Impact of prehensor stiffness on quality of haptic feedback in body-powered upper limb prostheses

by Anton Filatov

Institution: Colorado School of Mines
Year: 2016
Keywords: Haptic feedback; Upper limb prostheses; Stiffness discrimination; Disturbance detection
Posted: 02/05/2017
Record ID: 2135551
Full text PDF: http://hdl.handle.net/11124/170475


The design goal of a prosthesis is to replicate or replace the lost functionality of the missing limb. However, currently available upper limb prosthetic devices generally fall short of this goal, and as a consequence are often abandoned or rejected by the users. Potentially contributing to this shortfall is the lack of stiffness modulation in commercial prostheses  – whereas intact individuals are able to adapt the stiffness properties of their limbs to the requirements of specific tasks, prosthesis users do not have this option. One particularly troubled area of prosthesis functionality is the quality of haptic feedback provided by the prosthesis to the user. Body-powered prostheses, which account for the majority of all upper limb prostheses in use, provide a degree of haptic feedback by default, a consequence of the direct mechanical coupling between the actuating joint and the prehensor. The usually static prehensor stiffness is part of this mechanical linkage, and may be affecting the quality of the feedback transmitted to the user. Is it possible to improve the performance of the user in feedback-dependent tasks by modulating the prehensor stiffness of a body-powered prosthesis depending on the nature of the task? This work, attempts to answer this question based on the results of three successive human subject studies, in which able-bodied volunteers used a prosthesis emulator system to complete feedback-dependent tasks at various prehensor stiffness settings. The main conclusions may be summarized as 1) prehensor stiffness has a quantifiable and significant impact on the quality of haptic feedback provided to a prosthesis user, 2) the optimal prehensor stiffness varies depending on the task, and 3) the users are aware of the impact of prehensor stiffness on their performance, and are able to make informed, task-dependent adjustments. These three conclusions serve as an endorsement for the inclusion of prehensor stiffness modulation as a control modality in the design of future prosthesis, which may result in increased user satisfaction and capability. Advisors/Committee Members: Celik, Ozkan (advisor), Johnson, Kathryn E. (committee member), Silverman, Anne K. (committee member), Bach, Joel M. (committee member).