The design, development, and validation of a heuristic preprocessing approach to constraint-based reactive scheduling

This research evaluates domain specific heuristics for preprocessing inputs to constraint-based, reactive shop-floor scheduling systems. The heuristics are empirically tested for their effects on schedule stability.; General purpose preprocessing heuristics have been developed in the artificial intelligence literature to increase the efficiency of backtrack search in constraint satisfaction problems. The domain specific heuristics evaluated in this study are adaptations of priority sequencing rules found in the operations management literature. Findings indicate that the adapted priority sequencing rules are insignificant in the context of constraint-based, reactive, shop-floor scheduling systems. This result implies that further research is needed to identify new rules that are effective in maximizing schedule stability in these systems.; A testbed software system encompassing a data generation module, a heuristic preprocessor module, and a constraint-based rescheduler module is used in performing sampling and treatment functions. This testbed system can be used in future research to investigate the impact of different heuristics.