CAVECORE: Continuous, Automated Validation, and Evaluation of Cognitive Robots in Open-Ended Environments

CAVECORE (Continuous, Automated Validation, and Evaluation of Cognitive Robots in Open-Ended Environments) is a Marie Skłodowska-Curie Doctoral Network that trains the next generation of researchers to advance cognitive robotics—robots that can interact, learn, and adapt to open-ended real-world environments. The network addresses one of the central challenges in AI-enabled robotics: evaluating and validating the quality, safety, and reliability of such robots in a systematic and trustworthy manner.

Doctoral Candidates in CAVECORE will follow an interdisciplinary training program that combines expertise in robotics and AI (e.g., learning-enabled manipulation, control, and human-robot interaction) with cutting-edge validation and evaluation methods (e.g., automated testing, diagnosis, analysis, and AI metrology). They will develop novel approaches that make evaluability-by-design a core part of the robot development process, ensuring that cognitive robots provide continuous evidence of their performance and can adapt to evolving societal, ethical, and legal requirements such as those set out in the EU AI Act.