Part selection in a dynamic job-shop type flexible manufacturing system (FMS)
The research investigates the part selection problem which is defined as dynamic determination of one part among all available parts in a dynamic job-shop type FMS. A heuristic part selection algorithm, namely the state dependent algorithm, is developed by applying dynamic information in part selection to achieve the objective of increasing productivity. The earliest-due-date part selection algorithm with the objective of meeting due dates is also investigated for the part selection problem. A hypothetical FMS is formulated for the investigation. Computer simulation is used for evaluation. The first-in-first-out part selection algorithm is utilized as a comparative algorithm. Also evaluated is the state dependent algorithm under various scheduhng conditions formed by a set of robot scheduling rules and a set of machine scheduling rules. The analysis of results is conducted by computing absolute and relative improvements and by performing statistical tests for significant difference in absolute improvement for criteria based on job completion times, in-process inventory, and utilization, and criteria based on job due dates. Investigative results verify the effectiveness of the state dependent algorithm and the earliest-due-date algorithm for increasing productivity and for meeting due dates in part selection, respectively, and the effectiveness of the earliest due date scheduling rule for increasing productivity in scheduling. The research is also carried out to investigate the integration of part selection and scheduling in the dynamic job-shop type FMS. A heuristic robot scheduling rule is developed for the integration, which uses dynamic information in robot scheduling. Results reveal the effectiveness of the integration by utilizing the state dependent algorithm, the heuristic robot scheduling rule, and the earliest due date machine scheduling rule in comparison to other integration cases investigated and the comparative non-integration case. A good combination of part selection and scheduling is also identified which gives good FMS performance regarding both increasing productivity and meeting due dates in comparison to the comparative case. Recommendations for practical applications are made based on the investigative results. Future research ideas are also discussed.