Classification of human-robot team interaction paradigms

Human-robot team interaction is challenging in terms of system complexity and control synthesis. Classifying different interaction paradigms between a human and a robot team eases the formal analysis. The challenge is to classify the paradigms appropriately, w.r.t. the setting and the task to be performed. In this paper three interaction paradigms are formally defined and analyzed using controllability. It is shown that a straightforward classification of interaction paradigms, based on the mapping properties of the input space to the tangent of the state space is possible. Specific examples in the human-robot team interaction for cooperative manipulation tasks validate the proposed classification methods.