Intelligent decision support for flexible manufacturing: Design and implementation of a knowledge-based simulator

There is considerable research in decision making for the flexible manufacturing systems (FMS) domain. Much of it tends to be fragmentary due to differences in assumptions, constraints, modeling techniques and solution strategies. This paper suggests a common basis for support of FMS decision making as an attempt to alleviate these problems. It describes the architecture of an intelligent knowledge-based simulator KBSim, that provides systematic FMS research capability. KBSim is applied to an industrial FMS scheduling problem to reduce both mean flow time and tardiness when compared to several common scheduling heuristics. It is also used in a research-oriented modified job shop scheduling application. In both cases, it outperformed traditional decision making heuristics. Its efficiency, ease of use, and portability suggest that KBSim will prove useful in the automation of adaptive system control, facilitating periodic review of FMS decisions, and giving management a competitive edge.