Knowledge that Transforms

In most, if not all, of the standard examples of management science and Operations Research projects, it appears that somewhere along the line some part of the research effort has been devoted to a conceptualization or formalization of the system or process being studied. This conceptualization is very often called a model and may consist of a set of statements or may in fact be a physical representation of the system. It will be supposed that the term model is familiar enough to be used in a reasonably unambiguous way. The purpose of this paper is to define and discuss a problem related to models. It is felt that this is a serious problem, a problem which is becoming increasingly critical for the development of a unified science of management.

The transportation problem has offered two mathematical facets: (1) as a specialized type of linear programming problem, (2) as a method of representation of some combinatorial problems. In this paper a third aspect of the mathematical properties of the transportation problem is developed. It is shown that the same mathematical framework can be extended beyond pair-wise connections, to the determination of optimum linked paths over a series of points. This extension although viewed here as a linear programming problem, takes advantage of the combinatorial aspect of the transportation problem, and applications may arise which, like the assignment problem, appear to be combinatorial problems, but which can be solved by linear programming.

Article II of the Constitution of The Institute of Management Sciences provides that: “The objects of The Institute shall be to identify, extend, and unify scientific knowledge that contributes to the understanding and practice of management.” This evening, I shall offer a few personal views and convictions concerning possible paths through the future to these goals.

Many marketing problems which promise to be amenable to the techniques of operations research have apparently not been subjected to systematic analysis. This article is a first attempt at an analysis of one such area—the number of items stocked by a retailer and its relation to his sales, his costs, and his profits.

I am going to take advantage of my brief time with you here to see if we can establish the basic meaning of a Science of Management; and so to reveal the intent which has guided the steps of many who have chosen to search through the complex of social, political and economic variables which both describe the direction in which understanding may be found and inhibit the progress toward it.

This paper describes the application of the technique of simple “marginal analysis” to an Air Force problem and suggests that a similar application may be useful in many industrial situations. Examples of such industrial problems are also presented.

Most everyone would agree that optimal management decisions can only be made on the basis of sharp information. However, this question has not received much specific attention in the literature, largely because of a preoccupation with formal theories and models. This is quite understandable for, until comparatively recently, there was little theory in the field of management science. This is no longer the case, and the theories have far outrun data-gathering procedures. Indeed, the theories themselves suggest that better data-gathering procedures are needed than have heretofore been available. Improved theories and models simply demand improved sources and types of data. This need is particularly pressing in the field of work measurement. Stable and precise estimates of production rates would be highly useful, for example, in working out an optimal production policy, particularly with respect to planning and scheduling. In addition to this, stable and precise estimates of production rates are highly useful for implementing production policy at local levels. The success of a production policy depends on the quality of local implementation. It is also local implementation that gives flexibility and, hence, realism to the results obtained from broad formal models. Formal results would otherwise degenerate to elegant cookbook recipes which might easily turn out to be worse than intuitive solutions arrived at from “off-the-shelf” models.

This year the development of electronic digital computing machines has reached the point where American industry can turn from analysis and speculation to actual trial of the effectiveness of these machines in general business use. In these circumstances, it would seem necessary to look to the fundamentals right now, early in the growth period. Most of the business applications of computers, which we have planned up to this point, have led in one direction: greater speed in the processing of business data. To achieve this goal, we have aimed at more and more complete office automation. Probably that was a sound choice of a starting point. For business became interested in the electronic computer when it was pointed out that the machine can effect great savings in the processing of data. Hence this application of the computer was the foot in the door needed to get this industry off to a sound start financially.

In a recent article (Salveson, M. E. 1955. The assembly line balancing problem. J. Indust. Engrg. VI 18–25.), M. E. Salveson discussed various problems concerned with setting up production lines. The purpose of the present paper is to set forth an enumeration procedure for solving one of these problems. The reader is referred to Salveson's article for a detailed discussion of the circumstances in which this problem arises and for remarks concerning the assumptions in its statement.

An important managerial problem in manufacturing is to determine the optimum amount and mix of products that can be manufactured from a given set of facilities, such as machine tools. A method is discussed here for handling this problem more effectively in one instance than has been possible. It makes use of several standard procedures, but adapts them and adds others to accommodate the unavoidable difficulty of discreteness in real production situations. The traditional formulation of the economic lot size is shown to be incorrect in the general case and a preferred formulation is presented for the situation studied here. A method of optimizing the amount of in-process inventory to carry in the circumstances of this study also is given.