OpenHand is the Somatosensory companion to Open Muscle. As a few articles have aptly pointed out, the open muscle data was always designed with machine learning in mind. At its inception, the Open Muscle Project was modeled to be a part of a suite of sensors to turn muscle contractions in the forearm into commands that would control a prosthetic hand.
Open Hand is currently being developed to be low cost and to provide accurate training data for Open Muscle. The primary sensors being utilized are hall effect sensors with force resistive materials being used as supplemental data. Each finger will have two points of monitored articulation save for the thumb which will have three to four.
We aim to meet the following criteria whilst designing Open Hand:
- Fit adult hand size ranging from small to medium-large
- Have a 1 day or 16 hour operating battery life
- Low construction cost
- 3D printed structure with off the shelf sub-components