Integrating digital assistant with human operators for realizing process monitoring in manual machine tool operations

The human skill sets are critical in monitoring and identifying in-process faults during manual machine operations. The shortage of skilled human operators, lack of consistency in decision-making, and slower response result in lower productivity and part quality. This paper presents the digital assistant as a decision support system to improve operators’ perceptions about tool wear state and chatter onset while running the manually operated machines. The digital assistant acquires real-time process information using Acoustic Emission and Accelerometer sensors integrated with the data acquisition elements. The tool wear and chatter information is extracted from the sensor data using a decision-making module combining Root Mean Square and Support Vector Machine (SVM) with a quadratic kernel. The SVM classifier is trained using a learning-through-demonstration approach to digitize the expertise of a skilled operator. The decision-making module is integrated with the Human Machine Interface (HMI) unit to display real-time process information for appraising machine operators. The prediction abilities of the digital assistant are corroborated by performing machining experiments for different tool-work material combinations on the conventional manually operated engine lathe. The studies showed that the digital assistant can effectively capture process fault information and complement the decision-making abilities of human operators.