Reactive Search for Flexible Job Shop Scheduling

In the classical job shop scheduling problem (JSSP), n jobs are processed to completion on m unrelated machines. Each job requires processing on each machine exactly once. For each job, technology constraints specify a complete, distinct routing which is fixed and known in advance. Processing times are sequence-independent, fixed, and known in advance. Each machine is continuously available from time zero, and operations are processed without preemption. The objective is to minimize the maximum completion time (makespan). The flexible job shop (FJS) scheduling problem extends JSSP by assuming that, for each given operation, there exists at least one instance of the machine type necessary to perform it. We are thus confronted with two problems, that of assigning each operation to an appropriate machine, and that of sequencing the operations on each machine. The flexible-routing job shop (FRJS) scheduling problem, or job shop with multipurpose machines, extends JSSP by assuming that a machine may be capable of performing more than one type of operation. (For a given operation, there must exist at least one machine capable of performing it.) FRJS approximates a flexible manufacturing environment with numerically controlled work centers equipped with interchangeable tool magazines. This report investigates a reactive search (RS) strategy for FJS and FRJS, and compares it to the authors’ earlier dynamic, adaptive tabu search (TS) method. “Proof-ofconcept” results for problems constructed from three difficult JSSP instances are presented. MS of October 1997 Originally presented as “New Tabu Search Results for the Flexible Job Shop Problem”. INFORMS, Dallas, October 1997 (invited session ). Revised October 1998