D2=10, S1-D3=20, S2-D1=20, S2-D2=10 and

Minimum Transportation Cost is: 100



Transshipment Problem

x11=60, x12=10, x15=20, x22=50, x23=40, x33=40, x34=20, x44=60, x55=60 and
Minimum Transshipment Cost is: 100
Q6. 150 dresses of V and 50 dresses of Z by Dress Maker A


250 dresses of Y by Dress Maker B



150 dress of Z by Dress Maker C



100 dress of W and 75 dresses of X by Dress Maker D




Maximum Profit is: 1687.50


1.13 Further References

Hamdy A Taha. 1999. Introduction to Operations Research, PHI Limited, New Delhi.

Mustafi, C.K. 1988. Operations Research Methods and Practice, Wiley Eastern Ltd., New Delhi.

Mittal, K.V. 1976. Optimization Methods in Operations Research and Systems Analysis, Wiley Eastern
Ltd., New Delhi.










94
MBA-H2040 Quantitative Techniques for Managers






















UNIT II












ON
S

2 ASSIGNMENT PROBLEM

S
E



L




LESSON STRUCTURE




2.1 Introduction



2.2 Assignment Problem Structure and



Solution



2.3 Unbalanced Assignment Problem



2.4 Infeasible Assignment Problem



2.5 Maximization in an Assignment



Problem



2.6 Crew Assignment Problem



2.7 Summary



2.8 Key Solutions



2.9 Self Assessment Questions



2.10 Key Answers



2.11 Further References













95

MBA-H2040 Quantitative Techniques for Managers
























Objectives
After Studying this lesson, you should be able
to:

Assignment Problem Formulation
How to solve the Assignment Problem
How to solve the unbalanced problem

using appropriate method

Make appropriate modification when

some problems are infeasible

Modify the problem when the objective is

to maximize the objective function

Formulate and solve the crew assignment

problems









96
MBA-H2040 Quantitative Techniques for Managers




2.1 Introduction

The Assignment Problem can define as follows:

Given n facilities, n jobs and the effectiveness of each facility to each job, here the problem is to assign

each facility to one and only one job so that the measure of effectiveness if optimized. Here the

optimization means Maximized or Minimized. There are many management problems has a assignment

problem structure. For example, the head of the department may have 6 people available for assignment

and 6 jobs to fill. Here the head may like to know which job should be assigned to which person so that

all tasks can be accomplished in the shortest time possible. Another example a container company may

have an empty container in each of the location 1, 2,3,4,5 and requires an empty container in each of the

locations 6, 7, 8,9,10. It would like to ascertain the assignments of containers to various locations so as

to minimize the total distance. The third example here is, a marketing set up by making an estimate of

sales performance for different salesmen as well as for different cities one could assign a particular

salesman to a particular city with a view to maximize the overall sales.



Note that with n facilities and n jobs there are n! possible assignments. The simplest way of

finding an optimum assignment is to write all the n! possible arrangements, evaluate their total cost and

select the assignment with minimum cost. Bust this method leads to a calculation problem of formidable

size even when the value of n is moderate. For n=10 the possible number of arrangements is 3268800.


2.2 Assignment Problem Structure and Solution




The structure of the Assignment problem is similar to a transportation problem, is as follows:

Jobs









1

2



...



n

1

c

1

11 c12 ... c1n











2

c

1

21 c21 ... c2n







. . . . .

Workers





. . . .







. . . .





n

c

1

n1 cn2 ... cnn


1 1 ... 1



97

MBA-H2040 Quantitative Techniques for Managers
The element cij represents the measure of effectiveness when ith person is assigned jth job. Assume that

the overall measure of effectiveness is to be minimized. The element xij represents the number of ith

individuals assigned to the jth job. Since ith person can be assigned only one job and jth job can be

assigned to only one person we have the following




xi1 + xi2 + ................ + xin = 1, where i = 1, 2, . . . . . . . , n




x1j + x2j + ................ + xnj = 1, where j = 1, 2, . . . . . . . , n


and the objective function is formulated as



Minimize c11x11 + c12x12 + ........... + cnnxnn






xij 0




The assignment problem is actually a special case of the transportation problem where m = n and

ai = bj = 1. However, it may be easily noted that any basic feasible solution of an assignment problem

contains (2n ? 1) variables of which (n ? 1) variables are zero. Because of this high degree of degeneracy

the usual computation techniques of a transportation problem become very inefficient. So, hat a separate

computation technique is necessary for the assignment problem.



The solution of the assignment problem is based on the following results:




"If a constant is added to every element of a row/column of the cost matrix of an assignment

problem the resulting assignment problem has the same optimum solution as the original assignment

problem and vice versa". ? This result may be used in two different methods to solve the assignment

problem. If in an assignment problem some cost elements cij are negative, we may have to convert them

into an equivalent assignment problem where all the cost elements are non-negative by adding a suitable

large constant to the cost elements of the relevant row or column, and then we look for a feasible

solution which has zero assignment cost after adding suitable constants to the cost elements of the

various rows and columns. Since it has been assumed that all the cost elements are non-negative, this

assignment must be optimum. On the basis of this principle a computational technique known as

Hungarian Method is developed. The Hungarian Method is discussed as follows.


Hungarian Method:




The Hungarian Method is discussed in the form of a series of computational steps as follows, when the

objective function is that of minimization type.


Step 1:



98
MBA-H2040 Quantitative Techniques for Managers
From the given problem, find out the cost table. Note that if the number of origins is not equal to the
number of destinations then a dummy origin or destination must be added.


Step 2:
In each row of the table find out the smallest cost element, subtract this smallest cost element from each
element in that row. So, that there will be at least one zero in each row of the new table. This new table
is known as First Reduced Cost Table.

Step 3:
In each column of the table find out the smallest cost element, subtract this smallest cost element from
each element in that column. As a result of this, each row and column has at least one zero element. This
new table is known as Second Reduced Cost Table.

Step 4:
Now determine an assignment as follows:


1. For each row or column with a single zero element cell that has not be assigned or

eliminated, box that zero element as an assigned cell.

2. For every zero that becomes assigned, cross out all other zeros in the same row and for

column.

3. If for a row and for a column there are two or more zero and one can't be chosen by

inspection, choose the assigned zero cell arbitrarily.

4. The above procedures may be repeated until every zero element cell is either assigned

(boxed) or crossed out.


Step 5:
An optimum assignment is found, if the number of assigned cells is equal to the number of rows (and
columns). In case we had chosen a zero cell arbitrarily, there may be an alternate optimum. If no
optimum solution is found i.e. some rows or columns without an assignment then go to Step 6.

Step 6:
Draw a set of lines equal to the number of assignments which has been made in Step 4, covering all the
zeros in the following manner



1. Mark check () to those rows where no assignment has been made.




2. Examine the checked () rows. If any zero element cel occurs in those rows, check () the

respective columns that contains those zeros.



3. Examine the checked () columns. If any assigned zero element occurs in those columns,

check () the respective rows that contain those assigned zeros.



4. The process may be repeated until now more rows or column can be checked.



5. Draw lines through all unchecked rows and through all checked columns.


Step 7:
Examine those elements that are not covered by a line. Choose the smallest of these elements and
subtract this smallest from all the elements that do not have a line through them.



99

MBA-H2040 Quantitative Techniques for Managers



Add this smallest element to every element that lies at the intersection of two lines. Then the

resulting matrix is a new revised cost table.


Example 2.1:

Problem
A work shop contains four persons available for work on the four jobs. Only one person can work on
any one job. The following table shows the cost of assigning each person to each job. The objective is to
assign person to jobs such that the total assignment cost is a minimum.


Jobs







1

2



3



4








20 25 22 28

A









15 18 23 17

Persons

B





19 17 21 24





C





25 23 24 24





D




Solution



As per the Hungarian Method


Step 1: The cost Table


Jobs







1

2



3



4






A

20 25 22 28









Persons

B

15 18 23 17









C

19 17 21 24









D

25 23 24 24

















100
MBA-H2040 Quantitative Techniques for Managers


Step 2: Find the First Reduced Cost Table


Jobs







1

2



3

4






A 0 5 2 8










0 3 8 2

Persons

B





2 0 4 7





C





2 0 1 1





D





Step 3: Find the Second Reduced Cost Table


Jobs







1

2



3

4






A 0 5 1 7










0 3 7 1

Persons

B





2 0 3 6





C





2 0 0 0





D




Step 4: Determine an Assignment

By examine row A of the table in Step 3, we find that it has only one zero (cell A1) box this zero and
cross out all other zeros in the boxed column. In this way we can eliminate cell B1.

Now examine row C, we find that it has one zero (cell C2) box this zero and cross out (eliminate) the
zeros in the boxed column. This is how cell D2 gets eliminated.

There is one zero in the column 3. Therefore, cell D3 gets boxed and this enables us to eliminate cell
D4.




Therefore, we can box (assign) or cross out (eliminate) all zeros.

The resultant table is shown below:








101

MBA-H2040 Quantitative Techniques for Managers





Jobs







1

2



3

4










A

0 5 1 7











Persons

B

0 3 7 1













C

2 3 6













D

2 0 0






Step 5:

0


The solution obtained in Step 4 is not optimal. Because we were able to make three assignments when
four were required.

Step 6:




0

Cover all the zeros of the table shown in the Step 4 with three lines (since already we made three
assignments).




Check row B since it has no assignment. Note that row B has a zero in column 1, therefore check

column1. Then we check row A since it has a zero in column 1. Note that no other rows and columns are

checked. Now we may draw three lines through unchecked rows (row C and D) and the checked column

(column 1). This is shown in the table given below:



Jobs







1

2



3

4










0

A 5 1 7










Persons

B

0 3 7 1













C

2 3 6













D

2 0 0




102

0
MBA-H2040 Quantitative Techniques for Managers




Step 7:
Develop the new revised table.

Examine those elements that are not covered by a line in the table given in Step 6. Take the smallest

element in this case the smallest element is 1. Subtract this smallest element from the uncovered cells

and add 1 to elements (C1 and D1) that lie at the intersection of two lines. Finally, we get the new

revised cost table, which is shown below:



Jobs







1

2



3

4










A 0 4 0 6










Persons

B

0 2 6 0













C

3 0 3 6













D

3 0 0 0

Step 8:






Now, go to Step 4 and repeat the procedure until we arrive at an optimal solution (assignment).


Step 9:


Determine an assignment


Examine each of the four rows in the table given in Step 7, we may find that it is only row C which has
only one zero box this cell C2 and cross out D2.

Note that all the remaining rows and columns have two zeros. Choose a zero arbitrarily, say A1 and box
this cell so that the cells A3 and B1 get eliminated.




Now row B (cell B4) and column 3 (cell D4) has one zero box these cells so that cell D4 is eliminated.






Thus, all the zeros are either boxed or eliminated. This is shown in the following table



103

MBA-H2040 Quantitative Techniques for Managers

Jobs







1

2



3

4










0

A 4 0 6










Persons

B

0 2 6













C

3 3 6

0













D

3 0 0







Since the number of assignments equal to the number of rows (columns), the assignment shown

in the above tale is optimal.

0




The total cost of assignment is: 78 that is A1 + B4 + C2 + D3












20 + 17 + 17 + 24 = 78


2.3 Unbalanced Assignment Problem

0


In the previous section we assumed that the number of persons to be assigned and the number of jobs

were same. Such kind of assignment problem is called as balanced assignment problem. Suppose if

the number of person is different from the number of jobs then the assignment problem is called as

unbalanced.

If the number of jobs is less than the number of persons, some of them can't be assigned any job. So that

we have to introduce on or more dummy jobs of zero duration to make the unbalanced assignment

problem into balanced assignment problem. This balanced assignment problem can be solved by using

the Hungarian Method as discussed in the previous section. The persons to whom the dummy jobs are

assigned are left out of assignment.

Similarly, if the number of persons is less than number of jobs then we have introduce one or more

dummy persons with zero duration to modify the unbalanced into balanced and then the problem is

solved using the Hungarian Method. Here the jobs assigned to the dummy persons are left out.





104
MBA-H2040 Quantitative Techniques for Managers
Example 2.2:

Solve the following unbalanced assignment problem of minimizing the total time for performing all the
jobs.

Jobs







1 2 3 4 5










A 5 2 4 2 5














B

2 4 7 6 6













Workers

C

6 7 5 8 7













D

5 2 3 3 4









E

8 3 7 8 6









F

3 6 3 5 7






Solution

In this problem the number of jobs is less than the number of workers so we have to introduce a dummy
job with zero duration.







105

MBA-H2040 Quantitative Techniques for Managers
The revised assignment problem is as follows:


Jobs







1 2 3 4 5 6










A 5 2 4 2 5 0














B

2 4 7 6 6 0













Workers

C

6 7 5 8 7 0













D

5 2 3 3 4 0









E

8 3 7 8 6 0









F

3 6 3 5 7 0







Now the problem becomes balanced one since the number of workers is equal to the number jobs. So
that the problem can be solved using Hungarian Method.

Step 1: The cost Table


Jobs







1 2 3 4 5 6










A 5 2 4 2 5 0










Workers

B

2 4 7 6 6 0













C

6 7 5 8 7 0













D

5 2 3 3 4 0









E

8 3 7 8 6 0









F

3 6 3 5 7 0




106
MBA-H2040 Quantitative Techniques for Managers

Step 2: Find the First Reduced Cost Table


Jobs







1 2 3 4 5 6










A 5 2 4 2 5 0










B

2 4 7 6 6 0













Workers

C

6 7 5 8 7 0













D

5 2 3 3 4 0









E

8 3 7 8 6 0









F

3 6 3 5 7 0







Step 3: Find the Second Reduced Cost Table





Jobs







1 2 3 4 5 6










A 3 0 1 0 1 0














B

0 2 4 4 2 0













Workers

C

4 5 2 6 3 0













D

3 0 0 1 0 0









E

6 1 4 6 2 0









F

1 4 0 3 3 0




107

MBA-H2040 Quantitative Techniques for Managers



Step 4: Determine an Assignment




By using the Hungarian Method the assignment is made as follows:






Jobs







1 2 3 4 5 6










A 3 0 1

0 1 0















B

2 4 4 2 0













Workers

C

4

0

5 2 6 3













D

3 0 0 1 0









E

6 1 4 6 2 0

0









F

1 4 3 3 0








0

Step 5:



The solution obtained in Step 4 is not optimal. Because we were able to make five assignments

when six were required.


Step 6:






Cover all the zeros of the table shown in the Step 4 with five lines (since already we made five

assignments).



Check row E since it has no assignment. Not

0 e that row B has a zero in column 6, therefore check

column6. Then we check row C since it has a zero in column 6. Note that no other rows and columns are
checked. Now we may draw five lines through unchecked rows (row A, B, D and F) and the checked
column (column 6). This is shown in the table given below:





108
MBA-H2040 Quantitative Techniques for Managers

Jobs







1 2 3 4 5 6










A 3 0 1

0 1 0















B

2 4 4 2 0













Workers

C

4

0

5 2 6 3













D

3 0 0 1 0









E

6 1 4 6 2 0

0









F

1 4 3 3 0








0

Step 7:


Develop the new revised table.




Examine those elements that are not covered by a line in the table given in Step 6. Take the

smallest element in this case the smallest element is 1. Subtract this smallest element from the
uncovered cells and add 1 to elements (A6, B6, D6 and F6) that lie at the intersection of two lines.
Finally, we get the new revised cost table, which is shown below:



Jobs







1 2 3 4 5 6

0



109

MBA-H2040 Quantitative Techniques for Managers









A 3 0 1 0 1 1














B

0 2 4 4 2 1













Workers

C

3 4 1 5 2 0













D

3 0 0 1 0 1









E

5 0 3 5 1 0









F

1 4 0 3 3 1

Step 8:






Now, go to Step 4 and repeat the procedure until we arrive at an optimal solution (assignment).



Step 9:


Determine an assignment




Jobs







1 2 3 4 5 6










0

A 3 0 1 1 1














B

2 4 4 2 1













Workers

C

3

0 4 1 5 2













D

3 0 0 1 1









E

5 3 5 1 0 0









F

1 4 3 3 1











0



110
MBA-H2040 Quantitative Techniques for Managers
Since the number of assignments equal to the number of rows (columns), the assignment shown in the
above tale is optimal.



Thus, the worker A is assigned to Job4, worker B is assigned to job 1, worker C is assigned to

job 6, worker D is assigned to job 5, worker E is assigned to job 2, and worker F is assigned to job 3.
Since the Job 6 is dummy so that worker C can't be assigned.



The total minimum time is: 14 that is A4 + B1 + D5 + E2 + F3












2 + 2 + 4 + 3 + 3 = 14




Example 2.3:



A marketing company wants to assign three employees viz. A, B, and C to four offices located at

W, X, Y and Z respectively. The assignment cost for this purpose is given in following table.



Offices







W

X

Y Z










A

160 220 240 200







Employees

B





100 320 260 160









C

100 200 460 250












Solution



Since the problem has fewer employees than offices so that we have introduce a dummy

employee with zero cost of assignment.




The revised problem is as follows:




Offices







W

X

Y Z



111

MBA-H2040 Quantitative Techniques for Managers









A

160 220 240 200







Employees

B





100 320 260 160









C

100 200 460 250









D

0 0 0 0







Now the problem becomes balanced. This can be solved by using Hungarian Method as in the

case of Example 2.2. Thus as per the Hungarian Method the assignment made as follows:



Employee A is assigned to Office X, Employee B is assigned to Office Z, Employee C is

assigned to Office W and Employee D is assigned to Office Y. Note that D is empty so that no one is
assigned to Office Y.




The minimum cost of assignment is: 220 + 160 + 100 = 480



2.4 Infeasible Assignment Problem



Sometimes it is possible a particular person is incapable of performing certain job or a specific

job can't be performed on a particular machine. In this case the solution of the problem takes into
account of these restrictions so that the infeasible assignment can be avoided.



The infeasible assignment can be avoided by assigning a very high cost to the cells where

assignments are restricted or prohibited. This is explained in the following Example 2.4.

Example 2.4:



A computer centre has five jobs to be done and has five computer machines to perform them.

The cost of processing of each job on any machine is shown in the table below.




Jobs







1 2 3 4 5



112
MBA-H2040 Quantitative Techniques for Managers









1

70 30 X 60 30











Computer 2

X 70 50 30 30

Machines











3

60 X 50 70 60













4

60 70 20 40 X









5

30 30 40 X 70









Because of specific job requirement and machine configurations certain jobs can't be done on

certain machines. These have been shown by X in the cost table. The assignment of jobs to the machines
must be done on a one to one basis. The objective here is to assign the jobs to the available machines so
as to minimize the total cost without violating the restrictions as mentioned above.



Solution




Step 1: The cost Table


Because certain jobs cannot be done on certain machines we assign a high cost say for example

500 to these cells i.e. cells with X and modify the cost table. The revised assignment problem is as
follows:


Jobs







1 2 3 4 5










1

70 30 500 60 30











Computer 2

500 70 50 30 30

Machines











3

60 500 50 70 60













4

60 70 20 40 500









5

30 30 40 500 70




113

MBA-H2040 Quantitative Techniques for Managers






Now we can solve this problem using Hungarian Method as discussed in the previous sections.




Step 2: Find the First Reduced Cost Table





Jobs







1 2 3 4 5










1

40 0 470 30 0











Computer 2

470 40 20 0 0

Machines











3

10 450 0 20 10













4

40 50 0 20 480









5

0 0 10 470 40











Step 3: Find the Second Reduced Cost Table


Jobs







1 2 3 4 5



114
MBA-H2040 Quantitative Techniques for Managers









1

40 0 470 30 0











Computer 2

470 40 20 0 0

Machines











3

10 450 0 20 10













4

40 50 0 20 480









5

0 0 10 470 40







Step 4: Determine an Assignment



Jobs







1 2 3 4 5










1

40 0 470 30 0











Computer 2

470 40 20 0

Machines











0

3

10 450 20 10













4

40 50 0 20 480









5

0 10

0

470 40







Step 5:



The solution obtained in Step 4 is not optimal. Because we were able to make four assignments

when five were required.


Step 6:


Cover all the zeros of the table shown in the Step 4 with four lines (since already we made four

assignments).

0







115

MBA-H2040 Quantitative Techniques for Managers

Check row 4 since it has no assignment. Note that row 4 has a zero in column 3, therefore check

column3. Then we check row 3 since it has a zero in column 3. Note that no other rows and columns are
checked. Now we may draw four lines through unchecked rows (row 1, 2, 3 and 5) and the checked
column (column 3). This is shown in the table given below:

Jobs







1 2 3 4 5










1

40 0 470 30 0











Computer 2

470 40 20 0

Machines











0

3

10 450 20 10













4

40 50 0 20 480









5

0 10

0

470 40






Step 7:


Develop the new revised table.




Examine those elements that are not covered by a line in the table given in Step 6. Take the

smallest element in this case the smallest element is 10. Subtract this smallest element from the
uncovered cells and add 1 to elements (A6, B6, D6 and F6) that lie at the intersection of two lines.
Finally, we get the new revised cost table, which is shown below:

Jobs







1 2 3 4 5







0





1

40 0 471 30 0











Computer 2

470 40 21 0 0

Machines











3

0 440 0 10 0













4

30 40 0 10 470









5

0 0 11 470 40

Step 8:






Now, go to Step 4 and repeat the procedure until we arrive at an optimal solution (assignment).



116
MBA-H2040 Quantitative Techniques for Managers

Step 9:


Determine an assignment



Jobs







1 2 3 4 5










1

40 0 471 30

0











Computer 2

470 40 21 0

Machines











3

440 0 10 0













4

30 40 10 470









0

5

0 11 470 40








Since the number of assignments equal to the number of rows (columns), the assignment shown

in the above tale is optimal.


0



Thus, the Machine1 is assigned to Job5, Machine 2 is assigned to job4, Machine3 is assigned to

job1, Machine4 is assigned to job3 and Machine5 is assigned to job2.



The minimum assignment cost is: 170




0

2.5 Maximization in an Assignment Problem



There are situations where certain facilities have to be assigned to a number of jobs so as to

maximize the overall performance of the assignment. In such cases the problem can be converted into a
minimization problem and can be solved by using Hungarian Method. Here the conversion of
maximization problem into a minimization can be done by subtracting all the elements of the cost table
from the highest value of that table.

Example 2.5:



Consider the problem of five different machines can do any of the required five jobs with

different profits resulting from each assignment as illustrated below:



Machines







1 2 3 4 5



117

MBA-H2040 Quantitative Techniques for Managers









1

40 47 50 38 50













2

50 34 37 31 46

Jobs











3

50 42 43 40 45













4

35 48 50 46 46









5

38 72 51 51 49







Find out the maximum profit through optimal assignment.

Solution



This is a maximization problem, so that first we have to find out the highest value in the table

and subtract all the values from the highest value. In this case the highest value is 72.



The new revised table is given below:



Machines







1 2 3 4 5










1

32 35 22 34 22













2

22 38 35 41 26

Jobs











3

22 30 29 32 27













4

37 24 22 26 26









5

34 0 21 21 23









This can be solved by using the Hungarian Method.




By solving this, we obtain the solution is as follows:



118
MBA-H2040 Quantitative Techniques for Managers









Jobs



Machines










1





3










2





5















3





1










4





4



5

2




The maximum profit through this assignment is: 264


2.6 Crew Assignment Problem






The crew assignment problem is explained with the help of the following problem

Problem:



A trip from Chennai to Coimbatore takes six hours by bus. A typical time table of the bus service

in both the direction is given in the Table 1. The cost of providing this service by the company based on
the time spent by the bus crew i.e. driver and conductor away from their places in addition to service
times. The company has five crews. The condition here is that every crew should be provided with more
than 4 hours of rest before the return trip again and should not wait for more than 24 hours for the return
trip. Also the company has guest house facilities for the crew of Chennai as well as at Coimbatore.



Find which line of service is connected with which other line so as to reduce the waiting time to

the minimum.


Table 1



Departure from Route Number

Arrival at

Arrival at

Route Number Departure from

Chennai

Coimbatore

Chennai

Coimbatore

06.00

1

12.00

11.30

a

05.30

07.30

2

13.30

15.00

b

09.00

11.30

3

17.30

21.00

c

15.00

19.00

4

01.00

00.30

d

18.30

00.30

5

06.30

06.00

e

00.00


Solution

For each line the service time is constant so that it does not include directly in the computation. Suppose

if the entire crew resides at Chennai then the waiting times in hours at Coimbatore for different route

connections are given below in Table 2.



If route 1 is combined with route a, the crew after arriving at Coimbatore at 12 Noon start at 5.30

next morning. Thus the waiting time is 17.5 hours. Some of the assignments are infeasible. Route c



119

MBA-H2040 Quantitative Techniques for Managers
leaves Coimbatore at 15.00 hours. Thus the crew of route 1 reaching Coimbatore at 12 Noon are unable

to take the minimum stipulated rest of four hours if they are asked to leave by route c. Hence 1-c is an

infeasible assignment. Thus it cost is M (a large positive number).






Table 2







Route a b c d e










1

17.5 21 M 6.5 12













2

16 19.5 M 5 10.5













3

12 15.5 21.5 M 6.5













4

4.5 8 4 17.5 23









5

23 M 8.5 12 17.5












Similarly, if the crews are assumed to reside at Coimbatore then the waiting times of the crew in

hours at Chennai for different route combinations are given below in Table 3.


Table 3







Route a b c d e










1

18.5 15 9 5.5 M













2

20 16.5 10.5 7 M













3

M 20.5 14.5 11 5.5













4

7.5 M 22 18.5 13









5

13 9.5 M M 18.5







120
MBA-H2040 Quantitative Techniques for Managers


Suppose, if the crew can be instructed to reside either at Chennai or at Coimbatore, minimum

waiting time from the above operation can be computed for different route combination by choosing the
minimum of the two waiting times (shown in the Table 2 and Table 3). This is given in the following
Table 4.

Table 4







Route a b c d e










1

17.5* 15 9 5.5 12*













2

16* 16.5 10.5 5* 10.5*













3

12* 15.5* 14.5 11 5.5













4

4.5* 8* 14* 17.5* 13









5

13 9.5 8.5* 12* 17.5*






Note: The asterisk marked waiting times denotes that the crew are based at Chennai; otherwise they are
based at Coimbatore.

Now we can solve the assignment problem (presented in Table 4) using Hungarian Method.

Step 1: Cost Table (Table 5)


Table 5







Route a b c d e










1

17.5* 15 9 5.5 12*













2

16* 16.5 10.5 5* 10.5*













3

12* 15.5* 14.5 11 5.5













4

4.5* 8* 14* 17.5* 13









5

13 9.5 8.5* 12* 17.5*







121

MBA-H2040 Quantitative Techniques for Managers
Step 2: Find the First Reduced cost table (Table 6)


Table 6







Route a b c d e










1

12 9.5 3.5 0 6.5













2

11 11.5 5.5 0 5.5













3

6.5 10 9 5.5 0













4

0 3.5 9.5 13 8.5









5

4.5 1 0 3.5 9






Step 3: Find the Second Reduced cost table (Table 7)



Table 7







Route a b c d e










1

12 8.5 3.5 0 6.5













2

11 10.5 5.5 0 5.5













3

6.5 9 9 5.5 0













4

0 2.5 9.5 13 8.5









5

4.5 0 0 3.5 9









122
MBA-H2040 Quantitative Techniques for Managers
Step 4: Determine an Assignment (Table 8)


Table 8







Route a b c d e










0

1

12 8.5 3.5 6.5













2

11 10.5 5.5 0 5.5













3

6.5 9 9 5.5













4

2.5 9.5 13 8.5









5

4.5 0 3.5 9 0






Step 5: The solution obtained in Step 4 is not optimal since the number of assignments are less than the
number of rows (columns).

Step 6: Check () row 2 since it

0 has no assignment and note that row 2 has a zero in column d, therefore

check () column d also. Then check row 1 since it has zero in column d. Draw the lines through the
unchecked rows and checked column using 4 lines (only 4 assignments are made). This is shown in
Table 9.

Table 9







Route a b

0 c d e










0

1

12 8.5 3.5 6.5













2

11 10.5 5.5 0 5.5













3

6.5 9 9 5.5













4

2.5 9.5 13 8.5









5

4.5 0 3.5 9 0










123

0

MBA-H2040 Quantitative Techniques for Managers

Step 7: Develop a new revised table (Table 10)



Take the smallest element from the elements not covered by the lines in this case 3.5 is the

smallest element. Subtract all the uncovered elements from 3.5 and add 3.5 to the elements lie at the
intersection of two lines (cells 3d, 4d and 5d). The new revised table is presented in Table 10.



Table 10







Route a b c d e










1

8.5 5 0 0 3













2

7.5 7 2 0 2













3

6.5 9 9 9 0













4

0 2.5 9.5 16.5 8.5









5

4.5 0 0 7 9







Step 8: Go to Step 4 and repeat the procedure until an optimal solution if arrived.

Step 9: Determine an Assignment (Table 11)


Table 11







Route a b c d e



124
MBA-H2040 Quantitative Techniques for Managers









1

8.5 5

0 0 3













2

7.5 7 2 2













3

6.5 9 9 9

0















4

2.5 9.5 16.5 8.5









5

4.5 0 7 9 0



The assignment



illustrated in the above

Table 11 is optimal since the number of assignments is equal to the number of rows (columns).



Thus, the routes to be prepared to achieve the minimum waiting time are as follows:

















1

0 ?

c, 2 ? d, 3 ? e, 4 ? a and 5 ? b


By referring Table 5, we can obtain the waiting times of these assignments as well as the residence
(guest house) of the crews. This is presented in the following Table 12.

Table 12

Routes

Residence of the Crew

Waiting Time

0

1 ? c

Coimbatore

9

2 ? d

Chennai

5

3 ? e

Coimbatore

5.5

4 ? a

Chennai

4.5

5 - b

Coimbatore

9.5


2.7 Summary

The assignment problem is used for the allocation of a number of persons to a number of jobs so that the

total time of completion is minimized. The assignment problem is said to be balanced if it has equal

number of person and jobs to be assigned. If the number of persons (jobs) is different from the number

of jobs (persons) then the problem is said to be unbalanced. An unbalanced assignment problem can be

solved by converting into a balanced assignment problem. The conversion is done by introducing

dummy person or a dummy job with zero cost.



Because of the special structure of the assignment problem, it is solved by using a special

method known as Hungarian Method.

2.8 Key Terms

Cost Table: The completion time or cost corresponding to every assignment is written down in a table
form if referred as a cost table.



125

MBA-H2040 Quantitative Techniques for Managers

Hungarian Method: is a technique of solving assignment problems.

Assignment Problem: is a special kind of linear programming problem where the objective is to
minimize the assignment cost or time.

Balanced Assignment Problem: is an assignment problem where the number of persons equal to the
number of jobs.

Unbalanced Assignment Problem: is an assignment problem where the number of jobs is not equal to
the number of persons.

Infeasible Assignment Problem: is an assignment problem where a particular person is unable to
perform a particular job or certain job cannot be done by certain machines.



2.9 Self Assessment Questions

Q1. A tourist company owns a one car in each of the five locations viz. L1, L2, L3, L4, L5 and a
passengers in each of the five cities C1, C2, C3, C4, C5 respectively. The following table shows the
distant between the locations and cities in kilometer. How should be cars be assigned to the passengers
so as to minimize the total distance covered.


Cities







C1

C 2

C3 C4 C5










L1 120 110 115 30 36










Locations

L2

125 100 95 30 16













L3

155 90 135 60 50













L4

160 140 150 60 60













L5

190 155 165 90 85






Q2. Solve the following assignment problem
1 2 3 4 5



126
MBA-H2040 Quantitative Techniques for Managers





1





Rs.3 Rs.8 Rs.2 Rs.10 Rs.3













2

Rs.8 Rs.7 Rs.2 Rs.9 Rs.7









3

Rs.6 Rs.4 Rs.2 Rs.7 Rs.5









4

Rs.8 Rs.4 Rs.2 Rs.3 Rs.5









5

Rs.9 Rs.10 Rs.6 Rs.9 Rs.10






Q3. Work out the various steps of the solution of the Example 2.3.

Q4. A steel company has five jobs to be done and has five softening machines to do them. The cost of
softening each job on any machine is given in the following cost matrix. The assignment of jobs to
machines must be done on a one to one basis. Here is the objective is to assign the jobs to the machines
so as to minimize the total assignment cost without violating the restrictions.



Jobs







1 2 3 4 5










1

80 30 X 70 30











Softening 2

70 X 60 40 30

Machines











3

X 80 60 80 70













4

70 80 30 50 X









5

30 30 50 X 80






Q5. Work out the various steps of the solution of the problem presented in Example 2.5.

Q6. A marketing manager wants to assign salesman to four cities. He has four salesmen of varying
experience. The possible profit for each salesman in each city is given in the following table. Find out an
assignment which maximizes the profit.




127

MBA-H2040 Quantitative Techniques for Managers

Cities







1 2 3 4










1

25 27 28 38











Salesmen 2

28 34 29 40













3

35 24 32 33













4

24 32 25 28






Q7. Shiva's three wife, Rani, Brinda, and Fathima want to earn some money to take care of personal
expenses during a school trip to the local beach. Mr. Shiva has chosen three chores for his wife:
washing, cooking, sweeping the cars. Mr. Shiva asked them to submit bids for what they feel was a fair
pay for each of the three chores. The three wife of Shiva accept his decision. The following table
summarizes the bid received.




Chores







Washing Cooking Sweeping

1 2 3










Rani 25 18 17










Wife's Brinda

17 25 15













Fathima

18 22 32











Q8. Solve the following problem


Office







O1 O2 O3 O4



128
MBA-H2040 Quantitative Techniques for Managers









E1

2600 3200 3400 3000











Employees E2

2000 4200 3600 2600













E3

2000 3000 5600 4000











Q9. The railway operates seven days a week has a time table shown in the following table. Crews
(Driver and Guard) must have minimum rest of six hours between trans. Prepare the combination of
trains that minimizes waiting time away from the city. Note that for any given combination the crew will
be based at the city that results in the smaller waiting time and also find out for each combination the
city where the crew should be based at.


Train No.

Departure at

Arrival at

Train No.

Departure at

Arrival at

Bangalore

Chennai

Chennai

Bangalore

101

7 AM

9 AM

201

9 AM

11 AM

102

9 AM

11 AM

202

10 AM

12 Noon

103

1.30 PM

3.30 PM

203

3.30 PM

5.30 PM

104

7.30 PM

9.30 PM

204

8 PM

10 PM


2.10 Key Solutions

Q1. L1 ? C1, L2 ? C3, L3 ? C2, L4 _ C4, L5 ? C5 and
Minimum Distance is: 450

Q2. 1 ? 5, 2 ? 3, 3 ? 2, 4 ? 4, 5 ? 1 and
Minimum Cost is: Rs.21


Q4. 1 ? 2, 2 ? 4, 3 ? 3, 4 ? 4, 5 ? 1 and
Minimum Assignment Cost is:

Q6. 1 - 1, 2 ? 4, 3 ? 3, 4 ? 2 and
Maximum Profit is: 139

Q7. Rani ? Cooking, Brinda ? Sweeping, Fathima ? Washing and
Minimum Bids Rate is: 51

Q8. E1 ? O2, E2 ? O4, E3 ? O1
Since E4 is empty, Office O3 cannot be assigned to any one.
Minimum Cost is: 7800

Q9.





Trains





Cities















201 ? 103





Bangalore







202 ? 104





Chennai



129

MBA-H2040 Quantitative Techniques for Managers






203 ? 101





Bangalore







204 ? 102





Bangalore



2.11 Further References

Hamdy A Taha, 1999. Introduction to Operations Research, PHI Limited, New Delhi.

Cooper, L and D. Steinberg, 1974. Methods and Applications of Linear Programmings, Saunders,
Philadelphia, USA.

Mustafi, C.K. 1988. Operations Research Methods and Practices, Wiley Eastern Limited, New Delhi.


















130
MBA-H2040 Quantitative Techniques for Managers



UNIT II










3 INTRODUCTION TO INVENTORY MANAGEMENT

ON
S



S
E



L





LESSON STRUCTURE




3.1 Introduction



3.2 Objectives of Inventory



3.3 Inventory is an Essential Requirement



3.4 Basic Functions of Inventory



3.5 Types of Inventory



3.6 Factors Affecting Inventory



3.7 Summary



3.8 Key Terms

Objectives

3.9 Self Assessment Questions

After Studying this lesson, you should be able

3.10 Further References

to:







Understand what is inventory



Describe various inventory concepts



Describe the objectives of inventory



Explain the functions of inventory



Describe requirements of inventory



Explain different types of inventory



Describe different factors affecting



inventory













































131

MBA-H2040 Quantitative Techniques for Managers
3.1 Introduction

Simply inventory is a stock of physical assets. The physical assets have some economic value, which

can be either in the form of material, men or money. Inventory is also called as an idle resource as long

as it is not utilized. Inventory may be regarded as those goods which are procured, stored and used for

day to day functioning of the organization.



Inventory can be in the form of physical resource such as raw materials, semi-finished goods

used in the process of production, finished goods which are ready for delivery to the consumers, human

resources, or financial resources such as working capital etc.



Inventories means measures of power and wealth of a nation or of an individual during centuries

ago. That is a business man or a nation's wealth and power were assessed in terms of grammes of gold,

heads of cattle, quintals of rice etc.



In recent past, inventories mean measure of business failure. Therefore, businessmen have

started to put more emphasis on the liquidity of assets as inventories, until fast turnover has become a

goal to be pursues for its own sake.



Today inventories are viewed as a large potential risk rather than as a measure of wealth due to

the fast developments and changes in product life. The concept of inventories at present has necessitated

the use of scientific techniques in the inventory management called as inventory control.

Thus, inventory control is the technique of maintaining stock items at desired levels. In other

words, inventory control is the means by which material of the correct quality and quantity is made

available as and when it is needed with due regard to economy in the holding cost, ordering costs, setup

costs, production costs, purchase costs and working capital.

Inventory Management answers two questions viz. How much to order? and when to order?

Management scientist insisting that the inventory is an very essential requirement. Why? This is

illustrated in the next section with the help of materials conversion process diagram.

3.2 Objectives of Inventory

Inventory has the following main objectives:


To supply the raw material, sub-assemblies, semi-finished goods, finished goods, etc. to

its users as per their requirements at right time and at right price.

To maintain the minimum level of waste, surplus, inactive, scrap and obsolete items.
To minimize the inventory costs such as holding cost, replacement cost, breakdown cost

and shortage cost.

To maximize the efficiency in production and distribution.
To maintain the overall inventory investment at the lowest level.
To treat inventory as investment which is risky? For some items, investment may lead to

higher profits and for others less profit.




132
MBA-H2040 Quantitative Techniques for Managers
3.3 Inventory is an Essential Requirement

Inventory is a part and parcel of every facet of business life. Without inventory no business activity can

be performed, whether it being a manufacturing organization or service organization such as libraries,

banks, hospitals etc. Irrespective of the specific organization, inventories are reflected by way of a

conversion process of inputs to outputs. The conversion process is illustrated in the figure 3.1 as given

below.



From the figure 3.1 we can see that there may be stock pints at three stages viz. Input,

Conversion Process and Output. The socks at input are called raw materials whereas the stocks at the

output are called products. The stocks at the conversion process may be called finished or semi-finished

goods or sometimes may be raw material depending on the requirement of the product at conversion

process, where the input and output are based on the market situations of uncertainty, it becomes

physically impossible and economically impractical for each stock item to arrive exactly where it is

required and when it is required.















Random Fluctuations






Conversion

Inputs

Inventory

process Inventory



Sock-points






















Comparing Actual Material





















Stock levels to planned levels

Material Sock-Points






Land















Outputs

Labour









Capital





Management















Feed Back



Product Inventory



Stock-points











Corrective action: More or less Stocks













Fig: 3.1 Materials Conversion Process



Even it is physically possible to deliver the stock when it is required, it costs more expensive. This is the

basic reason for carrying the inventory. Thus, inventories play an essential and pervasive role in any

organization because they make it possible:



133

MBA-H2040 Quantitative Techniques for Managers


To meet unexpected demand
To achieve return on investment
To order largest quantities of goods, components or materials from the suppliers at

advantageous prices

To provide reasonable customer service through supplying most of the requirements from

stock without delay

To avoid economically impractical and physically impossible delivering/getting right

amount of stock at right time of required

To maintain more work force levels
To facilitate economic production runs
To advantage of shipping economies
To smooth seasonal or critical demand
To facilitate the intermittent production of several products on the same facility
To make effective utilization of space and capital
To meet variations in customer demand
To take the advantage of price discount
To hedge against price increases
To discount quantity


3.4 Basic Functions of Inventory



The important basic function of inventory is

Increase the profitability- through manufacturing and marketing support. But zero

inventory manufacturing- distribution system is not practically possible, so it is
important to remember that each rupee invested in inventory should achieve a specific
goal. The other inventory basic functions are


Geographical Specialization

Decoupling

Balancing supply and demand and

Safety stock


Inventory Investment Alternative

Investment is most important and major part of asset, which should be required to produce a minimum

investment return. The MEC (Marginal Efficiency of Capital) concept holds that an organization should

invest in those alternatives that produce a higher investment return than capital to borrow. The following

figure 3.2 shows that investment alternative A on the MEC curve is acceptable.












134
MBA-H2040 Quantitative Techniques for Managers

100

)



t ( %

A

c

c

e

pt

e

d I

n

v

entory Investment

e
n





















s
t
m

ve

A

Cost of Capital


I
n





















n




O



r
n



t
u

Re















Reject Inventory Investment



0



20



40



60



80



100

Total Investment Alternatives (in %)

Fig 3.2 Typical MEC curve





The curve shows that about 20% of the inventory investment alternatives will produce a return

on investment above the capital cost.

Geographical Specialization

Another basic inventory function is to allow the geographical specialization individual operating units.

There is a considerable distance between the economical manufacturing location and demand areas due

to factors of production such as raw material, labour, water, power. So that the goods from various

manufacturing locations are collected at a simple warehouse or plant to assemble in final product or to

offer consumers a single mixed product shipment. This also provides economic specialization between

manufacturing and distribution units/locations of an organization.

Decoupling
The provision of maximum efficiency of operations within a single facility is also one of the important

basic functions of the inventory. This is achieved by decoupling, which is done by breaking operations

apart so that one operation(s) supply is independent of another(s) supply.



The decoupling function serves in two ways of purposes, they are

1. Inventories are needed to reduce the dependencies among successive stages of operations so that
shortage of materials, breakdowns or other production fluctuations at one stage do not cause later stage
to shut down. This is illustrated in the following figure 3.3 in an engineering unit.



Raw

D i

e c a

s ti n

g

D r il l i

n

g

D e -

b

ur

ni

n g

P a c

ki n

g Finished

Material





















Products





Inventory

Inventory





135

MBA-H2040 Quantitative Techniques for Managers

Fig: 3.3 Decoupling of operation using inventory



The figure shows that the de-burning, packing could continue to operate from inventories should

die-casting and drilling be shut down or they can be decoupled from the production processes that
precede them.



2. One organizational unit schedules its operations independently of another organizational unit.

For example: Consider an automobile organization, here assembly process can be schedule separately
from engine built up operation, and each can be decoupled from final automobile assembly operations
through in process inventories.

Supply and Demand Balancing

The function of Balancing concerns elapsed time between manufacturing and using the product.

Balancing inventories exist to reconcile supply with demand. The most noticeable example of balancing

is seasonal production and year round usage like sugar, rice, woolen textiles, etc. Thus the investment of

balancing inventories links the economies of manufacturing with variations of usage.

Safety Stock

The safety stock also called as buffer stock. The function of safety stock concerns short range variations

in either replacement or demand. Determination of the safety stock size requires a great deal of

inventory planning. Safety stock provides protection against two types of uncertainty, they are





1. Sales in excess of forecast during the replenishment period
2. Delays in replenishment

Thus, the inventories committed to safety stocks denote the greatest potential for improved

performances. There are different techniques are available to develop safety stocks.

3.5 Types of Inventory

Inventory may be classified into manufacturing, service and control aspects, which is illustrated in the
figure 3.4 as given below:



136
MBA-H2040 Quantitative Techniques for Managers


Raw Materials/Production Inventory

















Work-in-Process Inventory




Manufacturing M.R.O Inventory



Aspect



Finished Goods Inventory



















Miscellaneous Inventory












Lot Size Stocks





y



t
or
n







ve















Anticipation Stocks


I
n



Inventory Service Aspect



e
s

of



yp

T















Fluctuation Stocks

.
4




3



i
g:



F















Risk Stocks



















A-Items Inventory




















B-Items Inventory



Control Aspect



















C-Items Inventory



137

MBA-H2040 Quantitative Techniques for Managers



Each inventory type is discussed in detail are as follows:


Raw Material/Manufacturing Inventory

There are five types of Manufacturing Inventory, they are

Production Inventory

Items going to final product such a raw materials, sub-assemblies purchased from outside are called
production inventory.

Work-in-Process Inventory

The items in the form of semi-finished or products at different stage of production process are known as
work-in-process inventory.

M.R.O. Inventory

Maintenance, Repair and Operating supplies such as spare parts and consumable stores, which do not go
into final product but are consumed during the production process.

Finished Goods Inventory

Finished Goods Inventory includes the products ready for dispatch to the consumers or
distributors/retailers.

Miscellaneous Inventory

Items excluding those mentioned above, such as waste, scrap, obsolete, and un-saleable items arising
from the main production process, stationery used in the office and other items required by office,
factory and other departments, etc. are called miscellaneous inventory.

Service Inventory

The service inventory can be classified into four types, they are:

Lot Size Stocks

Lot size means purchasing in lots. The reasons for this is to

Obtain quantity discounts
Minimize receiving and handling costs
Reduce purchase and transport costs

For example: It would be uneconomical for a textile factory to buy cotton everyday rather than in
bulk during the cotton season.


Anticipation Stocks

Anticipation stocks are kept to meet predictable changes in demand or in availability of raw materials.



138
MBA-H2040 Quantitative Techniques for Managers
For example: The purchase of potatoes in the potato season for sale of roots preservation products
throughout the year.
Fluctuation Stocks

Fluctuation stocks are carried to ensure ready supplies to consumers in the face of irregular fluctuations
in their demands

Risk Stocks

Risk stocks are the items required to ensure that there is no risk of complete production breakdown. Risk
stocks are critical and important for production.

Control Inventory

A good way of examining an inventory control is: to make ABC classification, which is also known as

ABC analysis. ABC analysis means the "control" will be "Always Better" if we start with the ABC

classification of inventory.

The ABC concepts classifies inventories into three groups in terms of percentage of total value

and percentage of number of inventory items, this is illustrated in the figure 3.5 and 3.6 as given below.







Values

e




u





A

A



10% 70%


Val






of











t
age

A items

20% 20%

B

B







e
r
c
e
n



P



B i t e

m

s 70% 10%



C

i t

e

m s

C



C





Percentage of Numbers Numbers











Fig: 3.5 ABC Classification (Frequency Form)

Fig: 3.6 ABC Classification (Tabular Form)




The three groups of inventory items are called A-items group, B-items group, C-items group,

which are explained as follows:

A-items Group: This constitutes 10% of the total number of inventory items and 70% of total money
value for all the items.




139

MBA-H2040 Quantitative Techniques for Managers
B-items Group: This constitutes 20% of the total number of inventory items and 20 % of total money
value for all the items.

C-items Group: This constitutes 70% of the total number of inventory items and 10 % of total money
value for all the items. This is just opposite of A-items group.



The ABC classification provides us clear indication for setting properties of control to the items,

and A-class item receive the importance first in every respect such as tight control, more security, and

high operating doctrine of the inventory control.

The coupling of ABC classification with VED classification enhances the inventory control

efficiency. VED classification means Vital, Essential and Desirable Classification. From the above

description, it may be noted that ABC classification is based on the logic of proportionate value while

VED classification based on experience, judgment, etc. The ABC /VED classification is presented in the

following figure 3.7.





V

E

D

Total

A

3

5

2

10

B

5

7

8

20

C

10

40

20

70

Total

18

52

30

100













Fig: 3.7 ABC/VED Classifications

This is an example of a particular case

The values are expressed in percentage




Note that the total number of categories becomes nine.


3.6 Factors Affecting Inventory

The main problem of inventory control is to answer tow questions viz.

1. How much to order? and
2. When to order?

These questions are answered by developing a inventory model, which is based on the

consideration of the main aspects of inventory viz. demand and cost. There are many factors related with

these tow main factors (Demand and Cost). In this section we will discuss these different factors.

The different factors are:


Economic Parameters

Demand

Ordering Cycle
Delivery Lag

Time Horizon



140
MBA-H2040 Quantitative Techniques for Managers

Stages of Inventory

Number of Supply Echelons

Number and Availability of Items

Government's and Organization's Policy

Economic Parameters

There are different types of economic parameters, they are:

Purchase Price

Procurement Costs

Selling Price

Holding Costs

Shortage costs

Information Processing System Operating Costs

Purchase Cost

The cost of the item is the direct manufacturing cost if it is produced in in-house or the cost paid to the

supplier for the item received. This cost usually equal to the purchase price. When the marketing price

goes on fluctuating, inventory planning is based on the average price mostly it is called as a fixed price.

When price discounts can be secured or when large production runs may result in a decrease in the

production cost, the price factor is of special interest.

Procurement Costs

The costs of placing a purchase order is known as ordering costs and the costs of initial preparation of a

production system (if in-house manufacturing) is called as set up cost. These costs are called as

procurement cost, but these costs vary directly with each purchase order placed or with set up made and

are normally assumed independent of the quantity ordered or produced.



Procurement costs include costs of transportation of items ordered, expediting and follow up,

goods receiving and inspection, administration (includes telephone bills, computer cost, postage, salaries

of the persons working for tendering, purchasing, paper work, etc.), payment processing etc. This cost is

expressed as the cost per order/setup.

Holding Costs

The holding costs also called as carrying costs. The cost associated with holding/carrying of stocks is

called holding cost or carrying cost or possession cost. Holding costs includes handling/carrying cost,

maintenance cost, insurance, safety measures, warehouse rent, depreciation, theft, obsolescence, salaries,

interest on the locked money, etc. Thus, by considering all these elements the storage cost is expressed

either as per unit of item held per unit of time or as a percentage of average money value of investment

held. Therefore the size of all these holding costs usually increases or decreases in proportion to the

amount of inventory that is carried.

Shortage Costs



141

MBA-H2040 Quantitative Techniques for Managers
These costs are penalty costs as a result of running out of stock at the time of item is required. There are

different forms of shortage cost, which is illustrated in the following figure 3.8. One form of the

shortage costs is called as back order on the selling side or backlogging cost on the production side when

the unsatisfied demand can be satisfied at later stage that is consumers has to wait till they gests the

supply.



The second form of shortage costs is called as lost sales costs on the selling side or no

backlogging costs on the production side, when the unsatisfied demand is lost or the consumers goes

some where else instead of waiting for the supply.

Shortage Costs





Selling Side











Production Side






























Back order Costs

Lost Sales Costs





Backlogging

No Backlogging



Costs



Costs



Fig: 3.8 Nature of Shortage Costs



These includes the costs of production stoppage, overtime payments, idle machine, loss of

goodwill, loss of sales opportunity, special order at higher price, loss of profits etc.

Information System Operation Costs

Today there are more inventory records should be maintain in the organization, so that some person

must update the records either by hand or by using computer. If the inventory levels are not recorded

daily, this operating cost is incurred in obtaining accurate physical inventory record counts. The

operating costs are fixed.

Demand

A commodity demand pattern may be deterministic or probabilistic.

Deterministic Demand

In this case, the demand is assumed that the quantities of commodity needed over subsequent periods of

time are known with certainty. This is expressed over equal time periods in terms of known constant

demands or in terms of variable demands. The two cases are called as static and dynamic demands.

Probabilistic Demand

This occurs when requirements over a certain time period are not known with certainty but their pattern

can be denoted by a known probability distribution. In this case, the probability distribution is said to be



142
MBA-H2040 Quantitative Techniques for Managers
stationary or non-stationary over time periods. The terms stationary and non-stationary are equivalent to

the terms static and dynamic in the deterministic demand.


For a given time period the demand may be instantaneously satisfied at the beginning of the time period

or uniformly during that time period. The effect of uniform and instantaneous demand directly reflects

on the total cost of carrying inventory.


Ordering Cycle

The ordering cost is related with the inventory situation time measurement. An ordering cycle can be

identified by the time period between two successive placements of orders. The later may be initiated in

one of two ways as:

Periodic Review

Continuous Review


Periodic Review
In this case, the orders are placed at equally intervals of time.

Continuous Review

In this case, an inventory record is updated continuously until a certain lower limit is reached at this
point a new order is placed. Some times this is referred as the two-bin system.

Delivery Lag




The requirement of the inventory is felt and an order is placed, it may be delivered instantaneously or

some times it may be needed before delivery if affected. The time period between the placement of the

requisition for an item and its receipt for actual use is called as delivery lag. The delivery lags also

known as lead time.



There are four different types of lead time, they are

Administrative Lead Time

Transportation Lead Time

Suppliers Lead Time

Inspection Lead Time

The Inspection Lead Time and Administrative Lead Time can be fixed in nature, where as the

Transportation Lead Time and Suppliers Lead Time can never be fixed. It means generally the lead time

may be deterministic or probabilistic.

Time Horizon

Time horizon is, the planning period over which inventory is to be controlled. The planning period may

be finite or infinite in nature. Generally, inventory planning is done on annual basis in most of the

organizations.


143

MBA-H2040 Quantitative Techniques for Managers




Stages of Inventory

In the sequential production process, if the items/parts are stocked at more than one point they are called

multi-stage inventories. This is illustrated in the following figure 3.9.



Raw

D i

e c a

s ti n

g

D r il l i

n

g

D e -

b

ur

ni

n g

P a c

ki n

g Finished

Material





















Products




Inventory

Inventory



Fig: 3.9 Multistage Inventories

Number of Supply Echelons

Already we saw that there are several stocking points in the inventory system. These stocking points are

organized in such that one points act as a supply source for some other points. For example, the

production factories supplies the products to warehouse and the warehouse supplies to the retailer and

then to the consumers. In this the each level of movement of the product is called on echelon. This is

illustrated in the figure 3.10 given below.





Consumer










Consumer

Retailer









Consumer



Warehouse




Retailer

Production



Factory



Warehouse




Fig: 3.10 Multi Echelon Supply Systems


Number and Availability of Items

Due to different marker situations some times supply position is poorly affected, which in turn affects

the poison of inventory in the organizations.



Generally inventory includes more than one item. Therefore the number of items in inventory

affects the situation when these items complete for limited total capital or limited space.

Government's and Organization's Policy


144
MBA-H2040 Quantitative Techniques for Managers

There are different governments and as well as organization policies such as import and export,

availability of capital, land, labour, pollutions systems, etc. The government has laid down some policy

norms for items to be imported as well as for other items like highly inflammable, explosive and other

important materials. Similarly, an organization also has certain policies based on the availability of

capital, labour, etc. All these policies affect organization inventories level.



We have discussed different factors that affect the inventory in an organization (above). These

factors are responsible for the development of proper inventory system is called as characteristics of

inventory.


3.7 Summary



This lesson illustrated the introduction of inventory and inventory management/control. This lesson also

illustrated the objectives of inventory, inventory functions, inventory types and the various factors

affecting the inventory.

3.8 Key Terms




Carrying Cost: Cost of maintaining one unit of an item in the stock per unit of time (normally one
year). The carrying cost also called as Holding Cost.

Decoupling: Use of inventories to break apart operations so that one operations supply is independent of
another.

Backlog: Accumulation of unsatisfied demands

Delivery Lag: Time between the placing of an order for the item and receipt of the items for use.

ABC Classification: Classifications of inventories in terms of annual usage value in different categories
of high value (A), medium value (B) and low value (C).

VED Classification: Vital Essential Desirable Classification. This is based on experience/judgment.
VED classification when coupled with ABC classification enhances the inventory control efficiency.

3.9 Self Assessment Questions

Q1. What is inventory and their Objectives?
Q2. Discuss different types of inventory.
Q3. What are the major functions of inventory in an organization?
Q4. Explain different factors affect the inventory?
Q5. Some business peoples think that inventory as necessary evil while others think inventory as an
asset. What is your view?

3.10 Further References

Hamdy A Taha, 1999. Introduction to Operations Research, PHI Limited, New Delhi.



145

MBA-H2040 Quantitative Techniques for Managers
Mustafi, C.K. 1988. Operations Research Methods and Practices, Wiley Eastern Limited, New Delhi.
Levin, R and Kirkpatrick, C.A. 1978. Quantitative Approached to Management, Tata McGraw Hill,
Kogakusha Ltd., International Student Editon.
Levin and Rubin. Quantitative Techniques for Managers, PHI, New Delhi.

UNIT II










4 INVENTORY MODELS

ON
S



S
E



L





LESSON STRUCTURE




4.1 Introduction



4.2 Deterministic Inventory Model



4.3 Deterministic Single Item Inventory Model



4.3.1 Economic Order Quantity Model I



4.3.2 Economic Order Quantity Model II



4.3.3 Economic Production Quantity Model



4.3.4 Price Discounts Model



4.3.5 Dynamic Demand Models



4.4 Deterministic Multi Items Inventory Model



4.4.1 Unknown Cost Structure Model



4.4.2 Known Cost Structure Model

Objectives

4.5 Probabilistic Inventory Models

After Studying this lesson, you should be able

4.5.1 Single Period Probabilistic Model

to:

4.5.2 Single Period Discrete Probabilistic

Understand Simple Deterministic

Demand Model

Inventory Models

4.6 Summary

Understand Simple Probabilistic

4.7 Key Terms

Inventory Models

4.8 Self Assessment Questions

Develop Simple Deterministic Inventory

4.9 Further References

Models



Illustrate the use of Simple Inventory



Models in Practical Situations



Briefly explain Single Item Inventory



Models



Briefly explain Multi Item Inventory



Models



Describe static and dynamic Inventory



Models



Develop Single Period Probabilistic



Models

















146
MBA-H2040 Quantitative Techniques
for Managers

4.1 Introduction



The methodology for inventory situation modeling is based on four concepts, they are:



1. Examine the inventory situation, list characteristics and assumption related to the

inventory situation.

2. Develop the total annual relevant cost equation in narrative form as:


Total annual



Procurement

Carrying cost

Stock out costs



r e l

e

v

a

n t

c

o

s

t = I

t

e

m

c

o

s t +

c

o s t +

(

c

y

c

l

e

s t

o

c

k

s +

(cost/sales back



+ safety

order)



stocks)


3. Convert the total annual relevant cost equation from narrative form into the shorthand

logic of mathematics.

4. Optimize the cost equation by finding the optimum for how much to order (also called

order quantity), when to re-order (also called re-order point) and the total annual
relevant cost.


In general, the situation of inventory can be classified into tow types viz. deterministic and

stochastic.


Deterministic- in this variables are known with certainty
Stochastic ? in this variables are probabilistic

This lesson briefly outlines Deterministic Inventory Models and Probabilistic (Discrete
Demand Distribution Model) Inventory Models.

In this section we will discuss deterministic inventory models and later we will discuss
probabilistic inventory models (section 4.5).


4.2 Deterministic Inventory Models




There are different deterministic inventory models, they are:


a. Deterministic single item Inventory Models

i.

EOQ ? Economic Order Quantity Model ? I

ii.

EOQ ? Economic Order Quantity Model ? II (instantaneous supply
when shortages are allowed)

iii.

EPQ ? Economic Production Model (Gradual supply case and shortage
not allowed)

iv.

Price Discounts Model (instantaneous supply with no shortages)

v.

Dynamic Demand Models



b. Deterministic multi-item Inventory Models

i.

Unknown cost structure Model

ii.

Known cost structure Model






147

MBA-H2040 Quantitative Techniques
for Managers



4.3 Deterministic Single Item inventory Models

Inventory models with all the known parameters with certainty are known as deterministic

inventory model. In this section we will discuss the deterministic inventory models for sing

item.

4.3.1 Economic Order Quantity (EOQ) Model I

The EOQ concept applies to the items which are replenished periodically into inventory in

lots covering several periods' needs, subject to the following conditions:



Consumption of item or sales or usage is uniform and continuous

The item is replenished in lots or batches, either by manufacturing or by purchasing


Description




The EOQ model is described under the following situations:



a. Demand is deterministic and it is denoted by D units per year.
b. Price per Unit or cost of purchase is C.
c. Planning period is one year.
d. Ordering Cost or Procurement Cost or Replenishment Cost is Co. Suppose if the items

are manufactured it is known as set up cost.

e. Holding Cost (or carrying cost) is Ch per unit of item per one year time period. The

Ch is expressed either in terms of cost per unit per period or in terms of percentage
charge of the purchase price.

f. Shortage Cost (mostly it is back order cost) is Cs per unit per year.
g. Order Size is Q.
h. Cycle period of replenishment is t.
i. Delivery lad/Lead Time is L (expressed in units of time)

In this section we will discuss about instantaneous supply when shortages are not allowed.

That is whatever is demanded, is supplied immediately after the lead time. If we assume these

constraints, a graph of inventory against time will be look like a regular saw-tooth pattern as

given below (Fig: 4.1).



148
MBA-H2040 Quantitative Techniques
for Managers



Inventory




Inventory











Reorder Point







Q
Q/2














L





L

Time

Reorder Level
t t t



Fig: 4.1. Saw-Tooth Inventory Model



- In this model we assumed that the shortages are not allowed, it means that shortage

cost is prohibitive or Cs is too much large or infinite.

- Everything is so known and regular, there is no need of safety stock.
- Inventory will run out altogether just as the next lot is received.


The different levels of inventories for this model are fixed as follows:

- Minimum level = Safety Stock (Buffer Stock)
- Maximum level = Minimum level + EOQ
- Reorder level = Minimum level + Lead Time Consumption



-- In this case safety stock is not needed, so that safety stock is zero i.e. Minimum
level = 0.

-- Maximum Inventory is the ORDER SIZE (lot size).

-- Maximum Inventory is the ORDER SIZE (lot size).
Therefore, the average inventory per cycle = ?(Maximum level + Minimum level).

Here cycle is the intermittent pattern, in which inventory vary from maximum
to minimum and then back to maximum.




-- Maximum inventory is Q



Therefore, the average inventory per cycle = ?(Q+O) =Q/2, and the average

inventory is time

independent.






149

MBA-H2040 Quantitative Techniques
for Managers



In this case, the Total Annual Relevant Cost is as follows:



Total annual

Annual Purchase









=







+ Annual Ordering

relevant cost

Cost (PC) +



Cost (OC)

(TC)

Annual Carrying



Cost (CC)




Quantity Ordering Number of Carrying Cost


Average

= (Price/Unit) Purchased + Cost/Order Orders placed / + per unit
number of




Per Year







year







units carried























-------------- eq.1

Note that,
Number of Orders/Year = Annual Demand/Order Size
















= D/Q

Thus, the eq.1 is written as:





TC = CD + Co D/Q + Ch Q/2 ------------------- eq.2






The EOQ or Order Size is that quantity, which minimizes the Total Cost. Total Cost is the

sum of Fixed Cost and Variable Cost. The Fixed Cost (CD) is independent of Order Size

while the variable Cost is dependent on the Order Size (Q). Since, the fixed cost does not play

any in minimization or maximization process, only variable cost will be minimized here.




For the total cost to be minimum, the first order derivative of TC is zero, that is,




dTC/dQ = -CoD/Q2 +Ch/2 = 0 ----- eq.3 or

CoD/Q = ChQ/2 --------- eq.4










or



Annual Ordering Cost = Annual Carrying Cost ----------- eq.5


The eq.5 may also be obtained from the following fig: 4.2




150
MBA-H2040 Quantitative Techniques
for Managers





TOTAL COST
Total Cost Curve



Annual Carrying
Cost Curve






TC



































Annual Ordering Cost

Curve




















Annual Purchase Cost

Curve








Q*





ORDER SIZE = Q











Fig: 4.2. Inventory Control Cost Trade-offs




Now, if we examine the eq.2 that is the total cost equation, we obtain the relationship

between the fixed cost and variable cost. This relationship is shown in the above fig.4.2. Note

that the total cost curve has the lowest value just above the intersection of the ordering cost

curve and carrying cost curve, and also at the intersection annual cost is equal to the annual

carrying cost.


From the eq.4, now we will get

EOQ = Q* =

2CoD ----------- eq.6

Ch
The cycle period

t = Optimal Order Quantity or t* = Q* =

2Co ----------

eq.7
Annual Demand D ChD

N = Total number of orders per year , which is the reciprocal of cycle period (1/t*)

That is




N = D =

ChD ------------------ eq.8



151

MBA-H2040 Quantitative Techniques
for Managers

Q 2Co

The annual cost = TC = CD +

2CoChD ----------- eq.9


Lead Time Consumption = (Lead time in years) * (Demand Rate per year)


Minimum Level = O



Maximum Level = Q*



Reorder Level = LD




Let us see few example of this case.


Example 4.1

A manufacturer uses Rs.20, 000 worth of an item during the year. Manufacturer estimated the
ordering cost as Rs.50 per order and holding costs as 12.5% of average inventory value. Find
the optimal order size, number of orders per year, time period per order and total cost.

Solution




Given that:

D = Rs.20, 000
Co = Rs.50
Ch = 12.5% of average inventory value / unit


Total Cost = TC = 25D + (0.125) Q , where Q is order size in Rs.


Q D


By applying the equations (eq.6) to (eq.9), we will get Q*, t*, N

?




Q* =

2CoD





Ch




=

2 * 50 * 20000



= Rs.4000





0.125




t* =

2Co





ChD




=

2 * 50 = 1 years = 73 days





(0.125) * (20000) 5




N = 1 = 5

t*

Note: TC means in this case variable cost only





152
MBA-H2040 Quantitative Techniques
for Managers



TC* =

2 * 50 * 0.125 * 20000

= Rs.500


Therefore






Order Size = Q = Rs.4000
Number of order / year = N = 5



Time period / order = t* = 73 days



Total Cost = TC* = Rs.500

Example 4.2



A manufacturer uses an item at a uniform rate of 25,000 units per year. Assume that no
shortage is allowed and delivery is at an infinite rate. The ordering, receiving and hauling cost
is Rs.23 per order, while inspection cost is Rs.22 per order. Interest costs is Rs.0.056 and
deterioration and obsolescence cost is Rs.0.004 respectively per year for each item actually
held in inventory plus Rs.0.02 per year per unit based on the maximum number of units in
inventory.



Determine the EOQ. If lead time is 40 days, find reorder level.


Solution




Given that



Demand = D = 25000 units/year



Ordering Cost = Co = 23 + 22 = Rs.45 per order



Storage cost Ch = 0.056 + 0.004 = Rs0.060 (based on actual inventory (=average

inventory)


Storage cost Ch = Rs.0.02 per unit/year

(based on maximum inventory)




Total Variable Cost = TC = 25 * 25000 + 0.060 * Q + 0.02 * Q











Q

2











= 625000 + 0.1*Q







Q 2



Thus,





Q* =

2CoD

= Q* =

2*25*25000 =3535.5 units (3535 units

approximately)






Ch



0.1






Reorder level = L*D = 40*25000 = 2739.7 units











365










That is = 2740 units





Therefore





EOQ = Q* = 3535 units







Reorder level = 2740 units





153

MBA-H2040 Quantitative Techniques
for Managers



Look in to the fig.4.2; in this the total cost curve is almost flat near the minimum cost

point. This indicates that small variations in optimal order size will not change the total cost

appreciably. For this purpose we will examine the model of sensitivity in the next section.

Model Sensitivity

In order to examine sensitivity of the model, we compare the sensitivity of the total costs (TC)

for any operating system with the total variable costs for an optimal inventory system (TC*)

by using the ratio TC/TC*. To do this, we have to calculate TC/TC* as a function of Q/Q*.



154
MBA-H2040 Quantitative Techniques
for Managers

Therefore






CoD + ChQ CoD + ChQ*

TC = Q 2 Q* 2 ---------- eq.10





Now by substituting Q* =

2CoD/Ch into equation eq.10 and solving algebraically we get

the relationship as


TC = 1 Q* + Q ---------------------- eq.11
TC* 2 Q Q* , which is shown in the following figure fig.4.3






TC/TC*


2.0


1.5


1.0


0.5




0.5 1.0 1.5 2.0 2.5 3.0 Q/Q*






Fig: 4.3 Inventory Sensitivity (in case of simple lot size)




According to this if Q is off from optimal either direction by a factor of 2; costs are

increased by only 25%. This has an important practical implication.



The model sensitivity is explained with the help of the following example, so that we

will understand more.

Example 4.3

Consider the Example 4.1. The sensitivity of total cost if order size is Rs.4000, then we will
get that


TC/TC* = 1 4000 + 4000 = 1

2 4000 4000




155

MBA-H2040 Quantitative Techniques
for Managers

This indicates that even though order quantity deviates from optimal by Rs.4000 or 100%, the

costs are only 25% higher than the optimal. This excess cost of the non-optimal order quantity

can be found as:

Excess Cost (Marginal Cost) = 0.25(TC*) = 0.25(500) = Rs.125.

4.3.2 The EOQ Model II

Here we are going to discuss, Instantaneous Supply When shortages are allowed. In this case,

stock outs are permitted which means that shortage cost is finite or it is not more. The entire

Model I assumptions (a to i) are also good applicable here. The Inventory situation with

shortages is represented diagrammatically in the following figure fig.4.4.



Look the figure there are two triangles viz. ABC and CEF.

The triangle ABC denotes inventory, whereas

The triangle CEF denotes the shortage
I = inventory level
S = shortage level
Q = order size = I + S
Cycle period = t = t1 + t2


Where t1 is the proportion of cycle period for inventory holding





t2 is the time of stock out










Inventory Reorder Point










A



I




Q t2 t2







B t1 C

F t1 S






S











Shortage








D E










t t








Fig: 4.4 Inventory Situation with Shortages



Total Variable Cost = Annual Ordering Cost + Annual Holding Cost + Annual

Shortage Cost



= CoD + I2Ch + (Q-I)2Cs ---------------- eq.12



156
MBA-H2040 Quantitative Techniques
for Managers









Q 2Q 2Q



From this we will get















EOQ = Q* =

( 2CoD ) ( Co+Ch) -------- eq.13









Ch Cs





Inventory Level = I* =

( 2CoD ) ( Cs ) -------- eq.14









Ch Co+Ch




Shortage Level = Q* - I* ------------------ eq.15




Cycle Period = t* = Q* ----------------- eq.16







D




Number of Orders/Year = 1









t*




Therefore





Total Variable Cost = TC* =

2CoChCsD













(Ch + Cs) -------------- eq.18




Thus, if we compare the total variable cost of Model I and Model II we will see that







2CoChD >

2CoChCsD ---- This implies that the annual











(Ch + Cs)

Cost when shortage is allowed is less than the annual inventory management cost when
shortages are not allowed. That is shortage should be allowed whenever the shortage cost is
not very large for reducing the total cost.

Example 4.4



Consider the following problem.

Problem

The demand for an inventory item each costing Re5, is 20000 units per year. The ordering

cost is Rs.10. The inventory carrying cost is 30% based on the average inventory per year.

Stock out cost is Rs.5 per unit of shortage incurred. Find out various parameters.


Solution



Given that





Demand = D = 20000





Ordering Cost = Co = Rs.10





Carrying Cost = Ch = 30% of Re 5 = 30 * 5 = 1.5















100



157

MBA-H2040 Quantitative Techniques
for Managers





Stock out Cost = Cs = Rs.5


Now we have to determine the various parameter of EOQ Model II such as EOQ, Inventory
Level, Shortage Level, Cycle Period, number of orders/year and Total Cost.





EOQ = Q* =

(2CoD ) ( Co+Ch)



Ch Cs







=

2*10*20000 0.30 + 10 = 1657 units









0.30 5







Inventory Level = I* =

( 2CoD ) (Cs )



















Ch Co+Ch





Inventory Level = I* =

2*10*20000 5 = 804 units









0.30 10+0.30




Shortage Level = Q* - I* = 1657 ? 804 = 853 units




Cycle Period = t* = Q* = 1657 = 30.24 days = 30 days







D 20000




Number of Orders/Year = 1 = 1 = D = 20000 = 12 Orders/year









t* Q*/D Q* 1657




Total Cost =

2CoChCsD







(Ch + Cs)







=

2*10*0.30*5*20000 = RS.336.4







0.30+5



4.3.3 Economic Production Quantity (EPQ) Model

Here we will discuss about Graduate Supply case when Shortages are not allowed. EOQ

model is more common in retail situation, while economic production quantity EPQ is

basically associated with manufacturing environment. EPQ shows that over a period of time

inventory gradually built and the consumption go side by side where production rate is higher

than that of consumption rate.



158
MBA-H2040 Quantitative Techniques
for Managers



Assumption (a) to (i) of EOQ Model I also hold good for this model. In this model the

Order Size (Q) is taken as Production Size, the annual production rate is taken as P such that P

> D, otherwise, if P D, the item will be used as fast as it is produced. This situation is

illustrated in the following figure fig: 4.5.



159

MBA-H2040 Quantitative Techniques
for Managers
















Maximum Inventory Level














B


Rate of












Decrease of













Inventory=D






Rate of increase of

Inventory = P-D












t1 t2 t1 t2


A D C Time

t t










Fig: 4.5 Gradual Replacement Inventory Situation




In the above figure, t = Cycle Time









t1 = Production Time









t2 = Depletion Time and



t = t1 + t2 of maximum inventory level BD.



Production Time = t1 = Q

P


Cycle Time = t = Q







D



Maximum Inventory Level = BD = (P ? D) * t1











= (P ? D) * Q

















P



Minimum Inventory Level = 0




Average Inventory Carried = (P ? D) Q + O













P



(P ? D) Q











=

2

2P





160
MBA-H2040 Quantitative Techniques
for Managers

Total Variable Cost/Year = Annual Setup Cost + Annual Carrying Cost









CoD (P ? D) Q









=

+ Ch



----- eq.19











Q P 2



Thus,





EPQ = Q* =

2CoD

------------------------ eq.20











Ch(P-D)









P







Total Variable Cost = TC* =

2CoCh(P-D) D --------- eq.21

















P



The economic production quantity model (gradual supply case and shortage not

allowed) is explained with the help of following Example 4.5.

Example 4.5

Problem

An inventory item unit is used at the rate of 200/day, and can be manufactured at a rate of

700/day. It costs Rs.3000 to set up the manufacturing process and Rs.0.2 per unit per day held

in inventory based on the actual inventory any time. Assume that shortage is not allowed.



Find out the minimum cost and the optimum number of units per manufacturing run.


Solution


Given that



Demand, D = 200 units



Production, P = 700 units



Set up Cost, Co = Rs.3000



Holding Cost, Ch = Rs.0.2




? Minimum Cost = TC* =

2CoCh(P-D) D



























P











=

2*3000*0.2(700-200) (200)

= Rs.414











700






EPQ = Q* =

2CoD =

2CoD * P





161

MBA-H2040 Quantitative Techniques
for Managers







Ch(P-D) Ch(P-D)







P



















=

2*3000*200*700

= 2898.2 = 2898 units







0.2(700-200)





Minimum Cost = Rs.414



Therefore





Number of units per Manufacturing Run = 2898 units




4.3.4 Price Discounts Model

In this section we will discuss, instantaneous supply with no shortages. We know very well

that whenever we make bulk purchasing of items there may be some discount in price is

usually offered by the suppliers. As far as discount concern, there are two types:

1. Incremental Discount ? discount allowed only for items which are in excess of the

specified amount. In this case, all the prices offered in different slabs are applicable in
finding the total cost.

2. All units Discount ? discount allowed fro all the items purchased. In this, only one

price at any one slab is applicable for finding the total cost.



Here we are going to discuss only all units' discount type.
Advantages of Bulk Purchase



Buying in bulk may results in the following advantages:


less unit price
less ordering cost
cheaper transportation

fewer stock outs

sellers preferential treatment

Disadvantages of Bulk Purchase



Bulk purchase also has the following disadvantages in addition to the above

advantages:

high carrying cost
lower stock turnover
huge capital required

less flexibility
older stocks

heavy deterioration

heavy depreciation



162
MBA-H2040 Quantitative Techniques
for Managers


In case of purchased items, if the discounts are allowed, the price C may vary according to the
following

pattern:



C = P0 if purchase quantity Q = Q0 < q1


= P1 if purchase quantity q1 Q1 < q2

..............................................



...............................................




= Pn if purchase quantity qn < Qn



where Pj-1 is > Pj, for j = 1, 2, ........, n.



-------- eq.22

Pj = Price per unit for the jth lot size


Suppose, shortages are not permitted, the total cost per year is obtained by the following set of
relations:


TC(Q

Q

j) = DPj + CoD + 1 ChQj

, where qj

j<qj+1



----------- eq.23



and Ch=ipj, i being percentage change for j=0 to n. Since price (or unit cost) varies with
purchase size (Q), the fixed cost term CD in equation eq.23 can't be omitted for minimizing
the total cost (TC).


Equation eq.23 for quantity discounts are represented in the figure fig: 4.6.








Total
Cost
(TC)


















163

MBA-H2040 Quantitative Techniques
for Managers




O q1 q2

q3

q4 Quantity (Q)










Fig: 4.6 Quantity Discount




The heavy curve on the various price discounts shows feasible portion of the total cost

which is a step function. Therefore, for determining the overall optimum, the following steps

of procedure is adopted:


Step 1: Find EOQ for the lowest price




That is compute Q*n =

2CoD









iPn




If Q* q

n

n, the optimum order quantity is Q*n.




If Q*n < qn, then go to Step 2.



Step 2: Compute Q*n -1 =

2CoD for the next lowest price.







iPn-1





q



If Q*n -1

n-1, then compare the total cost TCn-1 for purchasing Q*n -1 with the total

cost TCn for purchasing quantity qn and select least cost purchase quantity.


If Q*n -1 < qn-1 then go to Step 3.

Step 3: Compute Q*n -2 =

2CoD for the next lowest price.







iPn-2




If Q*n -2 qn-2, then compare the total cost TCn-2, TCn-1 and TCn for purchase

quantities Q*n -2 , qn-1 and qn respectively and select the optimum purchase quantity.




If Q*n -1 < qn-2 then go to Step 4.


Step 4: Continue the procedure until Q*

q

n -j

n-j. Then compare the total costs TCn-j with TCn-

j+1, .... , TCn-1, TCn for purchase quantities Q*n -j , qn-j+1, .... , qn-1, qn respectively, and select
the optimum purchase quantity.



The price discount model is explained with the help of the following examples.


Example 4.6


Consider the following problem, which explains the price discount model.

Problem




Suppose, annual demand for an item is 1500 units, ordering cost is Rs.250, inventory carrying
charge is 12% of the purchase price per year and the purchase prices are:



164
MBA-H2040 Quantitative Techniques
for Managers






P1 = Rs.5

for purchasing Q1 < 250





P2 = Rs.4.25 for purchasing 250 Q2 < 500





P3 = Rs.3.75 for purchasing 500 Q3




Find out the optimum purchase quantity.


Solution


Given that



Demand, D = 1500 units



Ordering Cost, Co = Rs.250



Carrying Cost, Ch = 12% = i



This problem belongs to the Step 1. So that as per Step 1







Q*3 =

2CoD =

2*250*1500 = 1290 units





iPn

(0.12)(3.75)


Note that 1290 > 500, Optimum Purchase Quantity = 1290 units.

Example 4.7

Problem
Consider the problem in Example 4.6 with ordering cost of Rs.10 only. Find out the optimum
purchase quantity.

Solution

Given that


Demand, D = 1500 units



Ordering Cost, Co = Rs.10



Carrying Cost, Ch = 12% = i


In this case,




Q*3 =

2CoD =

2*10*1500

= 258 units





iPn

(0.12)(3.75)





Since 258 < 500 = q3, so that we may have to compute






Q*2 =

2CoD =

2*10*1500 = 242 units





iPn

(0.12)(4.25)





Since 242 < 250 = q2, next we may have to compute



165

MBA-H2040 Quantitative Techniques
for Managers



Q*1 =

2CoD =

2*10*1500 = 224 units





iPn

(0.12)(5)






Since 224 > 0, now we have to compare the total cost for purchasing Q*1 = 224, q2 =

250 and q3 = 500 units respectively.

From equation eq.23

TC (Qj) = DPj + CoD + 1 ChQj

, where qjQj<qj+1


We will get

TC1 (for purchasing 224) = 5*1500 + 10*1500 + 1 (0.12) (5)224 = 7634.2










224 2


TC2 (for Q2=250) = 4.25*1500 + 10*1500 + 1 (0.12) (4.25)250 = 6498.75










250 2


TC3 (for Q3=500) = 3.75*1500 + 10*1500 + 1 (0.12) (3.75)500 = 5767.5








500 2



The EPQ for his problem is Q*3 = 500 units


4.3.5 Dynamic Demand Models

In this model, assume that demand is known with certainty, and although may vary from one

period to the next period. There are five types of dynamic demand inventory models, they are:



i)

Production Inventory Model (Incremental Cost Method)

ii)

Dynamic Inventory Model (Prescribed Rule Method)

iii)

Dynamic Inventory Model (Fixed EOQ Method)



The above five dynamic demand models of inventory are discussed in details in the following
subsequent sections.

i) Production Inventory Model

This is also called as Incremental Cost Method. This situation is explained with the help of
the following example 4.8.

Example 4.8


Consider the following problem.


Problem

A production factory has a fixed weekly cyclic demand as follows:










166
MBA-H2040 Quantitative Techniques
for Managers

Days

Mon

Tue

Wed

Thu

Fri

Sat

Sun

Demand

9

17

2

0

19

9

14

in units


Here the policy is to maintain constant daily production seven days a week. Shortage cost is
Rs.4 per unit per day and storage cost depends upon the size of Q, the quantity carried, as
follows:


Cost per unit for one

=1

=4

=10

day(Rs.)

Q

1Q3

4Q20

20Q


And the chargers are based on the situation at the end of the day.

Determine the optimal starting stock level.

Solution


Let the production rate be the average of the total sales which is






= 9 + 17 + 2 + 0 + 19 + 9 + 14 = 10 units/days



Day

On hand

Demand

Inventory

Shortage

Carrying



stock start

cost (Rs.)

cost (Rs.)



of day



Mon

8

9

-1

4

0

7

Tue

-1+10=9

17

-8

32

0



Wed

-8+10+=2

2

0

0

0

Next

Thu

0+10=10

0

10

0

44

we will

Fri

10+10=20

19

1

0

1

find the

Sat

1+10=11

9

2

0

2

total

Sun

2+10=12

14

-2

8

0

weekly

Mon

-2+10=8

Total Cost 44 + 43 = 87

cost for
different starting stocks, which is illustrated in the following tables.






Table 4.1 shows for starting stock 8, Table 4.2 shows for starting stock 9 and Table

4.3 shows for starting stock 10.
















167

MBA-H2040 Quantitative Techniques

Day

On hand

Demand

Inventory

Shortage

Carrying

for Managers

stock start

cost (Rs.)

cost (Rs.)

of day



Mon

9

9

0

0

0



Tue

0+10=10

17

-7

28

0



Wed

-7+10+=3

2

1

0

1

Table

Thu

1+10=11

0

11

0

44

4.1

Fri

11+10=21

19

2

0

2

Cost

Sat

2+10=12

9

3

0

3

Analysi

Sun

3+10=13

14

-1

4

0

s with

Mon

-1+10=9

Total Cost 32 + 50 = 82

Initial

Stock 8





Day

On hand

Demand

Inventory

Shortage

Carrying



stock start

cost (Rs.)

cost (Rs.)



of day



Mon

10

9

1

0

1



Tue

1+10=11

17

-6

24

0



Wed

-6+10+=4

2

2

0

2



Thu

2+10=12

0

12

0

48



Fri

12+10=22

19

3

0

3



Sat

3+10=13

9

4

0

16



Sun

4+10=14

14

0

0

0



Mon

0+10=10

Total Cost 24 + 70 = 94





Table 4.2 Cost Analysis with Initial Stock 9


























Table 4.3 Cost Analysis with Initial Stock 10




Therefore






The optimal solution for starting stock of 9 units is on MONDAY.






Minimum Total Cost for this is Rs.82.





168
MBA-H2040 Quantitative Techniques
for Managers



Procedure for solving such incremental problem is cost analysis, which is self

explanatory.

ii) Dynamic Inventory Model (Prescribed Rule Method)

Some times, the organization dealing with inventory may prescribe some rule of procurement
of items of inventory. For example


- Procuring every month
- Procuring every three months



The prescribed rule method is explained with the help of the following Example 4.9.

Example 4.9
Consider the following problem.




169

MBA-H2040 Quantitative Techniques
for Managers

Problem

An organization estimates the demand of an item as follows:

Month

Jan

Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total

Demand 140 98

62

134 20

72

22

164

139 170 248

51

1320


The organization has decided that their ordering cost is Rs.54 and carrying charge per unit per

month is 2% at the end of each month. The cost of item per unit is Rs.20. Assume that the

supply is instantaneous, there in no stock outs and no lead time. Only full month requirement

is ordered.

Solution

The following table is used for determining the total cost for the policy of quarterly ordering.


Month

Jan

Feb Mar Apr May Jun

Jul

Aug Sep Oct

Nov Dec Total

Starting

0

160 62

0

92

72

0

303 139 0

299 51

--

Inventory





























300 --

--

226 --

--

325 --

--

469 --

--

1320

Replenishment



























140 98

62

134 20

72

22

164 139 170 248 51

1320

Requirements





























160 62

0

92

72

0

303 139 0

299 51

0

1178

Ending
Inventory







Table 4.4 Total Cost Analysis for Quarterly Policy


Therefore




Total Carrying Cost = 1178*(0.02)*(20) = Rs.471.2





Total Replenishment Cost = 4*54 = Rs.216





Total Annual Cost = 216 + 471.2 = Rs.687.2


iii) Dynamic Inventory Model (Fixed EOQ Method)




The Fixed EOQ Method is explained with the help of the following Example 4.10.

Example 4.10


Consider the problem of Example 4.9.


Problem


An organization estimates the demand of an item as follows:


Month

Jan

Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Total

Demand 140 98

62

134 20

72

22

164

139 170 248

51

1320





170
MBA-H2040 Quantitative Techniques
for Managers

The organization has decided that their ordering cost is Rs.54 and carrying charge per unit per

month is 2% at the end of each month. The cost of item per unit is Rs.20. Assume that the

supply is instantaneous, there in no stock outs and no lead time. Only full month requirement

is ordered.


Solution



Average Monthly Demand = Demand = 1320 = 110 units/month











Month 12







EOQ =

2*54*110 = 172 (approximately)









(0.02)*(20)

In this case the organization has to ordered full month requirement, therefore 172 lies between

140 and 248 units, but 172 is more closer to 140 than 248, so that the organization order one

month requirement at the beginning of the January.



Similarly, at the beginning of the February, the organization order two month

requirement.



The detailed result is illustrated in the following Table 4.5



Month

Jan

Feb Mar Apr May Jun

Jul

Aug Sep Oct

Nov Dec Total

Starting

0

0

62

0

92

72

0

164 0

0

0

0

--

Inventory





























140 160 --

226 --

--

186 --

139 170 248 51

1320

Replenishment





























140 98

62

134 20

72

22

164 139 170 248 51

1320

Requirements




























Ending

0

62

0

92

72

0

164 0

0

0

0

0

390

Inventory







Table 4.5 EOQ Total Cost Analysis




Therefore






Total Ordering Cost = 8*54 = Rs.432





Total Carrying Cost = 390*0.4 = Rs.156




Thus, the Total Cost is reduced if EOQ policy is used instead of three month

(quarterly) rule.

4.4 Deterministic Multi Item Inventory Models

When there is more than one item in the inventory is called as multi item inventory. Since this

contains more items, the inventory control requires special type of care. This type of



171

MBA-H2040 Quantitative Techniques
for Managers

inventory problems may different types of constraints like capital, cost structure, storage

space, purchasing load etc. As the number of constraints increase the problem becomes more

complex. In this section we will discuss some examples of this type of inventory.



There are two types of multi item inventory model which is based on the structure of

the cost, they are:

1. Model with Unknown Cost Structure
2. Model with Known Cost Structure


4.4.1 Unknown Cost Structure Model




In India most of the organizations do not maintain the proper inventory records which may

provide sufficient cost information to generate the two basic parameters viz. procurement cost

and carrying cost of inventory control. Some organizations in some situations have not

developed cost structure related to inventory control, but still they wish to minimize total cost

of inventory. There may be critical situations, in which an organization may need to take

immediate actions to improve the situation without considering the structure of cost.



The use of inventory models without cost information is impossible, but we will show

here that it is possible to get many of benefits of inventory techniques even when carrying

cost and ordering cost are not known. In such a problem, there are two different approaches,

they are:



1. Minimize the Total Carrying Cost while keeping the number of orders/year fixed

or

2. Minimize the Total Number of orders/year while keeping the same level of

inventory.



Minimize the Total Carrying Cost while keeping the number of orders/year fixed Model



As per the EOQ model




Q* = 2CoD

or

Q* = 2CoD * D





Ch





Ch


Suppose = 2Co = Constant, because ordering cost and carrying cost are deterministic
values.




Ch


Therefore

Q* = * D ------------ eq.24





The equation eq.24 tells that EOQ is proportional to the square root of demand for any

inventory item of control.




172
MBA-H2040 Quantitative Techniques
for Managers



For this equation, we get





= Q* = Q*D = D = Square root of Demand ----------- eq.25

D D (D/Q*) Number of orders/year



Since is constant for any single item of inventory, we take as the constant for all

the inventory items. Thus, we take


= D



or = Sum of square roots of Demand of each inventory item -

eq.26


(D/Q*) Sum of the number of orders/year for each item





This model is explained with the help of the following Example 4.11.

Example 4.11
Consider the following problem.

Problem

An organization has the following procurement pattern of six items irrespective of their
demand level. Reduce the inventory levels while keeping total number of orders/year fixed.


Item No.

Number of

Demand

Order Size

Average

Orders/Year (Rs.)

(Rs.)

Inventory

1

5

2000,000

500000

250000

2

5

700,000

260000

130000

3

5

100,000

23500

11750

4

5

9,000

10000

5000

5

5

5,000

700

350

6

5

2,700

500

250

Solution



First we find the value of = D

---------- eq.27









(D/Q*)




Therefore

D = 1414 + 836 + 316 + 95 + 70 + 52 = Rs.2783








(D/Q*) = 5 + 5 + 5 + 5 + 5 + 5 = 30




Thus, = D

= 2783 = 92.7





(D/Q*) 30



Now we will analyze the ordering quantity, which is illustrated in the following Table

4.6.






Demand

D



EOQ=

Number of Average

Item No.

(D) (Rs.)

D

orders/year Inventory =
=D/EOQ

EOQ/
Number of
orders/year

1

2000,000

1414 92.7

131077.8

15.258

4369

2

700,000

836

92.7

77497.2

9.0

2583

3

100,000

316

92.7

29293.2

3.41

976

4

9,000

95

92.7

8806.5

1.0

293



173

MBA-H2040 Quantitative Techniques
for Managers

5

5,000

70

92.7

6489

0.77

216

6

2,700

52

92.7

4820.4

0.56

160



29.998

8597










Table 4.6 Analysis of Ordering Quantity



Thus, according to the policy of the organization ordering for six items a year each

item, Total Average Inventory becomes Rs.397350.



But, as per the new schedule as obtained in the above Table , the Average Inventory is

Rs.8597 which is much less and at the same time total Number of Orders remains same.



Therefore, substantial savings can still be achieved when cost information is known.


Minimize the Total Number of orders/year (or purchasing workload) while keeping the

same level of inventory Model




We will know that




Q* = D or

= Q*









D




Now onwards, we will take = Q* ------- eq.28 for all inventory items.











D




This model is explained with the help of the previous Example 4.11.




Here we will discuss about how to minimize the number of orders or purchasing

workload.

Problem

An organization has the following procurement pattern of six items irrespective of their

demand level. Reduce the inventory levels while keeping total number of orders/year fixed.



Item No.

Number of

Demand

Order Size

Average

Orders/Year (Rs.)

(Rs.)

Inventory

1

5

2000,000

500000

250000

2

5

700,000

260000

130000

3

5

100,000

23500

11750

4

5

9,000

10000

5000

5

5

5,000

700

350

6

5

2,700

500

250


Solution



Here K= Sum of the order size





Sum of square root of Demand



174
MBA-H2040 Quantitative Techniques
for Managers


K= 50000+260000+23500+10000+700+500 = 794700 = 285.555
1414+836+316+95+70+52 2783








175

MBA-H2040 Quantitative Techniques
for Managers



Item

Demand

D



EOQ= D

Number of Average

No.

(D) (Rs.)

orders/year Inventory =
=D/EOQ

EOQ/2

1

2000,000

1414 285.555 403774.77

4.95

201887.385

2

700,000

836

285.555 238723.98

2.93

119361.99

3

100,000

316

285.555 90235.38

1.10

45117.69

4

9,000

95

285.555 27127.725

0.33

13563.8625

5

5,000

70

285.555 19988.85

0.25

9994.425

6

2,700

52

285.555 14848.86

0.18

7424.43





9.74

397349.8

EOQ=794700








Table 4.7 Analysis for Reducing Number of Orders


Thus, from the Table 4.7, it is obvious that the total purchasing workload has reduced by

35%. Therefore, there is a definite saving of cost by applying these methods for multiple

items even when cost information is not known.


4.4.2 Known Cost Structure Model

We may classify the models with known cost structures into two main types, they are:


Model without Limitations

Model with Limitations


Here we will discuss these models, if we know the complete cost structure. In this model let

us consider the following symbolic notation:





Dj

-

Demand





Coj

-

Ordering Cost





Chj

-

Holding Cost or Carrying Cost for jth item respectively.



Model without Limitations



In any situation the items may be purchased according to their individual economic order

quantities, if there are no restrictions for the items being storing.



In this case,




Total Variable Cost per annum for n-items can be represented as









n







TC = CojDj + ChjQj and -------------------eq.29







J=1 Qj

2





176
MBA-H2040 Quantitative Techniques
for Managers



The optimum Order Size for each item is







Qj* =

2CojDj

for j = 1,........., n ---------------- eq.30





Chj


Model with Limitations

We know that the different constraints on inventories are available capital, order size per year,

storage space, etc. In this case we will consider a single constraint for the discussion of this

model. Note that the constraints capital and storage space are interchangeable.


Let




Qj is the order quantity for item j





Fj is the storage space (or floor space) for one unit of item j







(or) Fj ? the capital requirement for one unit of item j, since the storage

space and

the capital are interchangeable.





F is the total available storage space (floor space) or available capital


Then the constraints are as follows:


n



F1Q1 + F2Q2 + ............ FnQn F -------------eq.31



j=1


Here the objective is to minimize the total inventory cost expressed by the equation eq.29. So

that in order to solve this problem first we have to convert the constraint problem into

unconstrained minimization problem and then the optimum result is obtained.


The problem is to minimize the function known as Lagrange Function:













n n



L(, Q

) =

1, Q2, ..., Qn

CojDj + ChjQj + ( FjQj - F) ---------eq.32









j=1 Qj 2 j=1





Then, the values of optimal sizes are








Qj* =

2CojDj

for j = 1,........., n ------------eq.33







(C +2 F

hj

j)





177

MBA-H2040 Quantitative Techniques
for Managers




The known cost structures model is explained with the following Example 4.12.



Example 4.12



Consider the following problem.


Problem

A retail shop purchases three items of inventory viz. A, B and C respectively. The shop works

on the limitations that the shop is not able to invest more than Rs.40000 at any time. The other

shop relevant information is given in the following table:



Item

A

B

C

Demand rage (units/year) ? Dj

20000

1000

2000

Purchase Cost/unit ? Cj

40

200

100

Ordering Cost/Order ? Coj

100

150

200

Holding or Carrying Cost - J

40%

40%

40%





Solution



In the absence of constraints, the Optimal Order Sizes are:





Q1 =

2*100*20000 = 500





(.40) (40)





Q2 =

2*150*1000 = 61





(.40)(200)











Q3 =

2*200*2000 = 141





(.40) (100)









Now we know the optimal sizes, so that with these optimal sizes we may determine

the maximum investment, that is


Maximum Investment = 500 * 40 + 61 * 200 + 141 * 100







= 20000 + 12200 + 14100








= Rs. 46300





178
MBA-H2040 Quantitative Techniques
for Managers

Note that the Maximum Investment Rs.46300 is greater than the allowable investment
capacity i.e. Rs.40000 in inventory.



Therefore, equation eq.33 is used with the following alteration





Qj* =

2*Coj*Dj for j = 1, 2, 3. Since, there are only three items. --- eq.34





Cj(J+2)




Note that here = is the solution of the equation.



3



Qj* =

2*Coj*Dj * Cj = 40000



j=1

Cj(J+2)








or





40*20000*100 +

150*1000*200 +

200*2000*100 = 40000



0.1+

0.1+

0.1+










or




= 0.16899919



Now we will substitute the value of in equation eq.34



Therefore, now the order size becomes



Q*1 = 368



Q*2 = 45



Q*3 = 104




That is,



If limitations are not imposed on the purchase of quantities,

The optimal Total Cost = 2*100*20000 + 2*150*1000 + 2*200*2000













500



61



141












= 8000 + 4918 + 5674












= Rs.18592

But, under the limitations,




179

MBA-H2040 Quantitative Techniques
for Managers



The optimal Total Cost = 2*100*20000 + 2*150*1000 + 2*200*2000









368+2*0.17*40 45+2*0.17*200 104+2*0.17*100










= 18934.19, which is higher than the total cost without

limitations.

Note:


Many times the application of the equation

Q*j =

2CojDj

for j = 1,........., n ------------eq.33







(Chj+2 Fj)


to find optimal order quantities under limitations does not help to obtain the result. Therefore,
we have to apply trial and error procedure in the following manner:


i)

First determine the EOQ's for all type of inventory items without considering the

limitation that is taking =0 and then find Qj. If these values satisfy the constraint

(eq.31), then this solution becomes optimal because the constraint is not active.

ii)

If the constraint is not satisfied by the values obtained under (i) above, we give

some value to (arbitrarily but institutively) for example say =h and solve for

Qjs. Qjs satisfy the constraint, these are optimal quantities. Otherwise, we

interpolate or extrapolate the value in between 0 and h or beyond h. With this
value of , the order sizes obtained wil be approximately optimal.


4.5 Probabilistic Inventory Models

In previous sections, we have discussed simple deterministic inventory models where each

and every influencing factor is completely known. Generally in actual business environment

complete certainty never occurs. Therefore, here we will discuss some practical situations of

inventory problems by relaxing the condition of certainty for some of the factors.



The major influencing factors for the inventory problems are Demand, Price and Lead

Time. There are also other factors like Ordering Cost, Carrying Cost or Holding Cost and

Stock out Costs, but their nature is not so much disturbing. Because of this their estimation

provides almost, on the average, as known as values. Even Price can also be averaged out to

reflect the condition of certainty. But there are situations where Price fluctuations are too

much in the market and hence they influence the inventory decisions. Similarly, the demand

variations or consumption variation of an item as well as the lead time variation influence the

overall inventory policy. In this section we will discuss single period probabilistic models.

4.5.1 Single Period Probabilistic Models




180
MBA-H2040 Quantitative Techniques
for Managers

Single Period Discrete Probabilistic Model deals with the inventory situation of the items like

perishable goods, seasonal goods and spare parts requiring one time purchase only. These

items demand may by discrete or continuous. In these models the lead time is very much

important because purchases are made only once.



In single period model, the problem is analyzed using incremental (or

marginal) analysis and the decision procedure consists of a sequence of steps. In such cases,

there are two types of cost involved. There are Under Stocking Cost and Over Stocking Cost.

These two costs describe opportunity losses incurred when the number of units stocked is not

exactly equal to the number of units actually demanded.


In this section we will use the following symbols:

D = Demand for each unit of item (or a random variable)
Q = Number of units stocked or to be purchased
C1 = Under Stocking Cost some times also known as over ordering cost. This is an

opportunity loss associated with each unit left unsold i.e.

C1 = S ? Ch/2 ? Cs



C2 = Over Stocking Cost some times also known as under ordering cost. This is an

opportunity loss due to not meeting the demand, i.e










C2 = C + Ch ? V

Where




C = cost/unit





Ch = carrying cost/unit for the entire period





Cs = shortage cost





V = salvage value





S = selling price





In this section we are going to discuss only discrete demand distribution.


4.5.2 Single Period Discrete Probabilistic Demand Model (Discrete Demand

Distribution)





Here we will discuss the following methods of solving the single period discrete probabilistic
demand.


a. Incremental Analysis Method
b. Payoff Matrix Method


a. Incremental Analysis Method



The Cost equation is developed as follows:




For any quantity in stock Q, only D units are consumed (or demanded). Then for

specific period of time, the cost associated with Q units in stock is either:




181

MBA-H2040 Quantitative Techniques
for Managers

i)

(Q-D)C2, where D is the number of units demanded or consumed is less that or
equal to the number of units Q in stock. That is D Q.


ii)

(D-Q)C1, where the number of units required is greater than the number of units Q
in stock. That is D > Q.



We know that D is a random variable, so its probability distribution of demand is known.

(D) represents the probability that the demand is D units, such that total probability is one.



















That is (0) + (1) + ................. + (D) + ............. = (D) = 1 ------- eq. p1
















D=0

The total expected cost is the sum of expected cost of under-stocking and over-stocking.
Therefore

The Total Expected Cost, say f(Q), is given by





Q

f(Q) = C2 (Q-D) (D) + C1 (D-Q) (D) -------------------- eq.p2


D=0 D=Q+1


Suppose, Q* is the optimal quantity stocked, then the total expected cost f(Q*) will be

minimum. Thus, if we stock one unit more or less than the optimal quantity, the total expected
cost will be higher than the optimal.


Thus,


Q*+1

f(Q*+1) = C2 (Q*+1-D) (D) + C1 (D-Q*-1) (D)





D=0 D=Q*+2







Q* Q*





= C2 (Q*-D) (D) + C2 (D-Q*) (D) + C2 (D) - C1

(D)




D=0 D=Q*+1 D=0

D=Q*+1

Q*




= f(Q*) + (C2+C1) (D) - C1 --------------- eq.p3









D=0




Thus,





f(Q*+1) ? f(Q*) = (C2+C1) (DQ*) ? C1 0 --------- eq.p4


Q*




where (DQ*) = (D) - is a cumulative probability.









D=0




Similarly, we obtain






f(Q*-1) ? f(Q*) = C1 ? (C2+C1) (DQ*-1) 0 ------eq.p5



182
MBA-H2040 Quantitative Techniques
for Managers




We will obtain the following from the above equations eq.p4 and eq.p5






(DQ*-1) C1 (DQ*) --------------------- eq.p6







C2+C1




Therefore, the optimal stock level Q* satisfies the relationship (eq.p6).




Note that for practical application of (eq.p6), the three step procedure is as follows:

Step 1:

From the data, prepare a table showing the probability (D), and cumulative probability
(DQ) for each reasonable value of D.

Step 2:


Calculate the ratio C1 which is called as service level.







C2+C1

Step 3:


Determine the value of Q which satisfies the inequality eq.p6.


This situation is explained with the following Example 4.13.

Example 4.13

An organization stocks seasonal products at the start of the season and cannot reorder. The

inventory item costs him Rs.35 each and he sells at Rs.50 each. For any item that cannot be

met on demand, the organization has estimated a goodwill cost of Rs.25. Any unsold item will

have a salvage value of Rs.20. Holding cost during the period is estimated to be 10% of the

price. The probability distribution of demand is as follows:



Units Stocked

2

3

4

5

6

Probability of

0.35

0.25

0.20

0.15

0.05

Demand (D=Q)





Determine the Optimum Number of Items to be stocked.


Solution



Now we have to follow the above sequence of steps.


Step 1:
We will prepare the Table 4.8 containing the data regarding demand distribution as
follows:




183

MBA-H2040 Quantitative Techniques
for Managers




Units Stocked

Probability of

Cumulative



Demand (D=Q) Probability (DQ)



2

0.35

0.35



3

0.25

0.60



4

0.20

0.80



5

0.15

0.95



6

0.05

1.00







Table 4.8 Probability Distribution of Demand


Step 2:


Calculate the ratio C1 which is called as service level.







C2+C1



We see that




S=50, C=35, Ch=0.1*35= 3.5, V=20, Cs=25




Therefore







C2 = C+Ch-V = 35+3.5-20 = 18.5








C1 = S-C-Ch + Cs = 50-35-3.5 + 25 = 38.25

2

2



Thus, C1 = 38.25 = 0.6740


C2+C1 18.5+38.25

Step 3:

Look into the Table 4.8, the ratio 0.6740 lies between cumulative probabilities of 0.60 and
0.80 which in turn reflect the values of Q as 3 and 4 (units stocked).



That is

(D3) = 0.60<0.6740<0.80 = (D4)



Therefore, the optimal number of units to stock is 4 units.


Cost of Under Stocking Estimation

Suppose, in the previous Example 4.13, the under stocking cost is not known, but the decision

maker policy is to maintain a stock level of say 5 units. We can determine for what values of

C1(under estimating cost) does Q*=5?




In this case, we have the following inequality:




(D4) C1 (D5)





C2+C1




That is

(D4) C1 (D5) or









18.5+C1










Or



184
MBA-H2040 Quantitative Techniques
for Managers






0.80 C1 0.95







18.5+C1




So that the Minimum Value of C1 is determine by letting








C1 = 0.80 or C1 = (0.80)(18.5) = Rs.74







18.5+C1 (1-0.80)




Similarly, the Maximum Value of C1 is determine by letting








C1 = 0.95 or C1 = (0.95)(18.5) = Rs.351.5







18.5+C1 (1-0.95)




Therefore



74C1351.5





Perishable Products Inventory

Many of the organization manage merchandise which contains negligible utility if it is not

sold almost immediately. The examples of such kind of products are newspaper, fresh

produce, printed programmes for special events and other perishable products. Generally such

inventory items have high mark-up. The major difference between the wholesale cost and the

retail price is due to the risk vendor faces in stocking the inventory. Vendor faces

obsolescence costs on the one hand and opportunity costs on the other.

All this kind of problems can be very easily solved with the help of the above

discussed model. This is explained in the following Example 4.14.


Example 4.14




A boy selling newspaper, he buys papers for Rs.0.45 each and sells them for Rs.0.70 each.

The condition here is the boy cannot return unsold newspapers. The following table shows the

daily demand distribution. If each days demand is independent of the pervious days demand,

how many news papers should he order each day?



Number of

240

250

260

270

280

290

300

310

320

330

Customers

Probability

0.01

0.03

0.06

0.10

0.20

0.25

0.15

0.10

0.05

0.05


Solution

Step 1:



185

MBA-H2040 Quantitative Techniques
for Managers

Prepare a following Table 4.9 showing the probability (D), and cumulative probability
(DQ) for each reasonable value of D.

Number of

240

250

260

270

280

290

300

310

320

330

Customers

Probability

0.01

0.03

0.06

0.10

0.20

0.25

0.15

0.10

0.05

0.05

Cumulative

0.01

0.04

0.10

0.20

0.40

0.65

0.80

0.90

0.95

1.00

Probability







Table 4.9 Probability Discrete Distribution of Demand

Step 2:






C1 = 0.25 = 0.25 = 0.357

C2+C1 0.45+0.25 0.70

Step 3:


Thus, the Value of Q such that





(DQ*-1) 0.357 (DQ*) this gives





Q* = 280


Therefore

The newspaper boy should buy 280 papers each day.


b. Payoff Matrix Method

The Payoff Matrix Method of single period Discrete Probabilistic Demand Model is explained

with the help of the following Example 4.15.


Example 4.15

Consider the example 4.13 i.e.


An organization stocks seasonal products at the start of the season and cannot reorder.

The inventory item costs him Rs.35 each and he sells at Rs.50 each. For any item that cannot

be met on demand, the organization has estimated a goodwill cost of Rs.25. Any unsold item

will have a salvage value of Rs.20. Holding cost during the period is estimated to be 10% of

the price. The probability distribution of demand is as follows:



Units Stocked

2

3

4

5

6

Probability of

0.35

0.25

0.20

0.15

0.05

Demand (D=Q)





Determine the Optimum Number of Items to be stocked.




In this case the organization has five reasonable courses of action. The organization

can stock the items from 2 to 6 units. There is no possible reason to stock more than 6 items



186
MBA-H2040 Quantitative Techniques
for Managers

since the organization can never sell more than 6 items and there is no possible reason for

ordering less than 2 items. Since there are five courses of action for stocking and five levels of

demand, it follows that there are 25 combinations of one course of action and one level of

demand. For these 25 combinations, we can determine the organization payoffs in the form of

payoff matrix.



As per the information of cost given in the problem, the payoffs are obtained for the

following two situations:

When demand is not more than the stock level

When demand is more than the stock quantity.


That is

Payoffs For

QD

Q<D

Item Cost

-35Q

-35Q

Sale of items

50D

50Q

Goodwill Cost

-

-25(D-Q)

Salvage Cost

20(Q-D)

-

Holding Cost

-3.5(Q-D)-3.5D/2

-3.5Q/2

Total payoff

-18.50Q+31.75D

38.25Q-25D



The payoff matrix will be 5X5. Each element of the matrix can be computed by above

total payoffs for demand less than, equal to, or greater than the order size (Q). When demand

is less than or equal to the order size, we have the following contributions to the payoff.



Here the organization buys the items for Rs.Rs.35Q and the organization sells D of

them for Rs.50D, the organization earns salvage of Rs.20 (Q-D) for unsold items, and the

organization incurs holding cost of Rs.(.10)(35)(Q-D) on unsold items an average holding

cost of (.10)(35)D/2 on the sold items during the period. Thus the total payoff becomes -

128.5Q+31.75D for demand less or equal to order size.



If demand is more than the order size, the contributory payoff will consist of the

following:


- Purchase Cost Rs.35Q
- Selling Profit of Rs.50Q
- Goodwill Cost Rs.25(D-Q) and
- Holding Cost of Rs. (0.10)(35)Q/2



Thus, the total payoff for demand more than order size is 38.25Q-25D.

The payoff matrix is as follows (Table 4.10):


Units Demanded D

2 3 4 5
6
Units stocked or 2 26.50 1.50 -23.50 -48.5 -



187

MBA-H2040 Quantitative Techniques
for Managers

73.5
Order Size (Q) 3 8.0 39.75 14.75 -10.25 -
35.25
4 -10.5 21.25 53.0 28.0
3.0
5 -29.0 2.75 34.5 66.25
41.25
6 -47.5 -15.75 16.0 47.75
79.5


Probability of Demand 0.35 0.25 0.20 0.15
0.05


Table 4.10 Payoff Matrix





Now we will determine the expected payoff for each order size or courses of action.

The procedure for computing the expected values is simple, as follows:



Procedure: For any given course of action multiply each possible payoff for that

course of action by the corresponding probability of the given level of demand and add all of
these products up.



Thus, for first course of action of order size 2 units, the expected value of payoff is:






(26.5)(0.35) + (1.5)(0.25) + (-23.5)(0.2) + (-48.5)(0.15) + (-73.5)(0.05) = Rs.-6




For order size 3 units





(8.0)(0.35) + (39.75)(0.25) + (14.75)(0.2) + (-10.25)(0.15) + (-35.25)(0.05) =

Rs.12.3875



For order size 4 units





(-10.5)(0.35) + (21.25)(0.25) + (53)(0.2) + (28)(0.15) + (3)(0.05) = Rs.20.2625




For order size 5 units





(-29)(0.35) + (2.75)(0.25) + (34.5)(0.2) + (66.25)(0.15) + (41.25)(0.05) =

Rs.9.4375



For order size 6 units





(-47.5)(0.35) + (-15.75)(0.25) + (16)(0.2) + (47.75)().15) + (79.5)(0.05) = Rs.-

6.225

Therefore, we compute all the expected values:


Order Size Q

2

3

4

5

6

Expected value

Rs.-6

Rs.12.3875 Rs.20.2625 Rs.9.4375

Rs.-

6.225





188
MBA-H2040 Quantitative Techniques
for Managers

Here the objective is to select course of action which provides the highest payoff.

Thus, the organization should order for 4 units for the highest expected payoff value of
Rs.20.2625.





Note: If we compare the two methods i.e. incremental analysis and payoff matrix, if we see

the solution that incremental analysis provides only the optimum level of purchase quantity

and does not indicate about the level of expected profit. But, the payoff matrix method

provides both the answers i.e. optimum purchase quantity as well as the optimum expected

profit.



The interesting is here, we may also convert the payoff matrix to opportunity cost

matrix, where the opportunity cost is, in short, a cost sustained because the decision taken is

not the best in terms of the level of demand which actually occurs.



The computation of opportunity cost matrix from the payoff matrix is very easy. Take

any column of the payoff matrix corresponding to a specific level of demand and select the

largest payoff if the payoffs are profits, the smallest payoff if the payoffs are costs. Then

subtract each payoff in the same column from the largest payoff to get the corresponding

opportunity costs in the case of profits. If it is cost, subtract the smallest payoff from each

payoff in the same column to get the opportunity costs.



In this Example 4.15, we may obtain the opportunity cost matrix as follows (Table

4.11):


Units Demanded D

2 3 4 5 6
Order size
2 0 38.25 76.5 114.75 153.0
3 18.5 0 38.25 76.5 114.75
4 37.0 18.5 0 38.25 76.5
5 55.5 37.0 18.5 0 38.0
6 74.0 55.5 37.0 18.5 0


Probability of
Demand 0.35 0.25 0.20 0.15 0.05









Table 4.11 Opportunity Cost Matrix




Now, we have to determine the expected opportunity costs for each alternative courses

of action. The objective is to select the course of action which provides minimum expected

opportunity costs.

Therefore,

The expected opportunity cost for the first alternative course of action of order size 2

is:



189

MBA-H2040 Quantitative Techniques
for Managers



(0)(.35) + (38.25)(.25) + (76.5)(.2) + (114.75)(0.15) + (153)(.05) = Rs.49.725



For order size 3 units


(18.5)(.35) + (0)(.25) + (38.25)(.2) + (76.5)(.15) + (114.75)(.05) = Rs.31.3375


For order size 4 units


(37)(.35) + (18.5)(.25) + (0)(.2) + (38.25)(.15) + (76.5)(.05) = Rs.27.1375


For order size 5 units


(55.5)(.35) + (37)(.25) + (18.5)(.2) + (0)(.15) + (38)(.05) = Rs.34.275


For order size 6 units


(74)(.35) + (55.5)(.25) + (37)(.2) + (18.5)(.15) + (0)(.05) = Rs.49.95


That is, the following are the obtained expected opportunity costs;


Order Size Q

2

3

4

5

6

Expected Cost

Rs.

Rs.31.3375 Rs.27.1375 Rs.34.275 Rs.49.95

49.725




Thus, the decision is to select the minimum expected cost is that, the organization

should store 4 units for the lowest cost of Rs.27.1375

Relationship between the Payoff Matrix and Opportunity Cost Matrix



Here, we may find a relationship between the payoff matrix and the opportunity cost

matrix, as follows:



Let





EOC = K ? EP



Where,





EOC = Expected Opportunity Cost





EP = Expected Payoff or profit





K = Constant or

K = (26.5)(.35) + (39.75)(.25) + (53)(.2) + (66.25)(.15) + (79.5)(.05)




= 43.735






That is K = sum of the expected value of the largest elements in each column

of the

payoff matrix.

= the expected value of the payoffs for all the best courses of action.
Or

The expected opportunity cost for a given courses of action= K(43.735)-Expected pay of for
each
courses of action -
(eq1)





190
MBA-H2040 Quantitative Techniques
for Managers



Thus, it is obvious from the above equation eq1 that the maximum value of EP will

simultaneously produce, the minimum of EOC. The two analyses namely, the payoff matrix

method and the opportunity cost matrix method produce the same result.



Suppose, if the original matrix is in terms of costs it can by similar reasoning, be

shown that the above relationship (eq1) will be of the following form:













EOC = EP ? K



In this case EP is in terms of costs.


4.6 Summary






In this lesson various deterministic inventory models have been developed for various

operating conditions. Here we discussed single item inventory and as well as multi item

inventory models. In this lesson we also discussed probabilistic discrete demand models for

single period inventory items.


4.7 Key Terms



Inventory ? stores of goods or stocks.



Ordering Cost - Cost involved in placing an order.



Procurement Cost ? Same as ordering cost.



Replenishment Cost ? Same as ordering cost.



Set up Cost ? Cost associated with the setting of machine for production.



Shortage Cost ? Costs associated with the demand when stocks have been depleted, it

is

generally called as back order costs.



Safety Stock ? Extra Stocks.



Perishable Product ? The inventories that deteriorate with time.



Deterministic Model ? An inventory model where all the factors are completely

known.


Discrete Probability Distribution ? A probability distribution in which the variable

is allowed

to take only limited number of values.



Expected Opportunity Cost ? Expected value of the variable indicating opportunity

costs.


Expected payoff ? Expected value of the variable indicating payoffs.



Expected Value ? The average value or mean.



Minimum Value ? This is also known as safety stock or the buffer stock.



Maximum Value ? Level of inventory beyond which inventory is not allowed.



Payoff ? The benefit which accrues form a given combination of decision alternative

courses of




action and state of nature.



Reorder Level ? The stock level which is sufficient for the lead time consumption,

and an order

is initiated when inventory dips to this level.



Under Stocking Cost ? Cost relating to the out of stock situation under the

probabilistic



191

MBA-H2040 Quantitative Techniques
for Managers

situation.



Over Stocking Cost ? This is the cost of keeping more units than demanded.



Opportunity Cost Matrix ? Matrix of opportunity Costs.










4.8 Self Assessment Questions

Q1. A Production unit uses Rs.10,000 worth of an item during the year. The production units
estimated the ordering cost as Rs.25 per order and holding cost as 12.5 percent of the average
inventory value. Determine the optimal order size, number of orders per year, time period per
order and total cost.

Answer


Order Size = Q* = Rs.2000



No. of orders per year = N = 5



Time period per order = t* = 73 days



Total Cost = TC* = Rs.250


Q2. The usage of an inventory item each costing Re 1, is 10000 units/year and the ordering
cost is Rs.10, carrying charge is 20% based on the average inventory per year, stock out cost
is Rs.5 per unit of shortage incurred. Determine EOQ, inventory level, shortage level, cycle
period, number of order per year and the total cost.

Answer


EOQ = Q* = 1020 units



Inventory Level = I* = 980 units



Shortage Level = 40 units



Cycle Period = t* = 37 days



No. of orders/year = 10



Total Cost = Rs.400


Q3. The demand for a unit of item is at the rate of 200 per day and can be produced at a rate
of 800 per day. It costs Rs.5000 to set up the production process and Rs.0.2 per unit per day
held in inventory based on the actual inventory any time. Assume that the shortage is not
allowed. Find out the minimum cost and the optimum number of units per production run.

Answer


Hint: D=200 P=800 Co=Rs.5000 Ch=Rs.0.2




Q* = 3651 units



TC* = Rs.547.7


Q4. The demand for an item is 2400 units per year. The ordering cost is Rs.100, inventory
holding cost is 24 percent of the purchase price per year. Determine the optimum purchase
quantity if the purchase prices are as follows:




192
MBA-H2040 Quantitative Techniques
for Managers





P1 = Rs.10 for purchasing Q1 < 500





P2 = Rs.9.25 for purchasing 500 Q2 < 750





P3 = Rs.8.75 for purchasing 750 Q3


Answer




Economic Purchase Quantity = EPQ = Q* = 750 units





Q5. A company follows the following procurement pattern of five items irrespective of their
level of demand. Reduce the inventory levels while keeping the same total number of orders
per annum.


Item

Demand/Year Number of

Order size ($)

Average

($)

orders/year

Inventory

1

1000,000

5

250,000

125,000

2

640,000

5

160,000

80,000

3

90,000

5

22,500

11,250

4

2,500

5

625

350

5

1,600

5

400

200


Answer




According to the company policy, ordering five times a year each item, total average

inventory becomes $.216800.

But after the analysis of ordering quantity the average inventory becomes $.7698.32,

which is much less, at the same time the number of orders almost remain same. Thus,

substantial savings can still be achieved when cost information is not known.


Q6. Suppose the carrying cost is 30% per unit/year, unit price is Rs.4, and the ordering cost is
Rs.30 per order for an item used in an organization in the following pattern:


Month

Jan

Feb

Mar

Apr

May

Jun

Jul

Aug

Sep

Oct

Nov

Dec

Demand 200

220

150

170

210

200

180

220

170

200

160

140


Determine the ordering schedule and the total inventory cost using the following method.


1. Prescribed Rule Method
2. Fixed EOQ Method


Q7. The following numbers indicate the annual values in dollars of some thirty inventory
items of materials selected at random. Carry out an ABC analysis and list out the values of
three items viz. A-items, B-items and C-items.


1

2

3

9

75

3

4

6

13

2

3

12



193

MBA-H2040 Quantitative Techniques
for Managers

100

2

7

40

15

55

1

12

25

15

8

10

1

20

30

1

4

5





194
MBA-H2040 Quantitative Techniques
for Managers

Q8. An organization stoking two products. The organization has limited storage space and
can't store more than 40 units. The following are the demand distribution for the products:


Product1 Product2
Demand Probability of Demand Demand Probability of Demand
10 0.10 10 0.05
20 0.20 20 0.20
30 0.35 30 0.30
40 0.25 40 0.13
50 0.10 50 0.10


If the inventory holding cost is Rs.10 (product 1) and Rs15 (product 2) per unit of the ending
inventories, the shortage costs are Rs.20 and Rs.50 per unit at the ending shortage for the first
and second products respectively. Determine the economic order quantities for both the
products.

4.9 Further References

Hamdy A Taha, 1999. Introduction to Operations Research, PHI Limited, New Delhi.

Mustafi, C.K. 1988. Operations Research Methods and Practices, Wiley Eastern Limited,
New Delhi.

Levin, R and Kirkpatrick, C.A. 1978. Quantitative Approached to Management, Tata
McGraw Hill, Kogakusha Ltd., International Student Edition.

Peterson R and Silver, E. A. 1979. Decision Systems for Inventory Management and
Production Planning.

Handley, G and T.N. Whitin. 1983. Analysis of Inventory Systems, PHI.

Starr, M.K. and D.W. Miller. 1977. Inventory Control Theory and Practice, PHI.





















195

MBA-H2040 Quantitative Techniques
for Managers





UNIT III



NETWORK PROBLEMS

Introduction

A network consists of several destinations or jobs which are linked with one another. A

manager will have occasions to deal with some network or other. Certain problems pertaining

to networks are taken up for consideration in this unit.



LESSON 1



SHORTEST PATH PROBLEM


LESSON OUTLINE

The description of a shortest path problem.

The determination of the shortest path.


LEARNING OBJECTIVES

After reading this lesson you should be able to

- understand a shortest path problem
- understand the algorithm for a shortest path problem
-

work out numerical problems


THE PROBLEM

Imagine a salesman or a milk vendor or a post man who has to cover certain previously earmarked places to

perform his daily routines. It is assumed that all the places to be visited by him are connected well for a suitable

mode of transport. He has to cover all the locations. While doing so, if he visits the same place again and again

on the same day, it will be a loss of several resources such as time, money, etc. Therefore he shall place a

constraint upon himself not to visit the same place again and again on the same day. He shall be in a position to

determine a route which would enable him to cover all the locations, fulfilling the constraint.



The shortest route method aims to find how a person can travel from one location to another, keeping

the total distance traveled to the minimum. In other words, it seeks to identify the shortest route to a series of

destinations.





EXAMPLE

Let us consider a real life situation involving a shortest route problem.

A leather manufacturing company has to transport the finished goods from the factory

to the store house. The path from the factory to the store house is through certain



196
MBA-H2040 Quantitative Techniques
for Managers

intermediate stations as indicated in the following diagram. The company executive wants to

identify the path with the shortest distance so as to minimize the transportation cost. The

problem is to achieve this objective.









95 Store house



2

4



Factory 40











40











35

65



70

6





















40

1









100

5















20

3

Linkages from Factory to Store house





The shortest route technique can be used to minimize the total distance from a node designated as the

starting node or origin to another node designated as the final node.



In the example under consideration, the origin is the factory and the final node is the store house.

STEPS IN THE SHORTEST ROUTE TECHNIQUE

The procedure consists of starting with a set containing a node and enlarging the set by choosing a node in each

subsequent step.

Step 1:

First, locate the origin. Then, find the node nearest to the origin. Mark the distance between the origin and the

nearest node in a box by the side of that node.

In some cases, it may be necessary to check several paths to find the nearest node.

Step 2:

Repeat the above process until the nodes in the entire network have been accounted for. The last distance placed

in a box by the side of the ending node will be the distance of the shortest route. We note that the distances

indicated in the boxes by each node constitute the shortest route to that node. These distances are used as

intermediate results in determining the next nearest node.

SOLUTION FOR THE EXAMPLE PROBLEM

Looking at the diagram, we see that node 1 is the origin and the nodes 2 and 3 are neighbours

to the origin. Among the two nodes, we see that node 2 is at a distance of 40 units from node

1 whereas node 3 is at a distance of 100 units from node 1. The minimum of {40, 100} is 40.

Thus, the node nearest to the origin is node 2, with a distance of 40 units. So, out of the two

nodes 2 and 3, we select node 2. We form a set of nodes {1, 2} and construct a path

connecting the node 2 with node 1 by a thick line and mark the distance of 40 in a box by the

side of node 2. This first iteration is shown in the following diagram.





40



197

MBA-H2040 Quantitative Techniques
for Managers









95 Store house



2

4







40











40

6

Factory









35



65 70





1















40

5









100

3















20

ITERATION No. 1




Now we search for the next node nearest to the set of nodes {1, 2}. For this purpose, consider those

nodes which are neighbours of either node 1 or node 2. The nodes 3, 4 and 5 fulfill this condition. We calculate

the following distances.

The distance between nodes 1 and 3 = 100.

The distance between nodes 2 and 3 = 35.

The distance between nodes 2 and 4 = 95.

The distance between nodes 2 and 5 = 65.

Minimum of {100, 35, 95, 65} = 35.

Therefore, node 3 is the nearest one to the set {1, 2}. In view of this observation, the set of nodes is enlarged

from {1, 2} to {1, 2, 3}. For the set {1, 2, 3}, there are two possible paths, viz. Path 1 2 3 and Path 1

3 2. The Path 1 2 3 has a distance of 40 + 35 = 75 units while the Path 1 3 2 has a distance of

100 + 35 = 135 units.

Minimum of {75, 135} = 75. Hence we select the path 1 2 3 and display this path by thick edges. The

distance 75 is marked in a box by the side of node 3. We obtain the following diagram at the end of Iteration

No. 2.





40







95 Store house



2

4

Factory





40











40

6











35

65

70





1















40

5









100

3















20

75



ITERATION No. 2

REPEATING THE PROCESS

We repeat the process. The next node nearest to the set {1, 2, 3} is either node 4 or node 5.

Node 4 is at a distance of 95 units from node 2 while node 2 is at a distance of 40 units

from node 1. Thus, node 4 is at a distance of 95 + 40 = 135 units from the origin.



198
MBA-H2040 Quantitative Techniques
for Managers



As regards node 5, there are two paths viz. 2 5 and 3 5, providing a link to the origin. We already

know the shortest routes from nodes 2 and 3 to the origin. The minimum distances have been indicated in boxes

near these nodes. The path 3 5 involves the shortest distance. Thus, the distance between nodes 1 and 5 is 95

units (20 units between nodes 5 and 3 + 75 units between node 3 and the origin). Therefore, we select node 5

and enlarge the set from {1, 2, 3} to {1, 2, 3, 5}. The distance 95 is marked in a box by the side of node 5. The

following diagram is obtained at the end of Iteration No. 3.



40







95 Store house



2

4







40











40

6











35

65

70





1















40

5



Factory 100

3















20 95



75



ITERATION No. 3



Now 2 nodes remain, viz., nodes 4 and 6. Among them, node 4 is at a distance of 135 units from the

origin (95 units from node 4 to node 2 + 40 units from node 2 to the origin). Node 6 is at a distance of 135

units from the origin (40 + 95 units). Therefore, nodes 4 and 6 are at equal distances from the origin. If we

choose node 4, then travelling from node 4 to node 6 will involve an additional distance of 40 units. However,

node 6 is the ending node. Therefore, we select node 6 instead of node 4. Thus the set is enlarged from {1, 2, 3,

5} to {1, 2, 3, 5, 6}. The distance 135 is marked in a box by the side of node 6. Since we have got a path

beginning from the start node and terminating with the stop node, we see that the solution to the given problem

has been obtained. We have the following diagram at the end of Iteration No. 4.



40







95 Store house



2

4







40











40

6











35

65

70





1















40

135

5



Factory

100

3















20

95



75

ITERATION No. 4

MINIMUM DISTANCE

Referring to the above diagram, we see that the shortest route is provided by the path 1 2
3 5 6 with a minimum distance of 135 units.





199

MBA-H2040 Quantitative Techniques
for Managers

QUESTIONS

1. Explain the shortest path problem.

2. Explain the algorithm for a shortest path problem

3. Find the shortest path of the following network:


30


3

5

40
40
30 50 30

1


45

6

25

35


2

4



4. Determine the shortest path of the following network:




2

16

5


7
9 7


1

4

15

6

8 4
25

3














LESSON 2

MINIMUM SPANNING TREE PROBLEM

LESSON OUTLINE

The description of a minimum spanning tree problem.

The identification of the minimum spanning tree.


LEARNING OBJECTIVES


After reading this lesson you should be able to

-

understand a minimum spanning tree problem

-

understand the algorithm for minimum spanning tree problem

-

locate the minimum spanning tree



200
MBA-H2040 Quantitative Techniques
for Managers

-

carry out numerical problems


Tree: A minimally connected network is called a tree. If there are n nodes in a network, it

will be a tree if the number of edges = n-1.



Minimum spanning tree algorithm

Problem : Given a connected network with weights assigned to the edges, it is required to

find out a tree whose nodes are the same as those of the network.



The weight assigned to an edge may be regarded as the distance between the two

nodes with which the edge is incident.

Algorithm:



The problem can be solved with the help of the following algorithm.



The procedure consists of selection of a node at each step.

Step 1: First select any node in the network. This can be done arbitrarily. We will start with

this node.

Step 2: Connect the selected node to the nearest node.

Step 3: Consider the nodes that are now connected. Consider the remaining nodes. If there is

no node remaining, then stop. On the other hand, if some nodes remain, among them find out

which one is nearest to the nodes that are already connected. Select this node and go to Step

2.



Thus the method involves the repeated application of Steps 2 and 3. Since the number

of nodes in the given network is finite, the process will end after a finite number of steps. The

algorithm will terminate with step 3.

How to break ties:

While applying the above algorithm, if some nodes remain in step 3 and if there is a tie in the

nearest node, then the tie can be broken arbitrarily.



As a consequence of tie, we may end up with more than one optimal solution.

Problem 1:

Determine the minimum spanning tree for the following network.

60

5

2

70

60 60 80

100

7



1

3

40 120 50



8



201

4

6

MBA-H2040 Quantitative Techniques
for Managers

80 50 60

30

90

Solution:

Step 1: First select node 1. (This is done arbitrarily)

Step 2: We have to connect node 1 to the nearest node. Nodes 2, 3 and 4 are adjacent to node
1. They are at distances of 60, 40 and 80 units from node 1. Minimum of {60, 40, 80} = 40.
Hence the shortest distance is 40. This corresponds to node 3. So we connect node 1 to node
3 by a thick line. This is Iteration No. 1.


60

5



2 70

60 60 80

100

7



1

3

40 120 50



8

80 50 60

30

90

6

4



Iteration No. 1

Step 3: Now the connected nodes are 1 and 3. The remaining nodes are 2, 4, 5, 6, 7 and 8.

Among them, nodes 2 and 4 are connected to node 1. They are at distances of 60 and 80 from

node 1. Minimum of {60, 80} = 60. So the shortest distance is 60. Next, among the nodes 2,

4, 5, 6, 7 and 8, find out which nodes are connected to node 3. We find that all of them are

connected to node 3. They are at distances of 60, 50, 80, 60, 100 and 120 from node 3.

Minimum of {60, 50, 80, 60, 100, 120} = 50. Hence the shortest distance is 50.

Among these nodes, it is seen that node 4 is nearest to node 3.

Now we go to Step 2. We connect node 3 to node 4 by a thick line. This is Iteration

No.2.



60

5

2

70

60 60 80

100

7



1

3



202
MBA-H2040 Quantitative Techniques
for Managers

40 120 50



8

80 50 60

30

90

6

4



Iteration No. 2

Next go to step 3.

Now the connected nodes are 1, 3 and 4. The remaining nodes are 2, 5, 6, 7 and 8.

Node 2 is at a distance of 60 from node 1. Nodes 5, 6, 7 and 8 are not adjacent to node 1. All

of the nodes 2, 5, 6, 7 and 8 are adjacent to node 3. Among them, nodes 2 and 6 are nearer to

node 3, with equal distance of 60.

Node 6 is adjacent to node 4, at a distance of 90. Now there is a tie between nodes 2

and 6. The tie can be broken arbitrarily. So we select node 2. Connect node 3 to Node 2 by

a thick line. This is Iteration No. 3.







60

5

2

70

60 60 80

100

7



1

3

40 120 50



8

80 50 60

30

90

6

4



Iteration No. 3



We continue the above process.

Now nodes 1, 2, 3 and 4 are connected. The remaining nodes are 5, 6, 7 and 8. None

of them is adjacent to node 1. Node 5 is adjacent to node 2 at a distance of 60. Node 6 is at a

distance of 60 from node 3. Node 6 is at a distance of 90 from node 4. There is a tie between



203

MBA-H2040 Quantitative Techniques
for Managers

nodes 5 and 6. We select node 5. Connect node 2 to node 5 by a thick line. This is Iteration

No. 4.



60

5

2

70

60 60 80

100

7



1

3

40 120 50



8

80 50 60

30

90

6

4



Iteration No. 4



Now nodes 1, 2, 3, 4 and 5 are connected. The remaining nodes are 6, 7 and 8. Among them,

node 6 is at the shortest distance of 60 from node 3. So, connect node 3 to node 6 by a thick

line. This is Iteration No. 5.



60

5

2

70

60 60 80

100

7



1

3

40 120 50



8

80 50 60

30

90

6

4



Iteration No. 5

Now nodes 1, 2, 3, 4, 5 and 6 are connected. The remaining nodes are 7 and 8. Among them,

node 8 is at the shortest distance of 30 from node 6. Consequently we connect node 6 to node

8 by a thick line. This is Iteration No. 6.





204
MBA-H2040 Quantitative Techniques
for Managers

60

5

2

70

60 60 80

100

7



1

3

40 120 50



8

80 50 60

30

90

6

4



Iteration No. 6

Now nodes 1, 2, 3, 4, 5, 6 and 8 are connected. The remaining node is 7. It is at the shortest

distance of 50 from node 8. So, connect node 8 to node 7 by a thick line. This is Iteration

No.7.



60

5

2

70

60 60 80

100

7



1

3

40 120 50



8

80 50 60

30

90

6

4



Iteration No. 7

Now all the nodes 1, 2, 3, 4, 5, 6, 7 and 8 are connected by seven thick lines. Since no node

is remaining, we have reached the stopping condition. Thus we obtain the following minimum

spanning tree for the given network.



60

5

2



60

7



1

3



205

8

MBA-H2040 Quantitative Techniques
for Managers

40 50



50 60

30





Minimum Spanning Tree





QUESTIONS

1. Explain the minimum spanning tree algorithm.

2. From the following network, find the minimum spanning tree.



75

6

2



80 55 90

100

1

3

70 40



25 60

5



30

4



3. Find the minimum spanning tree of the following network:



12

5



2 15

5 8

2 10 13

8

1

3

6



9 4 10

5

7



4

4



206
MBA-H2040 Quantitative Techniques
for Managers











207

MBA-H2040 Quantitative Techniques
for Managers

LESSON 3

PROJECT NETWORK



LESSON OUTLINE

The key concepts

Construction of project network diagram


LEARNING OBJECTIVES


After reading this lesson you should be able to

- understand the definitions of important terms
-

understand the development of project network diagram

-

work out numerical problems



KEY CONCEPTS

Certain key concepts pertaining to a project network are described below:

1. Activity

An activity means a work. A project consists of several activities. An activity takes time. It

is represented by an arrow in a diagram of the network. For example, an activity in house

construction can be flooring. This is represented as follows:



flooring

Construction of a house involves various activities. Flooring is an activity in this project. We

can say that a project is completed only when all the activities in the project are completed.

2. Event

It is the beginning or the end of an activity. Events are represented by circles in a project

network diagram. The events in a network are called the nodes. Example:

Start Stop







Punching

Starting a punching machine is an activity. Stopping the punching machine is another activity.

3. Predecessor Event

The event just before another event is called the predecessor event.

4. Successor Event

The event just following another event is called the successor event.

Example: Consider the following.



3



208

1

2

4

6
MBA-H2040 Quantitative Techniques
for Managers















In this diagram, event 1 is predecessor for the event 2.

Event 2 is successor to event 1.

Event 2 is predecessor for the events 3, 4 and 5.

Event 4 is predecessor for the event 6.

Event 6 is successor to events 3, 4 and 5.

5. Network

A network is a series of related activities and events which result in an end product or service.

The activities shall follow a prescribed sequence. For example, while constructing a house,

laying the foundation should take place before the construction of walls. Fitting water tapes

will be done towards the completion of the construction. Such a sequence cannot be altered.

6. Dummy Activity

A dummy activity is an activity which does not consume any time. Sometimes, it may be

necessary to introduce a dummy activity in order to provide connectivity to a network or for

the preservation of the logical sequence of the nodes and edges.



7. Construction of a Project Network

A project network consists of a finite number of events and activities, by adhering to a certain

specified sequence. There shall be a start event and an end event (or stop event). All the

other events shall be between the start and the end events. The activities shall be marked by

directed arrows. An activity takes the project from one event to another event.



An event takes place at a point of time whereas an activity takes place from one point

of time to another point of time.



CONSTRUCTION OF PROJECT NETWORK DIAGRAMS

Problem 1:

Construct the network diagram for a project with the following activities:

Activity

Name of

Immediate



209

MBA-H2040 Quantitative Techniques
for Managers

EventEvent

Activity

Predecessor
Activity


12

A

-

13

B

-

14

C

-

25

D

A

36

E

B

46

F

C

56

G

D



Solution:

The start event is node 1.

The activities A, B, C start from node 1 and none of them has a predecessor activity. A joins

nodes1 and 2; B joins nodes 1 and 3; C joins nodes 1 and 4. So we get the following:



A

2



1

3

B

C

4



This is a part of the network diagram that is being constructed.

Next, activity D has A as the predecessor activity. D joins nodes 2 and 5. So we get



A D

1

2

5

Next, activity E has B as the predecessor activity. E joins nodes 3 and 6. So we get



B E

1

3

6



Next, activity G has D as the predecessor activity. G joins nodes 5 and 6. Thus we obtain

D G

2

5

6



Since activities E, F, G terminate in node 6, we get



5

G



3

6

E

4



210
MBA-H2040 Quantitative Techniques
for Managers

F



6 is the end event.

Combining all the pieces together, the following network diagram is obtained for the given

project:

D

5



2

A G

Start event End event
B E

1

3

6




C F



4



We validate the diagram by checking with the given data.






Problem 2:
Develop a network diagram for the project specified below:

Activity

Immediate



Predecessor Activity



A

-

B

A

C, D

B

E

C

F

D

G

E, F



Solution:

Activity A has no predecessor activity. i.e., It is the first activity. Let us suppose that activity

A takes the project from event 1 to event 2. Then we have the following representation for A:





211

MBA-H2040 Quantitative Techniques
for Managers

A

1

2



For activity B, the predecessor activity is A. Let us suppose that B joins nodes 2 and 3. Thus

we get



A B

1

2

3



Activities C and D have B as the predecessor activity. Therefore we obtain the following:

C

B

4



2

3

D





5

Activity E has D as the predecessor activity. So we get



C E

3

4

6





Activity F has D as the predecessor activity. So we get



D F

3

5

6l



"

Activity G has E and F as predecessor activities. This is possible only if nodes 6 and 6l are

one and the same. So, rename node 6l as node 6. Then we get



D F

3

5

6!





and



4

E



6

7

G

5



212
MBA-H2040 Quantitative Techniques
for Managers

F



G is the last activity.

Putting all the pieces together, we obtain the following diagram the project network:



Start event C

4 E End event


A B G

1

2

3



6

7


D F


5



The diagram is validated by referring to the given data.

Note: An important point may be observed for the above diagram. Consider the following
parts in the diagram


C E

3

4

6





and



D F

3

5

6l



"



We took nodes 6 and 6l as one and the same. Instead, we can retain them as different nodes.

Then, in order to provide connectivity to the network, we join nodes 6l and 6 by a dummy

activity. Then we arrive at the following diagram for the project network:





4

Start event C E

6

G

1

2

3

A B dummy
activity

7


D F End event

5

6l



!



213

MBA-H2040 Quantitative Techniques
for Managers


QUESTIONS:


1. Explain the terms: event, predecessor event, successor event, activity, dummy

activity, network.

2. Construct the network diagram for the following project:

Activity

Immediate



Predecessor Activity

A

-

B

-

C

A

D

B

E

A

F

C, D

G

E

H

E

I

F, G

J

H, I



214
MBA-H2040 Quantitative Techniques
for Managers

LESSON 4



CRITICAL PATH METHOD (CPM)

LESSON OUTLINE

The concepts of critical path and critical activities

Location of the critical path

Evaluation of the project completion time



LEARNING OBJECTIVES


After reading this lesson you should be able to

-

understand the definitions of critical path and critical activities

- identify critical path and critical activities
- determine the project completion time



INTRODUCTION

The critical path method (CPM) aims at the determination of the time to complete a project

and the important activities on which a manager shall focus attention.



ASSUMPTION FOR CPM

In CPM, it is assumed that precise time estimate is available for each activity.



PROJECT COMPLETION TIME

From the start event to the end event, the time required to complete all the activities of the

project in the specified sequence is known as the project completion time.



PATH IN A PROJECT

A continuous sequence, consisting of nodes and activities alternatively, beginning with the

start event and stopping at the end event of a network is called a path in the network.



CRITICAL PATH AND CRTICAL ACTIVITIES

Consider all the paths in a project, beginning with the start event and stopping at the end

event. For each path, calculate the time of execution, by adding the time for the individual

activities in that path.





The path with the largest time is called the critical path and the activities along this

path are called the critical activities or bottleneck activities. The activities are called critical



215

MBA-H2040 Quantitative Techniques
for Managers

because they cannot be delayed. However, a non-critical activity may be delayed to a certain

extent. Any delay in a critical activity will delay the completion of the whole project.

However, a certain permissible delay in a non ?critical activity will not delay the completion

of the whole project. It shall be noted that delay in a non-critical activity beyond a limit

would certainly delay the completion the whole project. Sometimes, there may be several

critical paths for a project. A project manager shall pay special attention to critical activities.



Problem 1:

The following details are available regarding a project:

Activity

Predecessor

Duration (Weeks)



Activity


A

-

3

B

A

5

C

A

7

D

B

10

E

C

5

F

D,E

4



Determine the critical path, the critical activities and the project completion time.

Solution:

First let us construct the network diagram for the given project. We mark the time estimates

along the arrows representing the activities. We obtain the following diagram:





Start event B

3 D End event

5 10
A

1

2

3 F

5

6

C 4
7 E


4 5




Consider the paths, beginning with the start node and stopping with the end node. There are

two such paths for the given project. They are as follows:

Path I





216
MBA-H2040 Quantitative Techniques
for Managers

A B D F


1

2

3

5

6

3 5 10 4

with a time of 3 + 5 + 10 + 4 = 22 weeks.

Path II


A C E F


1

2

4

5

6

3 7 5 4

with a time of 3 + 7 + 5 + 4 = 19 weeks.

Compare the times for the two paths. Maximum of {22,19} = 22. We see that path I has the
maximum time of 22 weeks. Therefore, path I is the critical path. The critical activities are A,
B, D and F. The project completion time is 22 weeks.


We notice that C and E are non- critical activities.
Time for path I - Time for path II = 22- 19 = 3 weeks.

Therefore, together the non- critical activities can be delayed upto a maximum of 3

weeks, without delaying the completion of the whole project.

Problem 2:

Find out the completion time and the critical activities for the following project:


D

5



2 20

A 8 G 8

B E H 11 K 6

1

3

6

8

10



10 16 I 14 L 5



C 7 J

9

F

7 10

4

25

Solution:

In all, we identify 4 paths, beginning with the start node of 1 and terminating at the end node

of 10. They are as follows:

Path I


A D G K

1

2

5

8

10



217

MBA-H2040 Quantitative Techniques
for Managers


8 20 8 6

Time for the path = 8 + 20 + 8 + 6 = 42 units of time.

Path II


B E H K


1

3

6

8

10

10 16 11 6

Time for the path = 10 + 16 + 11 + 6 = 43 units of time.

Path III


B E I L


1

3

6

9

10

10 16 14 5

Time for the path = 10 + 16 + 14 + 5 = 45 units of time.


Path IV


C F J L


1

4

7

9

10

7 25 10 5

Time for the path = 7 + 25 + 10 + 5 = 47 units of time.


Compare the times for the four paths. Maximum of {42, 43, 45, 47} = 47. We see that the

following path has the maximum time and so it is the critical path:

C F J L


1

4

7

9

10

7 25 10 5


The critical activities are C, F, J and L. The non-critical activities are A, B, D, E, G, H, I and

K. The project completion time is 47 units of time.

Problem 3:

Draw the network diagram and determine the critical path for the following project:

Activity

Time estimate (Weeks)

1- 2

5

1- 3

6



218
MBA-H2040 Quantitative Techniques
for Managers

1- 4

3

2 -5

5

3 -6

7

3 -7

10

4 -7

4

5 -8

2

6 -8

5

7 -9

6

8 -9

4

Solution: We have the following network diagram for the project:


D

5



2 5

2 H
A
5

1 B E

6 I

8 K

3

6 7 5 4

9



3 C F 10 J

6

7

G

4

4

Solution:

We assert that there are 4 paths, beginning with the start node of 1 and terminating at the end

node of 9. They are as follows:



Path I

A D H K


1

2

5

8

9

5 5 2 4

Time for the path = 5 + 5 + 2 + 4 = 16 weeks.

Path II

B E I K


1

3

6

8

9



219

MBA-H2040 Quantitative Techniques
for Managers

6 7 5 4

Time for the path = 6 + 7 + 5 + 4 = 22 weeks.

Path III

B F J


1

3

7

9

6 10 6

Time for the path = 6 + 10 + 6 = 16 weeks.

Path IV

C G J


1

4

7

9

3 4 6

Time for the path = 3 + 4 + 6 = 13 weeks.

Compare the times for the four paths. Maximum of {16, 22, 16, 13} = 22. We see that the

following path has the maximum time and so it is the critical path:


B E I K


1

3

6

8

9

6 7 5 4
The critical activities are B, E, I and K. The non-critical activities are A, C, D, F, G, H and J.

The project completion time is 22 weeks.





220
MBA-H2040 Quantitative Techniques
for Managers

QUESTIONS:

1. Explain the terms: critical path, critical activities.

2. The following are the time estimates and the precedence relationships of the activities

in a project network:

Activity

IMMEDIATE

time estimate

Predecessor

(weeks)

Activity

A

-

4

B

-

7

C

-

3

D

A

6

E

B

4

F

B

7

G

C

6

H

E

10

I

D

3

J

F, G

4

K

H, I

2


Draw the project network diagram. Determine the critical path and the project completion
time.



221

MBA-H2040 Quantitative Techniques
for Managers

LESSON 5



PERT



LESSON OUTLINE

The concept of PERT

Estimates of the time of an activity

Determination of critical path

Probability estimates



LEARNING OBJECTIVES


After reading this lesson you should be able to

- understand the importance of PERT
- locate the critical path
- determine the project completion time
- find out the probability of completion of a project before a stipulated time



INTRODUCTION

Programme Evaluation and Review Technique (PERT) is a tool that would help a project

manager in project planning and control. It would enable him in continuously monitoring a

project and taking corrective measures wherever necessary. This technique involves

statistical methods.



ASSUMPTIONS FOR PERT

Note that in CPM, the assumption is that precise time estimate is available for each activity in

a project. However, one finds most of the times that this is not practically possible.



In PERT, we assume that it is not possible to have precise time estimate for each

activity and instead, probabilistic estimates of time alone are possible. A multiple time

estimate approach is followed here. In probabilistic time estimate, the following 3 types of

estimate are possible:

1. Pessimistic time estimate ( t )

p

2. Optimistic time estimate ( t )

o

3. Most likely time estimate ( t )

m

The optimistic estimate of time is based on the assumption that an activity will not involve

any difficulty during execution and it can be completed within a short period. On the other

hand, a pessimistic estimate is made on the assumption that there would be unexpected



222
MBA-H2040 Quantitative Techniques
for Managers

problems during the execution of an activity and hence it would consume more time. The

most likely time estimate is made in between the optimistic and the pessimistic estimates of

time. Thus the three estimates of time have the relationship

t t t .

o

m

p



Practically speaking, neither the pessimistic nor the optimistic estimate may hold in

reality and it is the most likely time estimate that is expected to prevail in almost all cases.

Therefore, it is preferable to give more weight to the most likely time estimate.



We give a weight of 4 to most likely time estimate and a weight of 1 each to the

pessimistic and optimistic time estimates. We arrive at a time estimate ( t ) as the weighted

e

average of these estimates as follows:

t 4 t t

o

m

p

t



e

6

Since we have taken 6 units ( 1 for t , 4 for t and 1 for t ), we divide the sum by 6. With

p

m

o

this time estimate, we can determine the project completion time as applicable for CPM.



Since PERT involves the average of three estimates of time for each activity, this

method is very practical and the results from PERT will be have a reasonable amount of

reliability.

MEASURE OF CERTAINTY

The 3 estimates of time are such that

t t t .

o

m

p

Therefore the range for the time estimate is t t .

p

o



The time taken by an activity in a project network follows a distribution with a

standard deviation of one sixth of the range, approximately.

t t

i.e., The standard deviation =

p

o





6

2

t t

and the variance = 2

p

o





6



The certainty of the time estimate of an activity can be analysed with the help of the

variance. The greater the variance, the more uncertainty in the time estimate of an activity.

Problem 1:

Two experts A and B examined an activity and arrived at the following time estimates.

Expert

Time Estimate



223

MBA-H2040 Quantitative Techniques
for Managers

t

t

t

o

m

p

A

4

6

8

B

4

7

10



Determine which expert is more certain about his estimates of time:

Solution:

2

t t



Variance ( 2

) in time estimates = p o



6

2

8 4

4

In the case of expert A, the variance =





6

9

2

10 4

As regards expert B, the variance =

1



6

So, the variance is less in the case of A. Hence, it is concluded that the expert A is more

certain about his estimates of time.

Determination of Project Completion Time in PERT

Problem 2:

Find out the time required to complete the following project and the critical activities:


Activity

Predecessor

Optimistic time

Most likely time

Pessimistic time

Activity

estimate (to days)

estimate (tm days)

estimate (tp days)

A

-

2

4

6

B

A

3

6

9

C

A

8

10

12

D

B

9

12

15

E

C

8

9

10

F

D, E

16

21

26

G

D, E

19

22

25

H

F

2

5

8

I

G

1

3

5

Solution:


From the three time estimates t , t and t , calculate t for each activity. We obtain the following table:

p

m

o

e



Activity

Optimistic

4 x Most likely

Pessimistic

to+ 4tm+ tp

Time estimate

time estimate

time estimate

time estimate

t 4 t t

(t

o

m

p

o)

(tp)

t



e

6

A

2

16

6

24

4

B

3

24

9

36

6

C

8

40

12

60

10

D

9

48

15

72

12

E

8

36

10

54

9

F

16

84

26

126

21



224
MBA-H2040 Quantitative Techniques
for Managers

G

19

88

25

132

22

H

2

20

8

30

5

I

1

12

5

18

3


Using the single time estimates of the activities, we get the following network diagram for the project.


B

3 D F H 6

6 12 21 5
A
4

1

2

C 10 E G

5 I

8

9 22 3



4

Consider the paths, beginning with the start node and stopping wit

7 h the end node. There are

four such paths for the given project. They are as follows:

Path I

A B D F H


1

2

3

5

6

8

4 6 12 21 5
Time for the path: 4+6+12+21+5 = 48 days.


Path II

A B D G I
1

2

3

5

7

8

4 6 12 6 3

6


Time for the path: 4+6+12+ 6+3 = 31 days.

Path III

A C E F H

1

2

4

5

6

8

4 10 9 21 5

3

7


Time for the path: 4+10+9+ 21+5 = 49 days.

Path IV

A C E G I
1

2

4

5

7

8

4 10 9 6 3

6


Time for the path: 4+10+9+ 6+3 = 32 days.

Compare the times for the four paths.
Maximum of {48, 31, 49, 32} = 49.
We see that Path III has the maximum time.
Therefore the critical path is Path III. i.e., 1 2 4 5 6 8.
The critical activities are A, C, E, F and H.
The non-critical activities are B, D, G and I.
Project time (Also called project length) = 49 days.

Problem 3:

Find out the time, variance and standard deviation of the project with the following time estimates in weeks:



225

MBA-H2040 Quantitative Techniques
for Managers



Activity

Optimistic time

Most likely time

Pessimistic time

estimate (to)

estimate (tm)

estimate (tp)

1-2

3

6

9

1-6

2

5

8

2-3

6

12

18

2-4

4

5

6

3-5

8

11

14

4-5

3

7

11

6-7

3

9

15

5-8

2

4

6

7-8

8

16

18


Solution:


From the three time estimates t , t and t , calculate t for each activity. We obtain the following table:

p

m

o

e




Activity

Optimistic

4 x Most likely Pessimistic

to+ 4tm+ tp

Time estimate

time estimate

time estimate time estimate

t 4 t t

(t

o

m

p

o)

(tp)

t



e

6

1-2

3

24

9

36

6

1-6

2

20

8

30

5

2-3

6

48

18

72

12

2-4

4

20

6

30

5

3-5

8

44

14

66

11

4-5

3

28

11

42

7

6-7

3

36

15

54

9

5-8

2

16

6

24

4

7-8

8

64

18

90

15


With the single time estimates of the activities, we get the following network diagram for the project.


C

3 F

12 11



2

D 5 G I

5

A 6 7
4


4




1 5 B H



8

E 15
9

Consider the paths, beginn

6 ing with the start node and stopping with the end node. There are three such paths for

7

the given project. They are as follows:


Path I


A C F I
1

2

3

5

8



226
MBA-H2040 Quantitative Techniques
for Managers

6 12 11 4
Time for the path: 6+12+11+4 = 33 weeks.


Path II

A D G I
1

2

4

5

8

6 5 7

3 4

Time for the path: 6+5+7+ 4= 22 weeks.


Path III

B E H


6

7



1 5 9 15

8

3


Time for the path: 5+9+15 = 29 weeks.

Compare the times for the three paths.
Maximum of {33, 22, 29} = 33.
It is noticed that Path I has the maximum time.
Therefore the critical path is Path I. i.e., 1 2 3 5 8
The critical activities are A, C, F and I.
The non-critical activities are B, D, G and H.
Project time = 33 weeks.

Calculation of Standard Deviation and Variance for the Critical Activities:

Critical

Optimistic

Most likely Pessimistic

Range

Standard

Variance

Activity

time

time

time

(tp - to)

deviation =

2

t t

estimate

estimate

estimate

t t

2

p

o





p

o

(t



o)

(tm)

(tp)





6

6

A: 12

3

6

9

6

1

1

C: 23

6

12

18

12

2

4

F: 35

8

11

14

6

1

1

I: 58

2

4

6

4

2/3

4/9


Variance of project time (Also called Variance of project length) =
Sum of the variances for the critical activities = 1+4+1+ 4/9 = 58/9 Weeks.
Standard deviation of project time = Variance = 58/9 = 2.54 weeks.

Problem 4
A project consists of seven activities with the following time estimates. Find the probability that the project will
be completed in 30 weeks or less.


Activity

Predecessor

Optimistic time

Most likely time

Pessimistic time

Activity

estimate (to days)

estimate (tm days)

estimate (tp days)

A

-

2

5

8

B

A

2

3

4

C

A

6

8

10

D

A

2

4

6

E

B

2

6

10

F

C

6

7

8

G

D, E, F

6

8

10

Solution:




227

MBA-H2040 Quantitative Techniques
for Managers

From the three time estimates t , t and t , calculate t for each activity. The results are furnished in the

p

m

o

e

following table:


Activity

Optimistic

4 x Most

Pessimistic time

to+ 4tm+ tp

Time estimate

time estimate

likely time

estimate (tp)

t 4 t t

(t

o

m

p

o)

estimate

t



e

6

A

2

20

8

30

5

B

2

12

4

18

3

C

6

32

10

48

8

D

2

16

6

24

4

E

2

24

10

36

6

F

6

28

8

42

7

G

6

32

10

48

8


With the single time estimates of the activities, the following network diagram is constructed for the project.




3



B 3 6 E

C F

4

8 7
A D G
5 4 8
1

2

5

6


Consider the paths, beginning with the start node and stopping with the end node. There are three such paths for
the given project. They are as follows:


Path I

A B E G


1

3

2

5

6

5 3 6

4 8

8


Time for the path: 5+3+6+8 = 22 weeks.

Path II

A C F G
1

2

4

5

6

5 8

4 7 8

8

Time for the path: 5+8+7+ 8 = 28 weeks.


Path III

A D G
1

5

2

6

5 4 8

4


Time for the path: 5+4+8 = 17 weeks.

Compare the times for the three paths.
Maximum of {22, 28, 17} = 28.
It is noticed that Path II has the maximum time.



228
MBA-H2040 Quantitative Techniques
for Managers

Therefore the critical path is Path II. i.e., 1 2 4 5 6.
The critical activities are A, C, F and G.
The non-critical activities are B, D and E.
Project time = 28 weeks.

Calculation of Standard Deviation and Variance for the Critical Activities:


Critical

Optimistic

Most likely

Pessimistic

Range

Standard

Variance

Activity

time

time

time estimate

(tp -to)

deviation =

2

t t

estimate

estimate

(t

2

p

o

p)

t t



p

o



(t



o)

(tm)



6

6



A: 12

2

5

8

6

1

1

C: 24

6

8

10

2

4





3

9

4

F: 45

6

7

8

1

1







3

9

2

G: 56

6

8

10

2

4





4

3

9


Standard deviation of the critical path = 2 = 1.414

The standard normal variate is given by the formula


Given value of t Expected value of t in the critical path

Z



SD for the critical path



30 28

So we get Z

= 1.414

1.414

We refer to the Normal Probability Distribution Table.
Corresponding to Z = 1.414, we obtain the value of 0.4207
We get 0.5 + 0.4207 = 0. 9207
Therefore the required probability is 0.92
i.e., There is 92% chance that the project will be completed before 30 weeks. In other words, the chance that it
will be delayed beyond 30 weeks is 8%


QUESTIONS:

1. Explain how time of an activity is estimated in PERT.
2. Explain the measure of certainty in PERT.
3. The estimates of time in weeks of the activities of a project are as follows:


Activity

Predecessor

Optimistic

Most likely

Pessimistic



Activity

estimate of time

estimate of time

estimate of time

A

-

2

4

6

B

A

8

11

20

C

A

10

15

20

D

B

12

18

24



229

MBA-H2040 Quantitative Techniques
for Managers

E

C

8

13

24

F

C

4

7

16

G

D,F

14

18

28

H

E

10

12

14

I

G,H

7

10

19


Determine the critical activities and the project completion time.

4. Draw the network diagram for the following project. Determine the time, variance and standard deviation of
the project.:


Activity

Predecessor

Optimistic

Most likely

Pessimistic



Activity

estimate of time

estimate of time

estimate of time

A

-

12

14

22

B

-

16

17

24

C

A

14

15

16

D

A

13

18

23

E

B

16

18

20

F

D,E

13

14

21

G

C,F

6

8

10





230
MBA-H2040 Quantitative Techniques
for Managers

5. Consider the following project with the estimates of time in weeks:


Activity

Predecessor

Optimistic

Most likely

Pessimistic



Activity

estimate of time

estimate of time

estimate of time

A

-

2

4

6

B

-

3

5

7

C

A

5

6

13

D

A

4

8

12

E

B,C

5

6

13

F

D,E

6

8

14


Find the probability that the project will be completed in 27 weeks.



231

MBA-H2040 Quantitative Techniques
for Managers



NORMAL DISTRIBUTION TABLE


Area Under Standard Normal Distribution




0.00

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0.09

0.0

0.0000 0.0040 0.0080 0.0120 0.0160 0.0199 0.0239 0.0279 0.0319 0.0359

0.1

0.0398 0.0438 0.0478 0.0517 0.0557 0.0596 0.0636 0.0675 0.0714 0.0753

0.2

0.0793 0.0832 0.0871 0.0910 0.0948 0.0987 0.1026 0.1064 0.1103 0.1141

0.3

0.1179 0.1217 0.1255 0.1293 0.1331 0.1368 0.1406 0.1443 0.1480 0.1517

0.4

0.1554 0.1591 0.1628 0.1664 0.1700 0.1736 0.1772 0.1808 0.1844 0.1879

0.5

0.1915 0.1950 0.1985 0.2019 0.2054 0.2088 0.2123 0.2157 0.2190 0.2224

0.6

0.2257 0.2291 0.2324 0.2357 0.2389 0.2422 0.2454 0.2486 0.2517 0.2549

0.7

0.2580 0.2611 0.2642 0.2673 0.2704 0.2734 0.2764 0.2794 0.2823 0.2852

0.8

0.2881 0.2910 0.2939 0.2967 0.2995 0.3023 0.3051 0.3078 0.3106 0.3133

0.9

0.3159 0.3186 0.3212 0.3238 0.3264 0.3289 0.3315 0.3340 0.3365 0.3389

1.0

0.3413 0.3438 0.3461 0.3485 0.3508 0.3531 0.3554 0.3577 0.3599 0.3621

1.1

0.3643 0.3665 0.3686 0.3708 0.3729 0.3749 0.3770 0.3790 0.3810 0.3830

1.2

0.3849 0.3869 0.3888 0.3907 0.3925 0.3944 0.3962 0.3980 0.3997 0.4015

1.3

0.4032 0.4049 0.4066 0.4082 0.4099 0.4115 0.4131 0.4147 0.4162 0.4177

1.4

0.4192 0.4207 0.4222 0.4236 0.4251 0.4265 0.4279 0.4292 0.4306 0.4319

1.5

0.4332 0.4345 0.4357 0.4370 0.4382 0.4394 0.4406 0.4418 0.4429 0.4441

1.6

0.4452 0.4463 0.4474 0.4484 0.4495 0.4505 0.4515 0.4525 0.4535 0.4545

1.7

0.4554 0.4564 0.4573 0.4582 0.4591 0.4599 0.4608 0.4616 0.4625 0.4633

1.8

0.4641 0.4649 0.4656 0.4664 0.4671 0.4678 0.4686 0.4693 0.4699 0.4706

1.9

0.4713 0.4719 0.4726 0.4732 0.4738 0.4744 0.4750 0.4756 0.4761 0.4767

2.0

0.4772 0.4778 0.4783 0.4788 0.4793 0.4798 0.4803 0.4808 0.4812 0.4817

2.1

0.4821 0.4826 0.4830 0.4834 0.4838 0.4842 0.4846 0.4850 0.4854 0.4857

2.2

0.4861 0.4864 0.4868 0.4871 0.4875 0.4878 0.4881 0.4884 0.4887 0.4890

2.3

0.4893 0.4896 0.4898 0.4901 0.4904 0.4906 0.4909 0.4911 0.4913 0.4916

2.4

0.4918 0.4920 0.4922 0.4925 0.4927 0.4929 0.4931 0.4932 0.4934 0.4936

2.5

0.4938 0.4940 0.4941 0.4943 0.4945 0.4946 0.4948 0.4949 0.4951 0.4952

2.6

0.4953 0.4955 0.4956 0.4957 0.4959 0.4960 0.4961 0.4962 0.4963 0.4964

2.7

0.4965 0.4966 0.4967 0.4968 0.4969 0.4970 0.4971 0.4972 0.4973 0.4974

2.8

0.4974 0.4975 0.4976 0.4977 0.4977 0.4978 0.4979 0.4979 0.4980 0.4981

2.9

0.4981 0.4982 0.4982 0.4983 0.4984 0.4984 0.4985 0.4985 0.4986 0.4986

3.0

0.4987 0.4987 0.4987 0.4988 0.4988 0.4989 0.4989 0.4989 0.4990 0.4990




















LESSON 6



EARLIEST AND LATEST TIMES



232
MBA-H2040 Quantitative Techniques
for Managers



LESSON OUTLINE

The concepts of earliest and latest times

The concept of slack

Numerical problems



LEARNING OBJECTIVES


After reading this lesson you should be able to

-

understand the concepts of earliest and latest times

- understand the concept of slack

- calculate the earliest and latest times

-

find out the slacks

-

identify the critical activities

-

carry out numerical problems


INTRODUCTION

A project manager has the responsibility to see that a project is completed by the stipulated date, without delay.

Attention is focused on this aspect in what follows.


Key concepts

Certain key concepts are introduced below.

EARLIEST TIMES OF AN ACTIVITY

We can consider (i) Earliest Start Time of an activity and (ii) Earliest Finish Time of an activity.

Earliest Start Time of an activity is the earliest possible time of starting that activity on

the condition that all the other activities preceding to it were began at the earliest possible

times.

Earliest Finish Time of an activity is the earliest possible time of completing that activity. It is given by

the formula.

The Earliest Finish Time of an activity = The Earliest Start Time of the activity + The estimated

duration to carry out that activity.

LATEST TIMES OF AN ACTIVITY



We can consider (i) Latest Finish Time of an activity and (ii) Latest Start Time of an activity.

Latest Finish Time of an activity is the latest possible time of completing that activity

on the condition that all the other activities succeeding it are carried out as per the plan of the

management and without delaying the project beyond the stipulated time.

Latest Start Time of an activity is the latest possible time of beginning that activity. It is given by the

formula

Latest Start Time of an activity = The Latest Finish Time of the activity - The estimated
duration to carry out that activity.



233

MBA-H2040 Quantitative Techniques
for Managers


TOTAL FLOAT OF AN ACTIVITY




Float seeks to measure how much delay is acceptable. It sets up a control limit for delay.



The total float of an activity is the time by which that activity can be delayed without delaying the

whole project. It is given by the formula

Total Float of an Activity = Latest Finish Time of the activity - Earliest Finish Time of that activity.

It is also given by the formula

Total Float of an Activity = Latest Start Time of the activity - Earliest Start Time of that
activity.

Since a delay in a critical activity will delay the execution of the whole project, the total

float of a critical activity must be zero.

EXPECTED TIMES OF AN EVENT

An event occurs at a point of time. We can consider (i) Earliest Expected Time of Occurrence of an event and

(ii) Latest Allowable Time of Occurrence an event.

The Earliest Expected Time of Occurrence of an event is the earliest possible time of

expecting that event to happen on the condition that all the preceding activities have been

completed.

The Latest Allowable Time of Occurrence of an event is the latest possible time of

expecting that event to happen without delaying the project beyond the stipulated time.

PROCUDURE TO FIND THE EARLIEST EXPECTED TIME OF AN EVENT



Step 1. Take the Earliest Expected Time of Occurrence of the Start Event as zero.



Step 2. For an event other than the Start Event, find out all paths in the network which
connect the Start node with the node representing the event under consideration.

Step 3. In the "Forward Pass" (i.e., movement in the network from left to right), find out
the sum of the time durations of the activities in each path identified in Step 2.

Step 4. The path with the longest time in Step 3 gives the Earliest Expected Time of
Occurrence of the event



Working Rule for finding the earliest expected time of an event:





For an event under consideration, locate all the predecessor events and identify their earliest

expected times. With the earliest expected time of each event, add the time duration of the

activity connecting that event to the event under consideration. The maximum among all these

values gives the Earliest Expected Time of Occurrence of the event.





234
MBA-H2040 Quantitative Techniques
for Managers

PROCUDURE TO FIND THE LATEST ALLOWABLE TIME OF AN EVENT


We consider the "Backward Pass" (i.e., movement in the network from right to left).

The latest allowable time of occurrence of the End Node must be the time of completion

of the project. Therefore it shall be equal to the time of the critical path of the project.



Step 1. Identify the latest allowable time of occurrence of the End Node.



Step 2. For an event other than the End Event, find out all paths in the network which
connect the End node with the node representing the event under consideration.



Step 3. In the "Backward Pass" (i.e., movement in the network from right to left),
subtract the time durations of the activities along each such path.



Step 4. The Latest Allowable Time of Occurrence of the event is determined by the path
with the longest time in Step 3. In other words, the smallest value of time obtained in
Step 3 gives the Latest Allowable Time of Occurrence of the event.


Working Rule for finding the latest allowable time of an event:





For an event under consideration, locate all the successor events and identify their latest

allowable times. From the latest allowable time of each successor event, subtract the time

duration of the activity that begins with the event under consideration. The minimum among

all these values gives the Latest Allowable Time of Occurrence of the event.

SLACK OF AN EVENT

The allowable time gap for the occurrence of an event is known as the slack of that event. It is given by the

formula

Slack of an event = Latest Allowable Time of Occurrence of the event - Earliest Expected

Time of Occurrence of that event.

SLACK OF AN ACTIVITY

The slack of an activity is the float of the activity.
Problem 1:

The following details are available regarding a project:

Activity

Predecessor

Duration (Weeks)

Activity

A

-

12



235

MBA-H2040 Quantitative Techniques
for Managers

B

A

7

C

A

11

D

A

8

E

A

6

F

B

10

G

C

9

H

D, F

14

I

E, G

13

J

H, I

16

Determine the earliest and latest times, the total float for each activity, the critical activities and the project
completion time.

Solution:

With the given data, we construct the following network diagram for the project.





3 F

10
B 7 H
D 14

5

A 8 J

12 E 16

1

2

7

8

6 I

C 11 13
G

6

9


Consider the paths, beginning

4 with the start node and stopping with the end node. There are

four such paths for the given project. They are as follows:



Path I


A B F H J

1

2

3

5

7

1

12 7 10 14 16

Time of the path = 12 + 7 + 10 + 14 + 16 = 59 weeks.

Path II


A D H J


1

2

5

7

8



236
MBA-H2040 Quantitative Techniques
for Managers


12 8 14 16

Time of the path = 12 + 8 + 14 + 16 = 50 weeks.




Path III


A E I J

1

2

6

7

8

12 6 13 16

Time of the path = 12 + 6 + 13 + 16 = 47 weeks.


Path IV


A C G I J
1

2

4

6

7

8


12 11 9 13 16

Time of the path = 12 + 11 + 9 + 13 + 16 = 61 weeks.

Compare the times for the four paths. Maximum of {51, 50, 47, 61} = 61. We see that the maximum time of a
path is 61 weeks.

Forward pass:

Calculation of Earliest Expected Time of Occurrence of Events


Node Earliest Time of Occurrence of Node
1

0

2

Time for Node 1 + Time for Activity A = 0 + 12 = 12

3

Time for Node 2 + Time for Activity B = 12 + 7 = 19

4

Time for Node 2 + Time for Activity C = 12 + 11 = 23

5

Max {Time for Node 2 + Time for Activity D,
Time for Node 3 + Time for Activity F}
= Max {12 + 8, 19 + 10} = Max {20, 29} = 29

6

Max {Time for Node 2 + Time for Activity E,
Time for Node 4 + Time for Activity G}
= Max {12 + 6, 23 + 9} = Max {18, 32} = 32

7

Max {Time for Node 5 + Time for Activity H,
Time for Node 6 + Time for Activity I}
= Max {29 + 14, 32 + 13} = Max {43, 45} = 45

8

Time for Node 7 + Time for Activity J = 45 + 16 = 61


Using the above values, we obtain the Earliest Start Times of the activities as follows:





Activity

Earliest Start Time



(Weeks)



237

MBA-H2040 Quantitative Techniques
for Managers

A

0

B

12

C

12

D

12

E

12

F

19

G

23

H

29

I

32

J

45


Backward pass:

Calculation of Latest Allowable Time of Occurrence of Events


Node Latest Allowable Time of Occurrence of Node
8

Maximum time of a path in the network = 61

7

Time for Node 8 - Time for Activity J = 61 -16 = 45

6

Time for Node 7 - Time for Activity I = 45 -13 = 32

5

Time for Node 7 - Time for Activity H = 45 -14 = 31

4

Time for Node 6 - Time for Activity G = 32 - 9 = 23

3

Time for Node 5 - Time for Activity F = 31- 10 = 21

2

Min {Time for Node 3 - Time for Activity B,
Time for Node 4 - Time for Activity C,
Time for Node 5 - Time for Activity D,
Time for Node 6 - Time for Activity E}
= Min {21 - 7, 23 - 11, 31 - 8, 32 - 6}
= Min {14, 12, 23, 26} = 12

1

Time for Node 2 - Time for Activity A = 12- 12 = 0


Using the above values, we obtain the Latest Finish Times of the activities as follows:









Activity

Latest Finish Time



(Weeks)

J

61

I

45

H

45

G

32

F

31



238
MBA-H2040 Quantitative Techniques
for Managers

E

32

D

31

C

23

B

21

A

12


Calculation of Total Float for each activity:

Activity

Duration

Earliest Start Earliest

Latest

Latest

Total Float = Latest

(Weeks)

Time

Finish Time

Start Time Finish

Finish Time - Earliest

Time

Finish Time

A

12

0

12

0

12

0

B

7

12

19

14

21

2

C

11

12

23

12

23

0

D

8

12

20

23

31

11

E

6

12

18

26

32

14

F

10

19

29

21

31

2

G

9

23

32

23

32

0

H

14

29

43

31

45

2

I

13

32

45

32

45

0

J

16

45

61

45

61

0


The activities with total float = 0 are A, C, G, I and J. They are the critical activities.
Project completion time = 61 weeks.






Problem 2:


The following are the details of the activities in a project:

Activity

Predecessor

Duration (Weeks)



Activity


A

-

15

B

A

17

C

A

21

D

B

19

E

B

22



239

MBA-H2040 Quantitative Techniques
for Managers

F

C, D

18

G

E, F

15


Calculate the earliest and latest times, the total float for each activity and the project completion time.


Solution:

The following network diagram is obtained for the given project.


E

3

22
B
A 17 D 19 G



1 15

2 15

5

6


C 21 F 18





4

Consider the paths, beginning with the start node and stopping with the end node. There are

three such paths for the given project. They are as follows:



Path I

A B E G



1

2

3

5

6

15 17 22 15
Time of the path = 15 + 17 + 22 + 15 = 69 weeks.

Path II

A B D F G



1

2

3

4

5

6

15 17 19 18 15

Time of the path = 15 + 17 + 19 + 18 + 15 = 84 weeks.


Path III


A C F G


1

2

4

5

6

15 21 18 15

Time of the path = 15 + 21 + 18 + 15 = 69 weeks.

Compare the times for the three paths. Maximum of {69, 84, 69} = 84. We see that the maximum time of a path

is 84 weeks.



240
MBA-H2040 Quantitative Techniques
for Managers

Forward pass:

Calculation of Earliest Time of Occurrence of Events


Node

Earliest Time of Occurrence of Node

1

0

2

Time for Node 1 + Time for Activity A = 0 + 15 = 15

3

Time for Node 2 + Time for Activity B = 15 + 17 = 32

4

Max {Time for Node 2 + Time for Activity C,
Time for Node 3 + Time for Activity D}
= Max {15 + 21, 32 + 19} = Max {36, 51} = 51

5

Max {Time for Node 3 + Time for Activity E,
Time for Node 4 + Time for Activity F}
= Max {32 + 22, 51 + 18} = Max {54, 69} = 69

6

Time for Node 5 + Time for Activity G = 69 + 15 = 84






Calculation of Earliest Time for Activities



Activity

Earliest Start Time



(Weeks)

A

0

B

15

C

15

D

32

E

32

F

51

G

69


Backward pass:
Calculation of the Latest Allowable Time of Occurrence of Events


Node

Latest Allowable Time of Occurrence of Node

6

Maximum time of a path in the network = 84

5

Time for Node 6 - Time for Activity G = 84 -15 = 69

4

Time for Node 5 - Time for Activity F = 69 - 18 = 51

3

Min {Time for Node 4 - Time for Activity D,
Time for Node 5 - Time for Activity E}
= Min {51 - 19, 69 - 22} = Min {32, 47} = 32

2

Min {Time for Node 3 - Time for Activity B,
Time for Node 4 - Time for Activity C}
= Min {32 - 17, 51 - 21} = Min {15, 30} = 15

1

Time for Node 2 - Time for Activity A = 15 - 15 = 0


Calculation of the Latest Finish Times of the activities

Activity

Latest Finish Time (Weeks)

G

84



241

MBA-H2040 Quantitative Techniques
for Managers

F

69

E

69

D

51

C

51

B

32

A

15



Calculation of Total Float for each activity:

Activity

Duration

Earliest Start Earliest

Latest

Latest

Total Float = Latest

(Weeks)

Time

Finish Time

Start Time Finish

Finish Time - Earliest

Time

Finish Time

A

15

0

15

0

15

0

B

17

15

32

15

32

0

C

21

15

36

30

51

15

D

19

32

51

32

51

0

E

22

32

54

47

69

15

F

18

51

69

51

69

0

G

15

69

84

69

84

0


The activities with total float = 0 are A, B, D, F and G. They are the critical activities.
Project completion time = 84 weeks.


Problem 3:

Consider a project with the following details:


Name of

Predecessor

Duration

Activity

Activity

(Weeks)

A

-

8

B

A

13

C

A

9

D

A

12

E

B

14

F

B

8

G

D

7

H

C, F, G

12

I

C, F, G

9

J

E, H

10

K

I, J

7


Determine the earliest and latest times, the total float for each activity, the critical activities, the slacks of the
events and the project completion time.




242
MBA-H2040 Quantitative Techniques
for Managers

Solution:

The following network diagram is got for the given project:

E


3

6

14
B 13 H 12
8 J 10
F
A C I K
9 9 7
8

1

2

5

7

8

D 12 G 7






4

Path I


A B E J K

1

2

3

6

7

8

8 13 14 10 7

Time of the path = 8 + 13 + 14 + 10 + 7 = 52 weeks.


Path II



A B F H J K
1

2

3

5

6

7

8

8 13 8 12 10 7

Time of the path = 8 + 13 + 8 + 12 + 10 + 7 = 58 weeks.

Path III



A B F I K

1

2

3

5

7

8

8 13 8 9 7

Time of the path = 8 + 13 + 8 + 9 + 7 = 45 weeks.
Path IV



A C H J K

1

2

5

6

7

8

8 9 12 10

5 7


Time of the path = 8 + 9 + 12 + 10 + 7 = 46 weeks.



243

MBA-H2040 Quantitative Techniques
for Managers



Path V


A C I K


1

2

5

7

8

8 9 9 7

5

6


Time of the path = 8 + 9 + 9 + 7 = 33 weeks.

Path VI


A D G H J K
1

2

4

5

6

7

8

8 12 7 12 10 7
Time of the path = 8 + 12 + 7 + 12 + 10 + 7 = 56 weeks.


Path VII


A D G I K

1

2

4

5

7

8

8 12 7 9 7

Time of the path = 8 + 12 + 7 + 9 + 7 = 43 weeks.

Compare the times for the three paths. Maximum of {52, 58, 45, 46, 33, 56, 43} = 58.
We see that the maximum time of a path is 58 weeks.




244
MBA-H2040 Quantitative Techniques
for Managers

Forward pass:

Calculation of Earliest Time of Occurrence of Events


Node Earliest Time of Occurrence of Node
1

0

2

Time for Node 1 + Time for Activity A = 0 + 8 = 8

3

Time for Node 2 + Time for Activity B = 8 + 13 = 21

4

Time for Node 2 + Time for Activity D = 8+ 12 = 20

5

Max {Time for Node 2 + Time for Activity C,
Time for Node 3 + Time for Activity F,
Time for Node 4 + Time for Activity G}
= Max { 8 + 9, 21 + 8 , 20 + 7 } = Max {17, 29, 27} = 29

6

Max {Time for Node 3 + Time for Activity E,
Time for Node 5 + Time for Activity H}
= Max {21 + 14 , 29 + 12} = Max {35, 41} = 41

7

Max {Time for Node 5 + Time for Activity I,
Time for Node 6 + Time for Activity J}
= Max {29 + 9, 41 + 10} = Max {38, 51} = 51

8

Time for Node 7 + Time for Activity J = 51+ 7 = 58


Earliest Start Times of the activities


Activity

Earliest Start Time



(Weeks)

A

0

B

8

C

8

D

8

E

21

F

21

G

20

H

29

I

29

J

41

K

51






Backward pass:

Calculation of Latest Allowable Time of Occurrence of Events


Node Latest Allowable Time of Occurrence of Node
8

Maximum time of a path in the network = 58

7

Time for Node 8 - Time for Activity K = 58 -7 = 51

6

Time for Node 7 - Time for Activity J = 51 -10 = 41



245

MBA-H2040 Quantitative Techniques
for Managers

5

Min {Time for Node 6 - Time for Activity H,
Time for Node 7 - Time for Activity I}
= Min {41 - 12, 51 - 9} = Min {29, 42} = 29

4

Time for Node 5 - Time for Activity G = 29 - 7 = 22

3

Min {Time for Node 5 - Time for Activity F,
Time for Node 6 - Time for Activity E}
= Min {29 - 8, 41 - 14} = Min {21, 27} = 21

2

Min {Time for Node 3 - Time for Activity B,
Time for Node 4 - Time for Activity D,
Time for Node 5 - Time for Activity C}
= Min {21 - 13, 22 - 12, 29 - 9}
= Min {8, 10, 20} = 8

1

Time for Node 2 - Time for Activity A = 8 - 8 = 0


Latest Finish Times of the activities


Activity

Latest Finish Time



(Weeks)

K

58

J

51

I

51

H

41

G

29

F

29

E

41

D

22

C

29

B

21

A

8





Calculation of Total Float for each activity:

Activity

Duration

Earliest Start Earliest

Latest

Latest

Total Float = Latest

(Weeks)

Time

Finish Time

Start Time Finish

Finish Time - Earliest

Time

Finish Time

A

8

0

8

0

8

0

B

13

8

21

8

21

0

C

9

8

17

20

29

12

D

12

8

20

10

22

2

E

14

21

35

27

41

6

F

8

21

29

21

29

0



246
MBA-H2040 Quantitative Techniques
for Managers

G

7

20

27

22

29

2

H

12

29

41

29

41

0

I

9

29

38

42

51

13

J

10

41

51

41

51

0

K

7

51

58

51

58

0


The activities with total float = 0 are A, B, F, H, J and K. They are the critical activities.
Project completion time = 58 weeks.

Calculation of slacks of the events


Slack of an event = Latest Allowable Time of Occurrence of the event - Earliest Expected
Time of Occurrence of that event.


Event

Earliest Expected Time

Latest Allowable Time

Slack of the

(Node)

of Occurrence of Event

of Occurrence of Event

Event

1

0

0

0

2

8

8

0

3

21

21

0

4

20

22

2

5

29

29

0

6

41

41

0

7

51

51

0

8

58

58

0






Interpretation:

On the basis of the slacks of the events, it is concluded that the occurrence of event 4 may be delayed upto a

maximum period of 2 weeks while no other event cannot be delayed.

QUESTIONS

1. Explain the terms: The earliest and latest times of the activities of a project.

2. Explain the procedure to find the earliest expected time of an event.

3. Explain the procedure to find the latest allowable time of an event.

4. What is meant by the slack of an activity? How will you determine it?

5. Consider the project with the following details:



activity

Duration (weeks)

12

1

23

3



247

MBA-H2040 Quantitative Techniques
for Managers

24

7

34

5

35

8

45

4

56

1



Determine the earliest and the latest times of the activities. Calculate the total float for

each activity and the slacks of the events.



248
MBA-H2040 Quantitative Techniques
for Managers

LESSON 7



CRASHING OF A PROJECT

LESSON OUTLINE

The idea of crashing of a project

The criterion of selection of an activity for crashing

Numerical problems



LEARNING OBJECTIVES


After reading this lesson you should be able to

- understand the concept of crashing of a project
- choose an activity for crashing

- work out numerical problems

THE MEANING OF CRASHING:

The process of shortening the time to complete a project is called crashing and is usually achieved by putting

into service additional labour or machines to one activity or more activities. Crashing involves more costs. A

project manager would like to speed up a project by spending as minimum extra cost as possible. Project

crashing seeks to minimize the extra cost for completion of a project before the stipulated time.

STEPS IN PROJECT CRASHING:

Assumption: It is assumed that there is a linear relationship between time and cost.

Let us consider project crashing by the critical path method. The following four-step

procedure is adopted.

Step 1: Find the critical path with the normal times and normal costs for the activities and

identify the critical activities.

Step 2: Find out the crash cost per unit time for each activity in the network. This is

calculated by means of the following formula.

Crash cos t

Crash cos t Normal cos t



Time period

Normal time Crash time



249

MBA-H2040 Quantitative Techniques
for Managers





Activity Cost





Crash time & Cost
Crash Cost





Normal Cost Normal time & Cost


Activity Time
Crash Time Normal Time


Step 3: Select an activity for crashing. The criteria for the selection is as follows:

Select the activity on the critical path with the smallest crash cost per unit time. Crash this

activity to the maximum units of time as may be permissible by the given data.



Crashing an activity requires extra amount to be spent. However, even if the company

is prepared to spend extra money, the activity time cannot be reduced beyond a certain limit in

view of several other factors.

In step 1, we have to note that reducing the time of on activity along the critical path

alone will reduce the completion time of a project. Because of this reason, we select an

activity along the critical path for crashing.

In step 3, we have to consider the following question:



If we want to reduce the project completion time by one unit, which critical activity

will involve the least additional cost?



On the basis of the least additional cost, a critical activity is chosen for crashing. If

there is a tie between two critical activities, the tie can be resolved arbitrarily.



Step 4: After crashing an activity, find out which is the critical path with the changed

conditions. Sometimes, a reduction in the time of an activity in the critical path may cause a

non-critical path to become critical. If the critical path with which we started is still the

longest path, then go to Step 3. Otherwise, determine the new critical path and then go to

Step 3.



Problem 1: A project has activities with the following normal and crash times and cost:



250
MBA-H2040 Quantitative Techniques
for Managers

Activity

Predecessor

Normal Time

Crash Time

Normal Cost Crash Cost



Activity

(Weeks)

(Weeks) (Rs.)

(Rs.)

A

-

4

3

8,000

9,000

B

A

5

3

16,000

20,000

C

A

4

3

12,000

13,000

D

B

6

5

34,000

35,000

E

C

6

4

42,000

44,000

F

D

5

4

16,000

16,500

G

E

7

4

66,000

72,000

H

G

4

3

2,000

5,000


Determine a crashing scheme for the above project so that the total project time is reduced by 3 weeks.

Solution:

We have the following network diagram for the given project with normal costs:


E G
C 6

3 7

5 H

7

A 4

4 4 4 3


4

2

1

5

B

8

D F

5

5
6

4

6



4


Beginning from the Start Node and terminating with the End Node, there are two paths for the network as
detailed below:




Path I:

A B D F
1

2

5

7

8

4 5 6 5

The time for the path = 4 + 5 + 6 + 5 = 20 weeks.

Path II:

A C E G H
1

2

3

5

7

8

4 4 6 7 4

The time for the path = 4 + 4 + 6 + 7 + 4 = 25 weeks.

Maximum of {20, 25} = 25.




251

MBA-H2040 Quantitative Techniques
for Managers

Therefore Path II is the critical path and the critical activities are A, C, E, G and H. The non-critical activities are
B, D and F.


Given that the normal time of activity A is 4 weeks while its crash time is 3 weeks. Hence the time of

this activity can be reduced by one week if the management is prepared to spend an additional amount. However,
the time cannot be reduced by more than one week even if the management may be prepared to spend more
money. The normal cost of this activity is Rs. 8,000 whereas the crash cost is Rs. 9,000. From this, we see that
crashing of activity A by one week will cost the management an extra amount of Rs. 1,000. In a similar fashion,
we can work out the crash cost per unit time for the other activities also. The results are provided in the
following table.



Activity

Normal

Crash

Normal

Crash Cost

Crash cost

Normal

Crash Cost



Time

Time Cost



-

Time -

per unit

Normal

Crash

time

Cost

Time

A

4

3

8,000

9,000

1,000

1

1,000

B

5

3

16,000

20,000

4,000

2

2,000

C

4

3

12,000

13,000

1,000

1

1,000

D

6

5

34,000

35,000

1,000

1

1,000

E

6

4

42,000

44,000

2,000

2

1,000

F

5

4

16,000

16,500

500

1

500

G

7

4

66,000

72,000

6,000

1

6,000

H

4

3

2,000

5,000

3,000

1

3,000



A non-critical activity can be delayed without delaying the execution of the whole project. But, if a

critical activity is delayed, it will delay the whole project. Because of this reason, we have to select a critical

activity for crashing. Here we have to choose one of the activities A, C, E, G and H The crash cost per unit time

works out as follows:


Rs. 1,000 for A; Rs. 1,000 for C; Rs. 1,000 for E; Rs. 6,000 for G; Rs. 3,000 for H.

The maximum among them is Rs. 1,000. So we have to choose an activity with Rs.1,000 as the crash cost

per unit time. However, there is a tie among A, C and E. The tie can be resolved arbitrarily. Let us select A for

crashing. We reduce the time of A by one week by spending an extra amount of Rs. 1,000.

After this step, we have the following network with the revised times for the activities:


E G
C 6

3 7

5 H

7

A 4

4 4 4 3


3

2

1

5

B

8

D F

5

5
6

4

6



4

The revised time for Path I = 3 + 5 + 6 + 5 = 19 weeks.
The time for Path II = 3 + 4 + 6 + 7 + 4 = 24 weeks.
Maximum of {19, 24} = 24.




252
MBA-H2040 Quantitative Techniques
for Managers

Therefore Path II is the critical path and the critical activities are A, C, E, G and H. However, the time for A

cannot be reduced further. Therefore, we have to consider C, E, G and H for crashing. Among them, C and E

have the least crash cost per unit time. The tie between C and E can be resolved arbitrarily. Suppose we reduce

the time of C by one week with an extra cost of Rs. 1,000.

After this step, we have the following network with the revised times for the activities:



253

MBA-H2040 Quantitative Techniques
for Managers


E G
C 6

3 7

5 H

7

A 3

4 4 4 3


3

2

1

5

B

8

D F

5

5
6

4

6



4

The time for Path I = 3 + 5 + 6 + 5 = 19 weeks.
The time for Path II = 3 + 3 + 6 + 7 + 4 = 23 weeks.
Maximum of {19, 23} = 23.
Therefore Path II is the critical path and the critical activities are A, C, E, G and H. Now the time for A or C

cannot be reduced further. Therefore, we have to consider E, G and H for crashing. Among them, E has the least

crash cost per unit time. Hence we reduce the time of E by one week with an extra cost of Rs. 1,000.

By the given condition, we have to reduce the project time by 3 weeks. Since this has been accomplished, we
stop with this step.

Result: We have arrived at the following crashing scheme for the given project:

Reduce the time of A, C and E by one week each.
Project time after crashing is 22 weeks.
Extra amount required = 1,000 + 1,000 + 1,000 = Rs. 3,000.

Problem 2:

The management of a company is interested in crashing of the following project by spending an additional

amount not exceeding Rs. 2,000. Suggest how this can be accomplished.


Activity

Predecessor

Normal Time

Crash Time

Normal Cost Crash Cost

Activity

(Weeks)

(Weeks) (Rs.)

(Rs.)

A

-

7

6

15,000

18,000

B

A

12

9

11,000

14,000

C

A

22

21

18,500

19,000

D

B

11

10

8,000

9,000

E

C, D

6

5

4,000

4,500



Solution:

We have the following network diagram for the given project with normal costs:





3

B D 11
12
A C E


1

2

4

5

7 22 6



254
MBA-H2040 Quantitative Techniques
for Managers



There are two paths for this project as detailed below:

Path I:

A B D E



1 7 2 12 11

3 6 4

5


The time for the path = 7 + 12 + 11 + 6 = 36 weeks.

Path II:

A C E


1

4

2

5

7 22 6

The time for the path = 7 + 22 + 6 = 35 weeks.

Maximum of {36, 35} = 36.

Therefore Path I is the critical path and the critical activities are A, B, D and E. The non-critical activity is C.

The crash cost per unit time for the activities in the project are provided in the following table.



255

MBA-H2040 Quantitative Techniques
for Managers


Activity

Normal

Crash

Normal

Crash Cost

Crash cost

Normal

Crash Cost



Time

Time Cost



-

Time -

per unit

Normal

Crash

time

Cost

Time

A

7

6

15,000

18,000

3,000

1

3,000

B

12

9

11,000

14,000

3,000

3

1,000

C

22

21

18,500

19,000

500

1

500

D

11

10

8,000

9,000

1,000

1

1,000

E

6

5

4,000

4,500

500

1

500




We have to choose one of the activities A, B, D and E for crashing. The crash cost per unit time is as

follows:

Rs. 3,000 for A; Rs. 1,000 for B; Rs. 1,000 for D; Rs. 500 for E.

The least among them is Rs. 500. So we have to choose the activity E for crashing. We reduce the time of E by

one week by spending an extra amount of Rs. 500.


After this step, we have the following network with the revised times for the activities:



B D

3 11

12
A C E


1

2

4

5

7 22 5


The revised time for Path I = 7 + 12 + 11 + 5 = 35 weeks.

The time for Path II = 7 + 22 + 5 = 34 weeks.

Maximum of {35, 34} = 35.

Therefore Path I is the critical path and the critical activities are A, B, D and E. The non-critical activity is C.

The time of E cannot be reduced further. So we cannot select it for crashing. Next B and have the smallest crash

cost per unit time. Let us select B for crashing. Let us reduce the time of E by one week at an extra cost of Rs.

1,000.

After this step, we have the following network with the revised times for the activities:





3

B D 11
11
A C E


1

2

4

5

7 22 5

The revised time for Path I = 7 + 11 + 11 + 5 = 34 weeks.




256
MBA-H2040 Quantitative Techniques
for Managers

The time for Path II = 7 + 22 + 5 = 34 weeks.

Maximum of {34, 34} = 34.

Since both paths have equal times, both are critical paths. So, we can choose an activity for crashing from either

of them depending on the least crash cost per unit time. In path I, the activities are A, B, D and E. In path II, the

activities are A, C and E.

The crash cost per unit time is the least for activity C. So we select C for crashing. Reduce the time of C by one

week at an extra cost of Rs. 500.

By the given condition, the extra amount cannot exceed Rs. 2,000. Since this state has been met, we stop with
this step.

Result: The following crashing scheme is suggested for the given project:
Reduce the time of E, B and C by one week each.

Project time after crashing is 33 weeks.

Extra amount required = 500 + 1,000 + 500 = Rs. 2,000.





257

MBA-H2040 Quantitative Techniques
for Managers

Problem 3:

The manager of a company wants to apply crashing for the following project by spending an additional amount

not exceeding Rs. 2,000. Offer your suggestion to the manager.


Activity

Predecessor

Normal Time

Crash Time

Normal Cost Crash Cost



Activity

(Weeks)

(Weeks) (Rs.)

(Rs.)

A

-

20

19

8,000

10,000

B

-

15

14

16,000

19,000

C

A

22

20

13,000

14,000

D

A

17

15

7,500

9,000

E

B

19

18

4,000

5,000

F

C

28

27

3,000

4,000

G

D, E

25

24

12,000

13,000


Solution:
We have the following network diagram for the given project with normal costs:

C


3

5

A 22 28
20 F
17

1

D
15

7

B E G 25

19


2

4

There are three paths for this project as detailed below:

Path I:
A C F


2

4



1 20 22 28

6


The time for the path = 20 + 22 + 28= 70 weeks.




258
MBA-H2040 Quantitative Techniques
for Managers

Path II:

A D G


1

2

5

6

20 17 25

The time for the path = 20 + 17 + 25= 62 weeks.

Path III:

B E G


1

3

5

6

15 19 25

The time for the path = 15+19 +25 = 69 weeks.
Maximum of {70, 62, 69} = 70.
Therefore Path I is the critical path and the critical activities are A, C and F. The non-critical activities are B, D,
E and G.
The crash cost per unit time for the activities in the project are provided in the following table
Activity

Normal

Crash

Normal

Crash Cost

Crash cost

Normal

Crash Cost



Time

Time Cost



-

Time -

per unit

Normal

Crash

time

Cost

Time

A

20

19

8,000

10,000

2,000

1

2,000

B

15

14

16,000

19,000

3,000

1

3,000

C

22

20

13,000

14,000

1,000

2

500

D

17

15

7,500

9,000

1,500

2

750

E

19

18

4,000

5,000

1,000

1

1,000

F

28

27

3,000

4,000

1,000

1

1,000

G

25

24

12,000

13,000

1,000

1

1,000




We have to choose one of the activities A, C and F for crashing. The crash cost per unit time is as

follows:

Rs. 2,000 for A; Rs. 500 for C; Rs. 1,000 for F.

The least among them is Rs. 500. So we have to choose the activity C for crashing. We reduce the time of C by
one week by spending an extra amount of Rs. 500.


After this step, we have the following network with the revised times for the activities:



C


3

5

A 21 28
20 F
17

1

D
15

7

B E G 25

19


2

4



259

MBA-H2040 Quantitative Techniques
for Managers

The revised time for Path I = 20 + 21 + 28= 69 weeks.

The time for Path II = 20 + 17 + 25= 62 weeks.

The time for Path III = 15+19 +25 = 69 weeks.
Maximum of {69, 62, 69} = 69.

Since paths I and III have equal times, both are critical paths. So, we can choose an activity for crashing from

either of them depending on the least crash cost per unit time.

In path I, the activities are A, C and F. In path III, the activities are B, E and G.

The crash cost per unit time is the least for activity C. So we select C for crashing. Reduce the time of C by one

week at an extra cost of Rs. 500.

After this step, we have the following network with the revised times for the activities:

C


3

5

A 20 28
20 F
17

1

D
15

7

B E G 25

19


2

4

The revised time for Path I = 20 + 20 + 28= 68 weeks.
The time for Path II = 20 + 17 + 25= 62 weeks.

The time for Path III = 15+19 +25 = 69 weeks.
Maximum of {68, 62, 69} = 69.

Therefore path III is the critical activities. Hence we have to select an activity from Path III for crashing. We see

that the crash cost per unit time is as follows:


Rs. 3,000 for B; Rs. 1,000 for E; Rs. 1,000 for G.

The least among them is Rs. 1,000. So we can select either E or G for crashing. Let us select E for crashing. We

reduce the time of E by one week by spending an extra amount of Rs. 1,000.

By the given condition, the extra amount cannot exceed Rs. 2,000. Since this condition has been reached, we

stop with this step.


Result: The following crashing scheme is suggested for the given project:

Reduce the time of C by 2 weeks and that of E by one week.

Project time after crashing is 67 weeks.

Extra amount required = 2 x 500 + 1,000 = Rs. 2,000.


QUESTIONS


1. Explain the concept of crashing of a project.
2. Explain the criterion for selection of an activity for crashing.







260
MBA-H2040 Quantitative Techniques
for Managers





261

MBA-H2040 Quantitative Techniques
for Managers

UNIT IV

GAME THEORY & GOAL PROGRAMMING



LESSON 1 BASIC CONCEPTS IN GAME THEORY



LESSON OUTLINE

Introduction to the theory of games
The definition of a game

Competitive game

Managerial applications of the theory of games
Key concepts in the theory of games

Types of games



LEARNING OBJECTIVES


After reading this lesson you should be able to

- understand the concept of a game

- grasp the assumptions in the theory of games

- appreciate the managerial applications of the theory of games
- understand the key concepts in the theory of games
- distinguish between different types of games


Introduction to game theory


Game theory seeks to analyse competing situations which arise out of conflicts of interest.

Abraham Maslow's hierarchical model of human needs lays emphasis on fulfilling the basic

needs such as food, water, clothes, shelter, air, safety and security. There is conflict of interest

between animals and plants in the consumption of natural resources. Animals compete among

themselves for securing food. Man competes with animals to earn his food. A man also

competes with another man. In the past, nations waged wars to expand the territory of their

rule. In the present day world, business organizations compete with each other in getting the

market share. The conflicts of interests of human beings are not confined to the basic needs

alone. Again considering Abraham Maslow's model of human needs, one can realize that

conflicts also arise due to the higher levels of human needs such as love, affection, affiliation,

recognition, status, dominance, power, esteem, ego, self-respect, etc. Sometimes one

witnesses clashes of ideas of intellectuals also. Every intelligent and rational participant in a

conflict wants to be a winner but not all participants can be the winners at a time. The

situations of conflict gave birth to Darwin's theory of the `survival of the fittest'. Nowadays



262
MBA-H2040 Quantitative Techniques
for Managers

the concepts of conciliation, co-existence, co-operation, coalition and consensus are gaining

ground. Game theory is another tool to examine situations of conflict so as to identify the

courses of action to be followed and to take appropriate decisions in the long run. Thus this

theory assumes importance from managerial perspectives. The pioneering work on the theory

of games was done by von Neumann and Morgenstern through their publication entitled `The

Theory of Games and Economic Behaviour' and subsequently the subject was developed by

several experts. This theory can offer valuable guidelines to a manager in `strategic

management' which can be used in the decision making process for merger, take-over, joint

venture, etc. The results obtained by the application of this theory can serve as an early

warning to the top level management in meeting the threats from the competing business

organizations and for the conversion of the internal weaknesses and external threats into

opportunities and strengths, thereby achieving the goal of maximization of profits. While this

theory does not describe any procedure to play a game, it will enable a participant to select the

appropriate strategies to be followed in the pursuit of his goals. The situation of failure in a

game would activate a participant in the analysis of the relevance of the existing strategies and

lead him to identify better, novel strategies for the future occasions.

Definitions of game theory


There are several definitions of game theory. A few standard definitions are presented below.

In the perception of Robert Mockler, "Game theory is a mathematical technique

helpful in making decisions in situations of conflicts, where the success of one part depends at

the expense of others, and where the individual decision maker is not in complete control of

the factors influencing the outcome".

The definition given by William G. Nelson runs as follows: "Game theory, more

properly the theory of games of strategy, is a mathematical method of analyzing a conflict.

The alternative is not between this decision or that decision, but between this strategy or that

strategy to be used against the conflicting interest".

In the opinion of Matrin Shubik, "Game theory is a method of the study of decision

making in situation of conflict. It deals with human processes in which the individual

decision-unit is not in complete control of other decision-units entering into the environment".

According to von Neumann and Morgenstern, "The `Game' is simply the totality of

the rules which describe it. Every particular instance at which the game is played ? in a

particular way ? from beginning to end is a `play'. The game consists of a sequence of

moves, and the play of a sequence of choices".



263

MBA-H2040 Quantitative Techniques
for Managers

J.C.C McKinsey points out a valid distinction between two words, namely `game' and

`play'. According to him, "game refers to a particular realization of the rules".

In the words of O.T. Bartos, "The theory of games can be used for `prescribing' how

an intelligent person should go about resolving social conflicts, ranging all the way from open

warfare between nations to disagreements between husband and wife".

Martin K Starr gave the following definition: "Management models in the competitive

sphere are usually termed game models. By studying game theory, we can obtain substantial

information into management's role under competitive conditions, even though much of the

game theory is neither directly operational nor implementable".

According to Edwin Mansfield, "A game is a competitive situation where two or more

persons pursue their own interests and no person can dictate the outcome. Each player, an

entity with the same interests, make his own decisions. A player can be an individual or a

group".

Assumptions for a Competitive Game



Game theory helps in finding out the best course of action for a firm in view of the

anticipated countermoves from the competing organizations. A competitive situation is a

competitive game if the following properties hold:

1. The number of competitors is finite, say N.

2. A finite set of possible courses of action is available to each of the N competitors.

3. A play of the game results when each competitor selects a course of action from the

set of courses available to him. In game theory we make an important assumption

that al the players select their courses of action simultaneously. As a result, no

competitor will be in a position to know the choices of his competitors.

4. The outcome of a play consists of the particular courses of action chosen by the

individual players. Each outcome leads to a set of payments, one to each player,

which may be either positive, or negative, or zero.

Managerial Applications of the Theory of Games




The techniques of game theory can be effectively applied to various managerial

problems as detailed below:

1) Analysis of the market strategies of a business organization in the long run.

2) Evaluation of the responses of the consumers to a new product.

3) Resolving the conflict between two groups in a business organization.

4) Decision making on the techniques to increase market share.



264
MBA-H2040 Quantitative Techniques
for Managers

5) Material procurement process.

6) Decision making for transportation problem.

7) Evaluation of the distribution system.

8) Evaluation of the location of the facilities.

9) Examination of new business ventures and

10) Competitive economic environment.

Key concepts in the Theory of Games




Several of the key concepts used in the theory of games are described below:

Players:



The competitors or decision makers in a game are called the players of the game.

Strategies:



The alternative courses of action available to a player are referred to as his strategies.

Pay off:



The outcome of playing a game is called the pay off to the concerned player.

Optimal Strategy:



A strategy by which a player can achieve the best pay off is called the optimal strategy for

him.

Zero-sum game:



A game in which the total payoffs to all the players at the end of the game is zero is referred

to as a zero-sum game.

Non-zero sum game:

Games with "less than complete conflict of interest" are called non-zero sum games. The

problems faced by a large number of business organizations come under this category. In

such games, the gain of one player in terms of his success need not be completely at the

expense of the other player.

Payoff matrix:



The tabular display of the payoffs to players under various alternatives is called the payoff

matrix of the game.

Pure strategy:



If the game is such that each player can identify one and only one strategy as the optimal

strategy in each play of the game, then that strategy is referred to as the best strategy for that

player and the game is referred to as a game of pure strategy or a pure game.





265

MBA-H2040 Quantitative Techniques
for Managers

Mixed strategy:



If there is no one specific strategy as the `best strategy' for any player in a game, then the

game is referred to as a game of mixed strategy or a mixed game. In such a game, each player

has to choose different alternative courses of action from time to time.

N-person game:



A game in which N-players take part is called an N-person game.

Maximin-Minimax Principle :

The maximum of the minimum gains is called the maximin value of the game and the

corresponding strategy is called the maximin strategy. Similarly the minimum of the

maximum losses is called the minimax value of the game and the corresponding strategy is

called the minimax strategy. If both the values are equal, then that would guarantee the best

of the worst results.

Negotiable or cooperative game:

If the game is such that the players are taken to cooperate on any or every action which may

increase the payoff of either player, then we call it a negotiable or cooperative game.

Non-negotiable or non-cooperative game:

If the players are not permitted for coalition then we refer to the game as a non-negotiable or

non-cooperative game.

Saddle point:

A saddle point of a game is that place in the payoff matrix where the maximum of the row

minima is equal to the minimum of the column maxima. The payoff at the saddle point is

called the value of the game and the corresponding strategies are called the pure strategies.

Dominance:

One of the strategies of either player may be inferior to at least one of the remaining ones.

The superior strategies are said to dominate the inferior ones.

Types of Games:

There are several classifications of a game. The classification may be based on various

factors such as the number of participants, the gain or loss to each participant, the number of

strategies available to each participant, etc. Some of the important types of games are

enumerated below.

Two person games and n-person games:

In two person games, there are exactly two players and each competitor will have a finite

number of strategies. If the number of players in a game exceeds two, then we refer to the

game as n-person game.



266
MBA-H2040 Quantitative Techniques
for Managers

Zero sum game and non-zero sum game:

If the sum of the payments to all the players in a game is zero for every possible outcome of

the game, then we refer to the game as a zero sum game. If the sum of the payoffs from any

play of the game is either positive or negative but not zero, then the game is called a non-zero

sum game

Games of perfect information and games of imperfect information:

A game of perfect information is the one in which each player can find out the strategy that

would be followed by his opponent. On the other hand, a game of imperfect information is the

one in which no player can know in advance what strategy would be adopted by the

competitor and a player has to proceed in his game with his guess works only.

Games with finite number of moves / players and games with unlimited number of

moves:

A game with a finite number of moves is the one in which the number of moves for each

player is limited before the start of the play. On the other hand, if the game can be continued

over an extended period of time and the number of moves for any player has no restriction,

then we call it a game with unlimited number of moves.

Constant-sum games:

If the sum of the game is not zero but the sum of the payoffs to both players in each case is

constant, then we call it a constant sum game. It is possible to reduce such a game to a zero-

sum game.

2x2 two person game and 2xn and mx2 games:

When the number of players in a game is two and each player has exactly two strategies, the

game is referred to as 2x2 two person game.

A game in which the first player has precisely two strategies and the second player has

three or more strategies is called an 2xn game.

A game in which the first player has three or more strategies and the second player has
exactly two strategies is called an mx2 game.

3x3 and large games:

When the number of players in a game is two and each player has exactly three strategies, we

call it a 3x3 two person game.

Two-person zero sum games are said to be larger if each of the two players has 3 or

more choices.



267

MBA-H2040 Quantitative Techniques
for Managers

The examination of 3x3 and larger games is involves difficulties. For such games, the

technique of linear programming can be used as a method of solution to identify the optimum

strategies for the two players.

Non-constant games :

Consider a game with two players. If the sum of the payoffs to the two players is not constant

in all the plays of the game, then we call it a non-constant game.



Such games are divided into negotiable or cooperative games and non-negotiable or

non-cooperative games.

QUESTIONS

1. Explain the concept of a game.

2. Define a game.

3. State the assumptions for a competitive game.

4. State the managerial applications of the theory of games.
5. Explain the following terms: strategy, pay-off matrix, saddle point, pure strategy and

mixed strategy.

6. Explain the following terms: two person game, two person zero sum game, value of a

game, 2xn game and mx2 game.





268
MBA-H2040 Quantitative Techniques
for Managers

LESSON 2

TWO-PERSON ZERO SUM GAMES



LESSON OUTLINE

The concept of a two-person zero sum game

The assumptions for a two-person zero sum game
Minimax and Maximin principles



LEARNING OBJECTIVES


After reading this lesson you should be able to

- understand the concept of a two-person zero sum game
- have an idea of the assumptions for a two-person zero sum game
- understand Minimax and Maximin principles
- solve a two-person zero sum game
- interpret the results from the payoff matrix of a two-person zero sum game



Definition of two-person zero sum game


A game with only two players, say player A and player B, is called a two-person zero sum

game if the gain of the player A is equal to the loss of the player B, so that the total sum is

zero.

Payoff matrix:

When players select their particular strategies, the payoffs (gains or losses) can be represented

in the form of a payoff matrix.



Since the game is zero sum, the gain of one player is equal to the loss of other and

vice-versa. Suppose A has m strategies and B has n strategies. Consider the following payoff

matrix.

Player B's strategies

B

B

B

1

2

n

A a

a

a

1

11

12

1n

Player A's strategies

A

a

a

a

1

21

22

2n







Am a

a

a

1

m

m2

mn

Player A wishes to gain as large a payoff a as possible while player B will do his best to

ij

reach as small a value a as possible where the gain to player B and loss to player A be (-

ij

a ).

ij



269

MBA-H2040 Quantitative Techniques
for Managers

Assumptions for two-person zero sum game:

For building any model, certain reasonable assumptions are quite necessary. Some

assumptions for building a model of two-person zero sum game are listed below.

a) Each player has available to him a finite number of possible courses of action.

Sometimes the set of courses of action may be the same for each player. Or, certain

courses of action may be available to both players while each player may have certain

specific courses of action which are not available to the other player.

b) Player A attempts to maximize gains to himself. Player B tries to minimize losses to

himself.

c) The decisions of both players are made individually prior to the play with no

communication between them.

d) The decisions are made and announced simultaneously so that neither player has an

advantage resulting from direct knowledge of the other player's decision.

e) Both players know the possible payoffs of themselves and their opponents.

Minimax and Maximin Principles

The selection of an optimal strategy by each player without the knowledge of the competitor's

strategy is the basic problem of playing games.



The objective of game theory is to know how these players must select their respective

strategies, so that they may optimize their payoffs. Such a criterion of decision making is

referred to as minimax-maximin principle. This principle in games of pure strategies leads to

the best possible selection of a strategy for both players.



For example, if player A chooses his ith strategy, then he gains at least the payoff min

a , which is minimum of the ith row elements in the payoff matrix. Since his objective is to

ij

maximize his payoff, he can choose strategy i so as to make his payoff as large as possible.

i.e., a payoff which is not less than max min a .

ij

1im

1 jn

Similarly player B can choose jth column elements so as to make his loss not greater than

min max a .

ij

1 jn 1im



If the maximin value for a player is equal to the minimax value for another player, i.e.

max min a V min max a

ij

ij

1im

1 jn

1 jn

1im

then the game is said to have a saddle point (equilibrium point) and the corresponding

strategies are called optimal strategies. If there are two or more saddle points, they must be

equal.



270
MBA-H2040 Quantitative Techniques
for Managers



The amount of payoff, i.e., V at an equilibrium point is known as the value of the

game.

The optimal strategies can be identified by the players in the long run.

Fair game:



The game is said to be fair if the value of the game V = 0.

Problem 1:

Solve the game with the following pay-off matrix.

Player B

Strategies

I

II

III

IV

V

1

2

5

3

6

7

Player A Strategies 2

4

6

8

1

6

3

8

2

3

5

4

4 15 14

18

12

20

Solution:

First consider the minimum of each row.

Row

Minimum Value

1

-3

2

-1

3

2

4

12

Maximum of {-3, -1, 2, 12} = 12

Next consider the maximum of each column.

Column

Maximum Value

1

15

2

14

3

18

4

12

5

20

Minimum of {15, 14, 18, 12, 20}= 12

We see that the maximum of row minima = the minimum of the column maxima. So the

game has a saddle point. The common value is 12. Therefore the value V of the game = 12.

Interpretation:

In the long run, the following best strategies will be identified by the two players:



271

MBA-H2040 Quantitative Techniques
for Managers

The best strategy for player A is strategy 4.

The best strategy for player B is strategy IV.

The game is favourable to player A.

Problem 2:

Solve the game with the following pay-off matrix

Player Y

Strategies

I

II

III

IV

V

1 9 12

7

14

26

Player X Strategies 2 25 35 20

28 30

3 7

6

8

3

2

4 8

11 13

2

1

Solution:

First consider the minimum of each row.

Row

Minimum Value

1

7

2

20

3

-8

4

-2



Maximum of {7, 20, ?8, -2} = 20

Next consider the maximum of each column.

Column

Maximum Value

1

25

2

35

3

20

4

28

5

30



Minimum of {25, 35, 20, 28, 30}= 20



It is observed that the maximum of row minima and the minimum of the column

maxima are equal. Hence the given the game has a saddle point. The common value is 20.

This indicates that the value V of the game is 20.

Interpretation.



272
MBA-H2040 Quantitative Techniques
for Managers

The best strategy for player X is strategy 2.

The best strategy for player Y is strategy III.

The game is favourable to player A.

Problem 3:

Solve the following game:

Player B

Strategies

I

II

III

IV

1 1 6

8

4

Player A Strategies 2 3 7

2 8

3 5

5 1

0

4 3 4

5

7

Solution

First consider the minimum of each row.

Row

Minimum Value

1

-6

2

-8

3

-5

4

-4



Maximum of {-6, -8, -5, -4} = -4

Next consider the maximum of each column.

Column

Maximum Value

1

5

2

-4

3

8

4

7



Minimum of {5, -4, 8, 7}= - 4

Since the max {row minima} = min {column maxima}, the game under consideration has a

saddle point. The common value is ?4. Hence the value of the game is ?4.

Interpretation.

The best strategy for player A is strategy 4.

The best strategy for player B is strategy II.



273

MBA-H2040 Quantitative Techniques
for Managers

Since the value of the game is negative, it is concluded that the game is favourable to

player B.

QUESTIONS

1. What is meant by a two-person zero sum game? Explain.

2. State the assumptions for a two-person zero sum game.

3. Explain Minimax and Maximin principles.

4. How will you interpret the results from the payoff matrix of a two-person zero sum

game? Explain.

5. What is a fair game? Explain.

6. Solve the game with the following pay-off matrix.

Player B

Strategies

I

II

III

IV

V

1

7

5

2

3

9

Player A Strategies 2 10

8

7

4

5

3

9

12

0

2

1

4 11

2 1

3

4

Answer: Best strategy for A: 2

Best strategy for B: IV

V = 4

The game is favourable to player A

7. Solve the game with the following pay-off matrix.

Player B

Strategies

I

II

III

IV

V

1

2

3

8

7

0

Player A Strategies 2

1

7 5 2

3

3

4

2

3

5

1

4

6

4

5

4

7

Answer: Best strategy for A: 3

Best strategy for B: II

V = -2

The game is favourable to player B

8. Solve the game with the following pay-off matrix.



274
MBA-H2040 Quantitative Techniques
for Managers

Player B

6

4

Player A



7 5

Answer: Best strategy for A: 2

Best strategy for B: II

V = 5

The game is favourable to player A

9. Solve the following game and interpret the result.

Player B

Strategies

I

II

III

IV

1 3

7

1

3

Player A Strategies 2 1

2

3

1

3 0

4

2

6

4 2

1

5

1

Answer: Best strategy for A: 3

Best strategy for B: I

V = 0

The value V = 0 indicates that the game is a fair one.

10. Solve the following game:

Player B

Strategies

I

II

III

1 1

8

2

Player A Strategies 2 3

5

6



3 2

2

1





Answer: Best strategy for A: 2

Best strategy for B: I

V = 3

The game is favourable to player A

11. Solve the game



275

MBA-H2040 Quantitative Techniques
for Managers

Player B

I

II

III

IV

1 4

1

2

0

Player A



2 3

5

9

2

3 2

8

0

11

Answer : V = -1



12. Solve the game

Player Y

I

II

III

IV

V

1 4

0

1

7

1

2 0

3

5 7

5

Player X



3 3

2

3

4

3

4 6

4

1

0

5

5 0

0

6

0

0

Answer : V = 2



13. Solve the game

Player B

I

II

III

IV

V

1 9

3

4

4

2

2 8

6

8

5 12

Player A



3 10 7 19 18 14

4 8

6

8 11

6

5 3 5 16 10

8

Answer : V = 7





276
MBA-H2040 Quantitative Techniques
for Managers

14. Solve the game

Player Y

17 10 12

5

4

8

2

5

6

7

6

9

7

6

9

2

3

1

Player X



10 11 14

8 13

8

20 18 17 10 15 17

12 11 15

9

5 11

Answer : V = 10



15. Solve the game

Player B

12 14

8

7

4

9

2 13

6

7

9

8

13

6

8

6

3

1

Player A



14

9 10

8

9

6

20 18 17 11 14 16

8 12 16

9

6 13

Answer : V = 11



16. Examine whether the following game is fair.

Player Y

6

4

3

2

Player X 3

5

0

8

7

2

6

5

Answer : V = 0. Therefore, it is a fair game.











277

MBA-H2040 Quantitative Techniques
for Managers

LESSON 3

GAMES WITH NO SADDLE POINT

LESSON OUTLINE

The concept of a 2x2 game with no saddle point

The method of solution


LEARNING OBJECTIVES


After reading this lesson you should be able to

- understand the concept of a 2x2 game with no saddle point
- know the method of solution of a 2x2 game without saddle point
- solve a game with a given payoff matrix
- interpret the results obtained from the payoff matrix



2 x 2 zero-sum game

When each one of the first player A and the second player B has exactly two strategies,

we have a 2 x 2 game.

Motivating point

First let us consider an illustrative example.

Problem 1:

Examine whether the following 2 x 2 game has a saddle point

Player B

3 5

Player A



4

2

Solution:

First consider the minimum of each row.

Row

Minimum Value

1

3

2

2

Maximum of {3, 2} = 3

Next consider the maximum of each column.

Column

Maximum Value

1

4

2

5

Minimum of {4, 5}= 4



278
MBA-H2040 Quantitative Techniques
for Managers



We see that max {row minima} and min {column maxima} are not equal. Hence the

game has no saddle point.

Method of solution of a 2x2 zero-sum game without saddle point

Suppose that a 2x2 game has no saddle point. Suppose the game has the following pay-off

matrix.

Player B

Strategy

a

b

Player A Strategy



c

d

Since this game has no saddle point, the following condition shall hold:

Max{Min{a, }

b , Min{c, d}} Min{Max{a, }

c , Max{b, d}}

In this case, the game is called a mixed game. No strategy of Player A can be called the

best strategy for him. Therefore A has to use both of his strategies. Similarly no strategy

of Player B can be called the best strategy for him and he has to use both of his

strategies.



Let p be the probability that Player A will use his first strategy. Then the

probability that Player A will use his second strategy is 1-p.

If Player B follows his first strategy

Expected value of the pay-off to Player A

Expected value of the pay-off to Player A

Expected value of the pay-off to Player A

{

} {

}

arising from his first strategy

arising from his second strategy



ap (

c 1 )

p

(1)

In the above equation, note that the expected value is got as the product of the corresponding

values of the pay-off and the probability.

If Player B follows his second strategy

Expected value of the bp d(1 p)

(2)

pay-off to Player A

If the expected values in equations (1) and (2) are different, Player B will prefer the minimum

of the two expected values that he has to give to player A. Thus B will have a pure strategy.

This contradicts our assumption that the game is a mixed one. Therefore the expected values

of the pay-offs to Player A in equations (1) and (2) should be equal. Thus we have the

condition



279

MBA-H2040 Quantitative Techniques
for Managers

ap c(1 p)

bp d (1 p)

ap bp

(1 p)[d c]

p(a b)

(d c) p(d c)

p(a b) p(d c)

d c

p(a b d c)

d c

d c



p

(a d ) (b c)

a d b c d c

1 p

(a d ) (b c)

a b

(a d ) (b c)

The number of times A

The number of times A

d c

a b

{

}:{

}

:



will use first strategy

will use second strategy

(a d ) (b c) (a d ) (b c)

The expected pay-off to Player A

ap c(1 p)
c p(a c)

(d c)(a c)

c (ad)(bc)

c(a d) (b c

) (d c)(a c)



(a d ) (b c)

2

2

ac cd bc c ad cd ac c )

(a d ) (b c)

ad bc

(ad)(bc)

Therefore, the value V of the game is

ad bc



(a d ) (b c)

To find the number of times that B will use his first strategy and second strategy:

Let the probability that B will use his first strategy be r. Then the probability that B will use

his second strategy is 1-r.

When A use his first strategy

The expected value of loss to Player B with his first strategy = ar

The expected value of loss to Player B with his second strategy = b(1-r)

Therefore the expected value of loss to B = ar + b(1-r)







(3)

When A use his second strategy

The expected value of loss to Player B with his first strategy = cr



280
MBA-H2040 Quantitative Techniques
for Managers



The expected value of loss to Player B with his second strategy = d(1-r)



Therefore the expected value of loss to B = cr + d(1-r)







(4)



If the two expected values are different then it results in a pure game, which is a contradiction.

Therefore the expected values of loss to Player B in equations (3) and (4) should be equal.

Hence we have the condition

ar b(1 r) cr d (1 r)

ar b br cr d dr

ar br cr dr d b

r(a b c d ) d b



d b

r a bc d

d b

(ad)(bc)

Problem 2:

Solve the following game

Y

2 5

X



4 1

Solution:

First consider the row minima.



Row

Minimum Value

1

2

2

1



Maximum of {2, 1} = 2

Next consider the maximum of each column.

Column

Maximum Value

1

4

2

5



Minimum of {4, 5}= 4

We see that



281

MBA-H2040 Quantitative Techniques
for Managers

Max {row minima} min {column maxima}

So the game has no saddle point. Therefore it is a mixed game.

We have a = 2, b = 5, c = 4 and d = 1.

Let p be the probability that player X will use his first strategy. We have

d c

p (a d) (b c)

1 4

(21)(54)

3



3 9

3

6

1

2

1

1

The probability that player X will use his second strategy is 1-p = 1-

=

.

2

2

ad bc

2 20

18

Value of the game V =

3 .

(a d ) (b c)

3 9

6

Let r be the probability that Player Y will use his first strategy. Then the probability that Y

will use his second strategy is (1-r). We have

d b

r (a d)(b c)

1 5

(21)(54)

4

39



4

6
2

3

2

1

1 r 1

3

3

Interpretation.

1

1

p : (1-p) =

:



2

2

Therefore, out of 2 trials, player X will use his first strategy once and his second strategy

once.



282
MBA-H2040 Quantitative Techniques
for Managers

2

1

r : (1-r) =

:

3

3

Therefore, out of 3 trials, player Y will use his first strategy twice and his second strategy

once.

QUESTIONS

1. What is a 2x2 game with no saddle point? Explain.

2. Explain the method of solution of a 2x2 game without saddle point.

3. Solve the following game

Y

1

2 4

X



3 7

1

1

Answer: p = , r =

, V = 6

3

4

4. Solve the following game

Y

5

4



X



9

3

4

1

Answer: p =

, r = , V = -1

7

3

5. Solve the following game

Y

10 4

X



6

8

1

1

Answer: p =

, r =

, V = 7

4

2

6. Solve the following game

Y

20

8

X



2 10

1

1

Answer: p =

, r =

, V = 9

2

12

7. Solve the following game

Y



283

MBA-H2040 Quantitative Techniques
for Managers

10

2

X



1

5

1

1

Answer: p = , r =

, V = 4

3

4

8. Solve the following game

Y

12

6

X



6

9

1

1

Answer: p = , r = , V = 8

3

3

9. Solve the following game

Y

10

8

X



8

10

1

1

Answer: p =

, r =

, V = 9

2

2

10. Solve the following game

Y

16

4

X



4

8

1

1

Answer: p =

, r =

, V = 7

4

4

11. Solve the following game

Y

11

5

X



7

9

1

7

Answer: p =

, r =

, V = - 2

2

16

12. Solve the following game Y

9

3

X



5

7

1

5

Answer: p =

, r =

, V = - 2

2

12





284
MBA-H2040 Quantitative Techniques
for Managers

LESSON 4

THE PRINCIPLE OF DOMINANCE



LESSON OUTLINE

The principle of dominance

Dividing a game into sub games



LEARNING OBJECTIVES


After reading this lesson you should be able to

- understand the principle of dominance
-

solve a game using the principle of dominance

- solve a game by dividing a game into sub games



The principle of dominance


In the previous lesson, we have discussed the method of solution of a game without a saddle

point. While solving a game without a saddle point, one comes across the phenomenon of the

dominance of a row over another row or a column over another column in the pay-off matrix

of the game. Such a situation is discussed in the sequel.



In a given pay-off matrix A, we say that the ith row dominates the kth row if

a a for all j = 1,2,...,n

ij

kj

and

a a for at least one j.

ij

kj

In such a situation player A will never use the strategy corresponding to kth row,

because he will gain less for choosing such a strategy.

Similarly, we say the pth column in the matrix dominates the qth column if

a a for all i = 1,2,...,m

ip

iq

and

a a for at least one i.

ip

iq

In this case, the player B will loose more by choosing the strategy for the qth column than by

choosing the strategy for the pth column. So he will never use the strategy corresponding to

the qth column. When dominance of a row ( or a column) in the pay-off matrix occurs, we can

delete a row (or a column) from that matrix and arrive at a reduced matrix. This principle of

dominance can be used in the determination of the solution for a given game.



285

MBA-H2040 Quantitative Techniques
for Managers

Let us consider an illustrative example involving the phenomenon of dominance in a game.

Problem 1:

Solve the game with the following pay-off matrix:

Player B

I

II

III

IV

1 4

2

3

6

Player A



2 3

4

7

5

3 6

3

5

4

Solution:

First consider the minimum of each row.

Row

Minimum Value

1

2

2

3

3

3



Maximum of {2, 3, 3} = 3

Next consider the maximum of each column.

Column

Maximum Value

1

6

2

4

3

7

4

6



Minimum of {6, 4, 7, 6}= 4

The following condition holds:

Max {row minima} min {column maxima}

Therefore we see that there is no saddle point for the game under consideration.

Compare columns II and III.



Column II

Column III

2

3

4

7

3

5



286
MBA-H2040 Quantitative Techniques
for Managers



We see that each element in column III is greater than the corresponding element in column

II. The choice is for player B. Since column II dominates column III, player B will discard

his strategy 3.

Now we have the reduced game

I

II

IV

1 4

2

6



2 3

4

5

3 6

3

4

For this matrix again, there is no saddle point. Column II dominates column IV. The choice

is for player B. So player B will give up his strategy 4

The game reduces to the following:

I

II

1 4 2



2 3

4

3 6 3

This matrix has no saddle point.



The third row dominates the first row. The choice is for player A. He will give up his

strategy 1 and retain strategy 3. The game reduces to the following:

3 4



6 3

a b

Again, there is no saddle point. We have a 2x2 matrix. Take this matrix as



c d

Then we have a = 3, b = 4, c = 6 and d = 3. Use the formulae for p, 1-p, r, 1-r and V.

d c

p (a d) (b c)

3 6

(33)(64)

3

610



3

4

3

4

3

1

1 p 1

4

4



287

MBA-H2040 Quantitative Techniques
for Managers

d b

r (a d)(b c)

3 4

(33)(64)

1

610



1

4
1

4

1

3

1 r 1

4

4

The value of the game

ad bc

V (a d)(b c)

3x3 4x6

4



15

4

15

4

3 1

1 3

Thus, X =

, , 0, 0

and Y =

, , 0,0

are the optimal strategies.

4 4

4 4

Method of convex linear combination

A strategy, say s, can also be dominated if it is inferior to a convex linear combination of

several other pure strategies. In this case if the domination is strict, then the strategy s can be

deleted. If strategy s dominates the convex linear combination of some other pure strategies,

then one of the pure strategies involved in the combination may be deleted. The domination

will be decided as per the above rules. Let us consider an example to illustrate this case.

Problem 2:

Solve the game with the following pay-off matrix for firm A:

Firm B

B

B

B

B

B

1

2

3

4

5

A 4

8

2

5

6

1

A

4

0

6

8

5

Firm A 2



A 2

6

4

4

2

3

A

4

3

5

6

3

4

A 4 1

5 7

3

5



288
MBA-H2040 Quantitative Techniques
for Managers

Solution:

First consider the minimum of each row.

Row

Minimum Value

1

-2

2

0

3

-6

4

-3

5

-1



Maximum of {-2, 0, -6, -3, -1} = 0

Next consider the maximum of each column.

Column

Maximum Value

1

4

2

8

3

6

4

8

5

6



Minimum of { 4, 8, 6, 8, 6}= 4

Hence,

Maximum of {row minima} minimum of {column maxima}.

So we see that there is no saddle point. Compare the second row with the fifth row. Each

element in the second row exceeds the corresponding element in the fifth row. Therefore, A

2

dominates A . The choice is for firm A. It will retain strategy A and give up strategy A .

5

2

5

Therefore the game reduces to the following.

B

B

B

B

B

1

2

3

4

5

A 4

8

2 5

6

1

A

4

0

6 8

5

2

A 2

6 4 4

2

3

A 4

3

5 6

3

4

Compare the second and fourth rows. We see that A dominates A . So, firm A will retain

2

4

the strategy A and give up the strategy A . Thus the game reduces to the following:

2

4



289

MBA-H2040 Quantitative Techniques
for Managers

B

B

B

B

B

1

2

3

4

5

A 4

8

2

5 6

1



A

4

0

6

8

5

2

A 2

6

4

4 2

3

Compare the first and fifth columns. It is observed that B1 dominates B5. The choice is for

firm B. It will retain the strategy B and give up the strategy B . Thus the game reduces to the

1

5

following

B

B

B

B

1

2

3

4

A 4

8

2 5

1



A

4

0

6 8

2

A 2

6 4 4

3

Compare the first and fourth columns. We notice that B1 dominates B4. So firm B will discard

the strategy B and retain the strategy B . Thus the game reduces to the following:

4

1

B

B

B

1

2

3

A 4

8

2

1



A

4

0

6

2

A 2

6

4

3

For this reduced game, we check that there is no saddle point.

Now none of the pure strategies of firms A and B is inferior to any of their other

strategies. But, we observe that convex linear combination of the strategies B and B

2

3

dominates B , i.e. the averages of payoffs due to strategies B and B ,

1

2

3

8 2 0 6 6

4

,

,

3,3,

5

2

2

2

dominate B . Thus B may be omitted from consideration. So we have the reduced matrix

1

1

B

B

2

3

A 8

2

1



A

0

6

2

A 6

4

3

Here, the average of the pay-offs due to strategies A and A of firm A, i.e.

1

2

8 0 2

6

,

4,

2 dominates the pay-off due to A . So we get a new reduced 2x2 pay-

2

2

3

off matrix.





290
MBA-H2040 Quantitative Techniques
for Managers

Firm B's strategy

B

B

2

3

Firm A's strategy A 8

2



1

A 0

6

2

We have a = 8, b = -2, c = 0 and d = 6.

d c

p (a d) (b c)

6 0

(68)(20)

6



16

3

8

3

5

1 p 1

8

8

d b

r (a d)(b c)

6 (2)

16

8



16

1

2

1

1

1 r 1

2

2

Value of the game:

ad bc

V (a d)(b c)

6 8

x 0x(2)



16

48

3

16

So the optimal strategies are

3 5

1 1

A = , , 0, 0, 0 and B = 0, , , 0, 0 .

8 8

2 2

The value of the game = 3. Thus the game is favourable to firm A.







291

MBA-H2040 Quantitative Techniques
for Managers

Problem 3:



For the game with the following pay-off matrix, determine the saddle point

Player B

I

II

III

IV

1 2 1

0

3



Player A



2

1

0

3

2

3 3

2

1

4

Solution:

Column II

Column III

1

1

0

0 1

2 0

3

3 0

3 2

1

1 2

The choice is with the player B. He has to choose between strategies II and III. He will lose

more in strategy III than in strategy II, irrespective of what strategy is followed by A. So he

will drop strategy III and retain strategy II. Now the given game reduces to the following

game.

I

II

IV

1 2

1

3



2

1

0

2

3 3

2

4

Consider the rows and columns of this matrix.

Row minimum:





I Row

:

-3





II Row

:

0



Maximum of {-3, 0, -3} = 0





III Row

:

-3

Column maximum:





I Column

:

2





II Column

:

0



Minimum of {2, 0, 4} = 0





III Column

:

4

We see that

Maximum of row minimum = Minimum of column maximum = 0.

So, a saddle point exists for the given game and the value of the game is 0.







292
MBA-H2040 Quantitative Techniques
for Managers

Interpretation:

No player gains and no player loses. i.e., The game is not favourable to any player. i.e. It is a

fair game.

Problem 4:



Solve the game

Player B

4 8 6

Player A 6

2 10



4 5 7

Solution:



First consider the minimum of each row.

Row

Minimum

1

4

2

2

3

4



Maximum of {4, 2, 4} = 4

Next, consider the maximum of each column.

Column

Maximum

1

6

2

8

3

10

Minimum of {6, 8, 10} = 6

Since Maximum of { Row Minima} and Minimum of { Column Maxima } are different, it

follows that the given game has no saddle point.

Denote the strategies of player A by A , A , A . Denote the strategies of player B by B , B , B .

1

2

3

1

2

3

Compare the first and third columns of the given matrix.

B

B

1

3

4

6

6

10

7

7



The pay-offs in B are greater than or equal to the corresponding pay-offs in B . The

3

1

player B has to make a choice between his strategies 1 and 3. He will lose more if he follows



293

MBA-H2040 Quantitative Techniques
for Managers

strategy 3 rather than strategy 1. Therefore he will give up strategy 3 and retain strategy 1.

Consequently, the given game is transformed into the following game:

B

B

1

2

A 4

8

1



A 6

2

2

A 4

5

3

Compare the first and third rows of the above matrix.

B

B

1

2

A 4

8

1

A 4

5

3

The pay-offs in A are greater than or equal to the corresponding pay-offs in A . The player

1

3

A has to make a choice between his strategies 1 and 3. He will gain more if he follows

strategy 1 rather than strategy 3. Therefore he will retain strategy 1 and give up strategy 3.

Now the given game is transformed into the following game.

B

B

1

2

A 4

8

1

A 6

2

2

It is a 2x2 game. Consider the row minima.

Row

Minimum

1

4

2

2



Maximum of {4, 2} = 4

Next, consider the maximum of each column.

Column

Maximum

1

6

2

8

Minimum of {6, 8} = 6

Maximum {row minima} and Minimum {column maxima } are not equal

Therefore, the reduced game has no saddle point. So, it is a mixed game

a b 4 8

Take



. We have a = 4, b = 8, c = 6 and d = 2.

c d 6 2

The probability that player A will use his first strategy is p. This is calculated as



294
MBA-H2040 Quantitative Techniques
for Managers

d c

p (a d) (b c)

2 6

(42)(86)

4

614

4 1

8 2

The probability that player B will use his first strategy is r. This is calculated as

d b

r (a d)(b c)

2 8

8



6

8

3

4

Value of the game is V. This is calculated as

ad bc

V (a d)(b c)

4x2 8x6

8



8 48

8

40

5

8

Interpretation

Out of 3 trials, player A will use strategy 1 once and strategy 2 once. Out of 4 trials, player B

will use strategy 1 thrice and strategy 2 once. The game is favourable to player A.

Problem 5: Dividing a game into sub-games

Solve the game with the following pay-off matrix.

Player B

1

2

3

I 4

6

3

Player A



II 3

3

4

III 2

3

4

Solution:



First, consider the row mimima.



295

MBA-H2040 Quantitative Techniques
for Managers

Row

Minimum

1

-4

2

-3

3

-3



Maximum of {-4, -3, -3} = -3

Next, consider the column maxima.

Column

Maximum

1

2

2

6

3

4

Minimum of {2, 6, 4} = 2

We see that Maximum of { row minima} Minimum of { column maxima}.

So the game has no saddle point. Hence it is a mixed game. Compare the first and third

columns.

I Column

III Column

4

3

4 3

3

4

3 4

2

4

2 4

We assert that Player B will retain the first strategy and give up the third strategy. We get the

following reduced matrix.

4

6

3

3



2

3



We check that it is a game with no saddle point.

Sub games

Let us consider the 2x2 sub games. They are:

4 6 4

6 3

3







3

3 2 3 2 3

First, take the sub game

4 6



3

3



296
MBA-H2040 Quantitative Techniques
for Managers



Compare the first and second columns. We see that 4 6 and 3 3 . Therefore, the

4

game reduces to . Since 4 3, it further reduces to ?3.

3

Next, consider the sub game

4

6



2 3

We see that it is a game with no saddle point. Take a = -4, b = 6, c = 2, d = -3. Then the

value of the game is

ad bc

V (a d)(b c)

(4)(3) (6)(2)



(4 3) (6 2)

0

3

3

Next, take the sub game

. In this case we have a = -3, b = 3, c = 2 and d = -3. The

2 3

value of the game is obtained as

ad bc

V (a d)(b c)

(3)( 3

) (3)(2)



(3 3) (3 2)

9 6

3



6

5

11

Let us tabulate the results as follows:

Sub game

Value



4 6



3

3

-3



4

6



0

2 3



3

3



3

2 3

-



11





The value of 0 will be preferred by the player A. For this value, the first and third

strategies of A correspond while the first and second strategies of the player B correspond to

the value 0 of the game. So it is a fair game.



297

MBA-H2040 Quantitative Techniques
for Managers

QUESTIONS

1. Explain the principle of dominance in the theory of games.

2. Explain how a game can be solved through sub games.

3. Solve the following game by the principle of dominance:

Player B

Strategies

I

II

III

IV

1 8

10

9

14

Player A Strategies 2 10

11

8

12

3 13

12

14

13

Answer: V = 12

4. Solve the game by the principle of dominance:

1 7 2

6

2 7



5 2 6

Answer: V = 4



5. Solve the game with the following pay-off matrix

6 3 1 0

3





3 2 4 2 1

3

2

11

Answer : p

, r

, V



5

5

5

6. Solve the game

8

7

6

1

2

12 10 12

0

4



1

4

6

8 14 16

4

7

70

Answer : p

, r

, V



9

9

9



298
MBA-H2040 Quantitative Techniques
for Managers

LESSON 5

GRAPHICAL SOLUTION OF A 2x2 GAME WITH NO SADDLE POINT



LESSON OUTLINE

The principle of graphical solution

Numerical example



LEARNING OBJECTIVES


After reading this lesson you should be able to

-

understand the principle of graphical solution

-

derive the equations involving probability and expected value

- solve numerical problems


Example: Consider the game with the following pay-off matrix.

Player B

2 5

Player A



4 1

First consider the row minima.

Row

Minimum

1

2

2

1



Maximum of {2, 1} = 2.

Next, consider the column maxima.

Column

Maximum

1

4

2

5



Minimum of {4, 5} = 4.

We see that Maximum { row minima} Minimum { column maxima }

So, the game has no saddle point. It is a mixed game.

Equations involving probability and expected value:

Let p be the probability that player A will use his first strategy.

Then the probability that A will use his second strategy is 1-p.



299

MBA-H2040 Quantitative Techniques
for Managers

Let E be the expected value of pay-off to player A.

When B uses his first strategy



The expected value of pay-off to player A is given by

E 2 p 4(1 p)

2 p 4 4 p









(1)

4 2 p

When B uses his second strategy



The expected value of pay-off to player A is given by

E 5 p 1(1 p)

5 p 1 p









(2)

4 p 1



300
MBA-H2040 Quantitative Techniques for Managers
Consider equations (1) and (2). For plotting the two equations on a graph sheet, get some points on them

as follows:

E = -2p+4

p

0

1

0.5

E

4

2

3



E = 4p+1

p

0

1

0.5

E

1

5

3



Graphical solution:

Procedure:

Take probability and expected value along two rectangular axes in a graph sheet. Draw two straight

lines given by the two equations (1) and (2). Determine the point of intersection of the two straight lines

in the graph. This will give the common solution of the two equations (1) and (2). Thus we would obtain

the value of the game.

Represent the two equations by the two straight lines AB and CD on the graph sheet. Take the

point of intersection of AB and CD as T. For this point, we have p = 0.5 and E = 3. Therefore, the

value V of the game is 3.































301

MBA-H2040 Quantitative Techniques for Managers





E




E=4P+1
D

A
4
T
3

2 C B
E=-2P+4
1
P

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

















Problem 1:

Solve the following game by graphical method.

Player B

18

2

Player A



6

4



Solution:

First consider the row minima.

Row

Minimum

1

- 18

2

- 4





302
MBA-H2040 Quantitative Techniques for Managers

Maximum of {-18, - 4} = - 4.

Next, consider the column maxima.

Column

Maximum

1

6

2

2



Minimum of {6, 2} = 2.

We see that Maximum { row minima} Minimum { column maxima }

So, the game has no saddle point. It is a mixed game.

Let p be the probability that player A will use his first strategy.

Then the probability that A will use his second strategy is 1-p.

When B uses his first strategy



The expected value of pay-off to player A is given by

E 18

p 6 (1 p)

18

p 6 6 p







(I)

24

p 6

When B uses his second strategy



The expected value of pay-off to player A is given by

E 2 p 4 (1 p)

2 p 4 4 p (II)
6 p 4

Consider equations (I) and (II). For plotting the two equations on a graph sheet, get some points on them

as follows:

E = -24 p + 6

p

0

1

0.5

E

6

-18 -6

E = 6p-4

p

0

1

0.5

E

-4

2

-1



Graphical solution:

Take probability and expected value along two rectangular axes in a graph sheet. Draw two straight lines

given by the two equations (1) and (2). Determine the point of intersection of the two straight lines in the



303

MBA-H2040 Quantitative Techniques for Managers
graph. This will provide the common solution of the two equations (1) and (2). Thus we would get the

value of the game.

Represent the two equations by the two straight lines AB and CD on the graph sheet. Take the

1

point of intersection of AB and CD as T. For this point, we have p =

and E = -2. Therefore, the

3

value V of the game is -2.




E







8

6
A
4

E=6P-4
2
0 D

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.
-2
C
-4

-6

-8
B




E=-24P+6











304
MBA-H2040 Quantitative Techniques for Managers
QUESTIONS

1. Explain the method of graphical solution of a 2x2 game.

2. Obtain the graphical solution of the game

10

6



8 12

1

Answer: p =

, V = 9

2

3. Graphically solve the game

4

10



8

6

1

Answer: p =

, V = 7

4

4. Find the graphical solution of the game

12

12



2

6

1

3

Answer: p =

, V =



4

2

5. Obtain the graphical solution of the game

10

6



8 12

1

Answer: p =

, V = 9

2

6. Graphically solve the game

3

5



5

1

3

7

Answer: p =

, V =



4

2















305

MBA-H2040 Quantitative Techniques for Managers

LESSON 6

2 x n ZERO-SUM GAMES

LESSON OUTLINE

A 2 x n zero-sum game

Method of solution

Sub game approach and graphical method

Numerical example


LEARNING OBJECTIVES

After reading this lesson you should be able to

- understand the concept of a 2 x n zero-sum game
- solve numerical problems


The concept of a 2 x n zero-sum game

When the first player A has exactly two strategies and the second player B has n (where n is three

or more) strategies, there results a 2 x n game. It is also called a rectangular game. Since A has

two strategies only, he cannot try to give up any one of them. However, since B has many

strategies, he can make out some choice among them. He can retain some of the advantageous

strategies and discard some disadvantageous strategies. The intention of B is to give as minimum

payoff to A as possible. In other words, B will always try to minimize the loss to himself.

Therefore, if some strategies are available to B by which he can minimize the payoff to A, then B

will retain such strategies and give such strategies by which the payoff will be very high to A.


Approaches for 2 x n zero-sum game

There are two approaches for such games: (1) Sub game approach and (2) Graphical approach.


Sub game approach

The given 2 x n game is divided into 2 x 2 sub games. For this purpose, consider all possible 2 x 2 sub

matrices of the payoff matrix of the given game. Solve each sub game and have a list of the values of

each sub game. Since B can make out a choice of his strategies, he will discard such of those sub games

which result in more payoff to A. On the basis of this consideration, in the long run, he will retain two

strategies only and give up the other strategies.



Problem

Solve the following game



306
MBA-H2040 Quantitative Techniques for Managers
Player B

8

2

6

9

Player A



3 5 10 2

Solution:

Let us consider all possible 2x2 sub games of the given game. We have the following sub games:



8

2

1.



3

5

8

6

2.



3 10

8

9

3.



3 2

2

6

4.



5 10

2

9

5.



5 2

6

9

6.



10 2

Let E be the expected value of the pay off to player A. Let p be the probability that player A will use his

first strategy. Then the probability that he will use his second strategy is 1-p. We form the equations for

E in all the sub games as follows:

Sub game (1)



Equation 1: E 8 p 3(1 p) 5 p 3



Equation 2: E 2

p 5(1 p) 7

p 5

Sub game (2)



Equation 1: E 8 p 3(1 p) 5 p 3



Equation 2: E 6

p 10(1 p) 1

6 p 10

Sub game (3)



Equation 1: E 8 p 3(1 p) 5 p 3



Equation 2: E 9 p 2(1 p) 7 p 2

Sub game (4)



Equation 1: E 2

p 5(1 p) 7

p 5



Equation 2: E 6

p 10(1 p) 1

6 p 10



307

MBA-H2040 Quantitative Techniques for Managers
Sub game (5)



Equation 1: E 2

p 5(1 p) 7

p 5



Equation 2: E 9 p 2(1 p) 7 p 2

Sub game (6)



Equation 1: E 6

p 10(1 p) 1

6 p 10



Equation 2: E 9 p 2(1 p) 7 p 2

Solve the equations for each sub game. Let us tabulate the results for the various sub games. We have

the following:

Sub game

p

Expected value E

1

1

23





6

6

2

1

14





3

3

3

1

11





2

2

4

5

10





9

9

5

3

7





14

2

6

8

102





23

23



Interpretation:

Since player A has only 2 strategies, he cannot make any choice on the strategies. On the other hand,

player B has 4 strategies. Therefore he can retain any 2 strategies and give up the other 2 strategies.

This he will do in such a way that the pay-off to player A is at the minimum. The pay-off to A is the

2

6

minimum in the case of sub game 4. i.e., the sub game with the matrix

.

5 10



Therefore, in the long run, player B will retain his strategies 2 and 3 and give up his strategies 1

5

and 4. In that case, the probability that A will use his first strategy is p =

and the probability that he

9

4

will use his second strategy is 1-p =

. i.e., Out of a total of 9 trials, he will use his first strategy five

9



308
MBA-H2040 Quantitative Techniques for Managers

10

times and the second strategy four times. The value of the game is

. The positive sign of V shows

9

that the game is favourable to player A.

GRAPHICAL SOLUTION:

Now we consider the graphical method of solution to the given game.


Draw two vertical lines MN and RS. Note that they are parallel to each other. Draw UV

perpendicular to MN as well as RS. Take U as the origin on the line MN. Take V as the origin on the

line RS.

Mark units on MN and RS with equal scale. The units on the two lines MN and RS are taken as

the payoff numbers. The payoffs in the first row of the given matrix are taken along the line MN while

the payoffs in the second row are taken along the line RS.

We have to plot the following points: (8, 3), (-2, 5), (-6, 10), (9, 2). The points 8, -2, -6, 9 are

marked on MN. The points 3, 5, 10, 2 are marked on RS.

Join a point on MN with the corresponding point on RS by a straight line. For example, join the

point 8 on MN with the point 3 on RS. We have 4 such straight lines. They represent the 4 moves of the

second player. They intersect in 6 points. Take the lowermost point of intersection of the straight lines. It

is called the Maximin point. With the help of this point, identify the optimal strategies for the second

player. This point corresponds to the points ?2 and ?6 on MN and 5 and 10 on RS. They correspond to

2

6

the sub game with the matrix

.

5 10

The points ?2 and ?6 on MN correspond to the second and third strategies of the second player.

Therefore, the graphical method implies that, in the long run, the second player will retain his strategies

2 and 3 and give up his strategies 1 and 4.

We graphically solve the sub game with the above matrix. We have to solve the two equations E

= -7 p + 5 and E = - 16 p + 10. Represent the two equations by two straight lines AB and CD on the

5

graph sheet. Take the point of intersection of AB and CD as T. For this point, we have p =

and E =

9

10

10

. Therefore, the value V of the game is

. We see that the probability that first player will use his

9

9

5

4

first strategy is p = and the probability that he will use his second strategy is 1-p =

.

9

9


M R

10 10




309

MBA-H2040 Quantitative Techniques for Managers
9 9


8 8

7
7
6

5 6

4 5

3
4
2
3

1
2

1

Maximin point
U 0 0 V


-1 -1

-2 -2

-3 -3

N S

E

10 C

9

8

7

6
A
5

4

3

2
T
1
P

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

B
D


310
MBA-H2040 Quantitative Techniques for Managers
E=-7P+5
E=-16P+10





E = - 7p+5 E = - 16p+10

p

0

1

0.5

p

0

1

0.5

E

5

- 2 1.5

E 10

- 6

2



QUESTIONS



1. Explain a 2 x n zero-sum game.
2. Describe the method of solution of a 2 x n zero-sum game.
3. Solve the following game:

Player B

10

2 6

Player A



1 5 8

1

Answer: p = , V = 4

3



LESSON 7

m x 2 ZERO-SUM GAMES



LESSON OUTLINE

An m x 2 zero-sum game

Method of solution

Sub game approach and graphical method
Numerical example



LEARNING OBJECTIVES


After reading this lesson you should be able to

- understand the concept of an m x 2 zero-sum game
-

solve numerical problems




The concept of an m x 2 zero-sum game

When the second player B has exactly two strategies and the first player A has m (where m is

three or more) strategies, there results an m x 2 game. It is also called a rectangular game. Since B

has two strategies only, he will find it difficult to discard any one of them. However, since A has



311

MBA-H2040 Quantitative Techniques for Managers
more strategies, he will be in a position to make out some choice among them. He can retain some

of the most advantageous strategies and give up some other strategies. The motive of A is to get as

maximum payoff as possible. Therefore, if some strategies are available to A by which he can get

more payoff to himself, then he will retain such strategies and discard some other strategies which

result in relatively less payoff.


Approaches for m x 2 zero-sum game

There are two approaches for such games: (1) Sub game approach and (2) Graphical approach.


Sub game approach


The given m x 2 game is divided into 2 x 2 sub games. For this purpose, consider all possible 2 x 2 sub

matrices of the payoff matrix of the given game. Solve each sub game and have a list of the values of

each sub game. Since A can make out a choice of his strategies, he will be interested in such of those

sub games which result in more payoff to himself. On the basis of this consideration, in the long run, he

will retain two strategies only and give up the other strategies.

Problem

Solve the following game:

Player B

Strategies

I

II

1

5

8

Player A Strategies

2

2

10



3

12

4

4

6

5

Solution:

Let us consider all possible 2x2 sub games of the given game. We have the following sub games:



5

8

7.



2

10

5

8

8.



12

4

5

8

9.



6

5

2

10

10.



12

4



312
MBA-H2040 Quantitative Techniques for Managers

2

10

11.



6

5

12

4

12.



6

5

Let E be the expected value of the payoff to player A. i.e., the loss to player B. Let r be the probability

that player B will use his first strategy. Then the probability that he will use his second strategy is 1-r.

We form the equations for E in all the sub games as follows:

Sub game (1)



Equation 1: E 5r 8(1 r) 3r 8



Equation 2: E 2

r 10(1 r) 1

2r 10

Sub game (2)



Equation 1: E 5r 8(1 r) 3

r 8



Equation 2: E 12r 4(1 r) 8r 4

Sub game (3)



Equation 1: E 5r 8(1 r) 3

r 8



Equation 2: E 6r 5(1 r) r 5





Sub game (4)



Equation 1: E 2

r 10(1 r) 1

2r 10



Equation 2: E 12r 4(1 r) 8r 4

Sub game (5)



Equation 1: E 2

r 10(1 r) 1

2r 10



Equation 2: E 6r 5(1 r) r 5

Sub game (6)



Equation 1: E 12r 4(1 r) 8r 4



Equation 2: E 6r 5(1 r) r 5

Solve the equations for each 2x2 sub game. Let us tabulate the results for the various sub games. We

have the following:

Sub game

R

Expected value E

1

2

22





9

3



313

MBA-H2040 Quantitative Techniques for Managers

2

4

76





11

11

3

3

23





4

4

4

3

32





10

5

5

5

70





13

13

6

1

36





7

7



Interpretation:

Since player B has only 2 strategies, he cannot make any choice on his strategies. On the other

hand, player A has 4 strategies and so he can retain any 2 strategies and give up the other 2 strategies.

Since the choice is with A, he will try to maximize the payoff to himself. The pay-off to A is the

5

8

maximum in the case of sub game 1. i.e., the sub game with the matrix

.

2

10



Therefore, player A will retain his strategies 1 and 2 and discard his strategies 3 and 4, in the

2

long run. In that case, the probability that B will use his first strategy is r =

and the probability that

9

7

he will use his second strategy is 1-r =

. i.e., Out of a total of 9 trials, he will use his first strategy two

9

times and the second strategy seven times.

22

The value of the game is

. The positive sign of V shows that the game is favourable to player A.

3

GRAPHICAL SOLUTION:

Now we consider the graphical method of solution to the given game.

Draw two vertical lines MN and RS. Note that they are parallel to each other. Draw UV
perpendicular to MN as well as RS. Take U as the origin on the line MN. Take V as the origin on the
line RS.

Mark units on MN and RS with equal scale. The units on the two lines MN and RS are taken as the
payoff numbers. The payoffs in the first row of the given matrix are taken along the line MN while
the payoffs in the second row are taken along the line RS.

We have to plot the following points: (5, 8), (-2, 10), (12, 4), (6, 5).The points 5, -2, 12, 6 are marked
on MN. The points 8, 10, 4, 5 are marked on RS.



314
MBA-H2040 Quantitative Techniques for Managers

Join a point on MN with the corresponding point on RS by a straight line. For example, join the

point 5 on MN with the point 8 on RS. We have 4 such straight lines. They represent the 4 moves of the

first player. They intersect in 6 points. Take the uppermost point of intersection of the straight lines. It is

called the Minimax point. With the help of this point, identify the optimal strategies for the first player.

This point corresponds to the points 5 and -2 on MN and 8 and 10 on RS. They correspond to the sub

5

8

game with the matrix

. The points 5 and -2 on MN correspond to the first and second

2

10

strategies of the first player. Therefore, the graphical method implies that the first player will retain his

strategies 1 and 2 and give up his strategies 3 and 4, in the long run.

We graphically solve the sub game with the above matrix. We have to solve the two equations E

= - 3 r + 8 and E = - 12 r + 10. Represent the two equations by two straight lines AB and CD on the

2

graph sheet. Take the point of intersection of AB and CD as T. For this point, we have r =

and E =

9

22

22

. Therefore, the value V of the game is

. We see that the probability that the second player will

3

3

2

7

use his first strategy is r =

and the probability that he will use his second strategy is 1-r =

.

9

9










315

MBA-H2040 Quantitative Techniques for Managers
M R

12 12

11 11

10 10

9 9

8 minimax 8

7 point 7

6 6

5 5

4 4

3 3

2 2


1 1

U 0 0 V

-1 -1

-2 -2

-3 -3

-4 -4

-5 -5

-6 -6

-7 -7

-8 -8

-9 -9

-10 -10
N S



316
MBA-H2040 Quantitative Techniques for Managers





E
10

9 C

8
A T
7
6

5

4 B
E= -3r+8
3

2

1
r

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0


D
E= -12r+10











E = - 3r+8 E = - 12r+10



p

0

1

0.5

p

0

1

0.5

E

8

5

6.5

E

10

- 2

4







317

MBA-H2040 Quantitative Techniques for Managers
QUESTIONS


1.

What is an m x 2 zero-sum game? Explain.

2.

How will you solve an m x 2 zero-sum game? Explain.

3.

Solve the following game:

Player B

Strategies

I

II

1

20

8

Player A Strategies

2

5

2



3

8

12

1

Answer: r =

, V = 11

4






318
MBA-H2040 Quantitative Techniques for Managers

LESSON 8



LINEAR PROGRAMMING APPROACH TO GAME THEORY



LESSON OUTLINE

4.

How to solve a game with LPP?

5.

Formulation of LPP

6.

Solution by simplex method



LEARNING OBJECTIVES


After reading this lesson you should be able to

- understand the transformation of a game into LPP

- solve a game by simplex method



Introduction


When there is neither saddle point nor dominance in a problem of game theory and the payoff matrix is

of order 3x3 or higher, the probability and graphical methods cannot be employed. In such a case, linear

programming approach may be followed to solve the game.

Linear programming technique:

A general approach to solve a game by linear programming technique is presented below. Consider the

following m n game:

Player B

Y

Y

Y

1

2

n

X a

a

a

1

11

12

1n

Player A X

a

a

a



2

21

22

2n





X a

a

a

m

1

m

m 2

mn

It is required to determine the optimal strategy for A = X , X ,...X

and B = Y ,Y ,...Y . First we

1

2

n

1

2

m

shall determine the optimal strategies of player B.



If player A adopts strategy X , then the expected value of loss to B is

1

a Y a Y ... a Y V ,

11 1

12 2

1n n

where V is the value of game. If A adopts strategy X , then the expected value of loss to B is

2



319

MBA-H2040 Quantitative Techniques for Managers

a Y a Y ... a Y V

21 1

22 2

2n n

and so on. Also we have

Y Y ... Y 1

1

2

n

and

Y 0 for all j.

j

Without loss of generality, we can assume that V > 0. Divide each of the above relation by V and let

Y

'

j

Y

.

j

V

Then we have

Yj

1

'

Y

j

V

V

From this we obtain

'

a Y a Y ' ... a Y ' 1,

11

1

12

2

1n

n

a Y ' a Y ' ... a Y ' 1,

21

1

22

2

2n

n



a Y ' a Y ' ... a Y ' 1

m1

1

m2

2

mn

n

and

1

'

Y Y ' ... Y '



1

2

n

V

with '

Y

0 for all j.

j

The objective of player B is to minimise the loss to himself . Thus the problem is to minimize V, or

1

equivalently to maximise

. Therefore, the objective of player B is to maximise the value of

V

'

Y Y ' ... Y ' subject to the m linear constraints provided above.

1

2

n



Statement of the problem:

Maximise: '

Y Y ' ... Y ' , subject to

1

2

n

'

a Y a Y ' ... a Y ' 1,

11

1

12

2

1n

n

a Y ' a Y ' ... a Y ' 1,

21

1

22

2

2n

n



a Y ' a Y ' ... a Y ' 1

m1

1

m2

2

mn

n

'

Y , Y ' ,...,Y ' 0.

1

2

n



320
MBA-H2040 Quantitative Techniques for Managers
We can use simplex method to solve the above problem. For this purpose, we have to introduce non-

negative slack variables s , s ,..., s to each of the inequalities. So the problem can be restated as

1

2

m

follows:

Restatement of the problem:



Maximise: '

Y Y ' ... Y ' 0s 0s ... 0s subject to

1

2

n

1

2

m

'

a Y a Y ' ... a Y ' s 1,

11

1

12

2

1n

n

1

a Y ' a Y ' ... a Y ' s 1,

21

1

22

2

2n

n

2



a Y ' a Y ' ... a Y ' s 1

m1

1

m2

2

mn

n

m

with Y Y ' V for all j and s 0, s 0,..., s 0 .

j

j

1

2

m

Thus we get the optimal strategy for player B to be (Y ,Y ,...,Y ).

1

2

n

In a similar manner we can determine the optimal strategy for player A.

Application:

We illustrate the method for a 2X2 zero sum game.

Problem 1:

Solve the following game by simplex method for LPP:

Player B

3 6

Player A



5 2

Solution :

Row minima :

I row

: 3







II row

: 2







Maximum of {3,2} = 3



Column maxima:

I column

: 5







II column

: 6







Minimum of {5,6} = 5

So, Maximum of {Row minima} Minimum of {Column maxima}.

Therefore the given game has no saddle point. It is a mixed game. Let us convert the given game into a

LPP.

Problem formulation:



321

MBA-H2040 Quantitative Techniques for Managers
Let V denote the value of the game. Let the probability that the player B will use his first strategy be r

and second strategy be s. Let V denote the value of the game.

When A follows his first strategy:

The expected payoff to A (i.e., the expected loss to B) = 3 r + 6 s.

This pay-off cannot exceed V. So we have

3 r + 6 s V









(1)

When A follows his second strategy:

The expected pay-off to A (i.e., expected loss to B) = 5 r + 2 s.

This cannot exceed V. Hence we obtain the condition

5 r + 2 s V









(2)

From (1) and (2) we have

r

s

3

6 1

V

V



r

s

and 5

2 1

V

V

r

s

Substitute

x,

y .

V

V

Then we have

3x 6 y 1

and 5x 2 y

1

where r and s are connected by the relation

r s 1.

r

s

1

i.e.,





V

V

V

1

i.e., x y



V

1

B will try to minimise V. i.e., He will try to maximise

. Thus we have the following LPP.

V

1

Maximize

x y ,

V

subject to the restrictions

3x 6 y 1,

5x 2 y 1,

x 0, y 0

Solution of LPP:



322
MBA-H2040 Quantitative Techniques for Managers
Introduce two slack variables s , s . Then the problem is transformed into the

1

2

following one:

1

Maximize

x y 0.s 0.s

1

2

V

subject to the constraints

3x 6 y 1.s 0.s 1,

1

2

5x 2 y 0.s 1.s 1,



1

2

x 0, y 0, s 0, s 0

1

2

Let us note that the above equations can be written in the form of a single matrix equation as



A X = B

x



3 6 1

0

y

1



where A =

, X = , B = .

5

2 0 1

s

1

1




s

2

The entries in B are referred to as the b ? values. Initially, the basic variables are s , s . We have the

1

2

following simplex tableau:



x

y

s

s

1

2

b ? value

s - row

1

3

6

1

0

1

s - row

2

5

2

0

1

1

Objective

-1

-1

0

0

0

function row

Consider the negative elements in the objective function row. They are ?1, -1. The absolute values are

1, 1. There is a tie between these coefficients. To resolve the tie, we select the variable x. We take the

new basic variable as x. Consider the ratio of b-value to x-value. We have the following ratios:

1

s - row

:

1

3

1

s - row

:

2

5

1 1

1

Minimum of {

, } = .

3 5

5

Hence select s as the leaving variable. Thus the pivotal element is 5. We obtain the following tableau

2

at the end of Iteration No. 1.



323

MBA-H2040 Quantitative Techniques for Managers



x

y

s

s

1

2

b ? value

s - row

24

3

2

1

0



1





5

5

5

x - row

2

1

1

1



0





5

5

5

Objective

3

1

1

0

0





function row

5

5

5



3

Now, the negative element in the objective function row is

. This corresponds to y. We take the

5

new basic variable as y. Consider the ratio of b-value to y-value. We have the following ratios:

2



5

1

s - row

:



1

24

12

5

1



5

1

x - row

:



2

2



5

1 1

1

Minimum of

,



12

2 12

24

Hence select s as the leaving variable. The pivotal element is

. We get the following tableau at the

1

5

end of Iteration No. 2.





x

y

s

s

1

2

b ? value

y - row

5

1

1

0

1

-



24

8

12

x - row

1

1

1

1

0 -







12

4

6

Objective

1

1

1

0

0







function row

8

8

4



Since both x and y have been made basic variables, we have reached the stopping condition.



324
MBA-H2040 Quantitative Techniques for Managers

1

1

1

1



The optimum value of

is

. This is provided by x =

and y =

. Thus the optimum

V

4

6

12

r

s

4

2

value of the game is obtained as V = 4. Using the relations

x,

y , we obtain r and

V

V

6

3

4

1

s

.

12

3

Problem 2:

Solve the following game:

Player B

2 5

Player A



4 1

Solution:

The game has no saddle point. It is a mixed game. Let the probability that B will use his first strategy

be r. Let the probability that B will use his second strategy be s. Let V be the value of the game.

When A follows his first strategy:



The expected payoff to A (i.e., the expected loss to B) = 2 r +5 s.

The pay-off to A cannot exceed V. So we have

2 r + 5 s V









(I)

When A follows his second strategy:



The expected pay-off to A (i.e., expected loss to B) = 4 r + s.

The pay-off to A cannot exceed V. Hence we obtain the condition

4 r + s V











(II)

From (I) and (II) we have

r

s

2

5 1

V

V



r

s

and 4

1

V

V

Substitute

r

s

x and

.

y

V

V

Thus we have

2x 5 y 1

and 4x y

1

where r and s are connected by the relation

r s 1.



325

MBA-H2040 Quantitative Techniques for Managers

r

s

1

i.e.,





V

V

V

1

i.e., x y



V

1

The objective of B is to minimise V. i.e., He will try to maximise

.

V

Thus we are led to the following linear programming problem:

1

Maximize

= x + y

V

subject to the constraints

2x 5y 1,

4x y 1,

x 0, y 0.

To solve this linear programming problem, we use simplex method as detailed below.

Introduce two slack variables s , s . Then the problem is transformed into the following one:

1

2

1

Maximize

x y 0.s 0.s

1

2

V

subject to the constraints

2x 5 y 1.s 0.s 1,

1

2

4x y 0.s 1.s 1,



1

2

x 0, y 0, s 0, s 0

1

2

We have the following simplex tableau:



x

y

s

s

b ? value

1

2

s - row

2

5

1

0

1

1

s - row

2

4

1

0

1

1

Objective

-1

-1

0

0

0

function row


Consider the negative elements in the objective function row. They are ?1, -1. The absolute value are 1,

1. There is a tie between these coefficients. To resolve the tie, we select the variable x. We take the

new basic variable as x. Consider the ratio of b-value to x-value. We have the following ratios:

1

s - row

:



1

2



326
MBA-H2040 Quantitative Techniques for Managers

1

s - row :



2

4

1 1

1

Minimum of {

,

} =



2 4

4

Hence select s as the leaving variable. Thus the pivotal element is 4. We obtain the following tableau

2

at the end of Iteration No. 1.



x

y

s

s

1

2

b ? value

s - row

9

1

1

1

0



1





2

2

2

x - row

1

1

1

1



0





4

4

4

Objective

3

1

1

0

0





function row

4

4

4



3

Now, the negative element in the objective function row is

. This corresponds to y. We take the

4

new basic variable as y. Consider the ratio of b-value to y-value. We have the following ratios:

1



2

1

s - row

:



1

9

9



2

1





x - row

: 4

1



1



4

1 1

Minimum of , 1

9 9

9

Hence select s as the leaving variable. The pivotal element is

. We get the following tableau at the

1

2

end of Iteration No. 2.





x

y

s

s

b ? value

1

2

y - row

2

1

1

0

1







9

9

9



327

MBA-H2040 Quantitative Techniques for Managers

x - row

1

5

2

1

0







18

18

9

Objective

1

1

1

0

0







function row

6

6

3



Since both x and y have been made basic variables, we have reached the stopping condition.

1

1



The optimum value of

is .

V

3

2

1



This is provided by x =

and y =

. Thus the optimum value of the game is got as V = 3.

9

9

r

s

6

2

3

1

Using the relations

,

x

y , we obtain r and s .

V

V

9

3

9

3



Problem 3:

Solve the following game by simplex method for LPP:

Player B

48

2

Player A



6

4

Solution :

Row minima :

I row

: -48







II row

: -4







Maximum of {-48, -4} = -4

Column maxima:

I column

: 6







II column

: 2







Minimum of (6, 2} = 2

So, Maximum of {Row minima} Minimum of {Column maxima}.

Therefore the given game has no saddle point. It is a mixed game. Let us convert the given game into a

LPP.

Problem formulation:

Let V denote the value of the game. Let the probability that the player B will use his first strategy be r

and second strategy be s. Let V denote the value of the game.

When A follows his first strategy:

The expected payoff to A (i.e., the expected loss to B) = - 48 r + 2 s.

This pay-off cannot exceed V. So we have



328
MBA-H2040 Quantitative Techniques for Managers

- 48 r + 2 s V









(1)'

When A follows his second strategy:



The expected pay-off to A (i.e., expected loss to B) = 6 r - 4 s.

This cannot exceed V. Hence we obtain the condition

6 r - 4 s V









(2)'

From (1)' and (2)' we have

r

s

48 2 1

V

V

and



r

s

6

4 1

V

V

r

s

Substitute

x,

y .

V

V

Thus we have

48x 2y 1

and 6x 4 y



1

where r and s are connected by the relation

r s 1.

r

s

1

i.e.,





V

V

V

1

i.e., x y



V

1

B will try to minimise V. i.e., He will try to maximise

. Thus we have the following LPP.

V

1

Maximize

x y ,

V

subject to the restrictions

48

x 2y 1,

6x 4 y 1,

x 0, y 0

Solution of LPP:

Introduce two slack variables s , s . Then the problem is transformed into the following one:

1

2

1

Maximize

x y 0.s 0.s

1

2

V

subject to the constraints



329

MBA-H2040 Quantitative Techniques for Managers

48

x 2y 1.s 0.s 1,

1

2

6x 4 y 0.s 1.s 1,

1

2

x 0, y 0, s 0, s 0

1

2

Initially, the basic variables are s , s . We have the following simplex tableau:

1

2







x

y

s

s

1

2

b ? value

s - row

1

-48

2

1

0

1

s - row

6

-4

0

1

1

2

Objective

-1

-1

0

0

0

function row



Consider the negative elements in the objective function row. They are ?1, -1. The absolute value are 1,

1. There is a tie between these coefficients. To resolve the tie, we select the variable x. We take the

new basic variable as x. Consider the ratio of b-value to x-value. We have the following ratios:

1

s - row :



1

48

1

s - row:



2

6

1

1

1

Minimum of {

,

} =

.

48

6

48

Hence select s as the leaving variable. Thus the pivotal element is - 48. We obtain the following

1

tableau at the end of Iteration No. 1.





x

y

s

s

1

2

b ? value

x - row

1

1

1

1





0



24

48

48

s - row

15

1

9

2

0



1



4

8

8

Objective

25

1

1

0



0



function row

24

48

48





330
MBA-H2040 Quantitative Techniques for Managers

25

Now, the negative element in the objective function row is

. This corresponds to y. We take the

24

new basic variable as y. Consider the ratio of b-value to y-value. We have the following ratios:

1



48

1

x - row

:



1

2



24

9

82

3

s - row

:





2

15

10



4

1

3

3

Minimum of ,





2 10

10

15

Hence select s as the leaving variable. The pivotal element is

. We get the following tableau at

2

4

the end of Iteration No. 2.





x

y

s

s

1

2

b ? value

x - row

1

1

1

1

0







480

90

30

y - row

1

4

3

0

1







30

15

10

Objective

1

5

1

0

0







function row

18

18

3



Since both x and y have been made basic variables, we have reached the stopping condition.

1

1

1

3



The optimum value of

is

. This is provided by x =

and y =

.

V

3

30

10

r

s

Thus the optimum value of the game is got as V = - 3. Using the relations

x,

y , we obtain

V

V

1

9

r

and s

.

10

10





331

MBA-H2040 Quantitative Techniques for Managers
Problem 4:

Transform the following game into an LPP:

1 8 3

6

4 5



0 1

2

Solution:

We have to determine the optimal strategy for player B. Using the entries of the given matrix, we obtain

the inequalities

r 8 s 3t V ,

6r 4 s 5t V ,

s 2t V

Dividing by V, we get

r

s

t

8

3

1,

V

V

V

r

s

t

6

4

5

1,

V

V

V

s

t

2

1

V

V

subject to the condition

r + s + t = V.

Consequently, we have

r

s

t

1



.

V

V

V

V

Substitute

r

s

t





= x,

= y,

= w.

V

V

V

Then we have the relations

x + 8 y + 3 w 1,

6 x + 4 y + 5 w 1,

y + 2 w 1.

1

r

s

t

We have to minimise V. i.e, We have to maximise

=

. i.e, We have to maximise x + y +

V

V

V

V

w.

Thus, the given game is transformed into the following equivalent LPP:

maximise x + y + w



332
MBA-H2040 Quantitative Techniques for Managers
subject to the restrictions

x + 8 y + 3 w 1,

6 x + 4 y + 5 w 1,

y + 2 w 1,

x 0, y 0, w 0.



QUESTIONS

1. Explain how a game theory problem can be solved as an LPP.

2. Transform the game

Y

Y

1

2

X a

a



1

11

12



X a

a

2

21

22



into an LPP.

3. Using simplex method for LPP, solve the following game:

6 2



3 5

1

1

Answer: r

, s

, V = 4

2

2

4. Solve the following game with LPP approach:

10

6



4 8

1

3

Answer: r

, s

, V = 7

4

4





333

MBA-H2040 Quantitative Techniques for Managers

LESSON 9

GOAL PROGRAMMING FORMULATION


LESSON OUTLINE

7.

Introduction to goal programming

8.

Formulation of goal programming problems


LEARNING OBJECTIVES


After reading this lesson you should be able to

- understand the importance of goal programming

- formulate goal programming problems

Introduction



Generally speaking, the objective of a business organization is to maximize profits and to minimize
expenditure, loss and wastage. However, a company may not always attempt at profit maximization
only. At times, a necessity may arise to pay attention to other objectives also. We describe some
such situations in the sequel.

A manufacturing organization may like to ensure uninterrupted supply of its products even if it
means additional expenditure for the procurement of raw materials or personal delivery of goods
during truckers' strike, etc. with the objective of assuring the good will of the customers.

A company may be interested in the full utilization of the capacity of the machines and therefore
mechanics may be recruited for attending to break downs of the machines even though the
occurrence of such break downs may be very rare.

A company, driven by social consciousness, may spend a portion of its profits on the maintenance of
trees, parks, public roads, etc. to ensure the safety of the environment, with the objective earning the
support of the society.

Another organization may have the objective of establishing brand name by providing high quality
products to the consumers and for this purpose it may introduce rigorous measures of quality checks
even though it may involve an increased expenditure.

While all the sales persons in a company are formally trained and highly experienced, the
management may still pursue a policy to depute them for periodical training in reputed institutes so
as to maximize their capability, without minding the extra expenditure incurred for their training.

A travel agency may be interested to ensure customer satisfaction of the highest order and as a
consequence it may come forward to operate bus services even to remote places at the normal rates,
so as to retain the customers in its fold.

A bank may offer services beyond normal working hours or on holidays even if it means payment of
overtime to the staff, in order to adhere to the policy of customer satisfaction on priority basis.

A business organization may accord priority for the welfare of the employees and so a major part of
the earnings may be apportioned on employee welfare measures.

A garment designer would like to be always known for the latest fashion and hence may spend more
money on fashion design but sell the products at the normal rates, so as to earn the maximum
reputation.



334
MBA-H2040 Quantitative Techniques for Managers

A newspaper may be interested in earning the unique distinction of `Reporter of Remote Rural
Areas' and so it may spend more money on journalists and advanced technology for communication.

The above typical instances go to show that the top level management of a business organization
may embark upon different goals in addition to profit maximization. Such goals may be necessitated
by external events or through internal discussion. At times, one such goal may be in conflict with
another goal.

`Goal programming' seeks to deal with the process of decision making in a situation of multiple
goals set forth by a business organization. A management may accord equal priority to different
goals or sometimes a hierarchy of goals may be prescribed on their importance. One has to strive to
achieve the goals in accordance with the priorities specified by the management. Sometimes the
goals may be classified as higher level goals and lower level goals as perceived by the
management and one would be interested in first achieving the higher order goals and afterwards
considering lower order goals.

Some of the goals that may be preferred by a business organization are : maximum customer
satisfaction, maximum good will of the customers, maximum utilization of the machine capacity,
maximum reliability of the products, maximum support of the society, maximum utilization of the
work force, maximum welfare of the employees, etc.

Since different goals of an organization are based on different units, the goal programming has a
multi-dimensional objective function. This is in contrast with a linear programming problem in
which the objective function is uni-dimensional.

Given a goal of an organization, one has to determine the conditions under which there will be
under-achievement and over-achievement of the goal. The ideal situation will be the one with neither
under-achievement nor over-achievement of the goal.

Formulation of Goal Programming Problems

In the sequel, we consider illustrative situations so as to explain the process of problem formulation
in goal programming.

Notations

If there is a single goal, we have the following notations:



Let Du denote the under-achievement of the goal.

Let Do denote the over-achievement of the goal.

If there are two goals, we have the following notations:

Denote the under-achievement and the over-achievement of one goal by Du1 and Do1
respectively.

Denote the under-achievement and the over-achievement of another goal by Du2 and Do2
respectively.



PROBLEM 1:

Alpha company is known for the manufacture of tables and chairs. There is a profit of Rs. 200 per table

and Rs. 80 per chair. Production of a table requires 5 hours of assembly and 3 hours in finishing. In

order to produce a chair, the requirements are 3 hours of assembly and 2 hours of finishing. The

company has 105 hours of assembly time and 65 hours of finishing. The company manager is



335

MBA-H2040 Quantitative Techniques for Managers
interested to find out the optimal production of tables and chairs so as to have a maximum profit of Rs.

4000. Formulate a goal programming problem for this situation.





Solution:

The manager is interested not only in the maximization of profit but he has also fixed a target of Rs.

4000 as profit. Thus, the problem involves a single goal of achieving the specified amount of profit.

Let D denote the under achievement of the target profit and let D be the over achievement.

u

o



The objective in the given situation is to minimize under achievement. Let Z be the objective

function. Then the problem is the minimization of Z D .

u

Formulation of the constraints:

Let the number of tables to be produced be x and let the number of chairs to be produced be Y.



Profit from x tables = Rs. 200 x



Profit from y chairs = Rs. 80 y

The total profit =

Profit from x tables and y chairs

+ under achievement of the profit target

? over achievement of the profit target

So we have the relationship 200 x + 80 y + D - D = 4000.

u

o

Assembly time:

To produce x tables, the requirement of assembly time = 5 x hours. To produce y

chairs, the requirement is 3 y hours. So, the total requirement is 5 x 3 y hours. But the available time

for assembly is 105 hours. Therefore constraint





5 x + 3 y 105

must be fulfilled.

Finishing time:

To produce x tables, the requirement of finishing time = 3 x. To produce y chairs, the requirement is 2 y .

So, the total requirement is 3 x+ 2 y. But the availability is 65 hours. Hence we have the restriction

3x 2 y 65 .

Non-negativity restrictions:

The number of tables and chairs produced, the under achievement of the profit target and the

over achievement cannot be negative. Thus we have the restrictions

x 0, y 0, D 0, D 0 .

u

o



336
MBA-H2040 Quantitative Techniques for Managers
Statement of the problem:

Minimize Z D

u

subject to the constraints

200 x 80 y D D 4000

u

o

5 x 3 y 105



3 x 2 y 65

x 0, y 0, D 0, D 0

u

o

Problem 2:

Sweet Bakery Ltd. produces two recipes A and B. Both recipes are made of two food stuffs I and II.

Production of one Kg of A requires 7 units of food stuff I and 4 units of food stuff II whereas for

producing one Kg of B, 4 units of food stuff I and 3 units of food stuff II are required. The company

has 145 units of food stuff I and 90 units of food stuff II. The profit per Kg of A is Rs. 120 while that of

B is Rs. 90. The manager wants to earn a maximum profit of Rs. 2700 and to fulfil the demand of 12

Kgs of A. Formulate a goal programming problem for this situation.



Solution:



The management has two goals.

1. To reach a profit of Rs. 2700

2. Production of 12 Kgs of recipe A.



Let D denote the under achievement of the profit target.

up

Let D denote the over achievement of the profit target.

op

Let D denote the under achievement of the production target for recipe A.

uA

Let D denote the over achievement of the production target for recipe A.

oA



The objective in this problem is to minimize the under achievement of the profit target and to

minimize the under achievement of the production target for recipe A.

Let Z be the objective function. Then the problem is the minimization of

Z D D .

up

uA



Constraints

Suppose the company has to produce x kgs of recipe A and y kgs of recipe B in order to achieve the two

goals.



337

MBA-H2040 Quantitative Techniques for Managers
Condition on profit:

Profit from x kgs of A = 120 x

Profit from y kgs of B = 90 y

The total profit

=

Profit from x kgs of A + Profit from y kgs of B

+ under achievement of the profit target

? over achievement of the profit target







=

120 x 90 y D D

up

op

Thus we have the restriction

120x 90 y D D 2700 .

up

op

Constraint for food stuff I:

7 x 4 y 145

Constraint for food stuff II:

4 x 3 y 90

Non-negativity restrictions:

x, y, D , D , D , D 0

up

op

uA

oA

Condition on recipe A:

The target production of A =

optimal production of A

+ under achievement in production target of A

? over achievement of the production target of A.

Thus we have the condition

x D D

12

uA

oA

Statement of the problem:

Minimize Z D D

up

uA

subject to the constraints

120 x 90 y D D 2700

up

op

x D D

12

uA

oA

7 x 4 y 145



4 x 3 y 90

x, y, D , D , D , D 0

up

op

uA

oA



QUESTIONS

1. Explain the necessity of a goal programming.



338
MBA-H2040 Quantitative Techniques for Managers

2. Describe some instances of goal programming.

3. Explain the formulation of a goal programming problem.













339

MBA-H2040 Quantitative Techniques for Managers

UNIT ? V

Lesson Outline

Introduction
Terminologies of Queueing System
Empirical Queueing Models
Simulation ? Introduction
Types of Simulation
Major Steps of Simulation
Replacement and Maintenance Analysis ? Introduction
Types of Maintenance
Types of Replacement Problem
Determination of Economic life of an Asset


Learning Objectives


After reading this lesson you should be able to


Understand the nature and scope of Queneing System.
Queueing models and the solution to queueing model problems.
Importance of Simulation
Need for Replacement and maintenance
Solution to problems involving economic life of an Asset.



340
MBA-H2040 Quantitative Techniques for Managers

5-1 ? Queueing Theory

5.1.1 Introduction:

A flow of customers from finite or infinite population towards the service facility forms a queue (waiting line) an account of

lack of capability to serve them all at a time. In the absence of a perfect balance between the service facilities and the

customers, waiting time is required either for the service facilities or for the customers arrival. In general, the queueing

system consists of one or more queues and one or more servers and operates under a set of procedures. Depending upon the

server status, the incoming customer either waits at the queue or gets the turn to be served. If the server is free at the time of

arrival of a customer, the customer can directly enter into the counter for getting service and then leave the system. In this

process, over a period of time, the system may experience " Customer waiting" and /or "Server idle time"



5.1.2 Queueing System:

A queueing system can be completely described by


(1) the input (arrival pattern)
(2) the service mechanism (service pattern)
(3) The queue discipline and
(4) Customer's behaviour


5.1.3. The input (arrival pattern)

The input described the way in which the customers arrive and join the system. Generally, customers arrive in a more or less

random manner which is not possible for prediction. Thus the arrival pattern can be described in terms of probabilities and

consequently the probability distribution for inter-arrival times (the time between two successive arrivals) must be defined.

We deal with those Queueing system in which the customers arrive in poisson process. The mean arrival rate is denoted by

.





341

MBA-H2040 Quantitative Techniques for Managers
5.1.4 The Service Mechanism:-

This means the arrangement of service facility to serve customers. If there is infinite number of servers, then all the

customers are served instantaneously or arrival and there will be no queue. If the number of servers is finite then the

customers are served according to a specific order with service time a constant or a random variable. Distribution of service

time follows `Exponential distribution' defined by





f(t) = e -t , t > 0



The mean Service rate is E(t) = 1/

5.1.5 Queueing Discipline:-

It is a rule according to which the customers are selected for service when a queue has been formed. The most common
disciplines are


1. First come first served ? (FCFS)
2. First in first out ? (FIFO)
3. Last in first out ? (LIFO)
4. Selection for service in random order (SIRO)


5.1.6 Customer's behaviour


1. Generally, it is assumed that the customers arrive into the system one by one. But in some cases, customers

may arrive in groups. Such arrival is called Bulk arrival.


2. If there is more than one queue, the customers from one queue may be tempted to join another queue because of

its smaller size. This behaviour of customers is known as jockeying.



3. If the queue length appears very large to a customer, he/she may not join the queue. This property is known as

Balking of customers.



4. Sometimes, a customer who is already in a queue will leave the queue in anticipation of longer waiting line.

This kind of departare is known as reneging.


5.1.7 List of Variables

The list of variable used in queueing models is give below:


n - No of customers in the system



C - No of servers in the system



Pn (t) ? Probability of having n customers in the system at time t.



Pn - Steady state probability of having customers in the
system



P0 - Probability of having zero customer in the system



Lq - Average number of customers waiting in the queue.



Ls - Average number of customers waiting in the system
(in the queue and in the service counters)
Wq - Average waiting time of customers in the queue.
Ws - Average waiting time of customers in the system
(in the queue and in the service counters)
- Arrival rate of customers

- Service rate of server

- Utilization factor of the server

eff - Effective rate of arrival of customers
M - Poisson distribution
N - Maximum numbers of customers permitted in the system. Also, it denotes the size of the calling source of the
customers.
GD - General discipline for service. This may be first in first ? serve (FIFS), last-in-first serve (LIFS) random order
(Ro) etc.




342
MBA-H2040 Quantitative Techniques for Managers
5.1.8 Traffic intensity (or utilization factor)

An important measure of a simple queue is its traffic intensity given by



Traffic intensity = Mean arrival time

=



(< 1)









Mean service time













and the unit of traffic intensity is Erlang


5.1.9 Classification of Queueing models

Generally, queueing models can be classified into six categories using Kendall's notation with six parameters to define a

model. The parameters of this notation are

P- Arrival rate distribution ie probability law for the arrival /inter ? arrival time.



Q - Service rate distribution, ie probability law according to

which the customers are being served.

R - Number of Servers (ie number of service stations)
X - Service discipline
Y - Maximum number of customers permitted in the system.
Z - Size of the calling source of the customers.

A queuing model with the above parameters is written as
(P/Q/R : X/Y/Z)


5.1.10 Model 1 : (M/M/1) : (GD/ / ) Model



In this model

(i)

the arrival rate follows poisson (M) distribution.

(ii)

Service rate follows poisson distribution (M)

(iii)

Number of servers is 1

(iv)

Service discipline is general disciple (ie GD)

(v)

Maximum number of customers permitted in the system is infinite ()

(vi)

Size of the calling source is infinite ()


The steady state equations to obtain, Pn the probability of having customers in the system and the values for Ls, Lq, Ws and
Wq are given below.

n= 0,1,2,---- where = <1

















Ls ? Average number of customers waiting in the system
(ie waiting in the queue and in the service station)





P

n

n =

(1-)




Ls =






1-




Lq

=

Ls ?













=

-



1 -



343

MBA-H2040 Quantitative Techniques for Managers



=

- (1 - )





1 -









Lq

=

2


Average

waiting



time of

1 -c

u stomers in the system

(in the que



ue an

d in the service station)

= Ws

= Ws

= Ls =









(1 - )



=

x

1

=

1



1 -







(1 - )

















(Since = )











= 1

-

=

1

-




Ws = 1

-





Wq = Average waiting time of customers in the
queue.


=

Lq / = [1 / ] [ 2 / [1- ]]







= 1 / [ 2 / [1- ]]











=

Since =













-






























Wq =









-

Example 1:

The arrival rate of customers at a banking counter follows a poisson distibution with a mean of 30 per hours. The service rate
of the counter clerk also follows poisson distribution with mean of 45 per hour.


a) What is the probability of having zero customer in the system ?
b) What is the probability of having 8 customer in the system ?
c) What is the probability of having 12 customer in the system ?
d) Find Ls, Lq, Ws and Wq


Solution


Given arrival rate follows poisson distribution with
mean =30

= 30 per hour

Given service rate follows poisson distribution with

mean = 45

= 45 Per hour

Utilization factor = /






= 30/45







= 2/3







= 0.67

a) The probability of having zero customer in the system




P0

= 0 (1- )







= 1-







= 1-0.67



344
MBA-H2040 Quantitative Techniques for Managers







= 0.33

b) The probability of having 8 customers in the system




P8

= 8 (1- )







= (0.67)8 (1-0.67)







= 0.0406 x 0.33







= 0.0134

Probability of having 12 customers in the system is




P12

= 12 (1- )







= (0.67)12 (1-0.67)







= 0.0082 x 0.33 = 0.002706

Ls

=



=

0.67





1 -

1-0.67



=

0.67

= 2.03

0.33

= 2 customers




Lq

= 2



=

(0.67)2

=

0.4489

1-





1-0.67



0.33

=

1.36

=

1 Customer




Ws

=

1

=

1

=

1







-



45-30



15













=

0.0666 hour




Wq

=





=

0.67

=

0.67

-





45-30



15


=

0.4467 hour

Example 2 :




At one-man barbar shop, customers arrive according to poisson dist with mean arrival rate of 5 per hour

and the hair cutting time was exponentially distributed with an average hair cut taking 10 minutes. It is

assumed that because of his excellent reputation, customers were always willing to wait. Calculate the

following:



(i)

Average number of customers in the shop and the average numbers waiting for a haircut.

(ii)

The percentage of time arrival can walk in straight without having to wait.

(iii)

The percentage of customers who have to wait before getting into the barber's chair.

Solution:-



Given mean arrival of customer = 5/60 =1/12
and mean time for server



=

1/10

= /

= [1/12] x 10

=

10 /12











=

0.833

(i)

Average number of customers in the system (numbers in the queue and in the service station)







Ls

= / 1-

= 0.83 / 1- 0.83













= 0.83 / 0.17













= 4.88









= 5 Customers

(ii)

The percentage of time arrival can walk straight into barber's chair without waiting is


Service utilization

= %









= / %









= 0.833 x 100









=83.3



345

MBA-H2040 Quantitative Techniques for Managers

(iii)

The percentage of customers who have to wait before getting into the barber's chair = (1-)%





(1-0.833)%

=

0.167 x 100







=

16.7%

Example 3 :

Vehicles are passing through a toll gate at the rate of 70 per hour. The average time to pass through the
gate is 45 seconds. The arrival rate and service rate follow poisson distibution. There is a complaint
that the vehicles wait for a long duration. The authorities are willing to install one more gate to reduce
the average time to pass through the toll gate to 35 seconds if the idle time of the toll gate is less than
9% and the average queue length at the gate is more than 8 vehicle, check whether the installation of the
second gate is justified?

Solutions:-


Arrival rate of vehicles at the toll gate

=

70 per hour



Time taken to pass through the gate =

45 Seconds





Service rate

= 1 hours













45 seconds











= 3600/45

= 80



= 80 Vehicles per hour







Utilization factor = /













= 70 / 80













= 0.875



(a) Waiting no. of vehicles in the queue is Lq





Lq

= 2 / 1 -

=

(0.875)2













1-0.875











=

0.7656













0.125











=

6.125











= 6 Vehicles



(b) Revised time taken to pass through the gate

=30 seconds





The new service rate after installation of an
additional gate = 1 hour/35 Seconds

= 3600/35







=

102.68 Vehicles / hour





Utilization factor = / = 70 / 102.86







= 0.681











Percentage of idle time of the gate = (1-)%











= (1-0.681)%











= 0.319%











= 31.9











= 32%




This idle time is not less than 9% which is expected.




Therefore

the installation of the second gate is not justified since the average waiting number of vehicles in

the queue is more than 8 but the idle time is not less than 32%. Hence idle time is far greater than the number. of vehicles
waiting in the queue.



5.1.11 Second Model (M/M/C) : (GD/ / )Model



The parameters of this model are as follows:



(i)

Arrival rate follows poisson distribution

(ii)

Service rate follows poisson distribution

(iii)

No of servers is C'.

(iv)

Service discipline is general discipline.



346
MBA-H2040 Quantitative Techniques for Managers

(v)

Maximum number of customers permitted in the system is infinite


Then the steady state equation to obtain the probability of having n customers in the system is



Pn

= n Po



,

o n C

n!





= n Po

for n > c Where / c < 1





C n-c C!









Where [ / c] < 1 as = /

C-1









P0

={[ n/n!] + c / (c! [1 - /c])}-1

n = 0



where c! = 1 x 2 x 3 x ................. upto C
Lq

= [ c+1 / [c-1! (c - )] ] x P0



= (c Pc) / (c - )2

Ls

= Lq + and Ws = Wq + 1 /

Wq

= Lq /


Under special conditions Po

= 1 - and Lq = C+! / c 2 Where <1 and

Po = (C-) (c ? 1)! / c c
and L



q =

/ (c- ), where / c < 1


Example 1:

At a central warehouse, vehicles are at the rate of 24 per hour and the arrival rate follows poisson distribution. The unloading
time of the vehicles follows exponential distribution and the unloading rate is 18 vehicles per hour. There are 4 unloading
crews. Find

(i)

Po and P3

(ii)

Lq, Ls, Wq and Ws

Solution:


Arrival rate = 24 per hour



Unloading rate = 18 Per hour



No. of unloading crews C=4





= /

= 24 / 18=1.33



C-1









(i) P0

={[ n/n!] + c / (c! [1 - /c])}-1
n = 0
3

={[ (1.33)n/n!]+ (1.33)4 /(4! [1 - (1.33)/ 4])}-1


n = 0

={ (1.33)0 / = 0! + (1.33)1 / 1! + (1.33)2 / 2! + (1.33)3 / 3! +
(1.33)4 / 24! [1 - (1.33)/ 4] }-1
=[1 + 1.33 + 0.88 + 0.39 + 3.129/16.62] -1

=[3.60 + 0.19]-1 =

[3.79]-1

= 0.264

We know Pn

= ( n / n!) Po



for

0 n c



P3

= ( 3 / 3!) Po



Since 0 3 4





=

[(1.33)3 / 6 ] x 0.264





=

2.353 x 0.044





=

0.1035


(ii) Lq

=

C+1 X P0







(C ? 1)! (C-)2











=

(1.33)5

X

0.264











3! X (4 ? 1.33)2





=

(4.1616) X

0.264











6 X (2.77)2



347

MBA-H2040 Quantitative Techniques for Managers




=

(4.1616) X

0.264











46 .0374





=

1.099 / 46.0374





=

0.0239





=

0.0239 Vehicles

Ls

=

Lq +

=

0.0239 + 1.33









=

1.3539 Vehicles

Wq

=

Lq /



=

0.0239 /24









=

0.000996 hrs


Ws

=

Wq + 1 /

=

0.000996 + 1/18









=

0.000996 + 0.055555









=

0.056551 hours.



Example 2 :-

A supermarket has two girls ringing up sales at the counters. If the service time for each customer is exponential with mean 4
minutes and if the people arrive in poisson fashion at the rate of 10 per hour


a) What is the probability of having to wait for service?
b) What is the expected percentage of idle time for each girl?
c) If a customer has to wait, what is the expected length of his waiting time?


Solution:-

C-1









P0

={[ n/n!] + c / (c! [1 - /c])}-1
n = 0


Where = / given arrival rate = 10 per hour

= 10 / 60 = 1 / 6 per minute

Service rate = 4 minutes

= 1 / 4

person per minute

Hence = /

= (1 / 6) x 4

= 2 / 3

= 0.67

1









P0 ={[ n/n!]+(0.67)2 / (2! [1 - 0.67/2])}-1

n = 0

=[1 + ( / 1!) ] + 0.4489 / (2 ? 0.67)]-1
=[1 + 0.67 + 0.4489 / (1.33)]-1
=[1 + 0.67 + 0.34]-1
=[ 2.01]-1
= 1 / 2

The Probability of having to wait for the service is


P (w > 0)



=

c

X P0

c! [1 - /c]

=

0.67 2 X (1 / 2)
2! [1 ? 0.67 /2]



=

0.4489 / 2.66



=

0.168





b) The probability of idle time for each girl is






= 1- P (w > 0)



348
MBA-H2040 Quantitative Techniques for Managers






= 1-1/3







= 2/3



Percentage of time the service remains idle = 67% approximately




c) The expected length of waiting time (w/w>0)







=

1 / (c - )



=

1 / [(1 / 2) ? (1 / 6) ]

=

3 minutes

Examples 3 :

A petrol station has two pumps. The service time follows the exponential distribution with mean 4 minutes and cars arrive
for service in a poisson process at the rate of 10 cars per hour. Find the probability that a customer has to wait for service.
What proportion of time the pump remains idle?

Solution:

Given C=2

The arrival rate = 10 cars per hour.







= 10 / 60

= 1 / 6 car per minute







Service rate = 4 minute per cars.





Ie

= ? car per minute.



= /

= (1/6) / (1/4)







= 2 / 3







= 0.67

Proportion of time the pumps remain busy

= / c

= 0.67 / 2

= 0.33
= 1 / 3

The proportion of time, the pumps remain idle

=1 ? proportion of the pumps remain busy







=

1-1 / 3

= 2 / 3

C-1









P0

={[ n/n!] + c / (c! [1 - /c])}-1
n = 0


=[ ( 0.67)0 / 0!) + ( 0.67)1 / 1!) + ( 0.67)2 / 2!)[1- ( 0.67 / 2)1 ]-1
=[1 + 0.67 + 0.4489 / (1.33)]-1
=[1 + 0.67 + 0.33]-1
=[ 2]-1
=1 / 2

Probability that a customer has to wait for service




=

p [w>0]





=

c

x P0 =

(0.67)2 x 1/2







[c [1 - / c]





[2![1 ? 0.67/2]





=

0.4489



=

0.4489







1.33x2





2.66





=

0.1688

5.2 Simulation :

Simulation is an experiment conducted on a model of some system to collect necessary information on the behaviour of that
system.

5.2.1 Introduction :

The representation of reality in some physical form or in some form of Mathematical equations
are called Simulations .



Simulations are imitation of reality.


For example :



349

MBA-H2040 Quantitative Techniques for Managers

1. Children cycling park with various signals and crossing is a simulation of a read model traffic system
2. Planetarium
3. Testing an air craft model in a wind tunnel.


5.2.2. Need for simulation :

Consider an example of the queueing system, namely the reservation system of a transport corporation.

The elements of the system are booking counters (servers) and waiting customers (queue). Generally

the arrival rate of customers follow a Poisson distribution and the service time follows exponential

distribution. Then the queueing model (M/M/1) : (GD/ / ) can be used to find the standard results.

But in reality, the following combinations of distributions my exist.



1. Arrived rate does not follow Poisson distribution, but the service rate follows an exponential distribution.

2. Arrival rate follows a Poisson distribution and the service rate does not follow exponential distribution.



3. Arrival rate does not follows poisson distribution and the service time also does not follow exponential distribution.

In each of the above cases, the standard model (M/M/1) : (G/D/ / ) cannot be used. The last resort to find the
solution for such a queueing problem is to use simulation.


5.2.3. Some advantage of simulation :


1. Simulation is Mathematically less complicated
2. Simulation is flexible
3. It can be modified to suit the changing environments.
4. It can be used for training purpose
5. It may be less expensive and less time consuming in a quite a few real world situations.


5.2.4. Some Limitations of Simulation :


1. Quantification or Enlarging of the variables maybe difficult.
2. Large number of variables make simulations unwieldy and more difficult.
3. Simulation may not. Yield optimum or accurate results.
4. Simulation are most expensive and time consuming model.
5. We cannot relay too much on the results obtained from simulation models.


5.2.5. Steps in simulation :


1. Identify the measure of effectiveness.
2. Decide the variables which influence the measure of effectiveness and choose those variables, which affects the

measure of effectiveness significantly.

3. Determine the probability distribution for each variable in step 2 and construct the cumulative probability

distribution.

4. Choose an appropriate set of random numbers.
5. Consider each random number as decimal value of the cumulative probability distribution.
6. Use the simulated values so generated into the formula derived from the measure of effectiveness.
7. Repeat steps 5 and 6 until the sample is large enough to arrive at a satisfactory and reliable decision.


5.2.6. Uses of Simulation

Simulation is used for solving

1.Inventory Problem
2. Queueing Problem
3. Training Programmes etc.

Example :




350
MBA-H2040 Quantitative Techniques for Managers
Customers arrive at a milk booth for the required service. Assume that inter ? arrival and service time

are constants and given by 1.5 and 4 minutes respectively. Simulate the system by hand computations

for 14 minutes.


(i) What is the waiting time per customer?
(ii) What is the percentage idle time for the facility?
(Assume that the system starts at t = 0)

Solution :

First customer arrives at the service center at t = 0
His departure time after getting service = 0 + 4 = 4 minutes.
Second customer arrives at time t = 1.5 minutes
he has to wait = 4 ? 1.5 = 2.5 minutes.
Third customer arrives at time t = 3 minutes
he has to wait for = 8-3 = 5 minutes
Fourth customer arrives at time t = 4.5 minutes and he has to wait for 12 ? 4.5 = 7.5 minutes.
During this 4.5 minutes, the first customer leaves in 4 minutes after getting service and the second customer is getting
service.
Fifth customer arrives at t = 6 minutes
he has to wait 14 ? 6 = 8 minutes
Sixth customer arrives at t = 7.5 minutes
he has to wait 14 ? 7.5 = 6.5 minutes
Seventh customer arrives at t = 9 minutes
he has to wait 14 ? 9 = 5 minutes
During this 9 minutes the second customer leaves the service in 8th minute and third person is to get service in 9th minute.
Eighth customer arrives at t = 10.5 minutes
he has to wait 14 ? 10.5 = 3.5 minutes
Nineth customer arrives at t = 12 minutes

he has to wait 14 ? 12 = 2 minutes
But by 12th minute the third customer leaves the Service
10th Customer arrives at t = 13.5 minutes
he has to wait 14-13.5 = 0.5 minute

From this simulation table it is clear that


(i) Average waiting time for 10 customers


= 2.5+5+7.5+8+6.5+5.0+3.5+2+0.5







10



= 40.5 = 4.05

10
(ii) Average waiting time for 9 customers who are in waiting for service

40.5

= 4.5 minutes.

9
But the average service time is 4 minutes which is less that the average waiting time, the percentage of idle time for service =
0%

Exercise :

1. The arrival rate of customers at a banking counter follows a poisson distribution with a mean of 45
per hors. The service rate of the counter clerk also poisson distribution with a mean of 60 per hours.


(a) What is the probability of having Zero customer in the system (P0).
(b) What is the probability of having 5 customer in the system (P5).
(c) What is the probability of having 10 customer in the system (P10).
(d) Find Ls, Lq, Ws and Wq



351

MBA-H2040 Quantitative Techniques for Managers

2. Vehicles pass through a toll gate at a rate of 90 per hour. The average time to pass through the gate is
36 seconds. The arrival rate and service rate follow poisson distribution. There is a complaint the
vehicles wait for long duration. The authorities are willing to install one more gate to reduce the
average time to pass through the toll gate to 30 seconds if the idle time of the toll gate is less than 10%
and the average queue length at the gate is more than 5 vehicles. Vehicle whether the installation of
second gate is justified?

3. At a central ware house, vehicles arrive at the rate of 24 per hours and the arrival rate follows poisson
distribution. The unloading time of the vehicles follows exponentional distribution and the unloading
rate is 18 vehicles per hour. There are 4 unloading crews. Find the following.


a) P0 and P3
b) Lq, Ls, Wq and Ws


4.

Explain Queneing Discipline

5.

Describe the Queueing models (M/M/1) : (GD/ /)



and (M/M/C) : (GD/ / )


6.

Cars arrive at a drive-in restaurant with mean arrival rate of 30 cars per hors and the service rate

of the cars is 22 per hors. The arrival rate and the service rate follow poisson distribution. The number
parking space for cars is only 5. Find the standard results.
(Ans Lq = 2.38 cars, Ls = 3.3133 Cars, Wq = 0.116 hors and
Ws = 0.1615 hors)

7.

In a harbour, ship arrive with a mean rate of 24 per week. The harbour has 3 docks to handle

unloading and loading of ships. The service rate of individual dock is 12 per week. The arrival rate and
the service rate follow poisson distribution. At any point of time, the maximum No. of ships permitted
in the harbour is 8. Find P0, Lq, Ls, Wq, Ws
(Ans P0 = 0.1998, Lq = 1.0371 ships, Ls = 2.9671 ships,
Wq = 0.04478 week and Ws = 0.1281 week)

8.

Define simulation and its advantages.

9.

Discuss the steps of simulation.


5.3. Replacement models

5.3.1. Introduction:

The replacement problems are concerned with the situations that arise when some items such as men,

machines and usable things etc need replacement due to their decreased efficiency, failure or

breakdown. Such decreased efficiency or complete breakdown may either be gradual or all of a sudden.




If a firm wants to survive the competition it has to decide on whether to replace the out dated

equipment or to retain it, by taking the cost of maintenance and operation into account. There are two

basic reasons for considering the replacement of an equipment.

They are



352
MBA-H2040 Quantitative Techniques for Managers

(i)

Physical impairment or malfunctioning of various parts.

(ii)

Obsolescence of the equipment.



The physical impairment refers only to changes in the physical condition of the equipment itself.

This will lead to decline in the value of service rendered by the equipment, increased operating cost of

the equipments, increased maintenance cost of the equipment or the combination of these costs.

Obsolescence is caused due to improvement in the existing Tools and machinery mainly when the

technology becomes advanced therefore, it becomes uneconomical to continue production with the same

equipment under any of the above situations. Hence the equipments are to be periodically replaced.

Some times, the capacity of existing facilities may be in adequate to meet the current demand. Under

such cases, the following two alternatives will be considered.

1. Replacement of the existing equipment with a new one
2. Argument the existing one with an additional equipments.


5.3.2 Type of Maintenance


Maintenance activity can be classified into two types

i)

Preventive Maintenance

ii)

Breakdown Maintenance



Preventive maintenance (PN) is the periodical inspection and service which are aimed to detect

potential failures and perform minor adjustments a requires which will prevent major operating problem

in future. Breakdown maintenance is the repair which is generally done after the equipment breaks

down. It is offer an emergency which will have an associated penalty in terms of increasing the cost of

maintenance and downtime cost of equipment, Preventive maintenance will reduce such costs up-to a

certain extent . Beyond that the cost of preventive maintenance will be more when compared to the cost

of the breakdown maintenance.

Total cost = Preventive maintenance cost + Breakdown maintenance cost.




This total cost will go on decreasing up-to P with an increase in the level of maintenance up-to a

point, beyond which the total cost will start increasing from P. The level of maintenance corresponding

to the minimum total cost at P is the Optional level of maintenance this concept is illustrated in the

follows diagram



353

MBA-H2040 Quantitative Techniques for Managers


N






M









The points M and N denote optimal level of maintenance and optimal cost respectively


5.3.3 Types of replacement problem

The replacement problem can be classified into two categories.

i)

Replacement of assets that deteriorate with time (replacement due to gradual failure, due to wear
and tear of the components of the machines) This can be further classified into the following
types.



a)

Determination of economic type of an asset.



b)

Replacement of an existing asset with a new asset.

ii)

simple probabilistic model for assets which will fail completely (replacement due to sudden
failure).


5.3.4. Determination of Economic Life of an asset

Any asset will have the following cost components
i)

Capital recovery cost (average first cost), Computed form the first cost (Purchase price) of the
asset.

ii)

Average operating and maintenance cost.

iii)

Total cost which is the sum of capital recovery cost (average first cost) and average operating
and maintenance cost.


A typical shape of each of the above cost with respect to
life of the

asset is shown below
















354
MBA-H2040 Quantitative Techniques for Managers


From figure, when the life of the machine increases, it is clear that the capital recovery cost

(average first cost) goes on decreasing and the average operating and maintenance cost goes on

increasing. From the beginning the total cost goes on decreasing upto a particular life of the asset and

then it starts increasing. The point P were the total cost in the minimum is called the Economic life of

the asset. To solve problems under replacement, we consider the basics of interest formula.

Present worth factor denoted by (P/F, i,n). If an amount P is invested now with amount earning

interest at the rate i per year, then the future sum (F) accumulated after n years can be obtained.

P

-

Principal sum at year Zero

F

-

Future sum of P at the end of the nth year

i

-

Annual interest rate

n

-

Number of interest periods.



Then the formula for future sum F = P ( 1 + i ) n

P = F/(1 +i)n = Fx (present worth factor)

If A is the annual equivalent amount which occurs at the end of every year from year one through n
years is given by




A

=

P x i (1 +i)n









(1 +i)n - 1



=

P ( A / P, i, n )

=

P x equal payment series capital recovery factor


Example:

A firm is considering replacement of an equipment whose first cost is Rs. 1750 and the scrap value is
negligible at any year. Based on experience, it is found that maintenance cost is zero during the first
year and it increases by Rs. 100 every year thereafter.



(i) When should be the equipment replaced if

a) i = 0%
b) i = 12%

Solution :


Given the first cost = Rs 1750 and the maintenance cost is Rs. Zero during the first years and

then increases by Rs. 100 every year thereafter. Then the following table shows the calculation.

Calculations to determine Economic life

(a) First cost Rs. 1750

Interest rate = 0%



Summation Average cost Average first Average

Maintenan

End

of

of

of

cost

if total

cost

ce cost at

year (n)

maintenanc

maintenance

replaced

at through the

end of year

e

through the the

given given year



355

MBA-H2040 Quantitative Techniques for Managers

Cost

given year

year and

A

B (Rs)

C (Rs)

D (in Rs)

E (Rs)

F (Rs)

1750





C = B

C/A

D + E

A

1

0

0

0

1750

1750

2

100

100

50

875

925

3

200

300

100

583

683

4

300

600

150

438

588

5

400

1000

200

350

550

6

500

1500

250

292

542

7

600

2100

300

250

550

8

700

2800

350

219

569



The value corresponding to any end-of-year (n) in Column F represents the average total cost of

using the equipment till the end of that particulars year.


In this problem, the average total cost decreases till the end of the year 6 and then it increases.

Hence the optimal replacement period is 6 years ie the economic life of the equipment is 6 years.



(e) When interest rate i = 12%


When the interest rate is more than 0% the steps to get the economic life are summarized in the

following table.

Calculation to determine Economic life
First Cost = Rs. 1750

Interest rate = 12%

Mai

Present

Summation



nten

worth

as of present

(A/P,

En

Present

ance

beginning

worth

of

12%,n)

Annual

d

simulator

cost

of years

maintenanc

=i (1+i)n

equipment

of

maintena

at

(P/F,12v,n)

1



of e

costs

(1+i)n-1

total

cost

ye

nce

cost

end

maintenanc

through the



through the

ar

and first

of

e costs

given year



giver year

(n)

cost

year

G

s

A

B

C

D

E

F

G

H



C =

B

1

E+

BxC

D



FxG

(iR)

(1+12/100) n

Rs. 1750



1

0

0.8929

0

0

1750

1.1200

1960

2

100

0.7972

79.72

79.72

1829.72

0.5917

1082.6

3

200

0.7118

142.36

222.08

1972.08

0.4163

820.9

4

300

0.6355

190.65

412.73

2162.73

0.3292

711.9

5

400

0.5674

226.96

639.69

2389.69

0.2774

662.9

6

500

0.5066

253.30

892.99

2642.99

0.2432

642.7

7

600

0.4524

271.44

1164.43

2914.430

0.2191

638.5

8

700

0.4039

282.73

1447.16

3197.16

0.2013

680.7



356
MBA-H2040 Quantitative Techniques for Managers


Identify the end of year for which the annual equivalent total cost is minimum in column. In this

problem the annual equivalent total cost is minimum at the end of year hence the economics life of the

equipment is 7 years.


5.3.5.

Simple probabilistic model for items which completely fail


Electronic items like bulbs, resistors, tube lights etc. generally fail all of a sudden, instead of gradual

failure. The sudden failure of the item results in complete breakdown of the system. The system may

contain a collection of such items or just an item like a single tube-light. Hence we use some

replacement policy for such items which would minimize the possibility of complete breakdown. The

following are the replacement policies which are applicable in these cases.


i) Individual replacement policy :



Under this policy, each item is replaced immediately after failure.


ii) Group replacement policy :

Under group replacement policy, a decision is made with regard the replacement at what equal internals,

all the item are to be replaced simultaneously with a provision to replace the items individually which

fail during the fixed group replacement period.



Among the two types of replacement polices, we have to decide which replacement policy we

have to follow. Whether individual replacement policy is better than group replacement policy. With

regard to economic point of view. To decide this, each of the replacement policy is calculated and the

most economic one is selected for implementation.

Exercise :


1. List and explain different types of maintenance
2. Discuss the reasons for maintenance.
3. Distinguish between breakdown maintenance and preventive maintenance.
4. Distinguish between individual and group replacement polices.
5. A firm is considering replacement of an equipment whose first cost is Rs.4000 and the scrap

value is negligible at the end of any year. Based on experience, it has been found that the
maintenance cost is zero during the first year and it is Rs.1000 for the second year. It increase by
Rs.300 every years thereafter.

a) When should the equipment be replace if i = 0%
b) When should the equipment be replace if i = 12%

Ans . a) 5 years



b) 5 years



6. A company is planning to replace an equipment whose first cost is Rs.1,00,000. The operating

and maintenance cost of the equipment during its first year of operation is Rs.10,000 and it
increases by Rs. 2,000 every year thereafter. The release value of the equipment at the end of the



357

MBA-H2040 Quantitative Techniques for Managers

first year of its operation is Rs.65,000 and it decreases by Rs.10,000 every year thereafter. Find
the economic life of the equipment by assuming the interest rate as 12%.


[Ans : Economic life = 13 years and the corresponding annual equivalent cost = Rs. 34,510]


7. The following table gives the operation cost, maintenance cost and salvage value at the end of

every year of machine whose purchase value is Rs. 12,000. Find the economic life of the
machine assuming.

a) The interest rate as 0%
b) The interest rate as 15%



Operation cost at Maintenance cost Salvage value at the end

End

of the end of year at the end of year of year (Rs)

year

(Rs)

(Rs)

1

2000

2500

8000

2

3000

3000

7000

3

4000

3500

6000

4

5000

4000

5000

5

6000

4500

4000

6

7000

5000

3000

7

8000

5500

2000

8

9000

6000

1000


Ans :

a) Economic life of the machine = 2 years
b) Economic life of the machine = 2 years




































358
MBA-H2040 Quantitative Techniques for Managers



















359

This post was last modified on 14 March 2022