Anna Belardinelli, Chao Wang, Daniel Tanneberg, Stephan Hasler, Michael Gienger, "Train your robot in AR: investigating user experience with a continuously learning robot", Lifelong Learning and Personalization in Long-Term Human-Robot Interaction (LEAP-HRI), 2024.

Assistive robots that can be deployed in our homes will need to be understandable, operable, and teachable by non-expert users. This calls for an intuitive Human-Robot Interaction approach that is also safe and sustainable in the long term. Still, few studies have looked at repeated, unscripted interactions in a loosely supervised setting with a robot incrementally learning from the user and con- sequentially expanding its knowledge and abilities. In this study, we set out to test how the user’s experience and mental model of the robot evolve when spontaneously teaching it simple tasks in Augmented Reality (AR). Participants could freely access the AR glasses in a common office space and demonstrate physical skills in a virtual kitchen scene, while the holographic robot gave feedback about its understanding and could ask questions to generalize the acquired task knowledge. The robot learned the semantic effects of the demonstrated actions and upon request could reproduce those on observed or novel objects through generalization. Preliminary results show that users find the system engaging, understandable, and trustworthy, but with large variance on the last two constructs. Further analyses will assess how subjective measures can be cor- related to user behavior, to evaluate the relation between system understanding and teaching effectiveness.