On-chip tool wear estimation in micro-milling using artificial neural network

In micro milling, Tool wear estimation is crucial because it improves job surface quality and process integrity. This paper presents a simple approach that estimates the tool wear from acceleration data obtained during the micro milling process. The acceleration data for tool with different wear lengths, collected using a wireless-aided three-axis accelerometer sensor (presented in Arduino Nano 33 BLE Sense board, attached to a rotating micro-milling tool) and was preprocessed to predict the tool wear state of the tool. The tool wear length and acceleration data were used as the training dataset. This training dataset was subjected to end to end Deep Learning (ANN) based framework. Now the trained ANN model is classified in serial monitor of Arduino IDE for estimation of Tool Wear state. It was found that the setup accurately determines tool wear state of the tool during the micro milling process. All experiments were conducted in CSIR Lab.