A Hyper-Adaptive, Low-Cost, Toy-Like Prosthesis with a Compact Differential Mechanism and a Vision-Based Myoelectric Control Scheme

Children with congenital limb loss undergo a different prosthetic journey than adults who have suffered from traumatic limb loss. As the children are born without a limb, the use of a prosthesis can feel foreign, creating discomfort. This results in higher mean rejection rates (45% and 35% for body-powered and electric devices, respectively) for children, as opposed to adults (26% and 23%, respectively). This work presents a hyper-adaptive, toy-like prosthetic gripper that can take the role of a training prosthesis to motivate the child to learn muscle-based control in an organic setting through play. The gripper incorporates three hyper-adaptive fingers manufactured through hybrid deposition manufacturing (HDM) and includes a gear drive system allowing two fingers bases to rotate, thereby increasing the grasping area. The fingers are actuated through a compact, series-elastic differential mechanism that reduces the total number of required actuators to two. The developed gripper is operated using a vision-based myoelectric control framework that utilizes an RGB camera and a Convolutional Neural Network (CNN) for object detection and classification as well as for grasp selection and Electromyography (EMG) for grasp triggering. The efficiency of the gripper and the control framework was validated through a series of grasping experiments executed with daily life objects.