Felix Finkeldey achieves 3rd place at the Forum Junge Spitzenforschung 2026

Felix Finkeldey from the Department of Computer Science has won 3rd place at the Forum Junge Spitzenforschung 2026 with his LEAP project. The project is dedicated to a central challenge of industrial artificial intelligence: the reliable transfer of AI models from research into real production environments.

The focus is on the development of acoustic process monitoring for milling processes. A retrofittable audio box analyzes process noises in real time and detects unstable conditions such as chattering at an early stage. The aim is to make manufacturing processes more robust and efficient.

Industrial AI applications often face the challenge that large amounts of labeled data are required and models are difficult to transfer to new machines, tools and environments. LEAP addresses these hurdles with self-supervised learning and active learning approaches. This should reduce the effort required for data annotation and speed up the commissioning of new systems.

In addition, "ChatterCoach" is being developed as a learning platform that will make experience in machining available digitally and make it easier for new specialists in particular to get started.

The technological basis for LEAP is being developed as part of the ZIM project AudioIno. Participants include Lorenz Hoffmann GmbH, GFE Präzisionstechnik GmbH Schmalkalden and the Institute of Machining Technology.