Condition Monitoring using Machine Learning: A Review of Theory, Applications, and Recent Advances

Abstract

In modern industry, the quality of maintenance directly influences equipment's operational uptime and efficiency. Hence, based on monitoring the condition of the machinery, predictive maintenance can minimize machine downtime and potential losses. Throughout the field, machine learning (ML) methods have become noteworthy for predicting failures before they occur. However, the efficacy of the predictive maintenance strategy relies on selecting the appropriate data processing method and ML model. Existing surveys do not comprehensively inform users or evaluate the quality of the monitoring systems proposed. Hence, this survey reviews the recent literature on ML-driven condition monitoring systems that have been beneficial in many cases. Furthermore, in the reviewed literature, we provide an insight into the underlying findings on successful, intelligent condition monitoring systems. It is prudent to consider all factors when narrowing the search for the most effective model for a particular task. Therefore, the tradeoff between task constraints and the performance of each diagnostic technique are quantitively and comparatively evaluated to obtain the given problem's optimal solution.