Learning human-robot collaboration insights through the integration of muscle activity in interaction motion models

Abstract

Recent progress in human-robot collaboration (HRC) makes fast and fluid interactions possible. Methods like Interaction Probabilistic Movement Primitives (ProMPs) model human motion trajectories through motion capture systems. However, such presentation does not properly model tasks where the motion trajectories are similar. We propose to integrate the Electromyography (EMG) signals into the Interaction ProMPs framework. The contribution of this paper is the increased capacity to discern tasks that have similar trajectories but ones in which different tools are utilized and require the robot to adjust its pose for proper handling. Augmented Interaction ProMPs are used with an augmented vector involving muscle activity. Augmented trajectories are used to learn correlation parameters and robot motions are generated by finding a best fit task. Collaborative task scenarios with similar motions but different objects were used and compared. Integrating EMG signals into collaborative tasks significantly increases the ability to recognize nuances in the tasks.