| Job shops intrinsically have the flexibility necessary to achieve many of the goals of Just-in-Time (JIT) manufacturing, but they lack a coherent management approach integrating production decisions with these goals. In this work, a bi-directional heuristic-based algorithm for the scheduling of job shops with firm due dates is presented. The algorithm is applied to the job shop scheduling problem and its performance, with respect to mean tardiness and mean flowtime, is compared with that of more traditional scheduling approaches. Computational experience with the algorithm clearly shows that it outperforms traditional approaches.; While many scheduling techniques identify the bottleneck machine, our scheduling approach identifies critical jobs and/or critical operations, and schedules these jobs in either a forward or backward scheduling mode, whichever mode improves the job shop's performance. In addition, we demonstrate that schedules generated by the JIT Scheduling Algorithm have a more balanced workload and are therefore more robust than traditional nondelay schedules. This is because schedules generated by the algorithm contain embedded idle time which provides the job shop with the flexibility needed to handle changes and/or random disruptions.; Finally, we demonstrate how the JIT Scheduling Algorithm can be applied to job shop scheduling problems in which workload exceeds capacity for a predefined planning horizon. This problem occurs quite frequently in practice because unlike large manufacturers who usually have reasonable market projections, jobs shops have little to no demand forecast and must continually react to the incoming flow of orders. |
