| Resumo: | We present a control architecture for nonverbal HRI that allows an anthropomorphic assistant robot with a pro-active and anticipatory behaviour. The control architecture coordinates action and goal coordination between a motor impaired human and the robot as a dynamic process that combines contextual cues, shared task knowledge, and predicted outcomes of the human behaviour. The control architecture is formalized through a coupled system of dynamic neural fields, representing a distributed network of local but connected neural populations with specific functionalities. Each subpopulation encodes relevant information about action means, goals, and context as self-sustained activation patterns. These patterns are triggered by the input and evolve continuously in time under the influence of recurrent interactions. The architecture is validated in an assistive task where the robot acts as an assistant of a person with motor impairments. We show that the context dependent mapping from action observation onto appropriate complementary actions allows the robot to cope with dynamically changing situations. This includes adaptation to different users and mutual compensation of physical limitations.
|
|---|