Are you an EPFL student looking for a semester project?
Work with us on data science and visualisation projects, and deploy your project as an app on top of Graph Search.
We consider a three-stage dynamic flexible flow shop scheduling problem in which jobs of multiple types arrive dynamically over time, a quality feedback mechanism is present, and the setup timing and the process defect rate are closely related. At each machine in the second stage, a sequence-independent setup operation is necessary to changeover job types. Once a setup is done for a job type at a machine, the defect rate for the job type at the machine is reset to a low and stable phase which will be maintained for a predetermined time periods. However, after the phase, it turns to a relatively high and unstable phase. At the final inspection stage, jobs are inspected and the quality feedback will be given to the previous stage when the accumulative defect rate of each job type exceeds a certain tolerance level. To cope with the dynamic nature of the flexible flow shop scheduling problem, we propose two dispatching rule-based scheduling algorithms which consider the quality feedback as well as the real time shop information for the objectives of maximizing the quality rate and the mean tardiness of the finished jobs. The results of a series of simulation experiments will be given to evaluate the performance of the suggested algorithms. Since there have been few research on the shop floor scheduling problems with quality feedback, we expect that this research will make a contribution to the development of a practical real time scheduling methodology in multi-stage production systems with the consideration of the imperfect process quality.
Andreas Peter Burg, Robert Giterman, Reza Ghanaatian Jahromi, Andrea Bonetti