Industry 4.0 is accompanied by a variety of technologies that offer great potential for tomorrow’s electric motor production. In particular, data-driven approaches applying methods of artificial intelligence (AI) are increasingly coming into focus. So far, however, only a few scientific works have dealt with the potential of data-driven approaches in electric motor production. As a result, there is a lack of industry-specific best practices that electric motor manufacturers could use for orientation. Another reason for the hesitant adaptation of AI methods is the black-box character of the widely propagated deep learning algorithms as well as the often missing opportunity to integrate existing domain knowledge.
Therefore, the aim of the project E|KI-Opt is to exploit the potential of AI for process and production optimization in electric motor manufacturing by using data-driven approaches that are as explainable as possible as well as to derive best practices for the entire industry. On the one hand, cross-process relationships are to be detected and an AI model for predicting the electric motor’s quality based on material, process, and test data is to be derived. On the other hand, essential, quality-relevant sub-processes are to be optimized using AI as well. Building on the domain expertise of the participating plant manufacturer and operator as well as the AI expertise of the IT service provider and university, E|KI-Opt aims to create explainable and robust AI solutions that improve the production of electric motors and enrich the industry players with new AI-based business concepts.