Design Optimization of a Soft Robotic Rehabilitation Glove Based on Finger Workspace Analysis

The finger workspace is crucial for performing various grasping tasks. Thus, various soft rehabilitation gloves have been developed to assist individuals with paralyzed hands in activities of daily living (ADLs) or rehabilitation training. However, most soft robotic glove designs are insufficient to assist with various hand postures because most of them use an underactuated mechanism for design simplicity. Therefore, this paper presents a methodology for optimizing the design of a high-degree-of-freedom soft robotic glove while not increasing the design complexity. We defined the required functional workspace of the index finger based on ten frequently used grasping postures in ADLs. The design optimization was achieved by simulating the proposed finger-robot model to obtain a comparable workspace to the functional workspace. In particular, the moment arm length for extension was optimized to facilitate the grasping of large objects (precision disk and power sphere), whereas a torque-amplifying routing design was implemented to aid the grasping of small objects (lateral pinch and thumb-two-finger pinch). The effectiveness of the optimized design was validated through testing with a stroke survivor and comparing the assistive workspace. The observed workspace demonstrated that the optimized glove design could assist with nine out of the ten targeted grasping posture functional workspaces. Furthermore, the assessment of the grasping speed and force highlighted the glove's usability for various rehabilitation activities. We also present and discuss a generalized methodology to optimize the design parameters of a soft robotic glove that uses an underactuated mechanism to assist the targeted workspace. Overall, the proposed design optimization methodology serves as a tool for developing advanced hand rehabilitation robots, as it offers insight regarding the importance of routing optimization in terms of the workspace.