The PRISMA Hand I: A novel underactuated design and EMG/voice-based multimodal control

Abstract

The novel underactuated design of the PRISMA HAND I and its multimodal control based on integrated electromyographic signals and voice commands is presented. Due to limitations of EMG sensors and residual physiological signals of amputees, a smart solution utilizing a voice recognition module has been integrated into the EMG control to realize advanced intuitive commands and easy connection with the arm. Inspired by neuroscience studies on the human hand, two motor synergies have been implemented. The thumb adduction/abduction independent motion allows in-hand manipulation. The transmission system is based on whiffletree and pulleys for motion differentiation and on antagonistic elastic tendons. Using EMG and voice commands, the motors are moved together in a certain number of combination suitably and easily programmed using Arduino electronic prototyping. The use of a current sensor allows controlling the strength of the grasp to handle also fragile objects. The contributions of this work are: the bio-inspired design of kinematics and motion couplings by means of two motor synergies, the simple and efficient control interface allowing easy connection to the arm and intuitive use, finally, the cost reduction using economic hardware and mechanical components while preserving performance.