On Planning And Task Achievement Modalities For Human-Robot Collaboration

In this paper we present a robot supervision system designed to be able to execute collaborative tasks with humans in a flexible and robust way. Our system is designed to take into account the different preferences of the human partners, providing three operation modalities to interact with them. The robot is able to assume a leader role, planning and monitoring the execution of the task for itself and the human, to act as assistent of the human partner, following his orders, and also to adapt its plans to the human actions. We present several experiments that show that the robot can execute collaborative tasks with humans.