Communication interfaces and protocols for advanced control of high-end prosthetic hands

Abstract

State-of-the-art control of hand prostheses is conventionally done by using a simple threshold technique; one output from one input. Two input signals are necessary to control the two directions of one degrees of freedom (DoF), and to switch between different predefined functions either co-contraction or a physical switch is used. However, this serial operation is unnatural and slow. The purpose of this project was to develop advanced control systems for high-end prosthetic devices with the pattern recognition platform BioPatRec to allow more sophisticated control schemes. Two systems considering three different hands were developed: One system for the MyoHand VariPlus Speed or the SensorHand Speed from Ottobock, and another for the i-limb hand from Touch Bionics. In addition to these two systems, a previously proposed framework was further developed to specify a standard on how to interface a motor driven device with BioPatRec. This framework guide developers how to define their devices in BioPatRec and guide the development of the control firmware.