Abstract
- Continuous generating grinding is a highly productive and precise way to finish hardened gears. However, the dressing of the tool has a huge potential for optimization. Detecting the optimal point to stop the dressing process would save cost and reduce machine downtime. This paper provides a proof of concept of how machine learning models can achieve these goals based on acoustic emission data. Therefore, different ML-Classifiers were trained and compared. Particular emphasis was placed on the robustness of the models. In addition, the trade-off between fast and robust models was examined.