Predicting bottlenecks in manufacturing shops through capacity and demand observations from multiple perspectives

Uncertain factors in modern multi-variety and small-lot manufacturing make it extremely challenging to optimise and control the production process. Researchers propose a bottleneck-based optimisation method to reduce perplexity and enhance optimisation. Detecting bottlenecks is a crucial first step in this method and its accuracy has great impacts on production optimisation. This study proposes an independent bottleneck degree to describe the probability of a manufacturing cell becoming a system bottleneck, and model it using capacity and demand observations from the perspectives of capability, quality, and cost. Based on the independent bottleneck degree, we design a closed-loop multi-bottleneck prediction method, which can solve the responsibility cognisance problem resulting from correlation among manufacturing cells. Therefore, it can predict bottlenecks, especially multiple bottlenecks, accurately compared to existing methods.