SoC-Based Early Failure Detection System Using Deep Learning for Tool Wear

This study aims to implement a system-on-chip (SoC) detection system for tool wear monitoring and alarms high-precision machining processes. The proposed deep learning approach is trained by the collected sensors from real performed in three-axial computer numerical control (CNC) machine center combined with different conditions of spindle speed tightening torque. corresponding vibrational sound signals were accelerometer micro-electro-mechanical-system (MEMS) microphone, then flank was measured camera. In this study, designed early according ISO 8688-2:1989 standard. A model frequency spectrum inputs developed prediction. addition, treat variation detection, two sensor fusion approaches are presented implemented on an SoC-board (Pocket Beagle) landing cost reduction. average accuracies approximately 99.7% 87.75% single merged models, respectively. results demonstrated effectiveness performance approach.