A robust control approach to automated manufacturing systems allowing failures and reworks with Petri nets

The problem of automated manufacturing systems (AMSs) deadlock has received considerable attention recently. Most researchers have assumed that allocated resources do never fail. However, this is not the case in practice since some resources may fail unexpectedly. In our previous work, we just presumed unreliable resources and developed supervisory algorithms to ensure robust deadlock-free operation in the event of resource failure. In this paper, the failure includes not only failed resources but also failed products. For a product, it can have a temporary storage in a buffer when the product does not necessarily need to progress or an unreliable resource is not expectably well behaved. We propose a deadlock avoidance algorithm and a robust supervisory algorithm in the paradigm of buffer-supported system of sequential systems with shared resources and recovery (B-S4R2). The proposed algorithms can ensure the processes not requiring the failed resources still operate normally without any interruption and the failed products are recovered to continue their manufacturing process. Experimental results show that the proposed approach is effective as well as efficient in response to resource and product failures.