A proposal for improving production efficiency of existing machining line through a hybrid monitoring and optimisation process.

In a machining line with a Lean manufacturing environment, such as one used for mass production of automotive components, tool life is determined by a tool counter that is set on the safe side to avoid producing a failed product, so the cutting tools are typically underutilised. On the other hand, cutting parameters optimisation, which has a significant impact on tool life, is not feasible due to the rule that prohibits changing the cutting parameters during mass production to maintain the quality standard. To address both issues, this paper proposes a hybrid monitoring and optimisation process that can be carried out concurrently without interfering with production activities. The monitoring process involves two methods, namely threshold limit-based monitoring and machine learning-based monitoring, for ensuring product quality during the optimisation process. Meanwhile, Bayesian optimisation (BO) is used for the optimisation process due to its capability for autonomous search for the optimum condition of an unknown and expensive objective function. The case study demonstrated BO's ability to find the optimal condition with a small number of samples, while the machine learning method used can monitor tool wear and surface roughness with average errors of 5.5% and 7.7%, respectively. [ABSTRACT FROM AUTHOR]