OPERATIONS RESEARCH LECTURE NOTES

JANE ADUDA

November 6, 2013
Contents

LIST OF FIGURES v

LIST OF TABLES viii

COURSE OUTLINE viii

1 Introduction 1

1.1 History of operations Research . . . . . . . . . . . . . . . . . . . . . . . . . 3

1.2 The Nature of Operations Research . . . . . . . . . . . . . . . . . . . . . . 4

1.3 Operations Research Models . . . . . . . . . . . . . . . . . . . . . . . . . . 6

1.4 Solving the OR Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

2 Linear programming 11

2.1 Basic Assumptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

2.2 Mathematical Formulation of a LP model . . . . . . . . . . . . . . . . . . . 13

2.2.1 The General Linear Programming Model . . . . . . . . . . . . . . . 14

2.3 Resource Allocation Models . . . . . . . . . . . . . . . . . . . . . . . . . . 15

i
Operations Research Jane Akinyi Aduda

2.3.1 Maximization problems . . . . . . . . . . . . . . . . . . . . . . . . . 15

2.3.2 Minimization problems . . . . . . . . . . . . . . . . . . . . . . . . . 18

2.4 Solution Methods for L.P. Problems . . . . . . . . . . . . . . . . . . . . . . 20

2.4.1 Terminology for Solutions of the Model . . . . . . . . . . . . . . . . 21

3 Solving LP Problems 22

3.1 Graphical Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

3.2 The Simplex Computations . . . . . . . . . . . . . . . . . . . . . . . . . . 27

3.2.1 Algebraic determination of corner points . . . . . . . . . . . . . . . 29

3.2.2 The Simplex Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . 30

3.3 The Transportation Problem . . . . . . . . . . . . . . . . . . . . . . . . . . 39

3.3.1 Finding Basic Feasible Solution for Transportation Problem . . . . 43

3.3.2 Methods to find the bfs for a balanced TP . . . . . . . . . . . . . . 43

3.3.3 Northwest Corner Method (NWC) . . . . . . . . . . . . . . . . . . 44

3.3.4 Least-Cost Method . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

3.3.5 Vogel Approximation Method (VAM) . . . . . . . . . . . . . . . . . 46

3.3.6 Iterative Computations of the Transportation Algorithm . . . . . . 47

3.3.7 Maximization using the Transportation algorithm . . . . . . . . . . 55

3.4 The Assignment Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56

3.4.1 Unbalanced assignment model . . . . . . . . . . . . . . . . . . . . . 63

3.4.2 Maximization using the Assignment algorithm . . . . . . . . . . . . 66

ii
Operations Research Jane Akinyi Aduda

4 Inventory Models 68

4.1 Types of Inventory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70

4.2 ABC Classification of Inventories . . . . . . . . . . . . . . . . . . . . . . . 71

4.2.1 A -Class Items . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71

4.2.2 B -class items . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71

4.2.3 C -class items . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72

4.3 Lot/Order Size Model with no Shortages or Basic EOQ Model . . . . . . . 77

4.4 Derivation of Basic EOQ Model . . . . . . . . . . . . . . . . . . . . . . . . 77

5 Network Optimization Models 80

5.1 Terminologies used in Networks . . . . . . . . . . . . . . . . . . . . . . . . 81

5.2 The Shortest Path Problem . . . . . . . . . . . . . . . . . . . . . . . . . . 85

5.3 The Minimum Spanning Tree Problem . . . . . . . . . . . . . . . . . . . . 86

5.4 The Maximum Flow Problem . . . . . . . . . . . . . . . . . . . . . . . . . 87

5.5 The Minimum Cost Flow Problem . . . . . . . . . . . . . . . . . . . . . . . 89

5.6 PERT/CPM Models for Project Management . . . . . . . . . . . . . . . . 90

5.6.1 Basic difference between PERT and CPM . . . . . . . . . . . . . . 91

5.6.2 PERT/CPM Network Components And Precedence Relationship . 92

5.6.3 Critical Path Calculations . . . . . . . . . . . . . . . . . . . . . . . 95

5.6.4 Determination of the Critical Path . . . . . . . . . . . . . . . . . . 95

5.6.5 Project Management PERT . . . . . . . . . . . . . . . . . . . . . . 98

iii
Operations Research Jane Akinyi Aduda

6 Waiting Line Theory or Queuing Model 100

6.1 QUEUING SYSTEM OR PROCESS . . . . . . . . . . . . . . . . . . . . . 101

6.1.1 Input Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102

6.1.2 Service Mechanism or Service Facility . . . . . . . . . . . . . . . . . 103

6.2 QUEUING PROBLEMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105

6.3 SYMBOLS USED IN QUEUING MODELS . . . . . . . . . . . . . . . . . 106

6.3.1 Notations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107

iv
List of Figures

3.1 Graph of Marangi paint problem . . . . . . . . . . . . . . . . . . . . . . . 23

3.2 Graph For the X and Y Products . . . . . . . . . . . . . . . . . . . . . . . 25

3.3 Chicken Diet Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

3.4 Ozark Diet Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

3.5 Algebraic Determination of corner points . . . . . . . . . . . . . . . . . . . 30

3.6 Graphical Solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

4.1 ABC analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74

v
List of Tables

2.1 Data for a resource allocation Model . . . . . . . . . . . . . . . . . . . . . 14

2.2 Data for Marangi Paint Problem . . . . . . . . . . . . . . . . . . . . . . . 15

2.3 Data for Vitamin deficiency Problem . . . . . . . . . . . . . . . . . . . . . 18

2.4 Data for the Diet Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

3.1 Results for Marangi paint Problem . . . . . . . . . . . . . . . . . . . . . . 23

3.2 Data for products X and Y . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

3.3 Results for the X and Y products . . . . . . . . . . . . . . . . . . . . . . . 25

3.4 Ozark Diet Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

3.5 Results for Ozark Diet Problem . . . . . . . . . . . . . . . . . . . . . . . . 27

3.6 Corner Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

3.7 First Simplex Tableau . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

3.8 Feasibility condition 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

3.9 Second Simplex Tableau . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

3.10 Feasibility condition 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

3.11 Third Simplex Tableau . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

vi
Operations Research Jane Akinyi Aduda

3.12 Solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

3.13 Results for minimization Problem . . . . . . . . . . . . . . . . . . . . . . . 38

3.14 First Simplex Tableau (Dual) . . . . . . . . . . . . . . . . . . . . . . . . . 38

3.15 Second Simplex Tableau (Dual) . . . . . . . . . . . . . . . . . . . . . . . . 38

3.16 Third (optimal) Simplex Tableau (Dual) . . . . . . . . . . . . . . . . . . . 39

3.17 Shipping costs, Supply, and Demand for Powerco Example . . . . . . . . . 41

3.18 NWC Starting Solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

3.19 Least-Cost Method Starting Solution . . . . . . . . . . . . . . . . . . . . . 46

3.20 VAM Starting Solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

3.21 Z-row coefficients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

3.22 Z-row coefficients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

3.23 Z-row coefficients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

3.24 First iteration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

3.25 First iteration result . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

3.26 Second Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

3.27 Third Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

3.28 Fourth Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53

3.29 Assignment Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57

3.30 Assignment Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

3.31 Assignment Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

vii
Operations Research Jane Akinyi Aduda

3.32 Assignment Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

3.33 Assignment Model continued . . . . . . . . . . . . . . . . . . . . . . . . . . 62

3.34 Assignments solution 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

3.35 Assignments solution 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

3.36 Assignment Problem 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

3.37 Assignment Problem 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64

5.1 Table for park problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86

5.2 VIVA Company Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94

viii
Operations Research Jane Akinyi Aduda

HBC 2122: OPERATIONS RESEARCH I

PURPOSE: This course is aimed at providing students with concepts of Operations

Research and their applications in business and other related areas.

OBJECTIVES: At the end of this course the students should be able to:-

• Understand operations research techniques.

• Solve linear programming and optimization problems.

• Use operations research techniques to solve practical problems in business and

management.

SYLLABUS: History and nature of operations research. Linear Programming; simplex

method,solution and its interpretation. application areas. Transportation model;
using north-west method, least cost method, Vogel approximation method (VAM).
Assignment model; formulation solution. Inventory models: periodic model, quan-
tity models, basic economic order quantity, discounts, stock-out, buffer stock, ac-
tivity based costing analysis, Pareto analysis, just in time (JIT) systems. Net-
work model; deterministic, critical path analysis/ critical path method, probabilistic
model, programme evaluation review technique, crashing, resource levelling. queu-
ing model, single server and multi-server systems, Simulation: Introduction and ap-
plication in queuing and inventory models. Game theory; pure and mixed strategies,
SADDLE, dominance, graphical solution, solution by algebraic and linear program-
ming method.

PRE-REQUISITE: HBC 2211 MANAGEMENT MATHEMATICS II

COURSE OUTLINE: Week 1-4

• History and nature of operations research

• Linear Programming

(i) Simplex method,solution and its interpretation

ix
Operations Research Jane Akinyi Aduda

(ii) Application areas

(iii) Transportation model

(iv) Using northwest method,

(v) Least cost method

(vi) Vogel approximation method (VAM)

Week 5-10

• CAT I

• Assignment model; formulation and solution.

• Inventory models

(i) Periodic model,

(ii) Quantity models

(iii) Basic economic order quantity

(iv) Discounts, Stock-out, Buffer stock

(v) Activity based costing analysis

(vi) Pareto analysis

(vii) Just in time (JIT) systems

• Network models

(i) Deterministic, critical path analysis/ critical path method

(ii) Probabilistic model, programme evaluation review technique

(iii) Crashing, resource leveling

(iv) Queuing model, single server and multiserver systems

Week 11-14

• CAT II

• Simulation: Introduction and application in queuing and inventory models.

• Game theory
x
Operations Research Jane Akinyi Aduda

(i) Pure and mixed strategies

(ii) SADDLE

(iii) Dominance

(iv) Graphical solution

(v) Solution by algebraic and linear programming method.

TEACHING METHODOLOGIES: In order to achieve the learning objectives stated

above, the following learning and teaching methods will be used.

(i) Lectures

(ii) Independent Study

(iii) Group discussions

(iv) Tutorials

INSTRUCTIONAL MATERIALS: These will include tablet, white board, comput-

ers and teaching notes.

COURSE ASSESMENT: CATS & assignments 30%, Written examination 70%

REFERENCES:

1. Hamdy A. Taha, A.M. Natarajan, P. Balsubramanie and A. Tamilarasi. Operations

Research (An Introduction),(2008).

2. Terry Lucey. Quantitative Techniques, 6th edition, (2002).

3. N.K Tiwari, Shishir K. Shandilya Operations Research.

xi
Chapter 1

Introduction

Operational Research is an interdisciplinary branch of applied mathematics and formal

science that uses methods such as mathematical modelling, statistics, and algorithms to
arrive at optimal or near optimal solutions to complex problems. It is typically concerned
with optimizing the maxima (profit, assembly line performance, crop yield, bandwidth,
etc) or minima (loss, risk, etc.) of some objective function. Operations research helps
management achieve its goals using scientific methods.

According to the Operations Society of Great Britain, Operations Research is the

application of the methods of science to complex problems arising in the direction and
management of large systems of men, materials and money in industry, business, Govern-
ment and defence. The distinctive approach is to develop a scientific model of the system,
incorporating measurements of factors such as chance and risk, with which to predict and
compare the outcome of alternative decisions, strategies or controls. The purpose is to
help management to determine its policy and actions scientifically.

It provides a facility to a decision maker to evaluate the given problems, identify

alternative solutions, recognize the constraints and then assist the decision maker to have
the best possible solution available, which is known as the optimal solution.

It tries to avoid the dangers from taking decisions merely by guessing or by using thumb

1
Operations Research Jane Akinyi Aduda

rules. Management is a multidimensional and dynamic concept. It is multidimensional

in the sense that management problems and their solutions have consequences in several
dimensions, such as human, economic, social and political fields. Also, the manager
operates his system in an environment, which will never remain static, hence the dynamic
nature. Hence any manager, while making decisions, must consider all aspects in addition
to economic aspect, so that his solution should be useful in all aspects.

The general approach is to analyse the problem in economic terms and then implement
the solution if it is not aggressive or does not impact negatively on other aspects like
humanity, social and political factors.

In summary we can say operations research has the following features:-

• Operations Research uses Scientific Methods for making decisions.

• It is interdisciplinary approach for solving problems and it uses the knowledge and
experience of experts in various fields.

• While analysing the problems all aspects are considered and examined and analysed
scientifically for finding the optimal solution for the problem on hand.

• As operations research has scientific approach, it improves the quality of answers to

the problems.

• Operations research provides a scientific base for decision-making and provides sci-
entific substitute for judgement and intuition.

In other words, operations research can be said to have the following characteristics

1. Operations Research is an interdisciplinary team approach.

2. Operations Research increases the creative ability of the decision maker.

3. Operations Research is a systems approach.

2
Operations Research Jane Akinyi Aduda

1.1 History of operations Research

Since the advent of the industrial revolution, the world has seen a remarkable growth in
the size and complexity of organizations. The artisans small shops of an earlier era have
evolved into the billion-dollar corporations of today. An integral part of this revolutionary
change has been a tremendous increase in the division of labour and segmentation of
management responsibilities in these organizations. The results have been spectacular.

However, along with its blessings, this increasing specialization has created new prob-
lems, problems that are still occurring in many organizations. One problem is a tendency
for the many components of an organization to grow into relatively autonomous empires
with their own goals and value systems, thereby losing sight of how their activities and
objectives mesh with those of the overall organization. What is best for one component
frequently is detrimental to another, so the components may end up working at cross
purposes.

A related problem is that as the complexity and specialization in an organization in-

crease, it becomes more and more difficult to allocate the available resources to the various
activities in a way that is most effective for the organization as a whole. These kinds of
problems and the need to find a better way to solve them provided the environment for
the emergence of operations research (commonly referred to as OR).

The roots of OR can be traced back many decades, when early attempts were made
to use a scientific approach in the management of organizations. However, the beginning
of the activity called operations research has generally been attributed to the military
services early in World War II. Because of the war effort, there was an urgent need to
allocate scarce resources to the various military operations and to the activities within
each operation in an effective manner. Therefore, the British and then the U.S. military
management called upon a large number of scientists to apply a scientific approach to
dealing with this and other strategic and tactical problems. In effect, they were asked
to do research on (military) operations. These teams of scientists were the first OR

3
Operations Research Jane Akinyi Aduda

teams. By developing effective methods of using the new tool of radar, these teams were
instrumental in winning the Air Battle of Britain. Through their research on how to
better manage convoy and antisubmarine operations, they also played a major role in
winning the Battle of the North Atlantic. Similar efforts assisted the Island Campaign in
the Pacific. When the war ended, the success of OR in the war effort spurred interest in
applying OR outside the military as well. As the industrial boom following the war was
running its course, the problems caused by the increasing complexity and specialization in
organizations were again coming to the forefront. It was becoming apparent to a growing
number of people, including business consultants who had served on or with the OR teams
during the war, that these were basically the same problems that had been faced by the
military but in a different context. By the early 1950s, these individuals had introduced
the use of OR to a variety of organizations in business, industry, and government. The
rapid spread of OR soon followed.

1.2 The Nature of Operations Research

As its name implies, operations research involves research on operations. Thus, opera-
tions research is applied to problems that concern how to conduct and coordinate the
operations (i.e., the activities) within an organization. The nature of the organization is
essentially immaterial, and, in fact, OR has been applied extensively in such diverse areas
as manufacturing, transportation, construction, telecommunications, financial planning,
health care, the military, and public services, to name just a few. Therefore, the breadth
of application is unusually wide.

The research part of the name means that operations research uses an approach that
resembles the way research is conducted in established scientific fields.

To a considerable extent, the scientific method is used to investigate the problem

of concern. (In fact, the term management science sometimes is used as a synonym
for operations research.) In particular, the process begins by carefully observing and

4
Operations Research Jane Akinyi Aduda

formulating the problem, including gathering all relevant data.

The next step is to construct a scientific (typically mathematical) model that attempts
to abstract the essence of the real problem. It is then hypothesized that this model is
a sufficiently precise representation of the essential features of the situation that the
conclusions (solutions) obtained from the model are also valid for the real problem.

Next, suitable experiments are conducted to test this hypothesis, modify it as needed,
and eventually verify some form of the hypothesis. (This step is frequently referred to as
model validation.) Thus, in a certain sense, operations research involves creative scientific
research into the fundamental properties of operations. However, there is more to it than
this. Specifically, OR is also concerned with the practical management of the organization.
Therefore, to be successful, OR must also provide positive, understandable conclusions to
the decision maker(s) when they are needed.

One way of summarizing the usual (overlapping) phases of an OR study is the follow-
ing:

1. Define the problem of interest and gather relevant data.

2. Formulate a mathematical model to represent the problem.

3. Develop a computer-based procedure for deriving solutions to the problem from the
model.

4. Test the model and refine it as needed.

5. Prepare for the ongoing application of the model as prescribed by management.

6. Implement.

5
Operations Research Jane Akinyi Aduda

1.3 Operations Research Models

“The objective of Operations Research is to provide a scientific basis to the decision maker
for solving the problems involving the interaction of various components of an organization
by employing a team of scientists from various disciplines, all working together towards
finding a solution which is in the best interest of the organization as a whole. The best
solution thus obtained is known as optimal decision”.

Imagine that you have a 5-week business commitment between Nairobi (NBI) and
Kisumu (KSM). You fly out of Nairobi on Mondays and return on Wednesdays. A regular
round-trip ticket costs shs.8,000, but a 20% discount is granted if the dates on the ticket
span a weekend. A one-way ticket in either direction costs 75% of the regular price. How
should you buy the tickets for the 5-week period?

Consider this situation as a decision-making problem. Solving this requires the answer
to the following three questions:

1. What are the decision alternatives?

2. Under what restrictions is the decision made?

3. What is the appropriate objective criterion for evaluating the alternatives?

These three alternatives can be considered:

1. Buy five regular NBI-KSM-NBI for departure on Monday and return on Wednesday
of the same week.

2. Buy one NBI-KSM, four KSM-NBI-KSM that span weekends and one KSM-NBI.

3. Buy one NBI-KSM-NBI to cover Monday of the first week and Wednesday of the
last week and four KSM-NBI-KSM to cover the remaining legs. All tickets in this
alternative span at least one weekend.

6
Operations Research Jane Akinyi Aduda

The restriction on these options is that you should be able to leave NBI on Monday and
return on Wednesday of the same week.

An obvious objective criterion for evaluating the proposed alternative is the price of
the tickets. The alternative that yields the smallest cost would be the best. Specifically,
we have

• Alternative 1 cost = 5 × 8, 000 = kes 40, 000

• Alternative 2 cost = (0.75 × 2 × 8, 000) + (4 × 0.8 × 8, 000) = kes 37, 600

• Alternative 3 cost = 5 × 0.8 × 8, 000 = kes 32, 000

Option 3 looks like the best choice in this case.

Although this example illustrates the three main components of an OR model, situa-
tions differ in the details of how each component is developed and constructed.

Consider forming a maximum-area rectangle out of a piece of wire of length L inches.

Contrary to the example above, the number of alternatives are infinite (the length and
width of the rectangle can assume an infinite set of values). The alternatives of this
problem can be expressed as continuous (algebraic) variables. Let l= length and w=
width of the rectangle in inches.

The restrictions in the situation can be expressed verbally as:

1. Length + width of rectangle= Half the length of the wire.

2. Length and width cannot be negative.

Algebraically, these restrictions can be translated as;

2(l + w) = L

l ≥ 0, w ≥ 0
7
Operations Research Jane Akinyi Aduda

Finally, the objective of the problem is maximization of the area of the rectangle. If we
let z to be the area of the rectangle, then, the complete model becomes

maximize z = lw

subject to
2(l + w) = L

and
l, w ≥ 0

Based on these two examples, the general OR model can be given as

Maximize or minimize an Objective Function

subject to

Constraints

A solution of the model is feasible if it satisfies all the constraints, and optimal if, in
addition yo being feasible, it yields the best (maximum or minimum) value of the objective
function.

Problem: identify a fourth feasible alternative for the tickets problem.

1.4 Solving the OR Model

Generally, the type and complexity of the mathematical model dictates the nature of the
solution method. The following are some of the tools that can be used.

(i) Linear Programming Model- This model is used for resource allocation when the
resources are limited and there are number of competing candidates for the use of
resources. The model may be used to maximize the returns or minimize the costs.
Consider the following situations:
8
Operations Research Jane Akinyi Aduda

• A manufacturing company wants to use their resources optimally and produce

different quantities of each type of product, which yield different returns, so as
to maximize the returns. This is a maximization problem

• A company manufacturing different types of alloys requires to purchase the

three basic materials while at the same time maintaining the crucial percentage
of basic materials in each alloy. This is to be done at the minimum cost. This
is a minimization problem Both of these first two are resource allocation
models.

• Number of factories are manufacturing the same commodities in different ca-

pacities and the commodity is sent to various markets to meet the demands of
the consumers. When the cost of transportation is known, linear programming
is used to formulate a programme to distribute the commodity from factories
to markets at the least possible cost. The model used is transportation
model.

• When a company has number of orders on its schedule, which are to be pro-
cessed on the same machines and the processing time, is known, then we have
to allocate the jobs or orders to the machines, so as to complete all the jobs in
minimum time. This we can also solve using the Assignment model.

All the above-discussed models are Linear Programming Models.

(ii) Sequencing Model- When a manufacturing firm has some job orders, which can
be processed on two or more machines and the processing times of each job on each
machine is known, then the problem of processing in a sequence to minimize the
cost or time is known as Sequencing model.

(iii) Waiting Line Model or Queuing Model-A model used for solving a problem
where certain service facilities have to provide service to its customers, so as to avoid
lengthy waiting line or queue, so that customers will get satisfaction from effective
service and idle time of service facilities are minimized is waiting line model or
queuing model.
9
Operations Research Jane Akinyi Aduda

(iv) Replacement Model-Any capital item, which is continuously used for providing
service or for producing the product is subjected to wear and tear due to usage,
and its efficiency goes on reducing. This reduction in efficiency can be predicted
by the increasing number of breakdowns or reduced productivity. The worn out
parts or components are to be replaced to bring the machine back to work through
maintenance. A time is reached when even the maintenance cost becomes too high
and the manager feels replacement of the old machine by new one is the best thing
to do. These type of problems and can be solved using replacement models.

(v) Inventory Models-Any manufacturing firm has to maintain stock of materials for
its use. This stock of materials, which are maintained in stores, is known as in-
ventory. Inventory is one form of capital or money. The company has to maintain
inventory at optimal cost. There are different types of inventory problems, depend-
ing the availability and demand pattern of the materials. These can be solved by
the application of inventory models.

10
Chapter 2

Linear programming

A model, which is used for optimum allocation of scarce or limited resources to competing
products or activities under such assumptions as certainty, linearity, fixed technology, and
constant profit per unit, is a linear programming model.

Linear Programming is one of the most versatile, powerful and useful techniques for
making managerial decisions

Linear programming (LP) is a technique for optimization of a linear objective function,

subject to linear equality and linear inequality constraints. Informally, linear program-
ming determines the way to achieve the best outcome (such as maximum profit or lowest
cost) in a given mathematical model and given some list of requirements represented as
linear equations.

It is a mathematical technique concerned with the allocation of scarce resources. It

is a procedure that optimizes the value of some objective function (maximizing profits or
minimizing costs) when the factors involved (e.g. labour or machine hours) are subject
to some constraints (e.g. only 1,000 labour hours available in a week).

Problems to be solved must conform to the following:

1. The problem must be capable of being stated in numeric terms.

11
Operations Research Jane Akinyi Aduda

2. All factors involved in the problem must have a linear relationship.

3. The problem must permit a choice or choices between alternative courses of action.

4. There must be one or more restrictions in the factors involved.

The programming part of a LP refers to the solution method.

2.1 Basic Assumptions

Below are some of the important assumptions made when formulating a linear program-
ming model:

1. The decision maker here is completely certain (i.e., deterministic conditions) regard-
ing all aspects of the situation, such as availability of resources, profit contributions,
technology, courses of action and their consequences etc.

2. The relationship between variables in the problem and the resources available is lin-
ear. Here the term linearity implies proportionality and additivity. This assumption
is very useful as it simplifies modelling of the problem.

3. We assume here fixed technology i.e., production requirements are fixed during the
planning period and don’t change in the period.

4. The profit contribution of a product remains constant, irrespective of level of pro-

duction and sales.

5. The decision variables are continuous meaning companies can manufacture products
in fractional units e.g. 2.5 vehicles, 3.2 barrels of oil etc. This is referred to as the
assumption of perfect divisibility.

6. Only one decision is required for the planning period. This condition shows that the
linear programming model is a static model, implying that the linear programming
12
Operations Research Jane Akinyi Aduda

problem is a single stage decision problem. (Note: Dynamic Programming problem

is a multi-stage decision problem).

7. All variables are restricted to non-negative values (i.e., their numerical value will be
≥ 0).

2.2 Mathematical Formulation of a LP model

The following steps are followed:

1. Study the given situation, find the key decision to be made. Hence, identify the
decision variables(issues or factors in the problem whose values are to be determined)
in the problem, specifying their units of measurement.

2. Formulate the objective function to be optimized and express it as a linear function

of the decision variables (z=decision variables multiplied by their cost or profit
contributions).

3. Formulate the constraints of the problem (these are imposed by resource availability)
and express them as linear equalities or inequalities in terms of the decision variables.
These are restrictions on the decision variables which arise due to limited resources.

4. Add non-negativity restrictions.

The objective function, the set of constraints and the non-negativity restrictions together
form the LP model.

13
Operations Research Jane Akinyi Aduda

2.2.1 The General Linear Programming Model

In particular, this model is seeks to select the values for x1 , x2 , . . ., xn so as to

Optimize z = c1 x1 + c2 x2 + . . . + cn xn

Subject to a11 x1 + a12 x2 + . . . + a1n xn (≤, =, ≥) b1

a21 x1 + a22 x2 + . . . + a2n xn (≤, =, ≥) b2

.. (2.1)
.

am1 x1 + am2 x2 + . . . + amn xn (≤, =, ≥) bm

and x1 , x2 , . . . , xn ≥ 0

This can then be summarized as

n
X
Optimize z = cj x j
j=1
n
Subject to
X
aij xj (≤, =, ≥) bi , i = 1, 2, . . . , m (2.2)
j=1

and xj ≥ 0, j = 1, 2, . . . , n

or as in the table below

Table 2.1: Data for a resource allocation Model

Resource usage per unit of Activity
Resource 1 2 ... n Amount of
resource available
1 a11 a12 ... a1n b1
2 a21 a22 ... a2n b2
.. ..
. ... ... ... ... .
m am1 am2 ... amn bm
Contribution to z c1 c2 ... cn
per unit of activity

where xj s are the decision variables and aij s, bi s and cj s are constants.
14
Operations Research Jane Akinyi Aduda

2.3 Resource Allocation Models

2.3.1 Maximization problems

Here we mainly seek to maximize profits, revenue or output from a production process.

Example 2.1 (Marangi Paint Problem). The Marangi Paint company produces both
interior and exterior paints from two raw materials, M 1 and M 2. The following table
provides the basic data for the problem:

Table 2.2: Data for Marangi Paint Problem

Tons of raw material
Exterior paint Interior paint Maximum daily available (tons)
Raw material M 1 6 4 24
Raw material M 2 1 2 6
Profit per ton ($ 1000) 5 4

A market survey indicates that, the daily demand for interior paint cannot exceed that
of exterior paint by more than 1 ton. Also the maximum daily demand for interior paint
is 2 tons. Marangi paint company wants to determine the optimum (best) product mix
for interior and exterior paints that maximizes the daily profits.
Solution:

Step 1. Identify the decision variables. These are the amounts of exterior and interior
paints that need to be produced. Let x1 and x2 represent the tons of exterior and
interior paints to be produced daily respectively.

Step 2. Formulate the objective function. The company wants to maximize profits.
Letting z represent the total daily profit (in thousands of dollars), the objective
function can be given as,

maximize z = 5x1 + 4x2

15
Operations Research Jane Akinyi Aduda

Step 3. Formulate the constraints that restrict raw material usage and express the
demand. For raw material usage we have:-

(usage of raw materials by both paints) ≤ (maximum raw material availability)

From the data we have,

6x1 + 4x2 ≤ 24 Raw material M1

x1 + 2x2 ≤ 6 Raw material M2

For the demand restrictions we have the first one saying that the difference between
the daily production of interior and exterior paints, x2 − x1 , does not exceed 1 ton.
This translate to x2 − x1 ≤ 1. The second restriction stipulates that the maximum
daily demand for interior paint is limited to two tons, which translates to x2 ≤ 2.

Step 4. Add the non-negativity constraints.

Therefore the complete Marangi Model is:

maximize z = 5x1 + 4x2

subject to
6x1 + 4x2 ≤ 24

x1 + 2x2 ≤ 6

−x1 + x2 ≤ 1

x2 ≤ 2

and
x1 , x2 ≥ 0

Example 2.2 (KICOMI Garment Problem). The KICOMI retail store stocks two
types of shirts A and B, These are packed in attractive cardboard boxes. In one week the
store can sell a maximum of 400 shirts of type A and a maximum of 300 shirts of type B.
The storage capacity, however, is limited to a maximum of 600 of both types combined.
16
Operations Research Jane Akinyi Aduda

Type A shirt fetches a profit of Kshs. 20/- per unit and type B a profit of Kshs. 50/- per
unit. The store wants to establish how many of each type of shirt they need to stock per
week in order to maximize their total profit. Formulate a mathematical model for this
problem.
Solution:

Step 1. We require the decision variables with regard to the quantities we need to
stock weekly. Let x1 and x2 represent the numbers of shirts A and B to be stocked
weekly respectively

Step 2. The store is seeking to maximize weekly profits. The profit contributions for
types A and B are 20/- and 50/- per unit respectively, so if z represents the weekly
profits for KICOMI store, then the objective function becomes

z = 20x1 + 50x2

Step 3. We are limited by the sales and storage capacities so we consider the maximum
sales and the maximum storage. The sales and storage constraints are given as

x1 ≤ 400

x2 ≤ 300

x1 + x2 ≤ 600

Step 4. Finally
x1 , x2 ≥ 0

The complete model is therefore given as

maximize z = 20x1 + 50x2

subject to x1 ≤ 400

x2 ≤ 300

x1 + x2 ≤ 600

and x1 , x2 ≥ 0
17
Operations Research Jane Akinyi Aduda

2.3.2 Minimization problems

For minimization we are interested in minimizing costs be they production, storage etc
as well as time spent on production or deliveries and so on.

Example 2.3 (Vitamin Deficiency Problem). A patient consults a doctor to check

on his ill health. The Doctor finds him to be having deficiency of two vitamins, A and
D. The patient is advised to consume vitamins A and D regularly for some time so as to
regain his health. The doctor prescribes tonics I and II, both of which contain vitamins A,
and D in certain proportions. He is also advised to consume at least 40 units of vitamin
A and 50 units of vitamin D Daily. The costs of tonics I and II and the proportions of
vitamins A and D that they contain are given in the table below. Formulate the linear
programming model that minimizes the cost of tonics.

Table 2.3: Data for Vitamin deficiency Problem

Tonics
Vitamins I II Daily requirements
in units
A 2 4 40
D 3 2 50
Cost per unit in Kshs. 50 30

Solution:

Step 1. Let x1 and x2 represent the quantities of Tonic I and II that can be purchased

Step 2. Let the total cost of purchasing these tonics be z. The objective function
considering the costs becomes

minimize z = 50x1 + 30x2

Step 3. We are limited by the proportion of each vitamin in the tonics and their mini-

18
Operations Research Jane Akinyi Aduda

mum daily requirements. The constraints will then be given by

2x1 + 4x2 ≥ 40 Vitamin A constraint

3x1 + 2x2 ≥ 50 Vitamin D constraint

Step 4. Finally
x1 , x2 ≥ 0

In standard form, we have

minimize z =50x1 + 30x2

subject to 2x1 + 4x2 ≥ 40

3x1 + 2x2 ≥ 50

and x1 , x2 ≥ 0

Example 2.4 (Diet Problem). Polly wonders how much money she must spend on food
in order to get all the energy (2,000 kcal), protein (55 g), and calcium (800 mg) that she
needs every day. She chooses six foods that seem to be cheap sources of the nutrients and
collects her data in the following table of nutritive value per serving. She also decides to
impose servings-per-day limits on all foods. Present this problem in standard form.

Table 2.4: Data for the Diet Problem

Energy Protein Calcium Price per serving Servings at
Food Serving size (kcal) (g) (mg) (shillings) most per day
Oat meal 28g 110 4 2 3 4
Chicken 100g 205 32 12 24 3
Eggs 2 large 160 13 54 13 2
Whole milk 237 cc 160 8 285 9 8
Cherry pie 170 g 420 4 22 20 2
Pork with beans 260 g 260 14 80 19 2

Solution:
To design the most economical menu, she speculates about some as yet unspecified menu,
19
Operations Research Jane Akinyi Aduda

consisting of x1 servings of oatmeal, x2 servings of chicken, and so on. She wants to

find numbers x1 , x2 , . . . , x6 that satisfy her self-imposed servings-per-day limits, meet the
requirements for energy, protein, calcium, and yet minimize the cost. This diet problem
can be nicely formulated as follows.

minimize z = 3x1 + 24x2 + 13x3 + 9x4 + 20x5 + 19x6

subject to 0 ≤ x1 ≤ 4

0 ≤ x2 ≤ 3

0 ≤ x3 ≤ 2

0 ≤ x4 ≤ 8

0 ≤ x5 ≤ 2

0 ≤ x6 ≤ 2

and 110x1 + 205x2 + 160x3 + 160x4 + 420x5 + 260x6 ≥ 2000

4x1 + 32x2 + 13x3 + 8x4 + 4x5 + 14x6 ≥ 55

2x1 + 12x2 + 54x3 + 285x4 + 22x5 + 80x6 ≥ 800

and x1 , x2 , . . . , x6 ≥ 0

2.4 Solution Methods for L.P. Problems

Linear Programming, is a method of solving the types of problems in which two or more
candidates or activities compete to utilize the available limited resources, with a view of
optimizing the objective function of the problem. The objective may be to maximize the
returns or to minimize the costs. The various methods available to solve the problem are:

1. The Graphical Method:This method limits us to problems that have two decision
variables in the problem. (To deal with more decision variables by the graphical
method will become complicated, because we would have to deal with planes instead
of straight lines).

2. The Systematic Trial and Error method, where we go on giving various values
20
Operations Research Jane Akinyi Aduda

to variables until we get optimal solution. This method takes too much of time and
laborious.

3. The Algebraic method. In this method each decision variable is considered as a

vector and principles of vector algebra are used to get the optimal solution. This
method is also time consuming.

4. The Simplex method. When the problem is having more than two decision vari-
ables, simplex method is the most powerful method to solve the problem. It has a
systematic algorithm used to solve the problem.

2.4.1 Terminology for Solutions of the Model

In linear programming (and its extensions) any specification of values for the decision
variables (x1 , x2 , . . . , xn ) is called a solution, regardless of whether it is a desirable or
even an allowable choice. Different types of solutions are then identified by using an
appropriate adjective.

• A feasible solution is a solution for which all the constraints are satisfied.

• An infeasible solution is a solution for which at least one constraint is violated.

• The feasible region is the collection of all feasible solutions. However we can note
that it is possible for a problem to have no feasible solutions.

• An optimal solution is a feasible solution that has the most favourable value of the
objective function. Most problems will have just one optimal solution. However, it
is possible to have more than one. Another possibility is that of having no optimal
solutions.

• A corner-point feasible (CPF) solution is a solution that lies at a corner of the

feasible region. Note that if a problem has exactly one optimal solution, it must be
a CPF solution. If the problem has multiple optimal solutions, at least two must be
CPF solutions.
21
Chapter 3

Solving LP Problems

3.1 Graphical Method

Linear programming problems with two variables can be represented and solved graphi-
cally with ease. Though in real-life, the two variable problems are practised very little,
the interpretation of this method will help to understand the simplex method. The non
negativity constraint implies the decision variables must have positive values and always
lie in first quadrant of the graph. In graphical method, the inequalities (structural con-
straints) are considered to be equations. Consider the Marangi Paint Problem which was
summarized as

maximize z =5x1 + 4x2

subject to 6x1 + 4x2 ≤ 24

x1 + 2x2 ≤ 6

−x1 + x2 ≤ 1

x2 ≤ 2

and x1 , x2 ≥ 0

Graphically, this can be expressed as shown in figure 3.1.

To do this, we follow the following procedure:-

22
Operations Research Jane Akinyi Aduda

Figure 3.1: Graph of Marangi paint problem

• Plot the inequalities as equalities and shade off the unwanted region (use broken
lines for strict inequalities and bold lines for those “greater than or equal to”, or
“less than or equal to” situations.)

• Identify all the corner points, pick their coordinates and plug them into the objective
function.

• Pick the pair of points that give the optimal value of the objective function.

From this graph, we obtain the results as shown in table 3.1. From these results we see
that Marangi will maximize profits ($ 21,000) if they produce 3 tons of exterior paint and
1.5 tons of interior paint.

Table 3.1: Results for Marangi paint Problem

Corner Point Coordinates Value of Objective function
A (0,0) 0
B (0,1) 4
C (1,2) 13
D (2,2) 18
E (3,1.5) 21
F (4,0) 20
23
Operations Research Jane Akinyi Aduda

Consider the example below

Example 3.1. A company manufactures two products, X and Y by using three machines
A, B, and C. Machine A has 4 hours of capacity available during the coming week.
Similarly, the available capacity of machines B and C during the coming week is 24 hours
and 35 hours respectively. One unit of product X requires one hour of Machine A, 3
hours of machine B and 10 hours of machine C. Similarly one unit of product Y requires
1 hour, 8 hour and 7 hours of machines A, B and C respectively. When one unit of X is
sold in the market, it yields a profit of shs.50/- per product and that of Y is shs.70/- per
unit. Solve the problem by using graphical method to find the optimal product mix.
The details given in the problem is given in table 3.2:

Table 3.2: Data for products X and Y

Products
Machines (time required in hours) Available capacity in hours
X Y
A 1 1 4
B 3 8 24
C 10 7 35
Profit per unit in shs. 50 70

Let x and y be the units of X and Y to be produced by the company. The L.P. model is
then given by

maximize z =50x + 70y

subject to x + y ≤ 4

3x + 8y ≤ 24

10x + 7y ≤ 35

and x, y ≥ 0
The plot is as shown in figure 3.3.

The results are as shown in table 3.3 indicating they should produce 1.6 units of X
and 2.4 units of Y so as to maximize profits(shs.248) .
24
Operations Research Jane Akinyi Aduda

Figure 3.2: Graph For the X and Y Products

Table 3.3: Results for the X and Y products

Corner Point Coordinates Value of Objective function
A (0,0) 0
B (0,3) 210
C (1.6,2.4) 248
D (2.33,1.67) 233.33
E (3.48,0.03) 176

For minimization, Consider the problem below (The Chicken Diet Problem).

On a chicken farm, the poultry is given a healthy diet to gain weight. The chicken
have to consume a minimum of 15 units of Substance A and another 15 units of Substance
B. In the market there are only two classes of compounds: Type X, with a composition
of one unit of A to five units of B, and another type, Y, with a composition of five units
of A to one of B. The price of Type X is $10 and Type Y, $30. What are the quantities
of each type of compound that have to be purchased to cover the needs of the diet with
a minimal cost?

We can express this problem mathematically as follows:-

Let x1 and x2 represent the quantities of Compounds X and Y to be purchased

25
Operations Research Jane Akinyi Aduda

respectively.

minimize z =10x1 + 30x2

subject to x1 + 5x2 ≥ 15

5x1 + x2 ≥ 15

and x1 , x2 ≥ 0

This is a very straight forward problem with only two constraints and thus the solution
is at the intersection. The farm should purchase 2.5 units of each compounds at a cost of
$100 as shown in figure 3.3.

Figure 3.3: Chicken Diet Problem

Consider the following minimization problem

Example 3.2. Ozark farm uses at least 800 lb of special feed every day. This special
feed is a mixture of corn and soyabean meal with the following compositions: The dietary

Table 3.4: Ozark Diet Problem

Per lb of foodstuff
Feed stuff protein fibre Cost(shs/lb)
Corn 0.09 0.02 0.30
Soyabean meal 0.60 0.06 0.90

requirements of the special feed are at least 30% protein and at most 5% fibre. Ozark
Farm wishes to to determine the daily minimum-cost feed mix.
26
Operations Research Jane Akinyi Aduda

Let x1 and x2 represent the lb of corn and soyabean meal in the daily mix respectively.
The objective function seeks to minimize the total daily costs i.e.

minimize z = 0.3x1 + 0.9x2

Because Ozark Farm needs at least 800 lb of feed everyday, the associated constraint can
be expressed as
x1 + x2 ≥ 800

As for the protein and fibre dietary constraints we need at least 30% and 5% of the total
feed mix (x1 + x2 )lb, that is

0.09x1 + 0.60x2 ≥ 0.30(x1 + x2 ) which is also 0.21x1 − 0.30x2 ≤ 0

and 0.02x1 + 0.06x2 ≤ 0.05(x1 + x2 ) which is also 0.03x1 − 0.01x2 ≥ 0

Our problem can then be expressed as below:

minimize z =0.3x1 + 0.9x2

subject to x1 + x2 ≥ 800

0.21x1 − 0.30x2 ≤ 0

0.03x1 − 0.01x2 ≥ 0

and x1 , x2 ≥ 0
This is the graphical representation and the results are at point C which is the optimal
(cost minimizing) mix as seen in figure 3.4 as well as table 3.5

Table 3.5: Results for Ozark Diet Problem

Corner Point Coordinates Value of Objective function
B (200,600) 600
C (470.59,529.41) 437.65

3.2 The Simplex Computations

The development of the simplex method computations is facilitated by imposing two

requirements on the constraints of the problem. The two requirements help to standardize
27
Operations Research Jane Akinyi Aduda

Figure 3.4: Ozark Diet Problem

and streamline the calculations.

1. All constraints (with the exception of the non negativity of the variables) are equa-
tions with a non-negative right hand side.

2. All the variables are non-negative.

Converting inequalities into equations with non-negative RHS

• In (≤) constraints, the RHS represents a limit on the resource availability while the
LHS represents the usage of resources by the activities of the model. RHS-LHS=
unused or slack resources. To convert (≤) to =, add a slack variable to the LHS e.g.
in the Marangi problem, the inequality

6x1 + 4x2 ≤ 24 becomes 6x1 + 4x2 + s1 = 24, where s1 ≥ 0

and s1 represents the unused amount of M1.

• (≥) constraints set a lower limit on the activities of the LP model so that LHS-
RHS=surplus. The conversion of (≥) to = is achieved by subtracting a surplus
amount from the LHS of the inequality .e.g in the vitamin deficiency problem, the
28
Operations Research Jane Akinyi Aduda

inequality

2x1 + 4x2 ≥ 40 becomes 2x1 + 4x2 − S1 = 40, where S1 ≥ 0

NOTE:The equations must always have a non-negative RHS

3.2.1 Algebraic determination of corner points

In a set of m × n equations with m < n, we set n − m variables equal to zero and solve
the remaining equations for the remaining m variables. The resulting solution, if unique,
is a basic solution and must conform to one of the corner points of the solution space.
This means the maximum number of corner points is

n!
Cmn =
m!(n − m)!

Consider the following LP with two variables

maximize z =2x1 + 3x2

subject to 2x1 + x2 ≤ 4

x1 + 2x2 ≤ 5

and x1 , x2 ≥ 0

After converting the inequalities to equations we have

maximize z =2x1 + 3x2

subject to 2x1 + x2 + s1 = 4

x1 + 2x2 + s2 = 5

and x1 , x2 , s1 , s2 ≥ 0

This system therefore has m = 2 equations and n = 4 unknowns. We therefore set

n − m = 4 − 2 variables equal to zero and solve for the remaining 2 variables. We get the
graph in figure 3.5
The corner points can be summarized as in table 3.6
29
Operations Research Jane Akinyi Aduda

Figure 3.5: Algebraic Determination of corner points

Table 3.6: Corner Points

non basic Basic Basic Solution Corner points Feasible value of Obj. Func
(x1 , x2 ) (s1 , s2 ) (4,5) A Yes 0
(x1 , s1 ) (x2 , s2 ) (4,-3) C No -
(x1 , s2 ) (x2 , s1 ) (2.5,1.5) B Yes 7.5
(x2 , s1 ) (x1 , s2 ) (2,3) E Yes 4
(x2 , s2 ) (x1 , s1 ) (5,-6) F No -
(s1 , s2 ) (x1 , x2 ) (1,2) D Yes 8

3.2.2 The Simplex Algorithm

Rather than enumerate all the corner points as above, the simplex method only investi-
gates a select few. It focuses solely on CPF solutions. For any problem with at least one
optimal solution, finding one requires only finding a best CPF solution. Normally, the
simplex method starts from the origin and seeks to find the next variable that gives the
largest improvement to the objective function. The design of the simplex method calls
for increasing one variable at a time.

It deals with iterative process, which consists of first designing a Basic Feasible
Solution or a Programme and proceed towards the OPTIMAL SOLUTION and testing
each feasible solution for Optimality to know whether the solution on hand is optimal or
not. If not an optimal solution, redesign the programme, and test for optimality until the

30
Operations Research Jane Akinyi Aduda

test confirms OPTIMALITY. Hence we can say that the Simplex Method depends on two
concepts known as Feasibility and optimality. The iterative steps of the simplex method
are repeated until a finite optimal solution, if exists, is found. If no optimal solution, the
method indicates that no finite solution exists.

The Simplex procedure is as follows

Step 1: Determine the starting basic feasible solution.

Step 2: Select the entering variable using the optimality condition. This condition
states that the entering variable in a maximization (minimization) problem is the
non basic variable having the most negative (positive) coefficient in the z-row. Ties
are broken arbitrarily. The optimum is reached at the iteration where all the z- row
coefficients of the non-basic variables are non-negative (negative).

Step 3: Select the leaving variable using the feasibility condition. For both maxi-
mization and minimization problems, this condition states that the leaving variable
is the basic variable associated with the smallest non-negative ratio (intercept i.e.
ratio of the solution column to the constraint coefficients under the entering variable
(pivot column)). Ties are broken arbitrarily.

Step 4: Determine the new basic solution by using the appropriate Gauss- Jordan
Computations and go to step two.

The Gauss- Jordan row operations are

1. Pivot row:

(a) Replace the leaving variable in the Basic column with the entering variable.

(b) New pivot row=Current pivot row ÷ pivot element (the element at the inter-
section of the pivot row and pivot column)

2. All other rows, including z: New row = (Current row) - (pivot column coefficient)
× (New pivot row)
31
Operations Research Jane Akinyi Aduda

MAXIMIZATION CASE

Let us refer to the Marangi paint problem once again. We express the problem in this
form.

maximize z =5x1 + 4x2 + 0s1 + 0s2 + 0s3 + 0s4

subject to 6x1 + 4x2 + s1 = 24

x1 + 2x2 + s2 =6

−x1 + x2 + s3 =1

x2 + s4 = 2

and x1 , x2 , s1 , s2 , s3 , s4 ≥ 0

Next, we write the objective function as

z − 5x1 − 4x2 = 0

This way the initial simplex tableau can be given as:- The first iteration starts at the

Table 3.7: First Simplex Tableau

Basic z x1 x2 s1 s2 s3 s4 Solution
z 1 -5 -4 0 0 0 0 0 z-row
s1 0 6 4 1 0 0 0 24 s1 -row
s2 0 1 2 0 1 0 0 6 s2 -row
s3 0 -1 1 0 0 1 0 1 s3 -row
s4 0 0 1 0 0 0 1 2 s4 -row

origin (where x1 = x2 = 0), leaving us with only the slack variables as basic. This tableau
is not optimal and can be improved by increasing either x1 or x2 . Since in this tableau,
we have expressed out objective function as

z − 5x1 − 4x2 = 0

we pick the variable with the most negative coefficient as the next entering variable
(optimality condition), which in this case is x1 . The column corresponding to the entering
variable is called the pivot column.
32
Operations Research Jane Akinyi Aduda

To identify the leaving variable, we compute the intercepts (ratios) and pick the vari-
able with the minimum non-negative ratio. This identifies s1 as the leaving variable as
shown in table 3.8

Table 3.8: Feasibility condition 1

Basic Entering x1 Solution Ratio (or intercept)
24
s1 6 24 x1 = 6
= 4 minimum
6
s2 1 6 x1 = 1
=6
1
s3 -1 1 x1 = −1
= −1 ignore
2
s4 0 2 x1 = 0
= ∞ ignore

So for the next tableau, we swap x1 with s1 using the Gauss- Jordan operations.

1. Replace s1 in the basic column with x1 and calculate the coefficients of the new x1
as follows:
New x1 -row= Current s1 -row ÷6

1 2 1
= (0, 6, 4, 1, 0, 0, 0, 24) = (0, 1, , , 0, 0, 0, 4)
6 3 6

all other rows including z,

New row=(current row)-[(it’s pivot column coefficient)×(New pivot row)]

2. New z-row=current z-row-(-5)× new x1 -row, i.e.

2 1 2 5
(1, −5, −4, 0, 0, 0, 0, 0) − (−5) × (0, 1, , , 0, 0, 0, 4) = (1, 0, − , , 0, 0, 0, 20)
3 6 3 6

3. New s2 -row=current s2 -row-(1)× new x1 -row, i.e.

2 1 4 1
(0, 1, 2, 0, 1, 0, 0, 6) − (1) × (0, 1, , , 0, 0, 0, 4) = (0, 0, , − , 1, 0, 0, 2)
3 6 3 6

4. New s3 -row=current s3 -row-(-1)× new x1 -row, i.e.

2 1 5 1
(0, −1, 1, 0, 0, 1, 0, 1) − (−1) × (0, 1, , , 0, 0, 0, 4) = (0, 0, , , 0, 1, 0, 5)
3 6 3 6

33
Operations Research Jane Akinyi Aduda

Table 3.9: Second Simplex Tableau

Basic z x1 x2 s1 s2 s3 s4 Solution
z 1 0 − 23 5
6
0 0 0 20 z-row
2 1
x1 0 1 3 6
0 0 0 4 x1 -row
4
s2 0 0 3
− 61 1 0 0 2 s2 -row
5 1
s3 0 0 3 6
0 1 0 5 s3 -row
s4 0 0 1 0 0 0 1 2 s4 -row

5. New s4 -row=current s4 -row-(-1)× new x1 -row, i.e.

2 1
(0, 0, 1, 0, 0, 0, 1, 2) − (0) × (0, 1, , , 0, 0, 0, 4) = (0, 0, 1, 0, 0, 0, 1, 2)
3 6

This yields the second simplex tableau as shown in table 3.9 We can see that this tableau
is not optimal as the objective function can still be improved. The next entering variable
is x2 and using the feasibility condition we can identify the leaving variable to be s2 as
shown in table 3.10 So we repeat the Gauss- Jordan row operations and they produce the

Table 3.10: Feasibility condition 2

Basic Entering x2 Solution Ratio (or intercept)
2 2
x1 3
4 x2 = 4 ÷ 3
=6
4 4
s2 3
2 x2 = 2 ÷ 3
= 1.5 minimum
5 5
s3 3
5 x2 = 5 ÷ 3
=3
s4 1 2 x2 = 2 ÷ 1 = 2

following simplex tableau in 3.11.

1. New pivot x2 -row=current s2 -row ÷ 43

2. New z-row=current z-row −(− 23 )× New x2 -row

3. New x1 -row=current x1 -row −( 23 )× New x2 -row

4. New s3 -row=current s3 -row −( 35 )× New x2 -row

34
Operations Research Jane Akinyi Aduda

Table 3.11: Third Simplex Tableau

Basic z x1 x2 s1 s2 s3 s4 Solution
3 1
z 1 0 0 4 2
0 0 21 z-row
1
x1 0 1 0 4
− 12 0 0 3 x1 -row
x2 0 0 1 − 18 3
4
0 0 3
2
x2 -row
3
s3 0 0 0 8
− 54 1 0 5
2
s3 -row
1
s4 0 0 0 8
− 34 0 1 1
2
s4 -row

5. New s4 -row=current s4 -row −(1)× New x2 -row

Based on the optimality condition, none of the z-row coefficients associated with the non
basic variables s1 and s2 are negative, hence this last tableau is optimal and the solution
is interpreted as in table 3.12

Table 3.12: Solution

Decision variable Optimal value Recommendation
x1 3 Produce 3 tons of exterior paint daily
3
x2 2
Produce 1.5 tons of interior paint daily
z 21 Daily profit is $21,000

Therefore in summary, to solve a linear programming problem in standard form, use the
following steps.

1. Convert each inequality in the set of constraints to an equation by introducing slack

variables.

2. Create the initial simplex tableau.

3. Locate the most negative entry in the z row. The column for this entry is called
the entering column. (If ties occur, any of the tied entries can be used to determine
the entering column.)

4. Form the ratios of the entries in the solution-column with their corresponding posi-
tive entries in the entering column. The departing row corresponds to the smallest
35
Operations Research Jane Akinyi Aduda

non-negative ratio (If all entries in the entering column are 0 or negative, then there
is no maximum solution. For ties, choose either entry.) The entry in the departing
row and the entering column is called the pivot element.

5. Use elementary row operations so that the pivot element is 1, and all other entries
in the entering column are 0. This process is called pivoting.

6. If all entries in the bottom row are zero or positive, this is the final tableau. If not,
go back to Step 3.

7. If you obtain a final tableau, then the linear programming problem has a maximum
solution, which is given by the entry in the lower-right corner of the tableau.

MINIMIZATION CASE

Consider the minimization problem below

minimize C =0.12x1 + 0.15x2

subject to 60x1 + 60x2 ≥ 300

12x1 + 6x2 ≥ 36

10x1 + 30x2 ≥ 90

and x1 , x2 ≥ 0
The basic procedure used to solve such a problem is to convert it to a maximization
problem in standard form, and then apply the simplex method.

The first step in converting this problem to a maximization problem is to form the
augmented matrix for this system of inequalities. To this augmented matrix we add a
last row that represents the coefficients of the objective function, as follows.
..
 
60 60 . 300
..
 
 12 6 . 36 
 
..
 
 
 10 30 . 90 
 
 
 ... ... ... ...
..
 
0.12 0.15 . 0
36
Operations Research Jane Akinyi Aduda

Next, we form the transpose of this matrix by interchanging its rows and columns.
..
 
60 12 10 . 0.12
..
 
 
 60 6 30 . 0.15
 
 
... ... ... ... ... 
..
 
300 36 90 . 0

Note that the rows of this matrix are the columns of the first matrix, and vice versa.
Finally, we interpret the new matrix as a maximization problem as follows. (To do this,
we introduce new variables, y1 , y2 , and y3 .) We call this corresponding maximization
problem the DUAL of the original minimization problem.

maximize Z =300y1 + 36y2 + 90y3

subject to 60y1 + 12y2 + 10y3 ≤ 0.12

60y1 + 6y2 + 30y3 ≤ 0.15

and y1 , y2 , y3 ≥ 0

As it turns out, the solution of the original minimization problem can be found by applying
the simplex method to the new dual problem, as follows.

This problem in equation form is

maximize Z =300y1 + 36y2 + 90y3 + 0s1 + 0s2

subject to 60y1 + 12y2 + 10y3 + s1 = 0.12

60y1 + 6y2 + 30y3 + 0s1 + s2 = 0.15

and y1 , y2 , y3 , s1 , s2 ≥ 0

Graphically, this problem is solved as shown in figure 3.6

and the solution can be summarized as shown in table 3.13 from which we see that point
C is optimal with x1 = 3, x2 = 2 and Z = 0.66
Using the simplex algorithm, the initial solution is in table 3.14
The other iterations are as shown in table 3.15 and 3.16

Thus, the solution of the dual maximization problem is Z = 0.66. This is the same
value obtained by the graphical solution. The x-values corresponding to this optimal
37
Operations Research Jane Akinyi Aduda

Figure 3.6: Graphical Solution

Table 3.13: Results for minimization Problem

Corner Point Coordinates Value of Objective function
A (0,6) 0.90
B (1,4) 0.72
C (3,2) 0.66
D (9,0) 1.08

Table 3.14: First Simplex Tableau (Dual)

Basic z y1 y2 y3 s1 s2 Solution Ratio
z 1 -300 -36 -90 0 0 0 z-row
1
s1 0 60 12 10 1 0 0.12 s1 -row 500
1
s2 0 60 6 30 0 1 0.15 s2 -row 400

Table 3.15: Second Simplex Tableau (Dual)

Basic z y1 y2 y3 s1 s2 Solution Ratio
z 1 0 24 -40 5 0 3/5 z-row 0
y1 0 1 1/5 1/6 1/60 0 1/500 y1 -row 0.012
s2 0 0 -6 20 -1 1 0.03 s2 -row 0.0015

solution are obtained from the entries in the z-row corresponding to slack variable columns.
In other words, the optimal solution occurs when x1 = 3 and x2 = 2.

38
Operations Research Jane Akinyi Aduda

Table 3.16: Third (optimal) Simplex Tableau (Dual)

Basic z y1 y2 y3 s1 s2 Solution
z 1 0 12 0 3 2 0.66 z-row
y1 0 1 0.25 0 0.0025 -1/120 0.00175 y1 -row
y3 0 0 -3/10 1 -1/20 1/20 0.0015 s2 -row

The fact that a dual maximization problem has the same solution as its original
minimization problem is stated formally in a result called the von Neumann Duality
Principle, after the American mathematician John von Neumann (1903-1957).

3.3 The Transportation Problem

The general transportation problem is concerned (literally or figuratively) with distribut-

ing any commodity from any group of supply centres, called sources, to any group of
receiving centres, called destinations, in such a way as to minimize the total distribu-
tion cost. Each source has a certain supply of units to distribute to the destinations, and
each destination has a certain demand. A transportation problem basically deals with
the problem, which aims to find the best way to fulfil the demand of n demand points
using the capacities of m supply points. While trying to find the best way, generally a
variable cost of shipping the product from one supply point to a demand point or a similar
constraint should be taken into consideration.

These problems can be solved by the simplex method; however, specialized algorithms
have been developed which offer greater efficiency. The transportation problem has the
following characteristics:-

1. There is a set of m supply points from which goods are shipped. Supply point i can
supply at most si units. This is the amount of supply at source i

2. There is a set of n demand points to which the good is shipped. Demand point j
must receive at least dj units of the shipped good. This is the amount of demand
39
Operations Research Jane Akinyi Aduda

at destination j

3. Each unit produced at supply point i and shipped to demand point j incurs a
variable cost of cij .

The sources and destinations are represented by nodes. The routes linking the sources
and destinations are denoted by arcs. Arc (i, j) joining source i to destination j caries
two pieces of information.

1. The transportation cost per unit cij

2. The amount to be shipped xij

The objective of the model is to determine the unknowns xij that will minimize the total
transportation cost while satisfying all the supply and demand restrictions.

Example 3.3. Powerco has three electric power plants that supply the electric needs of
four cities.

• The associated supply of each plant and demand of each city is given in the table
3.17.

• The cost of sending 1 million kwh of electricity from a plant to a city depends on
the distance the electricity must travel.

Solution: Transportation tableau

A transportation problem is specified by the supply, the demand, and the shipping
costs. So the relevant data can be summarized in a transportation tableau. The trans-
portation tableau implicitly expresses the supply and demand constraints and the shipping
cost between each demand and supply point.

1. Decision Variable:
Since we have to determine how much electricity is sent from each plant to each
40
Operations Research Jane Akinyi Aduda

Table 3.17: Shipping costs, Supply, and Demand for Powerco Example
From Destination Supply
City 1 City 2 City 3 City 4 (Million kwh)
Plant 1 $8 $6 $10 $9 35
Plant 2 $9 $12 $13 $7 50
Plant 3 $14 $9 $16 $5 40
Demand 45 20 30 30
(Million kwh)

city;
xij = Amount of electricity produced at plant i and sent to city j
x14 = Amount of electricity produced at plant 1 and sent to city 4

2. Objective function
Since we want to minimize the total cost of shipping from plants to cities;
Minimize Z = 8x11 + 6x12 + 10x13 + 9x14 + 9x21 + 12x22 + 13x23 + 7x24 + 14x31 +
9x32 + 16x33 + 5x34

3. Supply Constraints:
Since each supply point has a limited production capacity;

x11 + x12 + x13 + x14 ≤ 35

x21 + x22 + x23 + x24 ≤ 50

x31 + x32 + x33 + x34 ≤ 40

4. Demand Constraints:

x11 + x21 + x31 ≥ 45

x12 + x22 + x32 ≥ 20

x13 + x23 + x33 ≥ 30

x14 + x24 + x34 ≥ 30

41
Operations Research Jane Akinyi Aduda

5. Sign Constraints:
Since a negative amount of electricity can not be shipped all xij ’s must be non
negative;
xij ≥ 0, (i = 1, 2, 3; j = 1, 2, 3, 4)

The mathematical formulation of the transportation model is to determine the non-

negative values of xij satisfying both availability (supply) and requirements (demand)
constraints and minimizing the total cost of transportation. So we have
m X
X n
minimize Z = xij cij
i=1 j=1
Xn
subject to xij = si , i = 1, 2, . . . , m
j=1
m
X
xij = dj , j = 1, 2, . . . , n
i=1

and xij ≥ 0, i = 1, 2, . . . , m; j = 1, 2, . . . , n

Balanced Transportation Problem

If Total supply equals to total demand, the problem is said to be a balanced transportation
problem:
m
X n
X
si = dj
i=1 j=1

• All constraints must be binding

• becomes relatively easy to find a basic feasible solution

• Simplex pivots do not involve multiplication, they reduce to additions and subtrac-
tions

• Therefore, it is desirable to formulate a transportation problem as a balanced trans-

portation problem

42
Operations Research Jane Akinyi Aduda

Balancing a Transportation problem if total supply exceeds total demand

If total supply exceeds total demand, we can balance the problem by adding a dummy
demand point. Since shipments to the dummy demand point are not real, they are
assigned a cost of zero.

Balancing a Transportation problem if total supply is less than total demand

If a transportation problem has a total supply that is strictly less than total demand the
problem has no feasible solution.

However, it is sometimes desirable to allow the possibility of leaving some demand

unmet:

• A penalty (cost) is often associated with the unmet demand

• To balance the problem, add a “dummy (or shortage) supply point”

3.3.1 Finding Basic Feasible Solution for Transportation Prob-

lem

Unlike other Linear Programming problems, a balanced Transportation Problem with m

supply points and n demand points is easier to solve, although it has m + n equality
constraints. The reason for that is, if a set of decision variables (xij s) satisfy all but one
constraint, the values for xij s will satisfy that remaining constraint automatically.

3.3.2 Methods to find the bfs for a balanced TP

There are three basic methods:

1. Northwest Corner Method

43
Operations Research Jane Akinyi Aduda

2. Minimum Cost Method

3. Vogel’s Method

3.3.3 Northwest Corner Method (NWC)

To find the bfs by the NWC method; Begin in the upper left (north-west) corner of the
transportation tableau and set x11 as large as possible (here the limitations for setting x11
to a larger number, will be the demand of demand point 1 and the supply of supply point
1. Your x11 value can not be greater than minimum of this 2 values).

1. Allocate as much as possible to the selected cell, and adjust the associated amount
of supply and demand by subtracting the allocated amount.

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

3. If exactly one row or column is left uncrossed out, stop. Otherwise move to the cell
to the right if a column has just been crossed out, or below if a row has been crossed
out. Go to step 1. This is shown in figure 3.18

This gives a starting basic solution of

Z = 35 × 8 + 10 × 9 + 12 × 20 + 13 × 20 + 16 × 10 + 5 × 30

= 280 + 90 + 240 + 260 + 160 + 150

= 1, 180

Note: The total number of allocations must be equal to m + n − 1, where m is the number
of rows and n is the number of columns, otherwise the solution will be degenerate.

44
Operations Research Jane Akinyi Aduda

Table 3.18: NWC Starting Solution

From Destination Supply
City 1 City 2 City 3 City 4 (Million kwh)
$8 $6 $10 $9
Plant 1 35 35
$9 $12 $13 $7
Plant 2 10 20 20 50
$14 $9 $16 $5
Plant 3 10 30 40
Demand 45 20 30 30
(Million kwh)

3.3.4 Least-Cost Method

The Northwest Corner Method dos not utilize shipping costs. It can yield an initial bfs
easily but the total shipping cost may be very high. The minimum cost method uses
shipping costs in order come up with a bfs that has a lower cost.

To begin the minimum cost method, first we find the decision variable with the smallest
shipping cost (xij ). Then assign xij its largest possible value, which is the minimum of
si and dj . After that, as in the Northwest Corner Method we should cross out row i and
column j and reduce the supply or demand of the non-crossed-out row or column by the
value of xij .

Then we will choose the cell with the minimum cost of shipping from the cells that do
not lie in a crossed-out row or column and we will repeat the procedure. This is shown
in figure 3.19 This gives a starting basic solution of

Z = 30 × 5 + 20 × 6 + 15 × 8 + 30 × 9 + 20 × 13 + 10 × 16

= 150 + 120 + 120 + 270 + 260 + 160

= 1, 080

45
Operations Research Jane Akinyi Aduda

Table 3.19: Least-Cost Method Starting Solution

From Destination Supply
City 1 City 2 City 3 City 4 (Million kwh)
$8 $6 $10 $9
Plant 1 15 20 35
$9 $12 $13 $7
Plant 2 30 20 50
$14 $9 $16 $5
Plant 3 10 30 40
Demand 45 20 30 30
(Million kwh)

3.3.5 Vogel Approximation Method (VAM)

VAM is an improved version of the Least-Cost Method, that generally, but not always,
produces better starting solutions.

1. Start by computing each row and column a penalty. The penalty will be equal to
the difference between the two smallest shipping costs in the row or column.

2. Identify the row or column with the largest penalty. Find the first basic variable
which has the smallest shipping cost in that row or column.

3. Then assign the highest possible value to that variable, and cross-out the row or
column as in the previous methods. Compute new penalties and use the same
procedure. If a row and a column are satisfied simultaneously, only one of the two
is crossed out, and the remaining row or column is assigned zero supply or demand.

If exactly one row or column with zero supply or demand remains uncrossed-out stop. If
one row (column) with positive supply (demand) remains uncrossed out, determine the
basic variables in the row (column) by the least cost method. Stop. This is shown in
figure 3.20. This gives a starting solution of
46
Operations Research Jane Akinyi Aduda

Table 3.20: VAM Starting Solution

From Destination Row Row Row Row Row Supply
City 1 City 2 City 3 City 4 plty 1 plty 2 plty 3 plty 4 plty 5 (M kwh)
Plant 1 $8 $6 $10 $9 2 2 2 1 2 35
10 25
Plant 2 $9 $12 $13 $7 2 3 3 3 1 50
45 5
Plant 3 $14 $9 $16 $5 4 5 40
10 30
plty 1 1 3 3 2
plty 2 1 3 3
plty 3 1 6 3
plty 4 1 3
plty 5 3
Demand 45 20 30 30
(M kwh)

Z = 30 × 5 + 10 × 9 + 10 × 6 + 45 × 9 + 25 × 10 + 5 × 13

= 150 + 90 + 60 + 405 + 250 + 65

= 1, 020

3.3.6 Iterative Computations of the Transportation Algorithm

After determining the starting solution (using any of the three methods discussed above),
we use the following algorithm to determine the optimal solution.

1. Use the simplex optimality condition to determine the entering variable as the cur-
rent non-basic variable that can improve the solution. If the optimality condition is
satisfied, stop, otherwise go to step 2.

47
Operations Research Jane Akinyi Aduda

2. Determine the leaving variable using the simplex feasibility condition. Change the
basis and return to step 1.

The determination of the entering variable from among the current non-basic variables
(those that are not part of the starting basic solution) is done by computing the non-basic
coefficients in the Z-row using the method of multipliers.

In the method of multipliers, we associate the multipliers ui and vj with row i and
column j of the transportation tableau. For each current basic variable xij ,

ui + vj = cij for each basic xij

Using the NWC starting solution, the initial basic feasible solution has 6 basic variables,
leading to 6 equations with 7 unknowns. To solve these equations, the method of multi-
pliers calls for arbitrary setting any u1 = 0 and then solving the remaining variables as
shown in table 3.21

Table 3.21: Z-row coefficients

Basic variable (u, v) equation Solution
x11 u1 + v1 = 8 set u1 = 0 ⇒ v1 = 8
x21 u2 + v1 = 9 if v1 = 8 ⇒ u2 = 1
x22 u2 + v2 = 12 if u2 = 1 ⇒ v2 = 11
x23 u2 + v3 = 13 if u2 = 1 ⇒ v3 = 12
x33 u3 + v3 = 16 if v3 = 12 ⇒ u3 = 4
x34 u3 + v4 = 5 if u3 = 4 ⇒ v4 = 1

To summarize, we have

u1 = 0, u2 = 1, u3 = 4, v1 = 8, v2 = 11, v3 = 12, v4 = 1

Next, we use ui and vj to evaluate the non-basic variables by computing

ui + vj − cij for each non-basicxij

The results of this evaluation are shown in table 3.22 The preceding information, together
48
Operations Research Jane Akinyi Aduda

Table 3.22: Z-row coefficients

Non-basic variable ui + vj − cij
x12 u1 + v2 − c12 = 0 + 11 − 6 = 5
x13 u1 + v3 − c13 = 0 + 12 − 10 = 2
x14 u1 + v4 − c14 = 0 + 1 − 9 = −8
x24 u2 + v4 − c24 = 1 + 1 − 7 = −5
x31 u3 + v1 − c31 = 4 + 8 − 14 = −2
x32 u3 + v2 − c32 = 4 + 11 − 9 = 6

Table 3.23: Z-row coefficients

Basic x11 x12 x13 x14 x21 x22 x23 x24 x31 x32 x33 x34
z 0 5 2 -8 0 0 0 -5 -2 6 0 0

with the fact that ui +vj −cij = 0 for each basic xij , is actually equivalent to computing the
Z-row of the simplex tableau as the following summary shows. Because the transportation
seeks to minimize costs, the entering variable is the one having the most positive coefficient
in the z-row. Thus x32 is the entering variable. On the transportation tableau, these
operations are shown as in table 3.24

Table 3.24: First iteration

v1 = 8 v2 = 11 v3 = 12 v4 = 1 Supply
$8 $6 $10 $9
u1 = 0 35 5 2 −8 35
$9 $12 $13 $7
u2 = 1 10 20 20 −5 50
$14 $9 $16 $5
u3 = 4 −2 6 10 30 40
Demand 45 20 30 30

Having identified x32 as the entering variable, we need to determine the leaving vari-
able. It means we want to ship through x32 to reduce the shipping costs. From table
49
Operations Research Jane Akinyi Aduda

3.24, if we ship θ though x32 , then the maximum value of θ is determined based on two
conditions.

1. Supply limits and demand requirements remain satisfied.

2. Shipments through all routes remain non-negative.

These two conditions determine the maximum value of θ and the leaving variable in the
following manner.

1. Construct a closed loop that starts and ends in the entering variable cell (3, 2). The
loop consists of connected horizontal and vertical segments only (no diagonals are
allowed). Except for the entering variable cell, each corner of the closed loop must
coincide with a basic variable. Exactly one loop exists for a given entering variable.

2. Next, assign the amount (the maximum value) θ to the entering variable cell (3, 2).
For the supply and demand limits to remain satisfied, we must alternate between
subtracting and adding the amount θ at the successive corners of the loop. (It
doesn’t matter whether the loop is traced in a clockwise or anti clockwise direction).
For θ to remain non-negative, the new values of the variables then remain non-
negative if
x21 = 20 − θ ≥ 0

x33 = 10 − θ ≥ 0

The corresponding maximum value of θ = 10

So we add 10 to cell (3, 2), subtract 10 from cell (3, 3), add 10 to cell (2, 3) and subtract
10 from cell (2, 2) and get back to cell (3, 2). The closed loop runs through

(3, 2) → (3, 3) → (2, 3) → (2, 2) → (3, 2)

as shown in figure 3.25. Shipping through (3, 2) saves 6 × 10 = 60 interms of shipping

costs (i.e. it’s coefficient multiplies by the cost in the cell.)
50
Operations Research Jane Akinyi Aduda

Table 3.25: First iteration result

v1 = 8 v2 = 11 v3 = 12 v4 = 1 Supply
$8 $6 $10 $9
u1 = 0 35 3 2 −8 35
$9 $12 $13 $7
u2 = 1 10 20 − θ 20 + θ −5 50
$14 $9 $16 $5
u3 = 4 −2 6θ 10 − θ 30 40
Demand 45 20 30 30

Table 3.26: Second Table

v1 = 8 v2 = 11 v3 = 12 v4 = 7 Supply
$8 $6 $10 $9
u1 = 0 35 5 2 −2 35
$9 $12 $13 $7
u2 = 1 10 10 30 1 50
$14 $9 $16 $5
u3 = −2 −8 10 −6 30 40
Demand 45 20 30 30

We can see that the leaving variable is x33 and the new allocation then becomes The new
cost then becomes

Z = 35 × 8 + 10 × 9 + 12 × 10 + 13 × 30 + 9 × 10 + 5 × 30

= 280 + 90 + 120 + 390 + 90 + 150

= 1, 120

However, this table is not optimal. This gives the new entering variable as x12 . Again For
θ to remain non-negative, the new values of the variables then remain nonnegative if

x11 = 35 − θ ≥ 0

x22 = 10 − θ ≥ 0
51
Operations Research Jane Akinyi Aduda

The corresponding maximum value of θ = 10

So we add 10 to cell (1, 2), subtract 10 from cell (1, 1), add 10 to cell (2, 1) and subtract
10 from cell (2, 2) and get back to cell (1, 2). The closed loop runs through

(1, 2) → (1, 1) → (1, 2) → (2, 2) → (1, 2)

.This should reduce the transport costs by 5 × 10 = 50 and x22 leaves. This gives table
3.27 The new cost then becomes

Table 3.27: Third Table

v1 = 8 v2 = 6 v3 = 12 v4 = 2 Supply
$8 $6 $10 $9
u1 = 0 25 10 2 −7 35
$9 $12 $13 $7
u2 = 1 20 −5 30 −4 50
$14 $9 $16 $5
u3 = 3 −3 10 −1 30 40
Demand 45 20 30 30

Z = 25 × 8 + 10 × 6 + 20 × 9 + 30 × 13 + 9 × 10 + 5 × 30

= 200 + 60 + 180 + 390 + 90 + 150

= 1, 070
However, this table is still not optimal. This gives the new entering variable as x13 . Again
For θ to remain non-negative, the new values of the variables then remain non-negative if

x11 = 25 − θ ≥ 0

x23 = 30 − θ ≥ 0

The corresponding maximum value of θ = 25

So we add 25 to cell (1, 3), subtract 25 from cell (1, 1), add 25 to cell (2, 1) and subtract
25 from cell (2, 3) and get back to cell (1, 3). The closed loop runs through

(1, 3) → (1, 1) → (2, 1) → (2, 3) → (1, 3)

52
Operations Research Jane Akinyi Aduda

.This should reduce the transport costs by 2 × 25 = 50 x11 leaves. This gives table 3.28
This table is optimal as there is no positive coefficient and the new and optimal cost then

Table 3.28: Fourth Table

v1 = 6 v2 = 6 v3 = 10 v4 = 2 Supply
$8 $6 $10 $9
u1 = 0 −2 10 25 −7 35
$9 $12 $13 $7
u2 = 3 45 −3 5 −2 50
$14 $9 $16 $5
u3 = 3 −5 10 −3 30 40
Demand 45 20 30 30

becomes

Z = 10 × 6 + 10 × 25 + 9 × 45 + 5 × 13 + 10 × 9 + 5 × 30

= 60 + 250 + 405 + 65 + 90 + 150

= 1, 020

Note: This is similar to the solution obtained by VAM. In all cases VAM gives the best
starting solution, and in most cases, this result is always the optimal.

Example 3.4.

The Hard-rock Concrete Company has plants in three locations and is currently working
on four major construction projects, each located at a different site. The shipping cost
per truckload of concrete in dollars, daily plant capacities, and daily project requirements
are provided in the accompanying table.

53
Operations Research Jane Akinyi Aduda

HH
H To
HH Project A Project B Project C Project D Plant Capacities
From H
HH
H
Plant 1 10 4 12 15 90
Plant 2 12 5 8 25 55
Plant 3 9 7 6 10 35
Project
Requirements 40 45 65 30 180

Formulate an initial feasible solution to Hard-rock’s transportation problem using VAM

and find the optimal solution.

solution Draw the cost matrix and calculate the row and column penalties at each
level as shown below. the penalties have been put in the last column and the last rows.
allocate the most at the cheapest route corresponding to the row or column with the
largest penalty and keep repeating the procedure until all allocations are made.

10 4 12 15 6 10 4 12 15 2
45 90 45 - 90
12 5 8 25 3 12 5 8 25 4
- 55 - - 55
9 7 6 10 1 9 7 6 10 3
- 35 - 30 35
40 45 65 30 180 40 45 65 30 180
1 1 2 5 1 - 2 5

In the first table, we allocate the most (45) in cell (1,2) and cross out the second column
and in the second table we allocate the most (30) in cell (3,4) and cross out the fourth
column.
Next we allocate the most (55) in cell (2,3) and cross out row two and then in the next
table allocate the most (5) in cell (3,3) and cross out row three. After this we are left
with only cell (1,1) and (1,3) and by the least cost method, we allocate (40) in (1,1) and
(5) in (1,3) and this completes the allocation.
54
Operations Research Jane Akinyi Aduda

10 4 12 15 2 10 4 12 15 2
45 - 90 40 45 5 - 90
12 5 8 25 4 12 5 8 25 -
- - 55 - 55 - - 55 - 55
9 7 6 10 3 9 7 6 10 3
- 30 35 - - 5 30 35
40 45 65 30 180 40 45 65 30 180
1 - 2 - 1 - 6 -

The next step is to check for optimality of this allocation using MODI method (also
known as the method of multipliers). So we calculate the ui ’s and vj ’s and the coefficients
for the non-basic variables.

v1 = 10 v2 = 4 v3 = 12 v4 = 16 Capacities
10 4 12 15
u1 = 0
40 45 5 -1 90
12 5 8 25
u2 = −4
-6 -5 55 -23 55
9 7 6 10
u3 = −6
-5 -9 5 30 35
Requirements 40 45 65 30 180

The optimal solution is such that x11 = 40, x12 = 45, x13 = 5, x23 = 55, x33 = 5 and
x34 = 30 and the total cost is (40×10)+(45×4)+(5×12)+(55×8)+(5×6)+(30×10) =
$1, 410

3.3.7 Maximization using the Transportation algorithm

For this, the problem has to be converted to a minimisation problem first. This is achieved
by subtracting the largest value of the matrix from all the other values in the matrix and
solving using the modified matrix
55
Operations Research Jane Akinyi Aduda

3.4 The Assignment Model

The assignment model is a special form of a linear programming model that is similar
to the transportation model. There are differences, however. It involves assignment of
people to projects, jobs to machines, workers to jobs and teachers to classes etc., while
minimizing the total assignment costs. One of the important characteristics of assignment
problem is that only one job (or worker) is assigned to one machine (or project). Hence
the number of sources are equal the number of destinations and each requirement and
capacity value is exactly one unit.

Application Areas of Assignment Problem.

The assignment problem finds applicability in various diverse business situations such as:-

1. In assigning machines to factory orders.

2. In assigning sales/marketing people to sales territories.

3. In assigning contracts to bidders by systematic bid-evaluation.

4. In assigning teachers to classes.

5. In assigning accountants to accounts of the clients.

6. In assigning police vehicles to patrolling areas.

In order to use this method, one needs to know only the cost of making all the possible
assignments. Each assignment problem has a matrix (table) associated with it. Normally,
the objects (or people) one wishes to assign are expressed in rows, whereas the columns
represent the tasks (or things) assigned to them. The number in the table would then be
the costs associated with each particular assignment. It may be noted that the assignment
problem is a variation of transportation problem with two characteristics.

56
Operations Research Jane Akinyi Aduda

1. the cost matrix is a square matrix, and

2. the optimum solution for the problem would be such that there would be only one
assignment in a row or column of the cost matrix .

Mathematical Statement of Problem

An assignment problem is a special type of linear programming problem where the ob-
jective is to minimize the cost or time of completing a number of jobs by a number of
persons. Furthermore, the structure of an assignment problem is identical to that of a
transportation problem. This problem can be expressed as in table 3.29 Where Cij is the

Table 3.29: Assignment Model

Jobs
Worker Job 1 Job 2 ... Job n
Worker 1 c11 c12 ... c1n
Worker 2 c21 c22 ... c2n
.. .. .. ... ..
. . . .
Worker n cn1 cn2 ... cnn

cost of performing the j th job by the ith worker and xij is the ith individual assigned to
j th job. The total cost is therefore

Total cost = x11 × C11 + x12 × C12 + . . . + xnn × Cnn .

This can also be written as:

57
Operations Research Jane Akinyi Aduda

n X
X n
minimize Z = xij cij
i=1 j=1
Xn
subject to xij = 1, ∀ i
j=1
Xm
xij = 1 ∀ j
i=1

and xij = 0 or 1

The assignment problem can be solved by the following four methods :

1. Enumeration method

2. Simplex method

3. Transportation method

4. Hungarian method

There are various ways to solve assignment problems. Certainly it can be formulated as
a linear program (as we saw above), and the simplex method can be used to solve it. In
addition, since it can be formulated as a network problem, the network simplex method
may solve it quickly. However, sometimes the simplex method is inefficient for assignment
problems (particularly problems with a high degree of degeneracy).

Enumeration method

In this method, a list of all possible assignments among the given resources and activities is
prepared. Then an assignment involving the minimum cost, time or distance or maximum
profits is selected. If two or more assignments have the same minimum cost, time or
distance, the problem has multiple optimal solutions. This method can be used only
if the number of assignments is less. It becomes unsuitable for manual calculations if
number of assignments is large.
58
Operations Research Jane Akinyi Aduda

The Hungarian Method

Although assignment problem can be solved using either the techniques of Linear Pro-
gramming or the transportation method, the assignment method is much faster and ef-
ficient. This method was developed by D. Konig, a Hungarian mathematician and is
therefore known as the Hungarian method of assignment problem.

The Hungarian Algorithm has been used with a good deal of success on many problems
and is summarized as follows.

Step 1 . Determine the cost matrix from the given problem.

(i) If the no. of sources is equal to no. of destinations, go to step 3.

(ii) If the no. of sources is not equal to the no. of destination, go to step2.

Step 2. Add a dummy source or dummy destination, so that the cost table becomes a
square matrix. The cost entries of the dummy source/destinations are always zero.

Step 3. Locate the smallest element in each row of the given cost matrix and then sub-
tract the same from each element of the row. These computations are referred to
as row reductions. In other words, the best course of action is determined for each
row, and the penalty or “lost opportunity” is developed for all other row values.

Step 4. In the reduced matrix obtained in the step 3, locate the smallest element of
each column and then subtract the same from each element of that column. Each
column and row now have at least one zero. These computations are called column
reductions.

Step 5. In the modified matrix obtained in the step 4, search for the optimal assignment
as follows:

(a) Examine the rows successively until a row with a single zero is found. Enclose
this value in a  and cross off (X) all other zeros in its column. Continue in
this manner until all the rows have been taken care of.
59
Operations Research Jane Akinyi Aduda

(b) Repeat the procedure for each column of the reduced matrix.

(c) If a row and/or column has two or more zeros and one cannot be chosen by
inspection then assign arbitrarily any one of these zeros and cross off all other
zeros of that row / column.

(d) Repeat (a) through (c) above successively until the chain of assigning  or
cross (X) ends.

Step 6. If the number of assignment  is equal to n (the order of the cost matrix), an
optimum solution is reached. If the number of assignment is less than n(the order
of the matrix), go to the next step.

Step 7. Draw the minimum number of horizontal and/or vertical lines to cover all the
zeros of the reduced matrix.

Step 8. Develop the new revised cost matrix as follows:

(a) Find the smallest element of the reduced matrix not covered by any of the lines.

(b) Subtract this element from all uncovered elements and add the same to all the
elements laying at the intersection of any two lines.

Step 9. Go to step 6 and repeat the procedure until an optimum solution is attained.

Example 3.5. The Kenya premier league has four football games on a particular night.
Their office wants to assign four teams of officials to the four games in a way that will
minimize the total distance travelled by the officials. The distances in miles for each team
of officials to each game location are shown in Table 3.30
The supply is always one team of officials, and the demand is for only one team of officials
at each game.

Obtain the first and second reduced cost matrices by performing row and column reduc-
tions as shown in the table below.

60
Operations Research Jane Akinyi Aduda

Table 3.30: Assignment Model

Game sites
Official I II III IV
A 210 90 180 160
B 100 70 130 200
C 175 105 140 170
D 80 65 105 120

First reduced cost matrix Second reduced cost matrix

   
120 0 90 70 105 0 55 15
   
   
 30 0 60 130  15 0 25 75
   
   
 70 0 35 65   55 0 0 10
   
15 0 40 55 0 0 5 0

Notice that as can be seen from the second matrix, the assignment of team A to Site II
means that no other team of officials can be assigned to that game. Once this assignment
is made, the zero in row B is infeasible, which indicates that there is not a unique optimal
assignment for team B. Therefore, Table does not contain an optimal solution.

Next, draw the minimum number of horizontal or vertical lines necessary to cross out
all zeros through the rows and columns of the table. The three lines indicate that there are
only three unique assignments, whereas four are required for an optimal solution. (Note
that even if the three lines could have been drawn differently, the subsequent solution
method would not be affected.)

Table 3.31: Assignment Model

105 0 55 15
15 0 25 75
55 0 0 10
0 0 5 0

Next, subtract the minimum value that is not crossed out (15) in Table 3.31 from
all values not crossed out. Then, add this minimum value to those cells where two lines
61
Operations Research Jane Akinyi Aduda

intersect. All the others remain unchanged. The second iteration for this model with the
appropriate changes is shown in Table 3.32.

Table 3.32: Assignment Model

Game sites
Official I II III IV
A 90 0 40 0
B 0 0 10 60
C 55 15 0 10
D 0 15 5 0

No matter how the lines are drawn in Table 3.32, at least four are required to cross
out all the zeros as shown in table 3.33). This indicates that four unique assignments can

Table 3.33: Assignment Model continued

90 0 40 0 90 0 40 0
0 0 10 60 0 0 10 60
or
55 15 0 10 55 15 0 10
0 15 5 0 0 15 5 0

be made and that an optimal solution has been reached. First, team A can be assigned
to either the Game Site II or IV. The assignments can then be made as follows

First possibility Second possibility

   
90 0 40 0
A 90 0
A 40 0
   
0 0
A 10 60   0
A 0 10 60 
   
   
   
 55 15 0 10   55 15 0 10 
   
0
A 15 5 0 0 15 5 0
A

These assignments and their respective distances are summarized as shown in tables 3.34
and 3.35

62
Operations Research Jane Akinyi Aduda

Table 3.34: Assignments solution 1 Table 3.35: Assignments solution 2

Assignment Distance Assignment Distance
A → II 90 A → IV 160
B→I 100 B → II 70
C → III 140 C → III 140
D → IV 120 D→I 80
450 miles 450 miles

Example 3.6. A plant has four operators to be assigned to four machines. The time
(minutes) required by each worker to produce a product on each machine is shown in the
following table. Determine the optimal assignment and compute total minimum time.

Table 3.36: Assignment Problem 1

Operator\ Machine A B C D
1 10 12 9 11
2 5 10 7 8
3 12 14 13 11
4 8 15 11 9

Example 3.7. Carolina Airlines, a small commuter airline in North Carolina, has six
flight attendants whom it wants to assign to six monthly flight schedules in a way that
will minimize the number of nights they will be away from their homes. The numbers
of nights each attendant must be away from home with each schedule are given in the
following table. Identify the optimal assignments that will minimize the total number of
nights the attendants will be away from home.

3.4.1 Unbalanced assignment model

Example 3.8.

The department of Commerce and Economics Studies plans to hold seminars on four con-
63
Operations Research Jane Akinyi Aduda

Table 3.37: Assignment Problem 2

Attendant\ Schedule A B C D E F
1 7 4 5 10 5 7
2 4 5 4 12 7 5
3 9 9 10 7 10 7
4 11 6 7 5 9 9
5 5 8 5 10 7 5
6 10 12 10 9 9 9

temporary topics-Leasing, Portfolio Management, Mergers and Acquisitions and Deriva-

tives. These seminars should be held once a week in the afternoons. However, scheduling
these seminars (one for each topic and not more than one seminar per afternoon) has to be
done carefully so that the number of students unable to attend is kept to a minimum. The
lecture timetable indicates that the number of students who cannot attend a particular
seminar on a specific day is as follows:

Topic
Day Leasing Portfolio Mergers & Derivatives
Management Acquisitions

Monday 50 40 60 20
Tuesday 40 30 40 30
Wednesday 60 20 30 20
Thursday 30 30 20 30
Friday 10 20 10 30

Find an optimal schedule of the seminars and the total number of students who unfortu-
nately will miss at least one seminar.

solution
This is an unbalanced minimization assignment problem. We first balance it by adding a
dummy topic.

64
Operations Research Jane Akinyi Aduda

Topic
Day Leasing Portfolio Mergers & Derivatives Dummy
Management Acquisitions 0

Monday 50 40 60 20 0
Tuesday 40 30 40 30 0
Wednesday 60 20 30 20 0
Thursday 30 30 20 30 0
Friday 10 20 10 30 0

Then obtain the second reduced cost matrix as the first one would look exactly like the
original cost matrix since each row now has a zero.

  Draw the minimum number

40 20 50 0 0
  of lines to cover the zeros
30 10 30 10 0
 
  40 20 50 0 0
30 10 30 10 0
 
50 0 20 0 0
 
 
20 10 10 10 0 50 0 20 0 0
  20 10 10 10 0
0 0 0 10 0 0 0 0 10 0

We have only 4 lines and we need 5 assignments so this is not optimal. Subtract the
minimum uncovered element(10) from all uncovered elements and add it to all the elements
lying in the intersections of any two lines. All the others remain unchanged. We obtain
the matrix below and draw the minimum number of lines covering the zeros again as
shown below.
 
30 10 40 0 0
  30 10 40 0 0
20 0 20 10 0 
 
  20 0 20 10 0
50 0 20 10 10
 
50 0 20 10 10
 

10 0 0 10 0 
 10 0 0 10 0
  0 0 0 20 10
0 0 0 20 10

We now have 5 lines and we need five assignments, so we start our assignment in row 3
as this is the first instance with only one zero and proceed to completion. The optimal
schedule is shown below.
65
Operations Research Jane Akinyi Aduda

Day Topic No. of students missing

 
30 10 40 0 0
A Monday Derivatives 20
 
 20 0
A 20 10 0 Tuesday No seminar 0
 
 
Wednesday Portfolio Management 20
 
 50 0 20 10 10 
 
 
 10 0
A 0 10 0
A  Thursday Mergers & acquisitions 20
 
0 0
A 0
A 20 10 Friday Leasing 10
Total 70

3.4.2 Maximization using the Assignment algorithm

For this, the problem has to be converted to a minimisation problem first. This is achieved
by subtracting the largest value of the matrix from all the other values in the matrix and
solving using the modified matrix

Example 3.9. The Kenya Navy wishes to assign four ships to patrol four sectors of the
Indian ocean. In some areas ships are to be on the outlook for illegal fishing boats, and
in other sectors to watch for enemy submarines, so the commander rates each ship in
terms of its profitable efficiency in each sector. These relative efficiencies are illustrated
in the table below. On the basis of the ratings shown, what patrol assignments should
the commander use to produce the greatest overall efficiencies.

Sector
Ship A B C D

1 20 60 50 55
2 60 30 80 75
3 80 100 90 80
4 65 80 75 70

solution
We first compute the opportunity costs (by subtracting all the values in the matrix from
the largest value in the matrix) to convert this problem to a minimization problem and
them minimize the opportunity costs and in so doing we will be maximizing the efficiencies.
66
Operations Research Jane Akinyi Aduda

The relative efficiency matrix is The opportunity cost matrix is

   
20 60 50 55 80 40 50 45
   
60 30 80 75 40 70 20 25
   
   
   
80 100 90 80 20 0 10 20
   
65 80 75 70 35 20 25 30

The second reduced matrix (every col-

The first reduced matrix (every row
umn entry minus column minimum) is
entry minus row minimum) is given as
given as
 
40 0 10 5  
  25 0
A 10 0
20 50 0 5 
   
5 50 0 0
A 
   
 
20 0 10 20 



  5 0 10 15 
15 0 5 10  
0 0
A 5 5

The minimum number of straight lines needed to cover all zeros in this total opportunity
cost matrix is four. Hence an optimal assignment can be made. The optimal assignment
is ship 1 to sector D, ship 2 to sector C, ship 3 to sector B, and ship 4 to sector A.

The overall efficiency, computed from the original efficiency data is therefore

Assignment Efficiency
Ship 1 to sector D 55
Ship 2 to sector C 80
Ship 3 to sector B 100
Ship 4 to sector A 65
Total efficiency 300

67
Chapter 4

Inventory Models

Inventories are materials stored, waiting for processing, or experiencing processing. They
are ubiquitous throughout all sectors of the economy. Observation of almost any company
balance sheet, for example, reveals that a significant portion of its assets comprises inven-
tories of raw materials, components and sub-assemblies within the production process,
and finished goods. Most managers don’t like inventories because they are like money
placed in a drawer, assets tied up in investments that are not producing any return and, in
fact, incurring a borrowing cost. They also incur costs for the care of the stored material
and are subject to spoilage and obsolescence.

The purpose of inventory theory is to determine rules that management can use to
minimize the costs associated with maintaining inventory and meeting customer demand.
Inventory models answer the following questions

1. When should an order be placed for a product?

2. How large should each order be?

In general, there are three main reasons to hold inventory:

• Economies of scale Usually, placing an order has a certain cost that is independent
of the quantities ordered. Thus, more frequent orders incur higher costs of ordering
68
Operations Research Jane Akinyi Aduda

paper work. This may even cause higher transportation costs because the cost of
transportation per unit is often smaller for larger orders. Economies of scale also
play an important role in those cases where buying in large quantities results in a
reduction of the unit price.

• Uncertainties As a product is converted from raw material to final product, vari-

abilities in these transformational processes lead to losses. Inventories help mitigate
the negative impact of uncertainties.

• Customer service levels While variabilities in demand and supply are inevitable, in-
ventories help buffer against these variations and ensure product availability. Con-
sequently, delays in satisfying customer demand are reduced (when out of stock)
and unnecessary loss of revenue is avoided while improving customer satisfaction.

There are a few reasons why minimal inventories should be held.

1. Inventories represent significant investment in capital and resources that may be

employed to better purposes.

2. Many products have limited life cycles after which they become obsolete. When
the product is perishable (e.g., fresh food items), excess inventories may have to be
thrown out after expiry. Even if the product is not perishable, markdowns may be
required to dispose of excess inventory, leading to lower revenues or losses.

3. Inventories may hide inefficiencies in the system. Reduction of inventories coupled

with better production processes form the basis of just-in-time systems

Virtually all businesses hold inventory and inventory management remains one of the
biggest business challenges. Many consider it to be a necessary evil; it helps ensure product
availability for customers but consumes valuable resources to procure and maintain.

69
Operations Research Jane Akinyi Aduda

4.1 Types of Inventory

For production systems there is a need for basically three types of inventories. These
are inventories of raw material, finished products, and work-in-process. Any of these
inventory systems could exist at any stage of the path of a product from its origin until
it reaches the hands of the final users. Note that if various stages of production are
performed at different locations, or sometimes even by different companies, work finished
in one location could be considered the final product for that location and raw material
for the next location downstream.

• Inventory of raw materials: The need for keeping an inventory of raw materials
arises because of several factors. As mentioned in the above example, it is often
not possible or economical to procure raw materials exactly when the need for them
arises. The cost factors involved include the cost of paperwork, transportation
costs, economies of scale, and unforeseeable events. Unpredictable events such as
international shortage of certain goods, a strike at the supplying plant, a dry season,
and the like, may interrupt the production process if a sufficient inventory of raw
materials is not on hand.

• Inventory of work-in-process: This category of materials refers to semi-finished

goods. These are the outputs of one stage of production that act as the input
to the following stage. The uncertainties involved in this case are less than the
raw material category but never non-existent. The uncertainties are less because
the capacities of each stage are known and controllable. Thus, the designer can
match the capacities. However, uncertainties are still there because of problems in
scheduling, unforeseeable breakdown of machinery, and so on.

• Inventory of the finished product: The inventory of the final product involves
even more variability than is experienced with raw material. The demand for the
final product is often uncertain, and the manufacturer has little control over it.
Another factor, which is important when considering final product inventories, is
70
Operations Research Jane Akinyi Aduda

the cost involved in producing various sized lots. This sometimes favours production
of larger than demanded lot sizes because of potential savings in the unit costs.

4.2 ABC Classification of Inventories

This is sometimes known as Always Better Control. This system of control is also known
as Selective Approach System. In ABC system of inventory control, the materials are
classified depending on their turnover and annual consumption cost.

4.2.1 A -Class Items

These items are less in number, but consume a large portion of the total inventory in-
vestment. Here, the annual consumption cost is more important than the unit cost of
the material. For example let us consider, two materials X and Y . The unit cost of X
is shs.1/- and annual consumption is 1000 units. The unit cost of Y is shs.200 and the
annual consumption is 3 units. Then annual consumption cost of X is shs.1000/- and
that of Y is shs.600/-. Here X is considered as high consumption cost material than Y .
Such items which consume nearly 70% of inventory cost are classified as ‘A’ -class items.

4.2.2 B -class items

There will be certain materials, whose total annual consumption cost will be somewhere
between 20 to 25% of total inventory investment. These items are labelled as ‘B’ -class
items. These items will form 60% of the number of items stored.

71
Operations Research Jane Akinyi Aduda

4.2.3 C -class items

The last class of items which are labelled as ‘C’ -class items, will be large in number may
be 30 to 35% of total number of items stored, but consumes only 5 to 10% total inventory
investment.

Hence we can say that ‘A’-Class items are less in number and consumes more money,
‘B’- Class items are medium in number and consumes 20 to 25% inventory investment and
‘C’-Class items are large in number and consumes only 5 to 10% of inventory investment.

Procedure for ABC analysis

1. List out all items in stores along with their unit price and annual consumption.

2. Calculate the annual consumption cost of each item, which is given by multiplying
the quantity consumed in the time period and the unit cost. If ‘q’ is the quantity
consumed in the time period and ‘p’ is the unit price, then annual consumption
value q × p = qp.

3. Rearrange the list in the descending order of the annual consumption cost. i.e.
highest cost at the top and next highest is the second and so on and the last item
is the lowest consumption value item.

4. Calculate the cumulative total of annual consumption value.

5. Find the parentage of each cumulative value with respect to the total cost of inven-
tory.

6. Mark a line at 70%, 90% and at 100%. All the items covered by 70% line are ‘A’
class items, those which are covered between 70% line and 90% line are ‘B’ class
items and those are covered by 90% and 100% are ‘C’ class items.

Example 4.1. The details of material stocked in a company are given below with the
unit cost and the annual consumption in shs.. Classify the material in to A class, B class
and C class by ABC analysis.
72
Operations Research Jane Akinyi Aduda

S.No. Item Code No. Annual consumption in shs. Unit price in shs..
1 501 30,000 10
2 502 280,000 15
3 503 3,000 10
4 504 110,000 5
5 505 4,000 5
6 506 220,000 10
7 507 15,000 5
8 508 80,000 5
9 509 60,000 15
10 510 8,000 10

Solution
First we find the annual usage value for each item (unit price×annual usage) and rank
them in descending order.

S.No. Item Annual consumption Unit price Annual Rank

Code No. in shs. in shs. Usage value
A B C D =B×C
1 501 30,000 10 300,000 6
2 502 280,000 15 4,200,000 1
3 503 3,000 10 30,000 9
4 504 110,000 5 550,000 4
5 505 4,000 5 20,000 10
6 506 220,000 10 2,200,000 2
7 507 15,000 5 75,000 8
8 508 80,000 5 400,000 5
9 509 60,000 15 900,000 3
10 510 8,000 10 80,000 7

Then we list the items in their descending order of annual consumption value, find the
cumulative value of annual consumption value and find the percentage of cumulative value
with respect to total inventory value. Draw lines at 70%, 90% and at 100%.
73
Operations Research Jane Akinyi Aduda

Rank Item annual Cumulative Cumulative Percentage Category

No. usage annual usage annual usage of items
percentage
1 502 4,200,000 42,000 48 10 A
2 506 2,200,000 64,000 73 20 A
3 509 900,000 73,000 83 30 B
4 508 550,000 78,500 90 40 B
5 504 400,000 82,500 94 50 B
6 501 300,000 85,500 98 60 B
7 510 80,000 86,300 98.6 70 C
8 507 75,000 87,050 99.4 80 C
9 503 30,000 87,350 99.6 90 C
10 505 20,000 87,550 100 100 C

So In A class we have 2 items consuming 73% of the amount and in B class, we have 4
items consuming 25% of the amount and in C class, we have 4 items consuming about
2% of the inventory investment. This is shown in the graph in figure 4.1

analysis.png

Figure 4.1: ABC analysis

Inventory models involve some or all of the following variables: We list the factors that
are important in making decisions related to inventories.
74
Operations Research Jane Akinyi Aduda

1. Ordering cost (c(z)) This is the cost of placing an order to an outside supplier
or releasing a production order to a manufacturing shop. The amount ordered is z
and the function c(z) is often non-linear.

2. Set-up cost (K) A common assumption is that the ordering cost consists of a fixed
cost, that is independent of the amount ordered, and a variable cost that depends
on the amount ordered. The fixed cost is called the set-up cost.

3. Product cost (c) This is the unit cost of purchasing the product as part of an
order. If the cost is independent of the amount ordered, the total cost is cz, where
c is the unit cost and z is the amount ordered. Alternatively, the product cost may
be a decreasing function of the amount ordered.

4. Holding cost (h) This is the cost of holding an item in inventory for some given unit
of time. It usually includes the lost investment income caused by having the asset
tied up in inventory. This is not a real cash flow, but it is an important component
of the cost of inventory. If c is the unit cost of the product, this component of the
cost is cα , where α is the discount or interest rate. The holding cost may also
include the cost of storage, insurance, and other factors that are proportional to the
amount stored in inventory.

5. Shortage cost (p) When a customer seeks the product and finds the inventory
empty, the demand can either go unfulfilled or be satisfied later when the product
becomes available. The former case is called a lost sale, and the latter is called a
backorder. Although lost sales are often important in inventory analysis, they are
not considered in this section, so no notation is assigned to it. The total backorder
cost is assumed to be proportional to the number of units back-ordered and the
time the customer must wait. The constant of proportionality is p, the per unit
backorder cost per unit of time.

6. Demand rate (d) This is the constant rate at which the product is withdrawn
from inventory. (units /time)

75
Operations Research Jane Akinyi Aduda

7. Lot/Order Size (Q) This is the fixed quantity received at each inventory replen-
ishment.

8. Order level (S) The maximum level reached by the inventory is the order level.
When back-orders are not allowed, this quantity is the same as Q. When back-orders
are allowed, it is less than Q.

9. Cycle time (τ ) The time between consecutive inventory replenishments is the cycle
time. For the models of this section τ = Q/d

10. Cost per time (T ) This is the total of all costs related to the inventory system
that are affected by the decision under consideration.

11. Optimal Quantities (Q∗ , S ∗ , τ ∗ , T ∗ ) The quantities defined above that maximize
profit or minimize cost for a given model are the optimal solution.

We make the following assumptions:

1. Repetitive Ordering: The ordering decision is repetitive in the sense that it is

repeated in a regular fashion.

2. Constant Demand: Demand is assumed to occur at a known, constant rate.

3. Constant Lead Time: The lead time for each order is a known constant, L. By
the lead time we mean the length of time between the instant when an order is
placed and the instant at which the order arrives.

4. Continuous Ordering: An order may be placed at any time. Inventory models

that allow this are called continuous review models. If the amount of on-hand
inventory is reviewed periodically and orders may be placed only periodically, we
are dealing with a periodic review model.

76
Operations Research Jane Akinyi Aduda

4.3 Lot/Order Size Model with no Shortages or Basic

EOQ Model

For the basic EOQ model to hold, certain assumptions are required:

1. Demand is deterministic and occurs at a constant rate.

2. In an order of any size (say q units is placed, an ordering and set-up cost K is
incurred.

3. The lead time for each order is zero. This means orders arrive on time.

4. No shortages are allowed.

5. The cost per unit-year of holding inventory is h. This implies that if I units are
held for T years, a holding cost of IT h is incurred.

Given these assumptions, the EOQ model determines an ordering policy that minimizes
the yearly sum of ordering cost, purchasing cost, and holding cost.

4.4 Derivation of Basic EOQ Model

We begin by making some simple observations.

• We should never place an order when I, the inventory level, is greater than zero; if
we place an order then the we are incurring an unnecessary holding cost.

• If I = 0, we must place an order to prevent a shortage.

• Each time an order is placed (when I = 0) we should order the same quantity, Q.

• T C(Q)= annual cost of placing orders + annual purchasing cost + annual holding
cost.
77
Operations Research Jane Akinyi Aduda

• To determine the annual holding cost, we need to examine the behaviour of I over
time.

Annual demand= d Demand rate= 1/d

Cycle length= Q/d and the number of cycles per year =1/Q/d = d/Q
Holding Cost per year=Holding Cost/cycle and the number of cycles per year = (0.5Q2 /d)h(d/Q) =
0.5hQ

A key concept in the study of EOQ models is the idea of a cycle.

• Definition: Any interval of time that begins with the arrival of an order and ends
the instant before the next order is received is called a cycle.

• The average inventory during any cycle is simply half of the maximum inventory
level attained during the cycle.

• This result will hold in any model for which demand occurs at a constant rate.

T c(Q) = Ordering Cost+Purchasing Cost+Holding Cost

T c(Q) = unit cost (by the number of orders)+unit cost (total amount)+Holding Cost
 
d hQ
=K + pd +
Q 2

(4.1)

and taking derivatives with respect to the order quantity we obtain

   
dT c(Q) d h d h
≡ 0 ⇐⇒ 0 = −K 2
+ 0 + ⇐⇒ K 2
=
dQ Q 2 Q 2
2Kd
⇐⇒ Q2 = (4.2)
rh
2Kd
=⇒ Q∗ =
h
78
Operations Research Jane Akinyi Aduda

The economic order quantity, or EOQ, minimizes T c(Q) thus, Q∗ does indeed minimize
total annual cost. Substituting the EOQ into the total cost expression gives

√
r r
hKd hKd
T∗ = + pd + = pd + 2hKd (4.3)
2 2

Example 4.2. Braneast Airlines uses 500 tail lights per year. Each time an order for
tail lights is placed, an ordering cost of $5 is incurred. Each light cost 40 cents, and the
holding cost is 8 cents/light/year. Assume that demand occurs at a constant rate and
shortages are not allowed. What is the EOQ? How many orders will be placed each year?
How much time will elapse between the placement of orders?

solution
We are given that K = $5, h = 0.08/light/year, and d = 500 lights/year. The EOQ is
r r
2Kd 2 × 5 × 500
Q∗ = = = 250
h 0.08

Hence, the airline should place an order for 250 tail lights each time that inventory reaches
zero.
The time between placement (or arrival) of orders is simply the length of a cycle. Since
the length of each cycle is , τ = Q∗ /d, the time between orders will be

Q∗ 250 1
τ= = = per year
d 500 2

Example 4.3. Paul Peterson is the inventory manager for Office Supplies, Inc., a large
office supply warehouse. The annual demand for paper punches is 20,000 units. The
ordering cost is $100 per order, and the carrying (holding) cost is $5 per unit per year.
Determine the EOQ

solution r r
∗ 2Kd 2 × 100 × 20, 000
Q = = = 894
h 5

79
Chapter 5

Network Optimization Models

Networks arise in numerous settings and in a variety of guises. Transportation, electri-

cal, and communication networks pervade our daily lives. Network representations also
are widely used for problems in such diverse areas as production, distribution, project
planning, facilities location, resource management, and financial planningto name just
a few examples. In fact, a network representation provides such a powerful visual and
conceptual aid for portraying the relationships between the components of systems that
it is used in virtually every field of scientific, social, and economic endeavor.

One of the most exciting developments in operations research (OR) in recent years has
been the unusually rapid advance in both the methodology and application of network
optimization models. A number of algorithmic breakthroughs have had a major impact,
as have ideas from computer science concerning data structures and efficient data manip-
ulation. Consequently, algorithms and software now are available and are being used to
solve huge problems on a routine basis that would have been completely intractable two
or three decades ago.

Many network optimization models actually are special types of linear programming
problems. For example, both the transportation problem and the assignment problem
have network representations.

80
Operations Research Jane Akinyi Aduda

Some applications of network models include

1. Designing a natural gas pipeline

2. Determining the maximum capacity of a pipeline

3. Determining the shortest route through cities

4. Determining the time schedule of activities in a project

5. Determining the minimum cost flow schedule from oil fields to refineries through a
pipeline network.

In this section, we discuss four important kinds of network problems and some basic ideas
of how to solve them. These are:-

(i) The shortest-path problem (algorithm)

(ii) The minimum spanning tree problem (algorithm)

(iii) The maximum flow problem (algorithm)

(iv) The minimum cost flow problem (algorithm)

Each of the first three types have a very specific structure that arises frequently in appli-
cations. The fourth type provides a unified approach to many other applications because
of its far more general structure.

5.1 Terminologies used in Networks

A network consists of a set of nodes linked by arcs or branches and the notation used is
(N,A) where N is the set of nodes and A is the set of arcs.

81
Operations Research Jane Akinyi Aduda

G
3 5

E F
1 C H

D
2 4

Where N = {1, 2, 3, 4, 5}
and A = {(1, 2), (1, 3), (2, 3), (2, 4)(2, 5), (3, 4), (3, 5), (4, 5)}

Arcs are labeled by naming the nodes at either end; for example, AB is the arc between
nodes A and B where in our case A could be 1 and B could be 2.

Associated with each network is a flow of some type through the arcs.

A directed arc is one which allows positive flow only in one direction and zero flow
in the opposite directione.g a one way street.

An undirected arc is one which allows flow in either direction e.g. a pipeline.

A directed network is a network that has only directed arcs.

An undirected network is a network that has only undirected arcs.

A path between two nodes is a sequence of distinct arcs connecting these nodes
through other nodes regardless of thedirection of the flow in ach arc.

A path forms a cycle or a loop if it connects a node to itself through other nodes.

Two nodes are said to be connected if the network contains at least one undirected
path between them. (Note that the path does not need to be directed even if the network
is directed.) A connected network is a network where every pair of nodes is connected.
82
Operations Research Jane Akinyi Aduda

A tree is a cycle free connected network comprised of a subset of all the nodes.

A spanning tree is one that connects all nodes in the network. If a connected net-
work has n nodes, every spanning tree has exactly n − 1 arcs

3 5

2 4

The example below will be used subsequently to illustrate the approach to the first three
of these problems.

Nairobi National Park has recently been set aside for a limited amount of sight-
seeing and backpack hiking. Cars are not allowed into the park, but there is a narrow,
winding road system for trams and for jeeps driven by the park rangers. This road system
is shown in the Figure below, where location O is the entrance into the park; other letters
designate the locations of ranger stations (and other limited facilities). The numbers give
the distances of these winding roads in miles. The park contains a scenic wonder at sta-
tion T. A small number of trams are used to transport sightseers from the park entrance
to station T and back. The park management currently faces three problems.

1. To determine which route from the park entrance to station T has the smallest
83
Operations Research Jane Akinyi Aduda

total distance for the operation of the trams.(This is an example of the shortest-
path problem)

2. Telephone lines must be installed under the roads to establish telephone communi-
cation among all the stations (including the park entrance). Because the installation
is both expensive and disruptive to the natural environment, lines will be installed
under just enough roads to provide some connection between every pair of stations.
The question is where the lines should be laid to accomplish this with a minimum
total number of miles of line installed. (This is an example of the minimum spanning
tree problem)

3. More people want to take the tram ride from the park entrance to station T than
can be accommodated during the peak season. To avoid unduly disturbing the
ecology and wildlife of the region, a strict ration has been placed on the number of
tram trips that can be made on each of the roads per day. (These limits differ for
the different roads, as we shall see later) Therefore, during the peak season, various
routes might be followed regardless of distance to increase the number of tram trips
that can be made each day. The question pertains to how to route the various trips
to maximize the number of trips that can be made per day without violating the
limits on any individual road. (This is an example of the maximum flow problem)

2 2 7

5 4 5
O B D T

4 3 7
1 1

4
C E

84
Operations Research Jane Akinyi Aduda

5.2 The Shortest Path Problem

Consider an undirected and connected network with two special nodes called the origin
and the destination. Associated with each of the links (undirected arcs) is a non-negative
distance. The objective is to find the shortest path (the path with the minimum total
distance) from the origin to the destination. The essence of this procedure is that it fans
out from the origin, successively identifying the shortest path to each of the nodes of
the network in the ascending order of their (shortest) distances from the origin, thereby
solving the problem when the destination node is reached. As a linear program, the
problem is given as:-

xij = the amount of flow in arc (i, j)


1, if arc (i, j) is the shortest route

=
0, otherwise


cij = the length of arc (i, j)

and the objective is to

X
maximize Z = cij xij
all defined arcs (i,j)

The constraints represent the conservation of flows equations at each node.

Applying this algorithm to the Nairobi National Park problem gives:-

After the work shown above, the shortest path from the destination to the origin can be
traced back through the last column of table 5.1 as either T → D → E → B → A → O
or T → D → B → A → O. Therefore, the two alternates for the shortest path from
the origin to the destination have been identified as O → A → B → E → D → T and
O → A → B → D → T , with a total distance of 13 miles on either path.

85
Operations Research Jane Akinyi Aduda

Table 5.1: Table for park problem

solved nodes closest total nth
directly connected connected distance nearest minimum last
n to unsolved nodes unsolved node involved node distance connection
1 O A 2 A 2 OA
2,3 O C 4 C 4 OC
A B 2+2=4 B 4 AB
A D 2+7=9
4 B E 4+3=7 E 7 BE
C E 4+4=8
A D 2+7=9
5 B D 4+4=8 D 8 BD
E D 7+1=8 D 8 ED
6 D T 8+5=13 T 13 DT
E T 7+7=14

5.3 The Minimum Spanning Tree Problem

It deals with linking nodes of a network directly or indirectly using the shortest total
length of connecting branches, e.g. the construction of paved roads between two towns
may pass through one or more other towns. The most economical design of the road
system would call for minimizing the total miles of paved roeds.

For the minimum spanning tree problem, if we consider the problem of the Nairobi
National Park, the required property is that the chosen links must provide a path between
each pair of nodes.

The minimum spanning tree problem can be summarized as follows.

1. You are given the nodes of a network but not the links. Instead, you are given the
potential links and the positive length for each if it is inserted into the network.
86
Operations Research Jane Akinyi Aduda

(Alternative measures for the length of a link include distance, cost, and time.)

2. You wish to design the network by inserting enough links to satisfy the requirement
that there be a path between every pair of nodes.

3. The objective is to satisfy this requirement in a way that minimizes the total length
of the links inserted into the network.

A network with n nodes requires only (n − 1) links to provide a path between each pair of
nodes. No extra links should be used, since this would needlessly increase the total length
of the chosen links. The (n − 1) links need to be chosen in such a way that the resulting
network (with just the chosen links) forms a spanning tree. Therefore, the problem is to
find the spanning tree with a minimum total length of the links. So the figure belows
illustrates this concept of a spanning tree for the Nairobi National Park problem. This
network is a feasible solution (with a value of 24 miles for the total length of the links)

2 2

4 5
O B D T

4 7

C E

5.4 The Maximum Flow Problem

Now recall that the third problem facing the Nairobi National Park management during
the peak season is to determine how to route the various tram trips from the park entrance
(station O) to the scenic wonder (station T) to maximize the number of trips per day.
87
Operations Research Jane Akinyi Aduda

(Each tram will return by the same route it took on the outgoing trip, so the analysis
focuses on outgoing trips only.) To avoid unduly disturbing the ecology and wildlife of
the region, strict upper limits have been imposed on the number of outgoing trips allowed
per day in the outbound direction on each individual road. For each road, the direction
of travel for outgoing trips is indicated by an arrow in the figue below. The number on
the arrow gives the upper limit on the number of outgoing trips allowed per day.

5 1 3

7 4 9
O B D T

4 5 1 6
2

4
C E

Given the limits, one feasible solution is to send 7 trams per day, with 5 using the route
O → B → E → T , 1 using O → B → C → E → T and 1 using O → B → C → E →
D → T . However, because this solution blocks the use of any routes starting with O → C
(because the E → T and E → D capacities are fully used), it is easy to find better feasible
solutions. Many combinations of routes (and the number of trips to assign to each one)
need to be considered to find the one(s) maximizing the number of trips made per day.
This kind of problem is called a maximum flow problem.

In general terms, the maximum flow problem can be described as follows.

1. All flow through a directed and connected network originates at one node, called the
source, and terminates at one other node, called the sink. (The source and sink in
88
Operations Research Jane Akinyi Aduda

the Nairobi National Park problem are the park entrance at node O and the scenic
wonder at node T, respectively.)

2. All the remaining nodes are transshipment nodes. (These are nodes A, B, C, D,
and E in the Seervada Park problem.)

3. Flow through an arc is allowed only in the direction indicated by the arrowhead,
where the maximum amount of flow is given by the capacity of that arc. At the
source, all arcs point away from the node. At the sink, all arcs point into the node.

4. The objective is to maximize the total amount of flow from the source to the sink.
This amount is measured in either of two equivalent ways, namely, either the amount
leaving the source or the amount entering the sink.

5.5 The Minimum Cost Flow Problem

The minimum cost flow problem holds a central position among network optimization
models, both because it encompasses such a broad class of applications and because it
can be solved extremely efficiently. Like the maximum flow problem, it considers flow
through a network with limited arc capacities. Like the shortest-path problem, it con-
siders a cost (or distance) for flow through an arc. Like the transportation problem or
assignment problem, it can consider multiple sources (supply nodes) and multiple desti-
nations (demand nodes) for the flow, again with associated costs. In fact, all four of these
previously studied problems are special cases of the minimum cost flow problem, as we
will demonstrate shortly. The reason that the minimum cost flow problem can be solved
so efficiently is that it can be formulated as a linear programming problem so it can be
solved by a streamlined version of the simplex method called the network simplex method.

The minimum cost flow problem is described below.

1. The network is a directed and connected network.

89
Operations Research Jane Akinyi Aduda

2. At least one of the nodes is a supply node.

3. At least one of the other nodes is a demand node.

4. All the remaining nodes are transshipment nodes.

5. Flow through an arc is allowed only in the direction indicated by the arrowhead,
where the maximum amount of flow is given by the capacity of that arc. (If flow
can occur in both directions, this would be represented by a pair of arcs pointing in
opposite directions.)

6. The network has enough arcs with sufficient capacity to enable all the flow generated
at the supply nodes to reach all the demand nodes.

7. The cost of the flow through each arc is proportional to the amount of that flow,
where the cost per unit flow is known.

8. The cost of the flow through each arc is proportional to the amount of that flow,
where the cost per unit flow is known.

9. The objective is to minimize the total cost of sending the available supply through
the network to satisfy the given demand. (An alternative objective is to maximize
the total profit from doing this.)

5.6 PERT/CPM Models for Project Management

A project such as construction of a bridge, highway, power plant, repair and maintenance
of an oil refinery or an air plane design, development and marketing a new product, re-
search and development etc., may be defined as a collection of interrelated activities (or
tasks) which must be completed in a specified time according to a specified sequence and
require resources such as personnel, money, materials, facilities etc.

90
Operations Research Jane Akinyi Aduda

The growing complexities of todays projects had demanded more systematic and more
effective planning techniques with the objective of optimizing the efficiency of executing
the project. Efficiency here implies effecting the utmost reduction in the time required
to complete the project while accounting for the economic feasibility of using available
resources.

Project management has evolved as a new field with the development of two ana-
lytic techniques for planning, scheduling and controlling projects. These are the Critical
Path Method (CPM) and the Project Evaluation and Review Technique (PERT). PERT
and CPM are basically time-oriented methods in the sense that they both lead to the
determination of a time schedule.

5.6.1 Basic difference between PERT and CPM

Though there are no essential differences between PERT and CPM as both of them
share in common the determination of a critical path and are based on the network
representation of activities and their scheduling that determines the most critical activities
to be controlled so as to meet the completion date of the project.

PERT (Program Evaluation Review Technique)

1. Since PERT was developed in connection with R and D work, it had to cope with the
uncertainties associated with R and D activities. In PERT, total project duration is
regarded as a random variable and therefore associated probabilities are calculated
so as to characterise it.

2. It is an event-oriented network because in the analysis of network emphasis is given

an important stages of completion of task rather than the activities required to be
performed to reach to a particular event or task.

91
Operations Research Jane Akinyi Aduda

3. PERT is normally used for projects involving activities of non-repetitive nature in

which time estimates are uncertain.

4. It helps in pin pointing critical areas in a project so that necessary adjustment can
be made to meet the scheduled completion date of the project.

CPM (Critical Path Method)

1. Since CPM was developed in connection with a construction project which consisted
of routine tasks whose resource requirements and duration was known with certainty,
it is basically deterministic.

2. CPM is suitable for establishing a tradeoff for optimum balancing between schedule
time and cost of the project.

3. CPM is used for projects involving activities of repetitive nature.

5.6.2 PERT/CPM Network Components And Precedence Re-

lationship

PERT/CPM networks consists of two major components as discussed below:

Events: which are points in time that signifiy the completion of some activities and the
beginning of new ones. The beginning and end points of an activity is thus described
by 2 events usually known as the Tail and head events. Events are commonly
represented by nodes in the network diagram. They do not consume time and
Resource.

Activities: which represent project operations or tasks to be conducted. An arc is

commonly used to represent an activity, with its head indicating the direction of
progress in the project. Activities originating from a certain event cannot start until
the activities terminating at the same event have been completed. They consume
time and Resource.
92
Operations Research Jane Akinyi Aduda

Events in the network diagram are identified by numbers. Numbers are given to events
such that arc head number must be greater than arc tail number. Activities are identified
by the numbers of their starting (tail) event and ending (head) event. An arc (i, j) extends
between two events, the tail event i representing the start and the head event j represents
the completion of the activity as shown below.

Activity
I J

The figure below shows another example, where activities (1, 3) and (2, 3) must be com-
pleted before activity (3, 4) can start.

3 4

The rules for constructing the network diagram are as follows:

1. Each activity is represented by one and only one arrow in the network.

2. No two activities can be identified by the same head and tail events.

3. To ensure the correct precedence relationship in the arrow diagram, the following
questions must be answered as every activity is added to the network:

(a) What activities must be completed immediately before these activity can start
?

(b) What activities must follow this activity ?

93
Operations Research Jane Akinyi Aduda

(c) What activity must occur concurrently with this activity ?

Example 5.1. VIVA company is planning to move their operations from Nairobi to
Kisumu. The manager in charge of planning the move, wants to ensure that everything
comes off according to plan, and making sure that the deadline is met. Table 5.2 shows
a list of all the activities the manager plans.

Table 5.2: VIVA Company Example

IMMEDIATE ACTIVITY
ACTIVITY DESCRIPTION PREDECESSORS TIME RESOURCES
A Select Office Site - 3
B Create Organizational and - 5
Financial Plan
C Determine Personnel Requirements B 3
D Design Facility A, C 4
E Construct Interior D 8
F Select Personnel to Move F 2
G Hire New Employees F 4
H Move Records, Key Personnel, etc. F 2
I Make Financial Arrangements B 5
with Institutions in Kisumu
J Train New Personnel H, E, G 3

NOTE The numbers assigned to the nodes are arbitrary. They are simply used to iden-
tify events and do not imply anything about precedence relationships. In the network
diagram each activity must start at the node in which its immediate predecessors ended.
Dummy activities can also be included to avoid problems of hanging events when drawing
the network diagram. A dummy activity in a project network analysis has zero duration.

The network diagram is as below:

94
Operations Research Jane Akinyi Aduda

D [4]design E [8] const.

2 5 7

A [3] select sight H [2] move

dummy J [3] train
dummy
F [2] select per.
1 4 6 G [4] hire

8 9
C [3]
B [5] plan
I [5] fin. arrgt.

5.6.3 Critical Path Calculations

The application of PERT/CPM should ultimately yield a schedule specifying the start
and completion time of each activity. The arc diagram is the first step towards achieving
that goal. The start and completion times are calculated directly on the arc diagrams
using simple arithmetic. The end result is to classify the activities as critical or non
critical. An activity is said to be critical if a delay in the start of that course makes a
delay in the completion time of the entire project. A noncritical activity is such that
the time between its earliest start and its latest completion time is longer than its actual
duration. A noncritical activity is said to have a slack or float time.

5.6.4 Determination of the Critical Path

A critical path defines a chain of critical activities that connects the start and end events
of the arc diagram. In other words, the critical path identifies all the critical activities of
the project. The critical path calculations include two phases.
95
Operations Research Jane Akinyi Aduda

The first phase is called the Forward Pass where all calculations begin from the start
node and move to the end node. At each node a number is computed representing the
earliest occurrence time of the corresponding event. These numbers are shown in squares
. In forward pass we note the number of heads joining the event. We take the maximum
earliest timing through these heads.

The second phase called the Backwards Pass, begins calculations from the “end”node
and moves to the “start”node. The number computed at each node is shown in a triangle
near end point which represent the latest occurrence time of the corresponding event.

Let ESi be the earliest start time of all the activities emanating from event i, i.e.
ESi represents the earliest occurrence time of event i, if i = 1 is the “start”event then
conventionally, for the critical path calculations, ESi = 0. If Dij be the duration of the
activity (i, j). Then the forward pass calculations are given by the formula:

ESi = max ESi + Dij

for all defined (i, j) activities with ESi = 0. Thus in order to compute ESj for event j,
ESi for the tail events of all the incoming activities (i, j) must be computed first. With
the computation of all ESj , the forward pass calculations are completed.

The backward pass starts from the “end”event. The objective of this phase to calculate
LCi , the latest completion time for all the activities coming into the event i. Thus if i = n
is the end event LCn = ESn initiates the backward pass. In general for any node i,

LCi = min LCj − Dij

for all defined activities are calculated, which ends the calculation of backward pass.

The critical path activities can now be identified by using the results of the forward
96
Operations Research Jane Akinyi Aduda

and backward passes. An activity (i, j) lies on the critical path if it satisfies the following
conditions.

1. ESi = LCi

2. ESj = LCj

3. ESj − ESi = LCj − LCi = Dij

These conditions actually indicate that there is no float or slack time between the earliest
stand and the latest start of the activity. Thus the activity must critical.

Example 5.2. Consider a network which stands from node 1 and terminate at node 6,
the time required to perform each activity is indicated on the arcs.

2 7 5

3 6
1 2 dummy 5 6

3 3 2

Let us start with forward pass with ES1 = 0.

Since there is only one incoming activity (1, 2) to event 2 with D12 = 3.

ES2 = ES1 + DS2 = 0 + 3 = 3.

Let us consider the end 3, since there is only one incoming activity (2, 3) to event 3, with
D23 = 3

ES3 = ES2 + D23 = 3 + 3 = 6.

97
Operations Research Jane Akinyi Aduda

To obtain ES4 , since there are two activities (3, 4) and (2, 4) to the event 4 with D24 = 2
and D34 = 0.

ES4 = maxi=2,3 {ESi + De4 }

= max {ES2 + D24 , ES3 + D34 }

= max {3 + 2, 6 + 0} = 6

Similary ES5 = 13 and ES6 = 19 Which completes the first phase.

In the second phase we have

LC6 = 19 = ES6

LC5 = 19 − 6 = 13

LC4 = minj=5,6 {LCj − D4j } = 6

LC3 = 6, LC2 = 3 and LC1 = 0 and hence activities (1, 2), (2, 3) (3, 4) (4, 5) (5, 6) are
critical and (2, 4) (4, 6), (3, 6), are non-critical. Thus the activities (1, 2), (2, 3) (3, 4) (4, 5)
(5, 6) define the critical path which is the longest possible time to complete the project.

5.6.5 Project Management PERT

The analysis in CPM does not take into the case where time estimates for the different
activities are probabilistic. Also it does not consider explicitly the cost of schedules. Here
we will consider both probability and cost aspects in project scheduling. Probability
considerations are incorporated in project scheduling by assuming that the time estimate
for each activity is based on 3 different values. They are:-

a= The optimistsic time, which will be required if the execution of the project goes
extremely well.

b= The pessimistic time, which will be required if everything goes bad.

98
Operations Research Jane Akinyi Aduda

m= The most likely time, which will be required if execution is normal.

The expected duration of each activity

a + b + 4m
D̄ =
6

This estimate can be used to study the single estimate D in the critical path calculation.
The variance of each activity denoted by V is defined by
 2
b−a
V =
6

The earliest expected times for the node i denoted by E(µi ) for each node i is obtained by
taking the sum of expected times of all activities leading to the node i, when more than
one activity leads to a node i, then greatest of all E(µi ) is chosen. Let µi be the earliest
occurrence time of the event i, we can consider µi as a random variable. Assuming that
all activities of the network are statistically independent, we can calculate the mean and
X
the variance of the µi as follows:- E[µi ] = ESi and V ar[µi ] = Vk , Where K defines the
k
activities along the largest path leading to i. For the latest expected time, we consider
the last node.

Note: The probability distribution of times for completing an event can be approxi-
mated by the normal distribution due to central limit theorem. Since µi represents the
earliest occurrence time, event will meet a certain schedule time STi (specified by an
analyst) with probability
 
µi − E[µi ] STi − E[µi ]
prob (µi ≤ STi ) = prob √ ≤ √
Vi Vi
= prob (Z ≤ Ki )

where
STi − E[µi ]
Z ∼ N (0, 1) and Ki = √
Vi

99
Chapter 6

Waiting Line Theory or Queuing

Model

Here customer or element represents a person or machine or any other thing, which is in
need of some service from servicing point and service represents any type of attention to
the customer to satisfy his need. For example,

1. Person going to hospital to get medical advice from the doctor is an element or a
customer,

2. A person going to railway station or a bus station to purchase a ticket for the journey
is a customer or an element,

3. A person at the ticket counter of a cinema hall,

4. A person at a grocery shop waiting to purchase consumables,

5. A bank pass book tendered to a bank clerk for withdrawal of money,

6. A machine break down and waiting for the attention of a maintenance crew.

7. Vehicles waiting at a traffic signal,

In the above cases, the service means,

100
Operations Research Jane Akinyi Aduda

1. Doctor is a service facility and medical care is a service,

2. Ticket counter is a service facility and issue of ticket is service.

3. Ticket counter is a service facility and issue of ticket is service.

4. Shop owner is a service facility and issue of items is service.

5. Bank clerk is a service facility and passing the cheque is service.

6. Maintenance crew is service facility and repairing the machine is service.

7. Traffic signals are service facility and control of traffic is service.

In general we can say that a flow of customers from infinite or finite population towards
the service facility forms a queue or waiting line on account of lack of capability to serve
them all at once.

Queues or waiting lines stand for the number of customers waiting to be served.
Queue does not include the customer being served. The process or system that performs
the services to the customer is termed as service channel or service facility.

6.1 QUEUING SYSTEM OR PROCESS

When we speak of queue, we have to deal with two elements, i.e. Arrivals and Service
facility. the entire queuing system can be completely described by:

(a) The input (Arrival pattern)

(b) The service mechanism or service pattern,

(c) The queue discipline and

(d) Customer behaviour.

101
Operations Research Jane Akinyi Aduda

Components of the queuing system are arrivals, the element waiting in the queue, the
unit being served, the service facility and the unit leaving the queue after service.

6.1.1 Input Process

This describes the way in which the customers arrive and join the system. In general
customer arrival will be in random fashion, which cannot be predicted, because the cus-
tomer is an independent individual and the service organization has no control over the
customer.

The input process is described by the following characteristics

Size of arrivals: This greatly depends on the nature of size of the population, which may
be infinite or finite. The arrival pattern can be more clearly described in terms of
probabilities and consequently the probability distribution for inter-arrival times i.e.
the time between two successive arrivals or the distribution of number of customers
arriving in unit time must be defined. Here we assume that customers arrive in
Poisson or Completely random fashion.

Inter-arrival time: The period between the arrival of individual customers may be con-
stant or may be scattered in some distribution fashion. Most queuing models assume
that the some inter-arrival time distraction applies for all customers throughout the
period of study. Here we assume distribution of service time follows a negative
exponential distribution

Capacity of the service system: In queuing context the capacity refers to the space
available for the arrivals to wait before taken to service. The space available may
be limited or unlimited. When the space is limited, length of waiting line crosses a
certain limit; no further units or arrivals are permitted to enter the system till some
waiting space becomes vacant.

Customer behaviour: The length of the queue or the waiting time of a customer or the
102
Operations Research Jane Akinyi Aduda

idle time of the service facility mostly depends on the behaviour of the customer.
Here the behaviour refers to the impatience of a customer during the stay in the
line. Customer behaviour can be classified as:

(i) Balking: This behaviour signifies that the customer does not like to join the
queue seeing the long length of it. This behaviour may effect in loosing a
customer by the organization. Always a lengthy queue indicates insufficient
service facility and customer may not turn out next time.

(ii) Reneging: In this case the customer joins the queue and after waiting for some
time looses his patience and leaves the queue. This behaviour of the customer
may also cause loss of customer to the organization.

(iii) Collusion: In this case several customers may collaborate and only one of
them may stand in the queue. One customer represents a group of customer.
Here the queue length may be small but service time for an individual will be
more. This may break the patience of the other customers in the waiting line
and situation may lead to any type of worst episode.

(iv) Jockeying: If there are number of waiting lines depending on the number of
service stations, for example gas stations, Cinema theatres, etc. A customer in
one of the queue after seeing the other queue length, which is shorter, with a
hope of getting the service, may leave the present queue and join the shorter
queue. Perhaps the situation may be that other queue which is shorter may be
having more number of Collaborated customers. In such case the probability
of getting service to the customer who has changed the queue may be very
less. Because of this character of the customer, the queue lengths may goes on
changing from time to time.

6.1.2 Service Mechanism or Service Facility

Service facilities are arranged to serve the arriving customer or a customer in the waiting
line is known as service mechanism. We consider
103
Operations Research Jane Akinyi Aduda

Service facility design: Arriving customers maybe asked to form a single line (Single
queue) or multi line (multi queue) depending on the service need. When they stand
in single line it is known as Single channel facility when they stand in multi lines it
is known as multi channel facility.

(i) Single channel queues: Only one unit can be served at a time, hence arriving
customers form a queue near the facility. The next element is drawn into
service only when the service of the previous customer is over.

(ii) Multi Channel queues: When the input rates increases, and the demand for
the service increases, the management will provide additional service facilities
to reduce the rush of customers or waiting time of customers. In such cases,
different queues will be formed in front of different service facilities.

Queue discipline or Service discipline: When the customers are standing in a queue,
they are called to service depending on the nature of the customer. The order in
which they are called is known as Service discipline. There are various ways in which
the customer called to serve. They are:

(i) First In First Out (FIFO) or First Come First Served (FCFS). We are quite
aware that when we are in a queue, we wish that the element which comes
should be served first, so that every element has a fair chance of getting service.
Moreover it is understood that it gives a good morale and discipline in the
queue.

(ii) Last in first out (LIFO) or Last Come First Served (LCFS). In this system, the
element that arrived last will have a chance of getting service first. In general,
this does not happen in a system where human beings are involved. But this is
quite common in Inventory system. This can also be written as First In Last
Out (FILO).

(iii) Service In Random Order (SIRO). In this case the items are called for service
in a random order. The element might have come first or last does not bother;
the servicing facility calls the element in random order without considering the
104
Operations Research Jane Akinyi Aduda

order of arrival. This may happen in some religious organizations but generally
it does not followed in an industrial / business system.

(iv) Service By Priority. Priority disciplines are those where any arrival is chosen
for service ahead of some other customers already in queue. As an example,
in a doctors clinic, when the doctor is treating a patient with stomach pain,
suddenly a patient with heart stroke enters the clinic, the doctor asks the
patient with stomach pain to wait for some time and give attention to heart
patient. This is the rule of priority.

6.2 QUEUING PROBLEMS

The most important information required to solve a waiting line problem is the nature
and probability distribution of arrivals and service pattern. The answer to any waiting
line problem depends on finding:

1. Queue length: The probability distribution of queue length or the number of persons
in the system at any point of time. Further we can estimate the probability that
there is no queue.

2. Waiting time: This is probability distribution of waiting time of customers in the

queue. That is we have to find the time spent by a customer in the queue before the
commencement of service, which is called his waiting time in the queue. The total
time spent in the system is the waiting time in the queue plus the service time. The
waiting time depends on various factors, such as:

(a) The number of units already waiting in the system,

(b) The number of service stations in the system,

(c) The schedule in which units are selected for service,

(d) The nature and magnitude of service being given to the element being served.

105
Operations Research Jane Akinyi Aduda

3. Service time: It is the time taken for serving a particular arrival.

4. Average idle time or Busy time distribution: The average time for which the system
remains idle. We can estimate the probability distribution of busy periods. If we
suppose that the server is idle initially and the customer arrives, he will be provided
service immediately. During his service time some more customers will arrive and
will be served in their turn according to the system discipline. This process will
continue in this way until no customer is left unattended to and the server becomes
free again after serving all the customers. At this stage we can conclude, that the
busy period is over.

6.3 SYMBOLS USED IN QUEUING MODELS

A queue is designated or described as shown below: A model is expressed as A/B/S :

(d/f ) where,

A: Arrival pattern of the units, given by the probability distribution of inter-arrival time
of units. For example, Poisson distribution, and inter arrival time is 1 minute or 10
units arrive in 30 minutes etc.

B: The probability distribution of service time of individual being actually served. For
example the service time follows negative exponential distribution and 10 units are
served in 10 minutes or the service time is 3 minutes, etc.

S: The number of service channels in the system. For example the item is served at one
service facility or the person will receive service at 3 facilities etc.

d: Capacity of the system. That is the maximum number of units the system can accom-
modate at any time. For example, the system has limited capacity of 40 units or
the system has infinite capacity etc.

f : The manner or order in which the arriving units are taken into service i.e. FIFO /
LIFO / SIRO /Priority.
106
Operations Research Jane Akinyi Aduda

6.3.1 Notations

X: Inter arrival time between two successive customers (arrivals).

Y : The service time required by any customer.

w: The waiting time for any customer before it is taken into service.

v: Time spent by the customer in the system.

n: Number of customers in the system, (in the waiting line) at any time, including the
number of customers being served.

pn (t) Probability that customers arrive in the system in time t

Φn (t): Probability that n units are served in time t.

U (T ): Probability distribution of inter arrival time p(t ≤ T ).

V (T ): Probability distribution of servicing time p(t ≤ T ).

F (N ): Probability distribution of queue length at any time p(N ≤ n).

En : Some state of the system at a time when there are n units in the system.

λn : Average number of customers arriving per unit of time, when there are already n
units in the system.

λ: Average number of customers arriving per unit of time.

µn : Average number of customers being served per unit of time when there are already
n units in the system.

µ: Average number of customers being served per unit of time.

1
: Inter arrival time between two arrivals.
λ
1
: Service time between two units or customers.
µ

107
Operations Research Jane Akinyi Aduda

λ
ρ= : System utility or traffic intensity which tells us how much time the system was
µ
3
utilized in a given time. For example given time is 8 hours and if ρ = , it means
8
to say that out of 8 hours the system is used for 3 hours and (8-3=5) 5 hours the
system is idle.

Formulae used

1. Average number of arrivals per unit of time = λ

2. Average number of units served per unit of time = µ

λ
3. Traffic intensity or utility ratio = ρ = with the condition (µ > λ)
µ
4. Probability that the system is empty = p0 = (1 − ρ).

5. Probability that there are n units in the system pn = ρn p0

ρ λ
6. Average number of units in the system E(n) = =
1−ρ (µ − λ)
ρ2 λ2
7. Average number of units in the waiting line = EL = =
(1 − ρ) µ(| −λ)
1 1
8. Average waiting time (mean time in the system) = Ew = =
(µ − λ) (1 − ρ)
9. Average length of waiting line with the condition that it is always greater than zero
ρ λ λ
= V (n) = = =
(1 − ρ)2 (µ − λ)2 µ(µ − λ)
1 1
10. Average time an arrival spends in the system = E(v) = =
µ(1 − ρ) (µ − λ)
11. System is busy = ρ

12. Idle time = (1 − ρ)

13. Probability that a customer has to wait on arrival ρ

108
Operations Research Jane Akinyi Aduda

Example 6.1. A T.V. Repairman finds that the time spent on his jobs have an expo-
nential distribution with mean of 30 minutes. If he repairs sets in the order in which they
come in, and if the arrival of sets is approximately Poisson with an average rate of 10 per
8 hour day, what is repairman’s expected idle time each day? How many jobs are ahead
of the average set just brought in?

Solution

This problem is Poisson arrival/Negative exponential service/single channel/infinite

capacity/ FIFO type problem.
10 5
Data: λ = 10 sets per 8 hour day = = sets per hour.
8 4
1 1
Given = 30 minutes, hence µ = × 60 = 2 2 sets per hour.
µ   30
λ 5/4
Hence, Utility ratio = ρ = = = 58 = 0.625. This means out of 8 hours, 5 hours
µ 2
the system is busy i.e. repairman is busy.
5 3
Probability that there is no queue = The system is idle = (1−ρ) = (1−ρ) = 1− = =
8 8
That is out of 8 hours the repairman will be idle for 3 hours.

Number of sets ahead of the set just entered = Average number of sets in system
λ ρ 0.625 5
= = = = ahead of jobs just came in.
(µ − λ) (1 − ρ) 1 − 0.625 3
Example 6.2. The arrivals at a telephone booth are considered to be following Poisson
distribution with an average time of 10 minutes between one arrival and the next. Length
of the phone call is assumed to be distributed exponentially with a mean of 3 minutes.

1. What is the probability that a person arriving at the booth will have to wait?

2. What is the average length of queue that forms from time to time?

Solution
1
Data: Time interval between two arrivals = 10 minutes = , Length of phone call
λ
1 1 1
= 3 minutes = . Hence λ = = 0.1 per minute and µ = = 0.33 per minute, and
µ 10 3
109
Operations Research Jane Akinyi Aduda

λ 0.10
ρ= = = 0.3.
µ 0.33

1. Any person who is coming to booth has to wait when there is somebody in the
queue. He need not wait when there is nobody in the queue i.e. the queue is empty.
Hence the probability of that an arrival does not wait = p0 = (1 − ρ).

Hence the probability that an arrival has to wait = 1-The probability that an arrival
does not wait =(1 − p0 )1 − (1 − ρ) = ρ = 0.3. That means 30% of the time the fresh
arrival has to wait. That means that 70% of the time the system is idle.

2. Average length of non-empty queue from time to time = (Average length of the
1
waiting line with the condition that it is always greater than zero) = = 1.43
1 − 0.3
persons.

Example 6.3. In a departmental store one cashier is there to serve the customers. And
the customers pick up their needs by themselves. The arrival rate is 9 customers for every
5 minutes and the cashier can serve 10 customers in 5 minutes. Assuming Poisson arrival
rate and exponential distribution for service rate, find:

1. Average number of customers in the system.

2. Average number of customers in the queue or average queue length.

3. Average time a customer spends in the system.

4. Average time a customer waits before being served.

9 10
Data: Arrival rate is λ = = 1.8 customers per minute. Service rate µ = = 2
5 5
λ 1.8
customers per minute. Hence ρ = = = 0.9
µ 2

ρ 0.9 0.9
1. Average number of customers in the system E(n) = = = =9
1−ρ 1 − 0.9 0.1
customers.
1 1
2. Average time a customer spends in the system = E(v) = = = 5
µ−λ 2 − 1.8
minutes.
110
Operations Research Jane Akinyi Aduda

λ2 λ
3. Average number of customers in the queue = E(L) = = (ρ) × =
µ(µ − λ) (µ − λ)
1.8
0.9 × = 8.1 customers.
2 − 1.8
ρ λ 0.9
4. Average time a customer spends in the queue = = = =
µ(1 − ρ) µ(µ − λ) 2(1 − 0.9)
0.9
= 4.5 minutes.
0.2
Example 6.4. A branch of a bank has only one typist. Since typing work varies in length
(number of pages to be typed), the typing rate is randomly distributed approximating a
Poisson distribution with a mean service rate of 8 letters per hour. The letter arrives at
a rate of 5 per hour during the entire 8-hour workday. If the typist is valued at shs.. 150
per hour, determine:

1. Equipment utilization,

2. The percent time an arriving letter has to wait,

3. Average system time, and

4. Average idle time cost of the typewriter per day.

Example 6.5. A product manufacturing plant at a city distributes its products by trucks,
loaded at the factory warehouse. It has its own fleet of trucks plus trucks of a private
transport company. This transport company has complained that sometimes its trucks
have to wait in line and thus the company loses money paid for a truck and driver of
waiting truck. The company has asked the plant manager either to go in for a second
warehouse or discount prices equivalent to the waiting time. The data available is:

Average arrival rate of all trucks = 3 per hour.

Average service rate is = 4 per hour.

The transport company has provided 40% of the total number of trucks. Assuming
that these rates are random according to Poisson distribution, determine:

1. The probability that a truck has to wait?

111
Operations Research Jane Akinyi Aduda

2. The waiting time of a truck that has to wait,

3. The expected waiting time of company trucks per day.

112