Operations Research (OR) is a multifaceted discipline focused on applying advanced analytical methods to enhance decision-making. Known by various names globally, such as "management science" (MS), "industrial engineering" (IE), and "decision science" (DS), the field often sees its terminology unified under "OR" for consistency.
Challenges in Definition: As a relatively new and evolving field, providing a definitive description of OR is complex. Its scope isn't fixed, and it continually integrates tools from diverse disciplines like mathematics, statistics, economics, psychology, and engineering. The primary objective of OR is to utilize these tools to improve decision-making, particularly in scenarios involving limited resources and incomplete information.
-
Decision-Making Support: OR employs mathematical and quantitative techniques to substantiate decisions, particularly in complex and high-stakes situations, such as urban transportation planning or product mix optimization in industries.
-
Interdisciplinary Tools: By combining tools from various disciplines, OR creates a unique knowledge set tailored for effective decision-making and resource allocation.
-
Scientific Methodology: OR is characterized by its use of scientific methods to model, analyze, and predict the behavior of systems involving humans and machines. This involves:
- Determining system behaviors.
- Analyzing these behaviors through appropriate models.
- Predicting future behaviors using these models.
Whole-Operation Analysis: Unlike other research and engineering fields, OR focuses on analyzing operations as a whole. This holistic approach was notably effective in military operations during World War II and has since been adapted for business and industrial applications, including inventory management, warehouse location optimization, and advertising strategies.
-
Operational Research Society of Great Britain (1962): OR is the application of modern science to complex problems in managing large systems of people, machines, materials, and money across various sectors. It involves creating scientific models to predict and compare outcomes of different decisions, aiding scientific policy and action determination.
-
Randy Robinson: OR improves operational, decision-making, and management effectiveness through scientific methods. This involves data analysis, mathematical modeling, and innovative approaches, often resulting in the development of related software, systems, and products.
-
Morse and Kimball: OR is a quantitative approach that provides executives with a scientific basis for decision-making regarding operations under their control.
-
Saaty: OR is the art of finding less bad answers to problems that otherwise would have worse answers, emphasizing the practical improvement of outcomes.
-
Miller and Starr: OR is applied decision theory that uses scientific, mathematical, or logical means to address executive decision problems, striving for rational solutions.
-
Pocock: OR is an applied science using analytical, mathematical, and quantitative methodologies to assess the implications of alternative management actions, providing a better basis for decision-making.
Operations Research is a dynamic, interdisciplinary field that applies scientific methods to enhance decision-making and resource allocation. By integrating diverse tools and focusing on whole-operation analysis, OR offers robust solutions to complex problems across various sectors, continually evolving to meet the challenges of modern management and decision-making.
Operations Research (OR) is a relatively new discipline that began systematically in the late 1930s in the UK. It wasn't a university subject 70 years ago, but today, it plays a crucial role in decision-making across various fields.
- 1936: The British Air Ministry established Bawdsey Research Station for radar experiments. This led to early OR work, integrating radar data with ground observations for fighter interception.
- 1937: The first major air-defense exercise used radar data, revealing the need for better tracking information.
- 1938: A second exercise highlighted the problem of coordinating information from multiple radar stations. A.P. Rowe from Bawdsey proposed a research program into the operational aspects of radar systems, coining the term "operational research."
- 1939: The last pre-war air defense exercise showed improvements due to OR contributions. The Stanmore Research Section was formed, later becoming the Operational Research Section (ORS).
- 1940: The Stanmore Research Section analyzed aircraft losses for Winston Churchill, influencing the decision not to send additional fighter squadrons to France, which contributed to the success of the Battle of Britain.
- 1941 onward: ORS was established in Coastal Command, conducting significant OR work, especially in anti-submarine warfare.
- 1951: The National Research Council of the USA formed a committee on OR, and Morse and Kimball published the first book on OR methods.
- 1952: The Operations Research Society of America was established.
- Industrial Applications: Success in the military led to OR being adopted by industrial managers for complex business problems. Today, OR is used in various government departments and industries worldwide.
- 1949: The first OR unit was established at the Regional Research Laboratory in Hyderabad.
- 1953: An OR unit was set up at the Indian Statistical Institute in Calcutta for national planning.
- 1955: The Operations Research Society of India was formed, becoming an early member of the International Federation of Operations Research Societies.
Operations Research began in the UK in the late 1930s, significantly contributing to military operations during World War II. Post-war, it expanded into industrial and governmental applications globally. India was an early adopter of OR, establishing units and societies to apply OR methods in various fields. Today, OR is a key subject in management and mathematics education.
The development process of Operations Research (OR) consists of six key steps, also known as phases. Here’s a simplified breakdown of each step:
-
Observe the Problem Environment
- Activities: Conduct conferences, site visits, research, and observations.
- Purpose: Gather sufficient information to understand and formulate the problem.
-
Analyze and Define the Problem
- Activities: Analyze the problem and clearly define it, including the objectives, uses, and limitations of the OR study.
- Purpose: Develop a clear understanding of the need for a solution and the nature of the problem.
-
Develop a Model
- Activities: Create a mathematical model representing the problem using equations and relationships. Test and modify the model as needed.
- Purpose: Develop a functional representation of the problem that can be used to find solutions.
-
Select Appropriate Data Input
- Activities: Collect and analyze internal and external data, gather opinions, and use computer data banks.
- Purpose: Ensure the model has the appropriate data inputs to operate and be tested effectively.
-
Provide a Solution and Test Its Reasonableness
- Activities: Use the model and input data to generate a solution. Test the solution to check for limitations and update the model if necessary.
- Purpose: Find a solution that aligns with the organization's objectives and test its feasibility.
-
Implement the Solution
- Activities: Address any behavioral issues and implement the solution.
- Purpose: Ensure the solution is properly implemented, leading to improved work quality and management support.
The six steps of OR development are: observing the problem environment, analyzing and defining the problem, developing a model, selecting appropriate data input, providing a solution and testing its reasonableness, and implementing the solution. Each step builds on the previous one to ensure a comprehensive and effective approach to solving complex problems.
| Process Activities | Process | Process Output |
|---|---|---|
| Site visits, Conferences, Observations, Research. | Step 1: Observe the problem environment. | Sufficient information and support to proceed |
| Define: Use, Objectives, Limitations | Step 2: Analyze and define the problem | Clear grasp of need for and nature of solution requested |
| Define interrelationships, Formulate equations, Use known O.R. Model, Search alternate Model | Step 3: Develop a Model | Models that works under stated environmental constraints |
| Analyze: internal-external data, facts Collect options, Use computer data banks | Step 4: Select appropriate data input | Sufficient inputs to operate and test model |
| Test the model find limitations update the model | Step 5: Provide a solution and test its reasonableness | Solution(s) that support current organizational goals |
| Resolve behavioral issues, Sell the idea, Give explanations, Management involvement | Step 6: Implement the solution | Improved working and Management support for longer run operation of model |
The key responsibility of manager is decision making. The role of the O.R. specialist is to help the manager make better decisions. Figure explains the relationship between the O.R. specialist and the manager/decision maker.
| STEPS IN PROBLEM RECOGNITION | INVOLVEMENT: O.R. SPECIALIST, MANAGER |
|---|---|
| Recognize from organizational symptoms thata problem exists. | Manager |
| Decide what variables are involved; statethe problem in quantitative relationships among the variables. | Manager and O.R. Specialist |
| Investigate methods for solving the problems as stated above; determine appropriate quantitative tools to be used | O.R. Specialist |
| Attempt solutions to the problems; find various solutions; state assumptions underlying these solutions; test alternative solutions. | O.R. Specialist |
| Determine which solution is most effective because of practical constraints within the organization; decide what the solution means for the organization. | Manager and O.R Specialist |
| Choose the solution to be used | Manager |
| Sell the decision to operating managers: Get their understanding and cooperation. | Manager and O.R Specialist |
Operations Research (OR) employs a variety of tools and techniques to solve complex problems. Here are some commonly used methods:
-
Linear Programming:
- Purpose: Optimize a criterion (e.g., profit, loss) within certain constraints.
- Description: The objective function and constraints are linear.
-
Game Theory:
- Purpose: Make decisions in conflicting situations involving multiple players.
- Description: Players' success is mutually exclusive, creating conflict.
-
Decision Theory:
- Purpose: Make decisions under certainty or probabilistic conditions about future outcomes.
-
Queuing Theory:
- Purpose: Minimize the cost of waiting in queues (e.g., customers, aircraft, jobs).
- Description: Balances waiting costs with service costs.
-
Inventory Models:
- Purpose: Minimize total inventory costs.
- Description: Addresses costs of purchasing, carrying, and stockouts.
-
Simulation:
- Purpose: Study problems by modeling processes and conducting trial-and-error experiments.
- Description: Used when actual experimentation isn't feasible.
-
Non-linear Programming:
- Purpose: Optimize when the objective function and constraints are non-linear.
- Description: Linear approximations can be used within a limited range.
-
Dynamic Programming:
- Purpose: Analyze multi-stage decision processes.
- Description: Each decision depends on previous decisions and external factors.
-
Integer Programming:
- Purpose: Optimize problems where variables must take integer values.
- Examples: Number of motors in an organization, passengers in an aircraft.
-
Markov Process:
- Purpose: Predict changes over time based on current state.
- Description: Transitions between states depend on the current state only.
-
Network Scheduling:
- Purpose: Plan, schedule, and monitor large projects.
- Techniques:
- PERT (Program Evaluation and Review Technique): Used when activity times are uncertain.
- CPM (Critical Path Method): Used when activity times are known accurately.
- Description: Identifies critical activities and paths using diagrams.
-
Information Theory:
- Purpose: Evaluate the effectiveness of information flow in systems.
- Description: Originated from electrical communication, used in communication networks and business organizations.
OR uses a diverse set of tools and techniques like linear programming, game theory, decision theory, queuing theory, inventory models, and simulation. Each method has a specific purpose and application, from optimizing resources and decision-making to scheduling projects and analyzing information flow.
Operations Research (OR) is widely used across various fields in business and government. Here are some typical applications:
- Effective assignment of audit teams
- Analysis of credit policies
- Cash flow planning
- Developing standard costs and costs for byproducts
- Planning delinquent account strategies
- Project scheduling, monitoring, and control
- Determining and deploying the proper workforce
- Resource allocation for projects
- Decisions on factory location and size
- Estimating the number of required facilities
- Hospital planning
- Designing international logistics systems
- Loading and unloading transportation
- Deciding warehouse locations
- Building cash management models
- Allocating capital among alternatives
- Financial planning models
- Investment and portfolio analysis
- Dividend policy making
- Inventory control
- Marketing balance projection
- Production scheduling and smoothing
- Allocating advertising budgets
- Timing product introductions
- Selecting product mix
- Choosing the most effective packaging
- Personnel planning and recruitment
- Skill balancing
- Scheduling training programs
- Designing effective organizational structures
- Optimal buying and reordering
- Materials transfer
- Controlling R&D projects
- Allocating R&D budgets
- Planning product introductions
Operations Research (OR) is valuable but has certain limitations, mostly related to model building, time, and cost issues. Key limitations include:
-
Distance Between OR Specialists and Managers
- OR specialists are often mathematicians or statisticians who might not fully understand business problems. Conversely, managers may find OR complex and hard to grasp.
-
Magnitude of Calculations
- OR aims to find optimal solutions by considering many factors. The number of calculations required can be enormous and typically requires computer processing.
-
Money and Time Costs
- Data used in OR models frequently changes, making it expensive to keep models updated. A good solution now might be better than a perfect solution later due to time constraints.
-
Non-Quantifiable Factors
- OR is effective only when all factors are quantifiable. It doesn't incorporate non-quantifiable factors like emotions or qualitative aspects.
-
Implementation Challenges
- Implementing decisions involves dealing with human relations and behavior, which can be complex. Psychological factors also play a crucial role in successful implementation.