The term "operations research" originated during World War II and precisely described the field at that time. Although operations research methods are now applied far more broadly than during the war, the name endures. For clarity, operations research can be defined as a scientific method for solving organizational management problems.
When solving any specific problem, the application of operations research methods involves:
- constructing mathematical, economic, or statistical models for decision-making and management problems in complex situations or under conditions of uncertainty;
- studying the interrelationships that determine the possible consequences of the decisions being made, as well as establishing effectiveness criteria that allow one to assess the relative advantage of one or another course of action.
It is sometimes supposed that the subject of operations research is the day-to-day management problems arising in the activities of one organization or another. In other words, it is considered to deal with management problems that arise in the course of performing certain daily, repetitive "operations." Operations research methods are indeed applied in solving certain problems of this type. These include, in particular, problems associated with production scheduling and inventory management, equipment operation and maintenance, and staffing.
However, operations research methods are often used to solve a different kind of management problem, one that has only an indirect relation to day-to-day operations. Problems of this kind are typically associated with planning. These include, in particular, problems of determining the product mix, developing long-term production expansion programs, designing a network of warehouse facilities in a wholesale distribution system, as well as problems of entering new areas of production or commercial activity through mergers with or acquisitions of other firms.
Over the last decade, the application of operations research methods has repeatedly demonstrated their great capabilities and high effectiveness in solving practical management problems.
The subject under consideration is best characterized as "analysis of managerial decisions." It is precisely the problem of decision-making (or choosing courses of action) that is central to all operations research. The analysis of managerial decisions involves decomposing a complex problem into sub-problems that are more amenable to logical and intuitive examination. The results of a thorough study of each sub-problem are appropriately synthesized, enabling a deeper understanding of the original problem as a whole. What, then, accounts for the emergence of complex management problems?
One reason is that in the modern economy, production-technical, market-commercial, and other factors are in complex mutual dependence. For example, the production plan of a given enterprise must take into account customer demand, raw material requirements, necessary working capital, equipment capacity, the probability of technical malfunctions, as well as production and technological constraints. Drawing up a plan that would be both realistic and economically advantageous is far from easy.
Another reason for the complexity of management problems arising in practical activity is that various divisions of the same organization may (perhaps not always fully consciously) pursue contradictory goals, responsibility for decisions and administrative authority are often widely dispersed across various structural units, and external economic factors on which the activity of the organization in question depends may contain elements of uncertainty.
The operations approach is directed toward improving the very procedures for developing managerial decisions. The degree of success of this approach is measured by the net profit obtained through the practical implementation of the results of the operations research study.
One should see the distinction between improving the procedures for making managerial decisions and improving the process of their execution — or, in other words, between a "good decision" and a "good practical result." For example, for certain reasons, the decision to place a bet at the races may be considered (from an economic or moral point of view) far from the best; however, if the bet is nonetheless placed, the result may turn out (in the case of a win) to be positive. Analysis aimed at developing the most well-founded decision is important because, in situations where the final result is not uniquely determined, the only way to influence the course of events is through the decision being made.
There are many methods for solving management problems, and most are similar. It may seem difficult to distinguish the operations approach from the methods used by production engineers, planning economists, accountants, or specialists in evaluating the economic effectiveness of management information systems. However, operations research methods possess a number of specific features.
For a given approach to solving a specific problem to be qualified as an operations approach, it must contain, in particular, the following elements:
- Decision-making orientation. The principal results of the analysis must have a direct and fully defined bearing on the choice of a course of action (i.e., strategy or tactics).
- Evaluation based on economic effectiveness criteria. Comparison of various possible courses of action must be based on quantitative assessments that unambiguously determine the utility of the expected outcome for the organization in question. Quantitative assessments for commercial firms usually involve using measurable quantities such as expenditures, revenues, cash on hand, and the rate of return on additional capital investment, among others. Market demand fluctuations must be appropriately quantified.
- The recommended solution must achieve an optimal "balance" taking into account all these frequently contradictory factors.
- Reliance on mathematical models. The procedures for handling the above-mentioned parameters must be defined with such precision that any specialist in systems analysis could interpret them completely unambiguously. In other words, working from the same data, different analyst-specialists must obtain identical results.
- The necessity of using a computer. This condition is by no means merely desirable. It should rather be considered necessary, which is dictated either by the complexity of the mathematical models used and the large volumes of data to be processed, or by the cumbersome computational procedures supporting one management and control system or another.