Concurrent Policy Blending and System Identification for Generalized Assistive Control
This work addresses the challenge of adaptive assistive robotics for individuals with motor impairments, presenting an incremental improvement over standard domain randomization methods.
The paper tackles the problem of enabling collaborative robots to perform complex tasks under varying system parameters by combining policy blending with system identification, resulting in generalized policies that are robust to parameter changes, as demonstrated on a robot-human itching task with motor impairments.
In this work, we address the problem of solving complex collaborative robotic tasks subject to multiple varying parameters. Our approach combines simultaneous policy blending with system identification to create generalized policies that are robust to changes in system parameters. We employ a blending network whose state space relies solely on parameter estimates from a system identification technique. As a result, this blending network learns how to handle parameter changes instead of trying to learn how to solve the task for a generalized parameter set simultaneously. We demonstrate our scheme's ability on a collaborative robot and human itching task in which the human has motor impairments. We then showcase our approach's efficiency with a variety of system identification techniques when compared to standard domain randomization.