Towards dynamic object manipulation with tactile sensing for prosthetic hands

Abstract

The manipulation of objects is an inherent capability of dexterous human hands. However, state-of-the-art prostheses are limited to only forming grasps and gestures due to the low information transfer rate in the human-machine interface. Additionally, in the prosthetic setting the hand is not privy to such global knowledge including the object's shape, weight, friction properties, etc. Existing techniques from robotics that can manipulate objects in this prosthetic setting require compensation terms which are counterproductive to the manipulation task. The incorporation of tactile sensing leads to a simpler control formulation and enables relaxation of several restrictive assumptions inherent in the robotics applications. In this work a novel control system is proposed that incorporates tactile sensing for object manipulation and optimal distribution of contact forces for prosthetic hands. Simulation results demonstrate the ability of the proposed control system to manipulate an object, while minimizing the use of frictional forces during manipulation.