Rendering Strategies for Underactuated Hand Exoskeletons

Underactuated exoskeletons offer improved simplicity, affordability, and movement adjustability based on interaction forces, but in turn they suffer from lack of controllability. In this letter, we propose different rendering strategies to improve the controllability of underactuated devices, with no prior knowledge about the task to be performed. Since underactuated devices cannot achieve all desired poses or render all desired forces, the proposed strategies replace these desired values with proxy values that the devices can satisfy. Strategies in the force and operational spaces aim to minimize the error between the desired and the actual output force or pose, respectively, during operation. We test these strategies first using an underactuated hand exoskeleton during a virtual grasping task. Then, we compare them to each other and to other existing methods by analyzing their control performance during a trajectory tracking task.