Human robot cooperation based on human intention inference

Abstract

In many human robot cooperation tasks, robotic agents are usually controlled by commands which are explicitly input by human. Although the explicit control method is straightforward and easy to implement, it will decrease the efficiency, fluency and adaptiveness of human robot cooperation. In this paper, we proposed an implicit control method that recognizes human intention and facilitates human robot cooperation. A finite state machine approach was used to implement human intention recognition algorithm. Four assembly configurations were designed and intention recognition experiments were carried out by using simulation and real data respectively. We have shown that our method can flexibly adapt to human intention even if the human operator retracts some assembly parts and changes mind during the tasks. Furthermore, the robustness of our method has also been demonstrated in noisy situations.