Flexible assembly scheduling using a knowledge-based approach

Abstract

Abstract This paper presents a knowledge-based scheduling system that generates a detailed production schedule for jobs in a flexible assembly system (FAS) by exploiting routing and operation flexibilities. Past research in flexible manufacturing did not take into account the flexibilities of such environments that allow alternative sequencing and routing of operations presumably to avoid the computational overhead that would result from these added dimensions of complexity. Moreover, jobs in an FAS require both machining and assembly operations. In our system, the resulting computational burden is reduced by using (1) a two-level hierarchical scheduling approach, (2) a search strategy that uses knowledge from the knowledge base to control the amount of search required, and (3) a computationally efficient post analysis of the schedule that often results in local improvements and is based on knowledge about solution structures stored in the knowledge base. The system has been implemented and tested using Common Lisp on a Macintosh SE/30 with 4 MB main memory. Computational experience with the system is reported.