Using timed event-graphs for the analysis of job-shop production systems

This paper studies the functioning of a job-shop-like production system for a repetitive processing of the jobs. A modeling base on deterministic timed Event-Graphs is developed to evaluate the dynamic performance of the system and in particular determine the conditions of functioning at maximal rate (i.e. maximum productivity). It is shown that, given any fixed sequencing of the jobs on the machines, it is always possible to fully utilize the bottleneck machines in steady-state with a finite number of jobs in-process. A heuristic algorithm is further developed to obtain a near-optimal sequencing and scheduling of the jobs, such that the bottleneck machines be fully utilized with the minimum work-in-progress (WIP) inventory. This condition guarantees an optimal functioning in case of Flexible Manufacturing Systems (FMS), since the resources used to transport the parts (such as pallets, carts, etc.) are minimized and the production rate is maximized.