Electric Vehicle Production: PEM Aims to Increase Data Competence


The chair "Production Engineering of E-Mobility Components" (PEM) of RWTH Aachen University has joined forces with Helmut Schmidt University (HMU) in Hamburg to launch the "NEED" research project. The aim of the joint project is to sustainably increase the data competence of young scientists in electric vehicle production.

  A tablet shows data in a production environment Copyright: © Adobe Stock – Panuwat (Balls)

The main aim is to find out whether interactions in the manufacturing processes of electric vehicles can be identified and quantified using digital methods, how conventional, engineering-based modeling approaches can be replaced by data-based analysis methods, and how the expertise gained can be disseminated and sustainably anchored in both the scientific and industrial environments.

Reducing scrap rates and increasing vehicle quality

The background is the challenge of being able to produce electric motors, batteries, and fuel cells for electric vehicles in a timely, economical and sustainable manner. "The production of these three core components is still characterized by interactions that occur both within individual process steps and across several process steps," says PEM head Professor Achim Kampker. "They have a significant influence on the scrap rate during production, on subsequent product quality and thus also on the service life of electric vehicles." In order to reduce the scrap rate and increase the quality of the vehicles, according to Kampker it is necessary to detect any defects that occur as early as possible and avoid corresponding faults in the future.

Conventional modeling reaches its limits

Up to now, the modeling of production processes and the associated interactions has mainly been carried out using conventional methods – such as simulations or statistical analyses. However, due to the high complexity and enormous amounts of data, conventional modeling methods have reached their limits, making artificial intelligence methods necessary.

Turning away from classical experiments

The expertise needed for this is to come from an exchange of knowledge between Helmut Schmidt University and the RWTH's PEM chair: While the Aacheners provide the HSU with know-how on electric vehicle component production, the University from Hamburg brings its expertise on artificial intelligence and "machine learning" to PEM. Ideally, the project should bring about a paradigm shift – away from classic experiments and toward greater use of existing or easily accessible data. "The associated automation in model creation will enable shorter development times and faster series start-up phases," explains Kampker, who, in addition to his expertise in electric mobility components and concepts in teaching and research, has many years of experience in the management of start-ups as well as established companies.

Further information on the project is provided here.