AI-Supported Process Monitoring in Machining

Abstract In this study, AI-supported anomaly detection methods in the milling of inhomogeneous sample materials are investigated. To simplify data generation, targeted boreholes were introduced into homogeneous material samples. Process data were collected by means of acceleration measurements on both the workpiece and tool sides and force measurements on the tool side. Implementing targeted feature extraction and applying feature-based machine learning algorithms achieved precise material classification and reliable differentiation between drilled and undrilled samples for process monitoring.