Decomposition of Asynchronous Automated Long Lines with Finite Buffer Capacity

Discrete asynchronous automated long production lines represent the majority of current industrial plants charac- terized by multi-stage serial layouts, high production volumes and inter-stage buffers with finite capacity. Perfor- mance evaluation techniques able to catch system-level behavior of asynchronous lines are therefore very relevant in practical applications. Current approximate analytical methods are able to provide good estimations, under some restrictive hypotheses. In this paper, a new model based on a two-level decomposition of the system, where each machine is characterized by deterministic processing time, a single exponential failure mode, and continuous mate- rials flow, is presented. The model is designed to better approximate, in comparison with existing models, the discrete flow by a continuous flow and in doing so introduces a new approach to the decomposition of continu- ous lines. Performance measures such as throughput and buffer level probability distribution are compared with the ones provided by discrete event simulation. The precision of the proposed method is preliminarily assed on some cases. The proposed method provides the foundations for further research on performance evaluation of more generalized discrete asynchronous systems.