Autonomous Slip-Prevention Grip Force Control and Its Potential in Shared Control of Robotic Prosthetic Hands.

This study aimed to develop and validate a novel autonomous slip-prevention grip force controller for a robotic prosthetic hand. The controller utilized an optical flow sensor to detect object slippage events, which triggered increases in the desired gripping force within a closed-loop force control system. We first validated the slip-prevention functionality on a bench test platform. Subsequently, the controller was integrated into a simple shared control framework to demonstrate its feasibility in enhancing the grasping function for prosthesis users. In this shared control framework, myoelectric position control (i.e., user control) was employed when the hand was not engaged in a grasp, while the intelligent autonomous grip force control was activated during contact with an object. The feasibility of the shared control was tested with human participants performing tasks involving the grasping, transferring, and releasing of various objects while using and wearing the prosthetic hand. Additionally, we compared the shared control to a myoelectric-only controller (without the autonomous grip force control) in the object manipulation task combined with a secondary cognitive task. The results demonstrated that the proposed automatic grip force controller effectively prevented slips and improved grasping stability under various perturbations. Compared to myoelectric-only control, the shared control reduced the user's cognitive effort required for object manipulation and lead to increased task success. While our designed slip-prevention grip force controller shows significant potential to assist the prosthesis users in improving grasping functionality, further exploration of other shared control paradigms is necessary to maximize user performance and acceptance in the future.