SMARTWEAR AI: INTELLIGENT TOOL CONDITION MONITORING IN HIGH-PRECISION MACHINING

Abstract

Accurate tool wear estimation is critical for high-precision machining to ensure product quality, minimize downtime, and optimize toolchange scheduling. This paper proposes a machine learning–based framework that integrates multi-sensor data, advanced feature engineering, and deep learning regression models to estimate tool flank wear in real time. Vibration, acoustic emission, cutting force, and spindle current signals are fused and processed to extract time- and frequency-domain features. A deep regression network is trained and validated on experimental machining datasets collected under varying cutting speeds, feeds, and depths of cut. The proposed approach achieves high estimation accuracy and low prediction latency compared to traditional empirical and classical machine learning methods. Results show significant reductions in estimation error and improved robustness across operating conditions. The framework supports predictive maintenance and can be embedded into CNC monitoring systems for real-time tool health management.