Artificial intelligence augmented digital twin for improving the machinability in a robotic carbon fiber reinforced plastics machining process

Abstract

Carbon fiber-reinforced plastics (CFRP) are widely used in the aerospace and automotive industries due to their high strength-to-weight ratio and excellent mechanical properties. Although CFRP is produced in a near-net shape, secondary machining processes, such as drilling, are essential to achieve precise tolerances. However, the anisotropic nature of the material makes machining difficult and often leads to defects such as delamination and uncut fibers. These challenges are further exacerbated by the use of industrial robots, as their lower stiffness compared to computer numerical control machines leads to increased vibrations and machining errors. To solve these problems, a innovative digital twin framework that integrates predictive modeling and real-time monitoring for robotic CFRP machining was developed in this study. A twin model synchronized with the robot predicts the stiffness of the tool-tip based on posture and enables optimal posture selection to improve stability. In addition, an integrated artificial intelligence monitoring system analyzes multi-sensor data in real time to predict machining defects. A data-driven infrastructure for seamless data acquisition, processing, and visualization has been established to improve defect detection and machining efficiency. The proposed framework increases the reliability of robotic CFRP machining by providing data-driven insights and enabling optimized machining strategies. By integrating predictive analytics and real-time monitoring, this study advances the intelligent manufacturing of high-precision CFRP.