Data-Driven Personalization of Body-Machine Interfaces to Control Diverse Robot Types

Body-machine interfaces for robotic teleoperation have been shown to improve user experience and performance. However, such interfaces must be tailored for each robot type and may require personalization to accommodate user's preferences. Here, we present a novel method to adaptively generate personalized body-machine interfaces from an operator's preferred body movements. The method captures individual motor synergies that are correlated to robot actions and translates them into control commands. The proposed method is validated on a set of users with varied behavioral patterns for teleoperating robots with diverse morphologies and degrees of freedom, such as a fixedwing drone, a quadrotor, and a robotic manipulator.