Dataset for AI-assisted detection of the wear level of a cutting tool on a CNC mill

Abstract

A METROM CNC mill was equipped with a microphone and an accelerometer to determine whether the used milling tool was in a good state or blunt. The measurement data was collected over several measurement series. The used tool was replaced at the beginning of each series with a new, similar tool. The working material was always steel. Machine parameters such as feed rate and rotational speed of the tool has been varied between different series. The machine was operated over a defined time frame while data was collected. After that, the machine was stopped, and the operator classified the quality of cut. The quality is directly related to the condition of the tool. The higher the wear, the worse the quality. Using this approach, we build a dataset consisting of different measurement series to show the degradation of a CNC cutter using vibration and acoustic measurements.

A METROM CNC mill was equipped with a microphone and an accelerometer to determine whether the used milling tool was in a good state or blunt. The measurement data was collected over several measurement series. The used tool was replaced at the beginning of each series with a new, similar tool. The working material was always steel. Machine parameters such as feed rate and rotational speed of the tool has been varied between different series. The machine was operated over a defined time frame while data was collected. After that, the machine was stopped, and the operator classified the quality of cut. The quality is directly related to the condition of the tool. The higher the wear, the worse the quality. Using this approach, we build a dataset consisting of different measurement series to show the degradation of a CNC cutter using vibration and acoustic measurements.