Scheduling jobs with uncertain setup times and sequence dependency

In many real-world job-shop situations, jobs are sequence dependent and the setup times for those jobs vary stochastically because of such random factors as crew skills, temporary shortage of equipment, tools and setup crews, and unexpected breakdowns of fixtures and tools during a setup operation. Assuming these random setup times to be fixed at, say, their expected values, which is often done in the extant literature, may lead to development of inefficient sequencing rules. The purpose of the present paper is to investigate the impact of setup-time variation on sequencing decisions, with normally-distributed setup times. Results show that setup-time variation has a negative impact on shop performance, but does not diminish the advantages of setup-conscious sequencing rules over conventional sequencing rules in dealing with setup times. A simulation model of a nine-machine job-shop is used in the investigation.