Skip to Main Content
In rehabilitation robotics, a strong coupling between human and robot entails high requirements for achieving mutual adaptation. The latter underlies the acceptance of the robotic device as an extension of the human body and promotes an efficient collaboration. We present automated metrics for quantifying models of human-robot interaction and the mutual adaptation based on the pattern of informational flow between the two participants in the interaction. These methods allow the robotic device to gain the ability to score the mutual adaptation and to implement strategies for increasing it, fostering the human-centered robot autonomy in rehabilitation robotics.