Towards a knowledge-based approach to space planning in industry

Abstract

Space planning involves typically, the organization of the space under a set of constraints to meet a given set of specifications. There are quite a few mathematical models describing the space planning problem. These abstract models aim at optimizing a certain predefined criterion function. However, these approaches need clear definition of the problem domain and precise measurement of the various parameters involved. In practice, none of these are feasible. Moreover, even when the parameters are identified, there remains the issue of search in a large problem space. In this paper, some of the well known approaches to the space planning problems are reviewed. The expert on the space planning is usually guided by the heuristics which are acquired through experience. This effort is towards building a knowledge-based system to capture such heuristics. The study describes how a knowledge management approach could be considered for solving space planning problems. In particular we look at this problem in the light of organizing in a shop floor in a typical wire rope industry. A lisp-based system is developed as a prototype. The paper includes a discussion of knowledge-based approach and optimizing techniques and concludes with a recommendation for judicious combination of the two.