Application of probabilistic methods in automated tool fault detection

This paper investigates the fault detection of automated tools in the production process, focusing on how the fault detection interval and the frequency of periodic tool replacement affect the expected cost of machining parts. The process of continuous machining of parts by automated tools is taken as the object of study, and an optimal strategy model for periodic inspection and replacement of equipment is developed based on a probabilistic approach. The optimal inspection intervals and strategies for periodic tool replacement are considered for two scenarios, namely tool-only failures and mixed tool and non-tool failures, when the benefits of the machining process are maximized. The model calculates probabilistically the probability of no failure before tool change and the probability of failure before tool change. It finds the optimal periodic inspection intervals and tool change intervals through optimisation using the expected cost of failure of a single part as an objective function. Second, simulation tests are performed using the Monte Carlo stochastic simulation method to verify the superiority of our model. Finally, a sensitivity analysis illustrates the strong robustness of the model.