PERSPECTIVES ON OPERATIONS-RESEARCH IN LONG-TERM PLANNING

Using two production allocation projects in the Philips Company as a vehicle for discussion, we are dealing with a number of fundamental questions in Operations Research. First, one has the choice between a continuum of decision alternatives and a finite, usually small, number of alternatives. The choice does not merely pose a technical problem. We explain our preference for a discrete instead of a continuous model, and we show the far-reaching consequences in the decision phase. Second, whereas Operations Research has been strong in the objective evaluation of the alternatives, it has generally ignored the subjective weighing. We sketch the potential of Operations Research in this area as well as the benefits of recent branches such as multi-criteria analysis and scenario analysis. This paper has been dedicated to Jacques Benders, at the occasion of his retirement from the Eindhoven University of Technology (June 1989). As a consultant of Philips Industries in Eindhoven, Benders cooperated during a long period of time with the present author on linear-programming applications in production allocation, trim-loss problems, traffic control, and on penalty-function methods for nonlinear programming (1963-1974).