A Probabilistic Framework for Semi-autonomous Robots Based on Interaction Primitives with Phase Estimation

This paper proposes an interaction learning method suited for semi-autonomous robots that work with or assist a human partner. The method aims at generating a collaborative trajectory of the robot as a function of the current action of the human. The trajectory generation is based on action recognition and prediction of the human movement given intermittent observations of his/her positions under unknown speeds of execution; a problem typically found when using motion capture systems in scenarios that lead to occlusion. Of particular interest, the ability to predict the human movement while observing the initial part of his/her trajectory allows for faster robot reactions, and as it will be shown, also eliminates the need of time-alignment of the training data. The method models the coupling between human-robot movement primitives and is scalable in relation to the number of tasks. We evaluated the method using a 7-DoF lightweight robot arm equipped with a 5-finger hand in a multi-task collaborative assembly experiment, also comparing results with our previous method based on timealigned trajectories.