Encoding Desired Postural Synergies in a Single-Actuator Soft Robotic Prosthetic Hand Through Finger Joint Stiffness Modulation

This paper presents the design and analysis of a single-actuator robotic prosthetic hand based on soft flexure joints, enabling the encoding of desired postural synergies of fingers through their joint stiffness modulation. Finger movements are coordinated using a simple whiffletree mechanism, which simultaneously pulls all tendons, with the synergy determined by the varying stiffness of the flexure joints. A forward quasistatic model is developed to predict joint angles based on tendon displacement driven by the actuator, assuming known joint stiffness values. Then, an inverse model is introduced to compute the required stiffness of finger joints for achieving a desired postural synergy. The proposed design is validated through simulations and real-world experiments on a 3Dprinted soft robotic prosthetic hand. The hand successfully demonstrates a postural synergy that produces both pinch and power grasp types. This approach provides a lightweight and mechanically simple solution for robotic prosthetic hands with single actuator, enabling desired postural synergy while maintaining functionality and adaptability for grasping tasks.