AI systems consist of a variety of AI models that are linked into a complete solution using rule-based algorithms.

For use in regulated environments, such as pharmaceutical packaging machinery and medical technology plant engineering, the decisions and limitations of AI systems must be transparent and explainable.

In the presentation, the development of AI systems using explainability methods will be demonstrated through an example. The results come from the BMBF-funded project Qua²ntum, in collaboration with Aimino Tech GmbH and the Fraunhofer Institute for Production Technology in Aachen.