The VisionAdapt transfer project focuses on the automatic detection of critical chips during turning, which cannot yet be reliably avoided and can lead to damage to tools or components. A special approach is 3-axis simultaneous turning, also known as High Dynamic Turning (HDT). An additional rotation axis allows the tool to be flexibly aligned, which enables more favorable chip shapes, less tool wear and higher productivity. To fully exploit this potential, the researchers are combining HDT with imaging chip shape detection and ML-supported process adaptation.

For this purpose, a camera system is installed directly in the machine, which continuously records the chips produced during machining. The image data is processed on an edge device and evaluated using specially developed ML models. In a first step, the system reacts to critical chip shapes with warnings or a machine stop. In a second stage, the machine should intervene independently and adapt the process in real time, for example by changing the feed rate or using oscillating tool movements for targeted chip breaking.

"With ML-based data evaluation, it will be possible to observe the chips in real time and adapt the turning process to the situation. This can further automate and accelerate production," says Prof. Dr.-Ing. Berend Denkena, Leibniz Universität Hannover. In addition to image data, machine control data and information on workpiece and tool geometries are also used to train the ML models. Over the course of the project, a prototype will be created that implements the process adaptations independently and is tested in a turning/milling center using practical use cases. If successful, this should result in a market-ready system for industrial series production.

"Our aim is to develop an innovative process that offers companies in the machining industry concrete added value. We are therefore ensuring the transfer between research and industry in two ways: firstly, we will develop future scenarios and derive future requirements for the process in order to take its long-term relevance into account during the development process. On the other hand, our practical partner will evaluate the marketability and contribute technical requirements from the perspective of industry," explains Prof. Dr.-Ing.