A rules-based scheduling system for flow type assembly

The scheduling system described in this paper is to be used for the short-term planning and control of assembly systems at the production activity control (PAC) level. PAC operates as close to real-time as possible, and therefore the scheduling system has to be capable of analysing the manufacturing system and generating a schedule in a short time frame. Because of this constraint, traditional scheduling techniques such as branch and bound, dynamic programming, etc., are inappropriate. The approach taken in this scheduling system is to combine heuristic rules with newly developed artificial intelligence techniques while incorporating insights into scheduling gained from a study of the just-in-time, optimized production technology and manufacturing resource planning approaches to production management. The resulting scheduling system is presented, and its implementation on a MicroVAX workstation using data from an electronics assembly facility is described.