Overview of Operations Research

Lecture notes
July 2024

1 Introduction
The term Operations Research (OR) was first coined by MC Closky and Trefthen in 1940 in a small town,
Bowdsey of UK. The main origin of OR was during the second world war – The military commands of UK
and USA engaged several inter-disciplinary teams of scientists to undertake scientific research into strategic
and tactical military operations. The name Operations Research (OR) was invented because the team was
dealing with research on military operations. The encouraging results obtained by British OR teams moti-
vated US military management to start with similar activities. Now OR activities has become universally
applicable to any area such as transportation, hospital management, agriculture, libraries, city planning,
financial institutions, construction management and so forth.

The wide scope of applications of operations research encouraged various organizations and individuals to
define it as follows:
• Operations research is the application of the methods of science to complex problems in the direction and
management of large systems of men, machines, 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 outcomes of
alternative decisions, strategies or controls. The purpose is to help management in determining its
policy and actions scientifically. - Operational Research Society, UK
• The application of the scientific method to the study of operations of large complex organizations or
activities. It provides top level administrators with a quantitative basis for decisions that will increase
the effectiveness of such organizations in carrying out their basic purposes. – Committee on OR
National Research Council, USA
• Operations research is essentially a collection of mathematical techniques and tools which in conjunction
with a system’s approach, are applied to solve practical decision problems of an economic or engineering
nature. – Daellenbach and George, 1978

• Operations research is the art of winning wars without actually fighting them.
As the discipline of operations research grew, numerous names such as Operations Analysis, Systems Anal-
ysis, Decision Analysis, Management Science, Quantitative Analysis, Decision Science were given to it. This
is because of the fact that the types of problems encountered are always concerned with ‘effective decision’.

2 Features of OR Approach
OR utilizes a planned approach following a scientific method and an interdisciplinary team, in order to
represent complex functional relationship as mathematical models, for the purpose of providing a quantitative
basis for decision-making and uncovering new problems for quantitative analysis. This definition implies
additional features of OR approach. The broad features of OR approach in solving any decision problem are
summarized as follows:

1
 Figure 1: Conceptual Framework of an Organization and OR Application Process

• Interdisciplinary Approach: For solving any managerial decision problem often an interdisciplinary
teamwork is essential. This is because while attempting to solve a complex management problem, one
person may not have the complete knowledge of all its aspects such as economic, social, political,
psychological, engineering, etc. Hence, a team of individuals specializing in various functional areas of
management should be organized so that each aspect of the problem can be analysed to arrive at a
solution acceptable to all sections of the organization.
• Scientific Approach: Operations research is the application of scientific methods, techniques and
tools to problems involving the operations of systems so as to provide those in control of operations
withoptimum solutions to the problems (Churchman et al.). The scientific method consists of observing
and defining the problem; formulating and testing the hypothesis; and analysing the results of the test.
The data so obtained is then used to decide whether the hypothesis should be accepted or not. If the
hypothesis is accepted, the results should be implemented, otherwise not.
• Holistic Approach: While arriving at a decision, an operations research team examines the relative
importance of all conflicting and multiple objectives. It also examines the validity of claims of various
departments of the organization from the perspective of its implications to the whole organization.
• Objective-oriented Approach: An operations research approach seeks to obtain an optimal solution
to the problem under analysis. For this, a measure of desirability (or effectiveness) is defined, based
on the objective(s) of the organization. A measure of desirability so defined is then used to compare
alternative courses of action with respect to their possible outcomes.

2
3 Models and Modelling in OR
The key to model-building lies in abstracting only the relevant variables that affect the criteria of the
measures-of-performance of the given system and in expressing the relationship in a suitable form. However,
a model should be as simple as possible so as to give the desired result. On the other hand, oversimplifying
the problem can also lead to a poor decision. Model enrichment is done by changing value of variables, and
relaxing assumptions. Refer to Table 1 and Table 2.

4 Methods for solving OR Models

In general, the following three methods are used for solving OR models, where values of decision variables
are obtained that optimize the given objective function (a measure of effectiveness).

• Analytical (Deductive) Method: In this method, classical optimization techniques such as calculus,
finite difference and graphs are used for solving an OR model. The analytical methods are non-iterative
methods to obtain an optimal solution of a problem.
• Numerical (Iterative) Method: When analytical methods fail to obtain the solution of a particular
problem due to its complexity in terms of constraints or number of variables, a numerical (or iterative)
method is used to find the solution. In this method, instead of solving the problem directly, a general
algorithm is applied for obtaining a specific numerical solution. The numerical method starts with
a solution obtained by trial and error, and a set of rules for improving it towards optimality. The
solution so obtained is then replaced by the improved solution and the process of getting an improved
solution is repeated until such improvement is not possible or the cost of further calculation cannot be
justified.
• Monte Carlo method: This method is based upon the idea of experimenting on a mathematical
model by inserting into the model specific values of decision variables for a selected period of time
under different conditions and then observing the effect on the criterion chosen. In this method,
random samples of specified random variables are drawn to know how the system is behaving for a
selected period of time under different conditions. The random samples form a probability distribution
that represents the real-life system and from this probability distribution, the value of the desired
random variable can then be estimated.

5 Phases of OR Approach
The most important feature of operations research is the use of the scientific method and the building of
decision models. The operations research approach to problem solving is based on these major phases:
• Defining the problem and gathering data: The first task is to study the relevant system and
develop a well-defined statement of the problem. This includes determining appropriate objectives,
constraints, interrelationships and alternative course of action. Ascertaining the appropriate objec-
tives is very important aspect of problem definition. OR team typically spends huge amount of time
in gathering relevant data to gain accurate understanding of problem and to provide input for next
phase. OR teams uses Data mining methods to search large databases for interesting patterns that
may lead to useful decisions.
• Formulating a mathematical model: This phase is to reformulate the problem in terms of math-
ematical symbols and expressions. The mathematical model of a business problem is described as
the system of equations and related mathematical expressions. The advantages of using mathematical
models are:
1. Describe the problem more concisely
2. Makes overall structure of problem comprehensible
3. Helps to reveal important cause-and-effect relationships

3
 4. Indicates clearly what additional data are relevant for analysis
• Deriving solutions from the model: This phase is to develop a procedure for deriving solutions
to the problem. A common theme is to search for an optimal or best solution. The main goal of OR
team is to obtain an optimal solution which minimizes the cost and time and maximizes the profit. To
obtain the solution, the OR team uses:

1. Heuristic procedure (designed procedure that does not guarantee an optimal solution) is used
to find a good sub-optimal solution.
2. Metaheuristics provides both general structure and strategy guidelines for designing a specific
heuristic procedure to fit a particular kind of problem.
3. Post-Optimality analysis is the analysis done after finding an optimal solution. It is also
referred as what-if analysis. It involves conducting sensitivity analysis to determine which param-
eters of the model are most critical in determining the solution.
• Testing the model and its solutions: After deriving the solution, it is tested as a whole for errors
if any. The process of testing and improving a model to increase its validity is commonly referred as
Model validation. The OR group doing this review should preferably include at least one individual
who did not participate in the formulation of model to reveal mistakes. A systematic approach to
test the model is to use Retrospective test. This test uses historical data to reconstruct the past and
then determine the model and the resulting solution. Comparing the effectiveness of this hypothetical
performance with what actually happened, indicates whether the model tends to yield a significant
improvement over current practice.

• Preparing to apply the model: After the completion of testing phase, the next step is to install a
well-documented system for applying the model. This system will include the model, solution procedure
and operating procedures for implementation. The system usually is computer-based. Databases and
Management Information System may provide up-to-date input for the model. An interactive computer
based system called Decision Support System is installed to help the manager to use data and models
to support their decision making as needed. A managerial report interprets output of the model and
its implications for applications.
• Implementation: The last phase of an OR study is to implement the system as prescribed by the
management. The success of this phase depends on the support of both top management and operating
management.

6 Techniques of OR
There is no unique set of problems that can be solved by using OR techniques. Several OR techniques can
be grouped into some basic categories as given below.

• Distribution (Allocation) Models: Distribution models are concerned with the allotment of avail-
able resources so as to minimise cost or maximize profit subject to the restrictions. Methods for solving
such type of problems are known as mathematical programming techniques. We distinguish between
linear and non-linear programming problems on the basis of linearity and non-linearity of the objective
function and/or constraints respectively. In linear programming problems, the objective function is lin-
ear and constraints are also linear inequalities/equations. Transportation and Assignment models can
be viewed as special cases of linear programming. These can be solved by specially devised procedures
called Transportation and Assignment Techniques.
• Job Sequencing Models: These models involve the selection of such a sequence of performing a series
of jobs to be done on service facilities (machines) that optimize the efficiency measure of performance
of the system. In other words, sequencing is concerned with such a problem in which efficiency measure
depends upon the order or sequence of performing a series of jobs.

4
 Figure 2: Methodology of Operations Research

• Network Models: These models are applicable in large projects involving complexities and inter-
dependencies of activities. Project Evaluation and Review Techniques (PERT) and Critical Path
Method CPM) are used for planning, scheduling and controlling complex project which can be char-
acterized as net-works.
• Replacement Models: These models deal with the determination of optimum replacement policy
in situations that arise when some items or machinery need replacement by a new one. Individual
and group replacement policies can be used in the case of such equipments that fail completely and
instantaneously.
• Production/Inventory Models: Inventory/Production models are concerned with the determina-
tion of the optimal (economic) order quantity and ordering (production) intervals considering the
factors such as-demand per unit time, cost of placing orders, costs associated with goods held up in
the inventory and the cost due to shortage of goods, etc. Such models are also useful in dealing with
quantity discounts and multiple products.
• Competitive Strategy Models (Games Theory): These models are used to determine the be-
haviour of decision-making under competition or conflict. Methods for solving such models have not
been found suitable for industrial applications, mainly because they are referred to an idealistic world
neglecting many essential features of reality.

• Waiting Line (or Queueing) Models: In queueing models an attempt is made to predict:
1. how much average time will be spent by the customer in a queue?
2. what will be an average length of waiting line or queue?
3. what will be the traffic intensity of a queueing system? etc.

The study of waiting line problems provides us methods to minimize the sum of costs of providing
service and cost of obtaining service which are primarily associated with the value of time spent by
the customer in a queue.

5
Category Description
Structure
• Physical model represents the physical appearance of the real object under
study, either reduced in size or scaled up. Physical models cannot be manipu-
lated and are not very useful for prediction.

– Iconic Models: An iconic model is a scaled (small or big in size) version

of the system. Such models retain some of the physical characteristics of
the system they represent. Eg. blueprints of homes, maps, etc.
– Analogue Models: An analogue model does not resemble physically the
system they represent, but retain a set of characteristics of the system.
Such models are more general than iconic models and can also be manip-
ulated. Eg. Organizational charts, graphs of time series, etc.

• Symbolic models These models use algebraic symbols (letters, numbers) and
functions to represent variables and their relationships for describing the prop-
erties of the system. Symbolic models are precise and abstract and can be
analysed by using laws of mathematics.
– Verbal Models: These models describe properties of a system in written
or spoken language. Eg. Written sentences, books, etc.
– Mathematical Models: These models use mathematical symbols, let-
ters, numbers and mathematical operators (+, –, ÷, ×) to represent re-
lationships among variables of the system to describe its properties or
behaviour. The solution to such models is obtained by applying suitable
mathematical technique.

Function
• Descriptive Models: These models are used to investigate the outcomes or
consequences of various alternative courses of action (strategies, or actions).
Since these models evaluate the consequence based on a given condition (or
alternative) rather than on all other conditions, there is no guarantee that an
alternative selected is optimal. Simulation is an example of a descriptive model
for conducting experiments with the systems based on given alternatives
• Predictive Models: These models represent a relationship between dependent
and independent variables and hence measure ‘cause and effect’ due to changes
in independent variables. These models do not have an objective function as
a part of the model of evaluating decision alternatives based on outcomes or
pay off values. Also, through such models decision-maker does not attempt
to choose the best decision alternative, but can only have an idea about the
possible alternatives available to him.

Table 1: Types of OR Models (1)

6
 Category Description
Time reference
• Static models: Static models represent a system at a particular point of time
and do not take into account changes over time. For example, an inventory
model can be developed and solved to determine an economic order quantity
assuming that the demand and lead time would remain same throughout the
planning period.
• Dynamic models: Dynamic models take into account changes over time, i.e.,
time is considered as one of the variables while deriving an optimal solution.
Thus, a sequence of interrelated decisions over a period of time are made to select
the optimal course of action in order to achieve the given objective. Dynamic
programming is an example of a dynamic model

Degree of Certainty
• Deterministic models: If all the parameters, constants and functional rela-
tionships are assumed to be known with certainty when the decision is made,
the model is said to be deterministic. Linear programming models are example
of deterministic models.
• Probabilistic (Stochastic) models: If at least one parameter or decision
variable is random (probabilistic or stochastic) variable, then the model is said to
be probabilistic. Since at least one decision variable is random, the independent
variable, which is the function of dependent variable(s), will also be random.
This means consequences (or payoff) due to certain changes in the independent
variable(s) cannot be predicted with certainty.

Quantification
• Heuristic models: If certain sets of rules (may not be optimal) are applied in
a consistent manner to facilitate solution to a problem, then the model is said
to be Heuristic.
• Analytical models: These models have a specific mathematical structure and
thus can be solved by the known analytical or mathematical techniques. Any
optimization model (which requires maximization or minimization of an objec-
tive function) is an analytical model.
• Simulation models: These models have a mathematical structure but can-
not be solved by the known mathematical techniques. A simulation model is
essentially a computer-assisted experimentation on a mathematical structure
of a problem in order to describe and evaluate its behaviour under certain as-
sumptions over a period of time. Simulation models are more flexible than
mathematical models and can, therefore, be used to represent a complex sys-
tem that cannot be represented mathematically. These models do not provide
general solution like those of mathematical models.

Table 2: Types of OR Models (2)