Impact of throughput-based objectives and machine grouping decisions on the short-term performance of flexible manufacturing systems

The most commonly-used objective in the literature for solving machine loading and grouping problems in flexible manufacturing systems is the maximization of steady-state throughput. In reality, orders arrive dynamically and the mix of products changes frequently, raising questions about the applicability of this objective in achieving shorter-term measures of the output rate. Moreover, the loading and grouping problems only determine which operations are to be performed on each machine, and whether any of the machines should be identically tooled so as to allow multiple routeings for some jobs. The actual short-term performance of the system also depends on scheduling and dispatching decisions. We study the impact of the various objectives ostensibly related to steady-state throughput and machine grouping decisions on the short-term makespan performance (in a static setting). Computational results suggest that minimizing the maximum percentage overload (relative to the optimal continuous workload allocation) across machine groups is an excellent objective. The results also indiate that reducing the number of machine groups and balancing workloads among the machines help to reduce makespan.