By loading the video you accept YouTube's privacy policy.
Learn more

Load video

Frederik Sailer and his team from the Fraunhofer Institute for Manufacturing Engineering and Automation will be demonstrating an AI-based visual inspection system at Control that can automatically detect a variety of surface defects. The special thing about this system is the fact that the Fraunhofer Institute has only used simulated images for the teach-in procedure Highly realistic 2D images are generated from a virtual 3D scene, which are astonishingly similar to real images. With the help of these 2D images, the AI system can also reliably detect surface defects such as scratches, dents and bumps in the real data. Here at the Control trade fair, it’s especially interesting for the Fraunhofer Institute to find out in which other sectors its technology can be put to use.