Enabling reliable Visual Quality Control in Smart Factories through TSN

Jens Popper, Carsten Harms, Martin Ruskowski, Isabel Rheinheimer

In: Roberto Teti , Doriana M. D'Addona (Hrsg.). Procedia CIRP Procedia CIRP. 88 - 13th CIRP Conference on Intelligent Computation in Manufacturing Engineering, 17-19 July 2019, Gulf of Naples, Italy Seiten 549-553 Elsevier B.V. 2020.


Deep learning has been particularly successful in many fields such as computer vision in recent years. However, only few applications of Deep Learning can be found in the manufacturing context. Potentially overloading a computer network with the large amounts of data as well as limited computing power represent a big obstacle, especially for production sensitive data. To make Deep Learning applicable in production, these problems are described and a solution utilizing Time-Sensitive Networking Standards and transfer learning is developed. Then an exemplary application for the visual control of workpieces in ongoing production is implemented in a test factory.

Deutsches Forschungszentrum für Künstliche Intelligenz
German Research Center for Artificial Intelligence