Design of a robotic hand and simple EMG input controller with a biologically-inspired parallel actuation system for prosthetic applications

Abstract

This paper presents the mechatronic design of a robotic hand for prosthetic applications. The main characteristic of this robotic hand is its biologically-inspired parallel actuation system, which is based on the behavior/strength space of the Flexor Digitorum Profundus (FDP) and the Flexor Digitorum Superficialis (FDS) muscles. The design separates the strength space of the FDS and FDP muscles into a lighter strength region where finer manipulation and general approach tasks are executed, and a higher strength region where the more robust grasps are achieved. Two parallel actuator types and kinematic structures are designed to complement the requirements of both strength space regions.This unique structure is intended to be driven by electromyographical (EMG) signals captured at the surface of the skin. The direct relation between signal and actuation system lends itself well to interpreting the EMG signals from the FDP and FDS muscles into effective task execution, with the goal of helping the user to achieve a good approximation of the full capabilities associated with the human hand, without compromising strength, dexterity, appearance, or weight; which are common issues associated with prosthetic hands. The designed finger’s capability of having a strength space similar to that of the FDS and FDP muscles is validated via direct inputs from a power supply and then via a controller using an actual EMG signal input from the human forearm. The controller is a simple feed forward system at this point in the research but provides the appropriate framework to integrate more elaborate control schemes and EMG signal conditioning as this portion of the research area matures.