Evaluation of Dispatching Rules Performance for a DJSSP Towards their Application in Industry 4.0

Dispatching rules (DRs) are very attractive heuristics for solving complex dynamic job-shop scheduling problems. DRs advantages can be summarized in their ability to make real-time scheduling decision and their ease of implementation. Many research works have been performed to design dispatching rules and to evaluate their performance regarding the main scheduling objectives (i.e. mean flow time, mean lateness, mean tardiness and so on). Despite their drawbacks in terms of system-state myopia, DRs become attractive again in the current context of Industry 4.0 which is characterized by highly dynamic production systems. The goal of this paper is to analyze commonly used dispatching rules in the literature to point out their strengths and weaknesses regarding their ability to fit with the context of Industry 4.0. We use discrete-event simulation for large scale dynamic job-shop scheduling problem to evaluate the performance of these rules regarding to the mean flow time objective.