A Conceptual Framework for Cyber-Physical Quality Monitoring System using Machine Learning

Abstract

In this paper, we propose a Cyber-Physical Quality System (CPQS) integrated framework that can predict, analyze, and validate the quality monitoring system in manufacturing with 95% accuracy in real-time using machine learning techniques. CPQS framework analyses real-time sensor networks and configures the importance of artificial intelligence-driven big data analytics for predicting the quality of cyber-physical production networks. Cyber-physical data like speed, feed, depth of cut, coolant temperature, vibrations, tangential cutting forces, and tool life for 400 parts were collected from the various sensors placed on Computerized Numerical Control (CNC) machines after doing modal analysis. Various machine learning techniques were used to predict the quality of the part wherein the inputs affecting it were predominately dominated by vibration and temperature. Extreme Gradient Boosting (XGB) machine learning techniques out of many could predict the quality of the part with 96.2%. accuracy. The caveat for the present results is that it has been tried out only for Titanium Alloy parts and the tool wear has been approximated using Taylor’s equation which can be enhanced by using image processing. The model deployed in real-time could produce defect-free parts quickly. This could reduce the cost of quality by 80%, thereby increasing the production line's productivity, quality, and efficiency