Remanufacturing: Fraunhofer IPK has developed an AI image recognition system for vehicle components.
Product variance – two generators with different part numbers are visually identical. © Fraunhofer IPK/Larissa Klassen

How can used vehicle parts be dealt with sensibly? A large number of used parts end up in the scrap yard every year. As anyone who has ever tinkered with an old car knows, it is more resource-efficient than recycling if alternators, starters and other parts are repaired and put back into circulation. This reduces the amount of waste, cuts the carbon footprint and extends the life of the parts. Remanufacturing – bringing used equipment up to new condition – can become a key component in meeting the Paris climate targets and on the way to a circular economy. According to a study by the VDI Centre for Resource Efficiency, up to 80% of manufacturing costs can be saved by remanufacturing used parts and up to 90% of material consumption can be reduced.

In order to increase the reuse rate, the Fraunhofer Institute for Production Systems and Design Technology IPK is developing an AI-based assistance system in the EIBA project to be able to identify used parts without QR or barcodes in an image-based and partially automated way. The system supports sorting and reading so that more used components can be remanufactured. Partners in the project, which is funded by the Federal Ministry of Education and Research BMBF, are Circular Economy Solutions GmbH, the Technical University of Berlin and the German Academy of Science and Engineering acatech.

“In the automotive industry, after the old part has been removed, it is assessed in the sorting centre on the basis of certain criteria whether it can be reused,” says Marian Schlüter from Fraunhofer IPK. “However, this is anything but trivial. Part numbers as the only visually reliable feature are no longer legible, scratched, painted over, or type plates have fallen off. So the worker sorts it out by mistake, it is purely recycled. This is exactly where the AI comes into play. It identifies the old parts independently of the part number based on their appearance and feeds them into a second life cycle.” Identification features such as weight, volume, shape, size and colour characteristics are used, but customer and delivery data also flow into the evaluation. Where the AI system fails with its image processing, the employee helps. According to Fraunhofer IPK, the flexible technology can be used for all kinds of dimensionally stable components. A study has shown a recognition accuracy of 98.9 %.


More on ndion

Discover more news on the topic of innovation.


Share this page on Social Media:

Print Friendly, PDF & Email