A computational study of a shifting bottleneck heuristic for multi-product complex job shops

Abstract

In this article, we present the results of a performance evaluation of a shifting bottleneck heuristic (SBH) applied to complex multi-product job shops. This type of job shop can be found in semiconductor wafer fabrication facilities (wafer fabs) that produce application-specific integrated circuits for a wide array of customers. The SBH decomposes the overall scheduling problem into scheduling problems for single tool groups. The solutions of these scheduling problems are connected via a disjunctive graph. We perform simulation experiments with (reference) models of wafer fabs to assess the performance of a rolling horizon approach in a dynamic environment. We compare the results of the SBH for multi-product situations with results for local and global dispatching rules. The SBH outperforms these dispatching rules with respect to total weighted tardiness. The application of the SBH provides the most benefit in the case of manufacturing systems that have to cope with many different products.