AI / Robotics
SMILE
Semantic Intuitive Skill Learning (SMILE) will endow robots with the capability for physical mediation, i.e., enable embodied AI systems to proactively and implicitly provide physical support in socially dynamic situations. Machines become physical enablers that co-exist with humans 24/7 and learn and grow within social groups.
We combine research in multi-modal imitation learning to grow system’s physical capabilities in interaction with the user. Semantic planning is integrated with learning and common sense to find creative solutions to novel problems, while explainable and ergonomic Human-Robot Interaction ensures easy and intuitive cooperation between user and system.
Using augmented reality, we research novel ways of interaction and knowledge sharing between humans and robots. Predictive ergonomics and biomechanics enables proactive physical robot support.