Buffer design for availability: a new simulative study in case of infant and random failures

In automatic production systems, the size of intermediate buffers significantly affects the success of the production system, especially when their functions is to protect the system from the inevitable processing time fluctuations caused by machine disruptions. In Battini et al. (2009), the authors provided a new experimental formula to define the optimal buffer size to maximise throughput without the need for oversized machines. The authors observed only 'wear-out failures', in which the failure rate increases with time. In order to better understand the potential benefits of previous results and extend their applications in industrial contexts, a new simulation study is now presented, with reference to all those interesting industrial cases in which production flow lines are largely affected by infant failures or random machine failures. The final results show a completely new experimental matrix that cover all the possible kinds of failures, able to work in different industrial settings and under different reliability performances.