Our goal in the interaction domain is to realize intuitive, empathetic interactions between humans and machines (such as driver assistance systems and robots) so that machines can harmoniously co-exist with humans in daily environments. We explore both verbal and physical ways for humans to seamlessly communicate their state to machines, just like humans communicate with each other about their intention, preference, and emotion. The machine would then be able to use this information to provide humans with not only physical support but also emotional support.

To this end, we conduct basic research toward estimating and calibrating various human states such as emotion, trust, intention, and satisfaction through non-intrusive sensing and actuation modalities. Major challenges include estimation of human internal states from indirect sensing, human modeling for predicting human responses, and human state control through external actions.

The core research areas related to interaction at HRI-US include Human Factors and Robotics. The Human Factors area explores novel interfaces, models, and metrics for human-machine interaction in both driver assistance and robotics applications. The Robotics area focuses on physical interactions between humans and robots in manipulation, communication, and navigation tasks.