Machine Learning in Electric Motor Production - Potentials, Challenges and Exemplary Applications

Abstract

Artificial intelligence entails a wide range of technologies, which provide great potential for tomorrow's electric motor production. Above all, data-driven techniques such as machine learning (ML) are increasingly moving into focus. ML provides systems the ability to automatically learn and improve from data without being explicitly programmed. However, the potential of ML has not yet been tapped by most electric motor manufacturers. Therefore, this paper aims to summarize potential applications of ML along the whole process chain. To do so, basic methods, potentials and challenges of ML are discussed first. Secondly, special characteristics of the application domain are outlined. Building on this, various ML approaches directly relating to electric motor production are presented. In addition, a selection of transferable approaches from related sectors is included, as many ML approaches can be used across industries. In conclusion, the given overview of different ML approaches helps practitioners to better assess the possibilities and limitations of ML. Moreover, it encourages the identification and exploitation of further ML use cases in electric motor production.