At CES 2019, a company called BrainRobotics was showing off what looked like a pretty impressive powered hand prostheses. It has individually movable fingers, which rely on eight separate EMG channels that sense remaining muscle movement for activation. Machine learning algorithms are supposedly in action within the device to regulate the hand’s action, so that movements are natural and intuitive.
Overall, the device is made from a modular design, which allows parts to be replaced easily and quickly without having to send it in to a special shop for repair. This is an important factor, as prosthetic hands undergo a great deal of wear and tear. To make the device fit ideally for every user, the socket can be 3D printed based on each patient’s stump so that as much comfort is achieved and the maximum usability can be derived from the prosthesis.