Analytical and Experimental Analysis for Position Optimization of a Grasp Assistance Supernumerary Robotic Hand

The reduction of muscular strength in elderly people and patients with neurological movement disorders drastically reduces the hand autonomy for performing several daily activities. This letter presents preliminary results towards a novel supernumerary hand system, i.e., the SoftHand X, that aims to augment the grasping capabilities of people with reduced hand functionalities. The system consists of an active soft robotic hand and a passive gravity compensation mechanism. A primary issue in the design of this supernumerary limb is its placement with respect to the user's arm, due to its profound effects in terms of expected kinematics and dynamics. This letter addresses this problem with a detailed analysis based on theoretical performance indexes and experimental proofs. Starting from 16 different positions, 2 optimal positions are found with multivariate optimization and then they are experimentally evaluated by 11 healthy subjects performing standard clinical tests. Both analytical and experimental results identify these two positions as the closest to those of the human hand, assumed as the optimum. Finally, a questionnaire and subjects' suggestion reveal one of them strongly favorite.