Exploration of lower limb body schema integration with respect to body-proximal robotics

Abstract

The integration of body-proximal robots into the body schema of their users is a crucial aspect for the acceptance of those artificial devices. Humans integrate information about such systems from visual, tactile, and proprioceptive perception. For the design of novel, human-oriented robotics, it is important to understand how this integration can be supported by the technical design of the robot. In this paper, first experiments aiming at the assessment of lower limb body schema integration during postural motions are performed. A robot is used to imitate the squatting movements of human subjects. In contrast to a previous implementation, human motion sensing is improved for real-time operation by using inertial measurement units. A first human-in-the-loop experiment indicates that this approach resolves limitations caused by delays of the previous RGB-D measurements. The integration into the body schema of the participants is assessed by means of a psychometric questionnaire. Experimental results regarding body schema integration suggest that for body-proximal robotics noticeably detached from human subjects, agency is the most stable of body schema variables and that to induce a sense of ownership and location, more feedback channels to the subjects might be required.