Data Driven Cutting Tool Fault Diagnosis System Using Machine Learning Approach: A Review

Abstract

Every industry is now turning towards industry 4.0. In this era, industry requires smart machine tools. But for small scale and medium scale industries, it is not affordable to buy new smart machineries. Therefore, the fault diagnosis system (FDS) has got unavoidable propensity in the machine of modern huge information and smart manufacturing. Simultaneously, it offers a solid answer for taking care of the mechanical machines & its cutting tools health status. Industry 4.0 and its key advances assume a fundamental part to make mechanical systems independent and along these lines make conceivable the automated data assortment from modern machines/cutting tools. In view of the gathered information, ML algorithms can be applied for automated shortcoming identification and finding. It is difficult to choose relevant machine learning (ML) procedures, kind of data, data size, and hardware to apply ML in mechanical systems. Determination of wrong FDS procedure, dataset, and data size may cause increase in downtime and infeasible for scheduled maintenance. Accordingly, this study aims to present, the brief review of literatures for investigation to find existing methodologies of ML and its applications, Supportability to develop novel system to diagnose faults in CNC hobbing cutter and to choose suitable ML methods for their required FDS.