Machine Learning applications for automatic detection of damage to material surfaces can improve product quality and reduce cost and time for the quality control process. Within the Fraunhofer Research Center for Machine Learning, scientists of Fraunhofer IAIS are working on expanding the application field of Machine Learning to make it more explainable and applicable to cases where there is a lack of suitable data.
One approach is Informed Machine Learning, which integrates expert knowledge, for example data from simulations or physical laws. In one project, the Fraunhofer IAIS team developed an image-based detection system for damage from hailstorms on vehicles. Insurance companies and their appraisers face the challenge of having to assess a large number of cases within a short time span. To facilitate the process, a mobile unit scans the car body’s damaged parts. Afterwards, the Fraunhofer Machine Learning algorithm detects, classifies and measures the damages automatically. The system is filed for patent application.
The technology is applicable to other areas of industrial quality assurance and damage control, where smooth and reflective surfaces are produced, processed or tested.
You are welcome to contact us to arrange a meeting at the fair.
Contact person exhibit 14: Ulrich Nütten, send e-mail