Comparison of conventional and dynamic FMS periodic loading rules

Abstract

Reports on a comparison by computer simulation between conventional periodic loading (PL) and job‐group loading (JGL). When conventional PL is used, the search of the best part input sequence must be performed in order to optimize performance of a flexible manufacturing system (FMS). JGL works, instead, as a dynamic rule for real time scheduling of FMS, defining a part releasing policy able to guarantee the reaching of a periodic steady state without non‐productive times on the bottleneck workstation. However, JGL does not assure, in some cases, the same performance arising from the optimal part input sequence of conventional PL, in terms of non‐productive times in FMS filling and emptying phases, work in progress and throughput time. The paper demonstrates that using any JGL rule or the best PL part input sequence gives rise to negligible differences in FMS performance. Furthermore, the dynamic capabilities of the JGL also allow for spontaneously restoring the FMS periodic steady state without non‐productive times after any transient, for instance when production mix changes occur.