New Techniques For Programmed Decision Making

Decision making

Techniques Decision Making

World War II brought large numbersof scientist traine in the use of mathematical tools into contact for the first time with  operational and managerial problems. Designers of military aircraft could not plan aircraft armament without making assumptions about the formation in which the planes would be flown and the strategy of their commitment to action. Mathematical economists responsible for material allocation had to come to grips with complex logistics system.

The need for solving these problems, coupled with the tools of quantitative analysis that the scientists and econometricians brought with them, have produced some new approaches to management decision making that are of fundamental importance.

Operation Research

Many people - notably some of  the pioneer operations researchers them-selves-have tried to define operations research. The net result is usually to identify it with scientific method or straight thinking applied to management problems, and to imply that it is some thing that can be done only by natural scientists. definitions of this kind unintentionally imperealistic , raise the hackies of those identified with the earlier phrase " scientific management ", who had thought that clear, scientific thinking is what they had always been doing.

Except in matters of degree ( example, The operations researchers tend to use rather high-powered mathematics), it is not clear that operations research embodies any philosophy different from that of scientific management. sharles babbage and frederick taylor will have to be made, retroactively, charter members of the operations research societies.

New Techniques

A more understandable and defensible definition of operations that, emerging out of the military needs of World War II, has brought the decision-making problems of management within the range of interest of large numbers of natural scientists and, particularly, of mathematicians and statisticians. The operation researchers soon joined forces with mathematical economists who had come into  the same area-to the mutual benefit of both groups. And by now there has been widespread fraternization between these exponents of the "new" scientific management and men trained in the earlier traditions of scientific management and industrial engneering. No meaningful line can be drawn any more to demarcate operations research from scientific management or scientific management from management science.

Along with some mathematical tools, which I shall discuss presently, operations research brough into management decision makinga point of view called the system approach. The system approach is no easier to define than operations research for it is a set of attitudes and a frame of mind rather than a definite and explicit theory. At its vaguest, it means looking at the whole problem-again, hardly a novel idea, not always a very helpful one. Somewhat more concretely, it means designing the components of a system and making individual decisions within it in the light of the implication of these decisions for the system as a whole. We now a little about how this might be done :

1. Economic analysis has something to say about rational behaviour in complex systems of interacting elements, and particularly  about the conditions under which the choices that are optimal for sub systems will or will not be optimal for a system as a whole. Economic analysis also has a great deal to say about the price system as a possible mechanism for decentralizing decision making.
   
   
2. Mathematical techniques have been developed and adapted by engineers and economists for analysing the dynamic behaviour of complex systems. under the labels of servomechanism theory and cybernetics, such techniques underwent rapid development at about the time of world war II. They have considerable usefulness in the design of dynamic systems. System design is such a modish, if not faddish, word at the moment that i don't want toexaggerate the amount of well-understood technique that stands behind it. Nevertheless, it is fair to say that we can approacht the design and analysis of large dynamic systems today with a good deal more sophistication than we could ten years ago.

The Mathematical Tool

Operations research progressed from the talking to the action stage by finding tools with which to solve concrete managerial problems. Among the tools, some of them relatively new, some of them already know to statisticians, mathematicians, game theory, and probability theory. Behind each of these formidable terms lies a mathematical  model for a range  of management problems. Linear programming, for example, can be used to provide a mathematical model for the operations of a gasoline refinery, or a commercial cattle-feed manufacturing operation. Dynamic programming can be used as a model for many inventory and production planning situations. Game theory models have been used in a wide variety of contexts-they have been, perhaps, the most versatile of all.

Whatever the spesific mathematical tool, the general recipe for using it in management decision making is something like this.

1. Construct a mathematical model that astisfies the conditions of the tool to be used and which, at the same time, mirrors the important factors in the management situasion tot be analysed.
2. Define he criterion function the measure that is to be used for comparing the relative merits of various possible coursess of action.
3. Obtain empirical estimates of the numerical parameters in the model that specify the particular, concrete situation to which it is to be applied.
4. Carry through the mathematical process of finding the course of action which, for the specified parameter values, maximizes the criterion function.

In any decision-making situation where we apply this recipe successfully, we have, in fact, constructed a program for the organization's decision. We have either annexed some decision that had been judgmental to the area of programmed decision making, or we have  replaced a rule-of-thumb program with a more sophisticated program that guarantees us optimal decision - optimal, that is, within the framework of the mathematical model.

Enter the Computer

It was an historical accident with large consequences that the same war which spawned operations research saw also the the birth of the modern digital computer as a practical device. The computer was conceived as a device for exploring by numerical analysis the properties of mathematical systems too large or too complex to be terated by known analytic mehtods. The system of differential equations that were arising in aerodynamics, meteorology and the design of nuclear reactors were obvious candidates for this treatment.

Of course, the bread-and-butter applications of computers to business decision making have had little to do with either mathematical models or simulation. They have had to do with automating a wohle host routine and repetitive data-processing activities that had for many years been highly programmed butnot nearly so completely automated. Through this development, large-scale data processing is becoming a factory operation, an operation that exceeds in degree of automation all but a very few manufacturing processes.

The revolution in Programmed Decision Making has by no means reached its limits, but we can now see its shape. The rapidity of change stems partly from the fact that there been not a single innovation but several related innovations, all of which contribute ta it.

1. The electronic computer is bringing about, with unexpected speed, a high level of automation in the routine, programmed decision making and data processing that was formerly the province of clerks.
2. The area of programmed decision making is being rapidly extended as we find ways to apply the tools of operations research to types of decisions that have up to  now been regarded as judgmental-particularly, but not exclusively, middle-management decisions in the area of manufacturing and warehousing.
3. The computer has extended the capability of the mathematical techniques to problems far too large to be handled by less automatic computing devices, and has further extended the range of programmable decision by contributing the new technique of simulation.
4. Companies are just beginning to discover ways of bringing together the first two of these developmentsz; of combining the mathematical techniques for making decision about aggregative middle management variables with the data-processing techniques for implementing these decision in detail at clerical levels.

Out of the combination of these four developments there is emerging the new picture of the data-processing factory for manufacturing, in a highly mechanized way, the organization's programmed decisions-just as the physical processing factory manufactures its products in a manner that becomes increasingly mechanized. The automated factory of the future will operate on the basis of programmed decisions produced in the automated office beside it.

So an article about the " New Tecniques For Programmed Decision Making" 

may be useful.

*sources and references; A general book on the history of the development of computers and on their use by management is John A. Postley, Computers and people (New York ; Mc Graw-Hill book company)

activity analysis of production and allocation ( John Wiley)

Share this post