Predicting the quality of products holds enormous potential for industry. The aim of the Bavarian research project in the AI production network (EBQuoPro) at Augsburg University of Applied Sciences (THA) is to use predictive quality algorithms to identify risks before actual series production.
The project aims to help avoid unplanned downtimes due to reactive maintenance. This enables companies to take measures at an early stage to improve the quality of their products, reduce rejects during production and increase their competitiveness. Traceability process data from electronics manufacturers is used for this. This data is to be analyzed using AI algorithms in order to find patterns and then incorporate this information into the inspection strategy. For example, those components that fail more frequently in different assemblies are tested more closely. It may also be possible to allow inspection gaps where there is no risk.
The research project is being carried out under the consortium leadership of Rohde & Schwarz Messgerätebau. The Rohde & Schwarz production site in Memmingen is contributing its many years of experience at the interface to product development (NPI) and providing its expertise in the assembly of complex modules and the systematic recording of large amounts of quality data. In addition to BMK and THA, the business informatics branch of the Fraunhofer Institute for Applied Information Technology (FIT) is also a research partner.
www.bmk-group.de/home
www.rohde-schwarz.com
www.fit.fraunhofer.de
