Configuration Design of Complex Integrated Manufacturing Systems

The configuration design of complex integrated manufacturing systems such as semiconductor wafer fabricaton plants is a multi-objective, multi-criterion design problem. This paper describes a qualitative reasoning methodology for the automated configuration design of manufacturing systems. The amounts, incremental qualitatives, and rates of change that make up the components of a qualitative model are first derived. They provide a formalised body of knowledge concerning the relationship between throughput, flow-time and use. A computation procedure and a system architecture that use an embedded queueing network program are then described. The methodology enhances the coherence of rule-based design knowledge, and can automatically generate a configuration design that satisfies specified performance requirements.