Sensor-Based Exoskeletal Mirroring System

Abstract

Upper-limb impairment after stroke or motor nerve injury most often leaves patients without fine control of thehand. This paper presents a low-cost, sensor-based exoskeletal hand mirroring system in which a measurement gloveworn on the intact hand records finger joint angles with magnetic rotary encoders, and a cable-driven exoskeletonreproduces those motions on the impaired hand. Unlike designs that couple the finger joints, the exoskeleton actuatesthe metacarpophalangeal (MCP) and proximal interphalangeal (PIP) joints of the index, middle, and ring fingers independently, plus a single-degree-of-freedom thumb, while keeping all actuators on the forearm. The glove tracked finger position with a mean absolute error of 1.19° and a lag of about 312 ms, introduced mainly by smoothing. The exoskeleton followed commanded motions with 72–140 ms of lag but had reduced joint angle accuracy for certain motions due to the nature of a differential string drive, with a mean absolute error of 15-25°. The exoskeleton exerted a maximum force per finger of around 1.2 N. The results establish the feasibility of independent dual-joint mirroring with actuation hardware kept off the hand.