Improving superfluous load avoidance release (SLAR): A new load-based SLAR mechanism

Abstract

A workload limit forms an essential part of most order release methods designed for high-variety make-to-order contexts, but this mechanism does not necessarily lead to the lowest possible direct load buffer. To counter this, the Superfluous Load Avoidance Release (SLAR) procedure was developed that avoided the use of a workload limit altogether. SLAR significantly improves performance compared to alternative release methods in the literature but has been criticized for being impractical as it can lead to uncontrolled loads at stations downstream in the routing of an order. This criticism can be overcome by introducing a workload limit. Using simulation, this study shows that introducing a limit to SLAR further reduces the superfluous direct load, specifically during high load periods. This not only controls the load at downstream stations but also yields a further reduction in the percentage of tardy jobs whilst maintaining SLAR's good mean tardiness and shop floor throughput time performance. Meanwhile, introducing an additional load-based trigger further improves mean tardiness performance. These results partly question one of SLAR's original design principles and extend the theory upon which SLAR is built, by linking each of SLAR's two release triggers to periods of low and of high load. Finally, by gaining control over the workload in periods of high load, the practical applicability of SLAR is substantially enhanced.