UNIT-1

Operational Research:
―Operations Research is the systematic, method oriented study of the basic structure,
characteristics, functions and relationships of an organization to provide the executive with a
sound, scientific and quantitative basis for decision making.‖ – E.L. Arnoff and M.J. Netzorg

Operations research (OR) is an analytical method of problem-solving and decision-making that is

useful in the management of organizations. In operations research, problems are broken down
into basic components and then solved in defined steps by mathematical analysis.

The process of operations research can be broadly broken down into the following steps:
1. Identifying a problem that needs to be solved.
2. Constructing a model around the problem that resembles the real world and variables.
3. Using the model to derive solutions to the problem.
4. Testing each solution on the model and analy
5. zing its success.
6. Implementing the solution to the actual problem.
Characteristics of operations research
There are three primary characteristics of all operations research efforts:

1. Optimization- The purpose of operations research is to achieve the best performance

under the given circumstances. Optimization also involves comparing and narrowing
down potential options.

2. Simulation- This involves building models or replications in order to try out and test
solutions before applying them.

3. Probability and statistics- This includes using mathematical algorithms and data to
uncover helpful insights and risks, make reliable predictions and test possible solutions.

Uses of operations research

Operations research can be applied to a variety of use cases, including:

 Scheduling and time management.

 Urban and agricultural planning.
 Enterprise resource planning (ERP) and supply chain management (SCM).
 Inventory management.
 Network optimization and engineering.
 Packet routing optimization.
 Risk management.

Characteristics (Features) of Operation Research:

Main characteristics of operations research (O.R.) are follows:

(i) Inter-Disciplinary Team Approach: This requires an inter-disciplinary team including

individuals with skills in mathematics, statistics, economics, engineering, material sciences,
computer etc.
(ii) Wholistic Approach to the System: While evaluating any decision, the important interactions
and their impact on the whole organisation against the functions originally involved are
reviewed.

(iii) Methodological Approach: O.R. utilises the scientific method to solve the problem

(iv) Objective Approach: O.R. attempts to find the best or optimal solution to the problem under
consideration, taking into account the goals of the organization.

Methodology of Operation Research:

Operation Research, is a scientific approach for decision-making, and therefore must follow
following steps:

1. Formulating the Problem: The problem must be first clearly defined. It is common to start
the O.R. study with tentative formulation of the problem, which is reformulated over and again
during the study. The study must also consider economical aspects.

While formulating the O.R. study, analyists must analyse following major components:

(i) The environment: Environment involves physical, social and economical factors which are
likely to affect the problem under consideration. O.R. team or analysts must study the
organisation contents including men, materials, machines, suppliers, consumers, competitors, the
government and the public.

(ii) Decision-makers: Operation analyst must study the decision-maker and his relationship to
the problem at hand.

(iii) Objectives: Considering the problem as whole, objectives should be defined.

(iv) Alternatives: The O.R. study determines as to which alternative course of action is most
effective to achieve the desired objectives. Expected reactions of the competitors to the
alternative must also be considered.

2. Deriving Solution: Models are used to determine the solution either by simulation or by
mathematical analysis. Mathematical analysis for deriving optimum solution includes analytical
or numerical procedure, and uses various branches of mathematics.

3. Testing the Model and Solution: A properly formulated and correctly manipulated model is
useful in predicting the effect of changes in control variables on the overall system effectiveness.
The validity of the solution is checked by comparing the results with those obtained without
using the model.

4. Establishing Controls over the Solution: The solution derived from a model remains
effective so long as the uncontrolled variables retain their values and the relationship. The
solution goes out of control, if the values of one or more variables vary or relationship between
them undergoes a change. In such circumstances the models need to be modified to take the
changes into account.

5. Implementing the Solution: Solution so obtained should be translated into operating

procedure to make it easily understandable and applied by the concerned persons. After applying
the solution to the system, O.R. group must study the response of the system to the changes
made.

Operation Research Models:

Operation Research model is an idealised representation of the real life situation and represents
one or more aspects of reality. Examples of operation research models are: a map, activity charts
balance sheets, PERT network, break-even equation, economic ordering quantity equation etc.
Objective of the model is to provide a means for analysing the behaviour of the system for
improving its performance.

Classification of Models:
Models can be classified on the basis of following factors:
1. By degree of Abstraction:
i. Mathematical models.
ii. Language models.
2. By Function:
i. Descriptive models.
ii. Predictive models.
iii. Normative models for repetitive problems.
3. By Structure:
i. Physical models.
ii. Analogue (graphical) models.
iii. Symbolic or mathematical models.
4. By Nature of Environment:
i. Deterministic models.
ii. Probabilistic models.
5. By the Time Horizon:
i. Static models.
ii. Dynamic models.

Characteristics of a Good Model:

i. Assumptions should be simple and few.
ii. Variables should be as less as possible.
iii. It should be able to assimilate the system environmental changes without change in its
framework.
iv. It should be easy to construct.

Constructing the Model:

A mathematical model is a set of equations in which the system or problem is described. The
equations represent objective function and constraints. Objective function is a mathematical
expressions of objectives (cost or profit of the operation), while constraints are mathematical
expressions of the limitations on the fulfillment of the objectives.

These expressions consist of controllable and uncontrollable variables.

The general form of a mathematical model is:

O = f (xi, yi)
Where O = Objective function
Xi = Controllable variables
yi = Uncontrollable variables
f = Relationship between O, and xi, yi.
Since model is only an approximation of the real situation, hence it may not include all the
variables.

Simplification in Operation Research Models:

While constructing the model, efforts should be made to simplify them, but only up to the extent
so that there is no significant loss of accuracy.

Some of the common simplifications are:

i. Omitting certain variables.
ii. Aggregating (or grouping) variables.
iii. Changing the nature of variables e.g., considering variables as constant or continuous.
iv. Changing relationship between variables i.e., considering them as linear or straight line.
v. Modify constraints.

Techniques of Operation Research:

Important techniques of Operation Research are being described hereunder:

(i) Inventory Control Models:

Operation Research study involves balancing inventory costs against one or more of the
following costs:
i. Shortage costs.
ii. Ordering costs.
iii. Storage costs.
iv. Interest costs.

This study helps in taking decisions about:

i. How much to purchase.
ii. When to order.
iii. Whether to manufacture or to purchase i.e., make and buy decisions.
The most well-known use is in the form of Economic Order Quantity equation for finding
economic lot size.

(ii) Waiting Line Models:

These models are used for minimizing the waiting time and idle time together with the costs
associated therewith.

Waiting line models are of two types:

(a) Queuing theory, which is applicable for determining the number of service facilities and/or
the timing of arrivals for servicing.

(b) Sequencing theory which is applicable for determining the sequence of the servicing.

(iii) Replacement Models:

These models are used for determining the time of replacement or maintenance of item,
which may either:

(i) Become obsolete, or

(ii) Become inefficient for use, and
 (iii) Become beyond economical to repair or maintain.
(iv) Allocation Models:

These models are used to solve the problems arising when:

(a) There are number of activities which are to be performed and there are number of alternative
ways of doing them,

(b) The resources or facilities are limited, which do not allow each activity to be performed in
best possible way. Thus these models help to combine activities and available resources so as to
optimise and get a solution to obtain an overall effectiveness.

Applications of Operation Research: These techniques are applied to a very wide range of
problems. Here only some of the common applications are being mentioned:

(i) Distribution or Transportation Problems:

In such problems, various centres with their demands are given and various warehouses with
their stock positions are also known, then by using linear programming technique, we can find
out most economical distribution of the products to various centres from various warehouses.

(ii) Product Mix:

These techniques can be applied to determine best mix of the products for a plant with available
resources, so as to get maximum profit or minimum cost of production.

(iii) Production Planning:

These techniques can also be applied to allocate various jobs to different machines so as to get
maximum profit or to maximise production or to minimise total production time.

(iv) Assignment of Personnel:

Similarly, this technique can be applied for assignment of different personnel with different
aptitude to different jobs so as to complete the task within a minimum time.

(v) Agricultural Production:

We can also apply this technique to maximise cultivator‘s profit, involving cultivation of number
of items with different returns and cropping time in different type of lands having variable
fertility.

(vi) Financial Applications:

Many financial decision making problems can be solved by using linear programming technique.

Some of them are:

(i) To select best portfolio in order to maximise return on investment out of alternative
investment opportunities like bonds, stocks etc. Such problems are generally faced by the
managers of mutual funds, banks and insurance companies.

(ii) In deciding financial mix strategies, involving the selection of means for financing firm,
projects, inventories etc.
Limitations of Operations Research:

i. These do not take into account qualitative and emotional factors.

ii. These are applicable to only specific categories of decision-making problems.
iii. These are required to be interpreted correctly.
iv. Due to conventional thinking, changes face lot of resistance from workers and sometimes
even from employer.
v. Models are only idealised representation of reality and not be regarded as absolute.

Linear Programming: Linear programming is a mathematical concept that is used to find the
optimal solution of the linear function. This method uses simple assumptions for optimizing the
given function. Linear Programming has a huge real-world application and it is used to solve
various types of problems.

Components of Linear Programming

A linear programming problem has two basic parts:
 First Part: It is the objective function that describes the primary purpose of the formation to
maximize some return or to minimize some.
 Second Part: It is a constant set, It is the system of equalities or inequalities that describe the
condition or constraints of the restriction under which Optimisation is to be accomplished.

Linear Programming Formula

A linear programming problem consists of,
 Decision variables
 Objective function
 Constraints
 Non-Negative restrictions

Decision variables are the variables x, and y, which decide the output of the linear
programming problem and represent the final solution.
The objective function, generally represented by Z, is the linear function that needs to be
optimized according to the given condition to get the final solution.
The restrictions imposed on decision variables that limit their values are called constraints.
Now, the general formula of a linear programming problem is,
Objective Function: Z = ax + by
Constraints: cx + dy ≥ e, px + qy ≤ r
Non-Negative restrictions: x ≥ 0, y ≥ 0

To Solve Linear Programming Problems

Before solving the linear programming problems first we have to formulate the problems
according to the standard parameters. The steps for solving linear programming problems are,
Step 1: Mark the decision variables in the problem.
Step 2: Build the objective function of the problem and check if the function needs to be
minimized or maximized.
Step 3: Write down all the constraints of the linear problems.
Step 4: Ensure non-negative restrictions of the decision variables.
Step 5: Now solve the linear programming problem using any method generally we use either
the simplex or graphical method.

Linear Programming Methods

We use various methods for solving linear programming problems. The two most common
methods used are,
 Simplex Method
 Graphical Method
Let‘s learn about these two methods in detail in this article,
Linear Programming Simplex Method
One of the most common methods to solve the linear programming problem is the simplex
method. In this method, we repeat a specific condition ‗n‘ number of times until an optimum
solution is achieved.
The steps required to solve linear programming problems using the simplex method are,
Step 1: Formulate the linear programming problems based on the given constraints.
Step 2: Convert all the given inequalities to equations or equalities of the linear programming
problems by adding the slack variable to each inequality where ever required.
Step 3: Construct the initial simplex table. By representing each constraint equation in a row
and writing the objective function at the bottom row. The table so obtained is called the
Simplex table.
Step 4: Identify the greatest negative entry in the bottom row the column of the element with
the highest negative entry is called the pivot column
Step 5: Divide the entries of the right-most column with the entries of the respective pivot
column, excluding the entries of the bottommost row. Now the row containing the least entry is
called the pivot row. The pivot element is obtained by the intersection of the pivot row and the
pivot column.
Step 6: Using matrix operation and with the help of the pivot element make all the entries in
the pivot column to be zero.
Step 7: Check for the non-negative entries in the bottommost row if there are no negative
entries in the bottom row, end the process else start the process again from step 4.
Step 8: The final simplex table so obtained gives the solution to our problem.

Linear Programming Graphical Method

As the name suggests this method uses graphs to solve the given linear programming
problems. This is the best method to solve linear programming problems and requires less
effort than the simplex method.
While using this method we plot all the inequalities that are subjected to constraints in the
given linear programming problems. As soon as all the inequalities of the given LPP are
plotted in the XY graph the common region of all the inequalities gives the optimum solution.
All the corner points of the feasible region are calculated and the value of the objective
function at all those points is calculated then comparing these values we get the optimum
solution of the LPP.
Example: Find the maximal and minimal value of z = 6x + 9y when the constraint
conditions are,
 2x + 3y ≤ 12
 x and y ≥ 0
 x+y≤5

Solution:
Step 1: First convert the inequations into normal equations. Hence the equations will be 2x+3y
= 0, x = 0, y = 0 and x + y = 5.
Step 2: Find the points at which 2x + 3y and x + y = 5 cut the x-axis and y-axis. To find the
point of intersection of the x-axis put y = 0 in the respective equation and find the point.
Similarly for y-axis intersection points put x = 0 in the respective equation.
Step 3: Draw the two lines cutting the x-axis and y-axis. We find that the two axes cut each
other at (3,2).
Step 4: For x ≥ 0 and y ≥ 0, we find that both inequations are followed. Hence the region will
include an area region enclosed by two axes and both lines including the origin. The plotted
region is shown below in the figure.
Step 5: Find Z for each point and maxima and minima.

Coordinates Z = 6x + 9y
(0,5) Z = 45
(0,4) Z = 36
(5,0) Z = 30
(6,0) Z = 36
(3,2) Z = 36
Hence, we find that Z = 6x + 9y is maximum at (0,5) and minimum at (5,0).
What is Sensitivity Analysis?
Sensitivity analysis determines how different values of an independent variable affect a
particular dependent variable under a given set of assumptions. In other words, sensitivity
analyses study how various sources of uncertainty in a mathematical model contribute to the
model's overall uncertainty. This technique is used within specific boundaries that depend on
one or more input variables.

Sensitivity analysis is used in the business world and in the field of economics. It is commonly
used by financial analysts and economists and is also known as a what-if analysis.

KEY TAKEAWAYS
 Sensitivity analysis determines how different values of an independent variable affect a
particular dependent variable under a given set of assumptions.
 This model is also referred to as a what-if or simulation analysis.
 Sensitivity analysis can be used to help make predictions in the share prices of publicly
traded companies or how interest rates affect bond prices.
 Sensitivity analysis allows for forecasting using historical, true data.
 While sensitivity analysis determines how variables impact a single event, scenario analysis
is more useful to determine many different outcomes for more broad situations.

How Sensitivity Analysis Works

Sensitivity analysis is a financial model that determines how target variables are affected based
on changes in other variables known as input variables. It is a way to predict the outcome of a
decision given a certain range of variables. By creating a given set of variables, an analyst can
determine how changes in one variable affect the outcome.

Both the target and input—or independent and dependent—variables are fully analyzed when
sensitivity analysis is conducted. The person doing the analysis looks at how the variables move
as well as how the target is affected by the input variable.

Usefulness of Sensitivity Analysis

Financial models that incorporate sensitivity analysis can provide management a range of
feedback that is useful in many different scenarios. The breadth of the usefulness of sensitivity
analysis includes but is not limited to:

 Understanding influencing factors. This includes what and how different external factors
interact with a specific project or undertaking. This allows management to better understand
what input variables may impact output variables.
 Reducing uncertainty. Complex sensitivity analysis models educate users on different
elements impacting a project; this in turn informs members on the project what to be alert for
or what to plan in advance for.
 Catching errors. The original assumptions for the baseline analysis may have had some
uncaught errors. By performing different analytical iterations, management may catch
mistakes in the original analysis.
 Simplifying the model. Overly complex models may make it hard to analyze the inputs. By
performing sensitivity analysis, users can better understand what factors don't actually matter
and can be removed from the model due to its lack of materiality.
 Communicating results. Upper management may already be defensive or inquisitive about
an undertaking. Compiling analysis on different situations helps inform decision-makers of
other outcomes they may be interested in knowing about.
 Achieving goals. Management may lay long-term strategic plans that must meet specific
benchmarks. By performing sensitivity analysis, a company can better understand how a
 project may change and what conditions must be present for the team to meet its metric
targets.

Sensitivity vs. Scenario Analysis

In finance, a sensitivity analysis is created to understand the impact a range of variables has on
a given outcome. It is important to note that a sensitivity analysis is not the same as a scenario
analysis. As an example, assume an equity analyst wants to do a sensitivity analysis and a
scenario analysis around the impact of earnings per share (EPS) on a company's relative
valuation by using the price-to-earnings (P/E) multiple.

The sensitivity analysis is based on the variables that affect valuation, which a financial model
can depict using the variables' price and EPS. The sensitivity analysis isolates these variables
and then records the range of possible outcomes.

On the other hand, for a scenario analysis, an analyst determines a certain scenario such as
a stock market crash or change in industry regulation. The analyst then changes the variables
within the model to align with that scenario. Put together, the analyst has a comprehensive
picture and now knows the full range of outcomes, given all extremes, and has an understanding
of what the outcomes would be, given a specific set of variables defined by real-life scenarios.

Advantages and Limitations of Sensitivity Analysis

Conducting sensitivity analysis provides a number of benefits for decision-makers. First, it acts
as an in-depth study of all the variables. Because it's more in-depth, the predictions may be far
more reliable. Secondly, It allows decision-makers to identify where they can make
improvements in the future. Finally, it allows for the ability to make sound decisions about
companies, the economy, or their investments.

There are some disadvantages to using a model such as this. The outcomes are all based on
assumptions because the variables are all based on historical data. Very complex models may be
system-intensive, and models with too many variables may distort a user's ability to analyze
influential variables.

Pros

 Provides management different output situations based on risk or changing variables

 May help management target specific inputs to achieve more specific results
 May easily communicate areas to focus on or greatest risks to control
 May identify mistakes in the original benchmark
 Generally reduces the uncertainty and unpredictability of a given undertaking

Cons

 Heavily relies on assumptions that may not become true in the future
 May burden computer systems with complex, intensive models
 May become overly complicated which distorts an analysts ability to
 May not accurately integrate independent variables (as one variable may not accurately
the impact of another variable)