Haptic Sensory Feedback for Improved Interface to Smart Prosthetics.

Abstract

Grip force feedback is not available in modern myoelectric upper-limb prostheses, yet its benefits are well known in object manipulation tasks performed through cable-driven body-powered prostheses. To evaluate the efficacy of grip force feedback in a myoelectric prosthesis, direct head-to-head comparisons should be made with body-powered prostheses, as well as with proposed designs that provide grip force feedback through haptic displays such as vibrotactile arrays. Direct comparisons, however, are difficult because myoelectric control for a trans-radial amputee uses residual muscles in the forearm, body-power generally refers interaction to the shoulder, and haptic displays often involve additional information encoding transformations. Currently, no unifying theory exists to cover both information encoding as well as the body part used for control or display. The work developed in this dissertation presents a systematic hypothesis-driven approach to evaluating both information encoding and body part used in the display of grip force feedback. Drawing upon principles from psychophysics, teleoperation, and sensory substitution, we use a series of human subject experiments to quantify the value of grip force feedback for an amputee wearing a trans-radial myoelectric prosthesis. Our findings demonstrate that both able-bodied individuals and amputees scale and coordinate their grip force for the anticipated weight of an object, that control and grip force feedback should be located on the same body site to improve stiffness recognition, and that grip force feedback is more useful than vision feedback in stiffness recognition through a prosthesis.