Using kinesthetic feedback to improve human-robotic control

点击翻译按钮获取中文摘要

A seamless integration of sensation and motor function allows humans to effortlessly move their limbs to interact with the environment. Crucial to control of a human arm is continuous sensory feedback encompassing vision, proprioception (the sense of where the limbs are in space) and kinesthesia (movement sense), which are rapidly integrated into motor execution. This communication is disrupted when there is a machine in the loop, as is required for advanced robotic control. Integrating relevant sensory feedback from a robotic machine into natural sensory-motor human processes should improve the controllability of these devices. The kinesthetic illusion (KI) is one method for eliciting movement sensation, which is critical to successful endpoint control of a real (or possibly a robotic) limb. The KI is a perceptual illusion whereby vibrating the muscle tendons creates the feeling as though the limb is moving, although it is stationary. My research will investigate whether the KI can be integrated into a human's control of a robotic system to enhance the accuracy of positioning and grasping tasks. In my first research theme, I will define practical approaches to reliably produce movement sensations in healthy adults by manipulating experimental factors such as vibration, skin stretch, muscle contraction, and visual input. This work will provide essential knowledge on how to optimize the experience of the kinesthetic illusion. My second research theme will explore how kinesthetic sensations can be integrated into functional control of a robotic system. I will investigate the ability to use the KI to enhance robotic control accuracy for positioning and gripping tasks. This work will inform my long-term research objective which is to implement new technical approaches to non-invasively interface with humans to consistently communicate the state of an external robotic device with complex joint movements. This information will provide new knowledge for the research community on manipulating the experience of the KI through external experimental factors, and result in the development of new techniques that will have relevance for ongoing research on the use of naturally integrated feedback for human-robotic control applications. By harnessing existing knowledge of the neurophysiology of sensory-motor integration, my research will provide a sound basis for practical translation of these techniques to provide effective feedback from robotic machines to human users.