TL;DR
This paper introduces and evaluates a novel modular soft actuator design for personalized hand rehabilitation gloves. By testing against individualized dummy fingers, they demonstrate that these modular actuators offer superior performance and better adaptability to user-specific finger sizes and stiffness compared to conventional whole-finger designs, paving the way for more effective and safer rehabilitation devices.
What Problem Does This Paper Address?
Hand rehabilitation after events like strokes is crucial, but existing rigid robotic solutions are often cumbersome and lack flexibility. Soft robotic gloves (SRGs) offer a promising alternative, but their effectiveness is significantly hampered by individual user differences. Specifically, variations in finger size, joint range of motion (ROM), and joint stiffness can lead to poor actuator performance, discomfort, and even injury. Current SRG designs largely overlook these individual variations, hindering their widespread adoption in clinical practice. This paper aims to systematically investigate the impact of these individual differences on actuator performance and propose a design that addresses them.
Methods
The researchers developed and compared two types of fiber-reinforced soft actuators:
- Modular Type: This design consists of individual actuator modules, each intended for a specific finger joint (DIP, PIP, MCP). These modules are connected by 3D-printed rigid connectors. The key innovation here is the ability to customize the overall glove length and joint spacing by simply changing these connectors, allowing for adaptation to different finger sizes.
- Conventional Type: This represents a more traditional whole-finger assistive actuator, designed to cover a larger portion of the finger.
To evaluate these actuators, they created custom-designed dummy index fingers that mimicked individual differences in:
- Finger Size: Three sizes (small, medium, large) were created based on anthropometric data.
- Joint Stiffness: The MCP joint was equipped with torsion springs to simulate low, mid, and high stiffness levels, reflecting pathological conditions like hyperextension often seen in stroke patients.
Experiments involved measuring joint ROM and torque when actuators were mounted on these individualized dummy fingers. They also characterized the actuators independently for bending angle and tip force.
Datasets
The study utilized anthropometric data from the "Human Hand Dimensions Data for Ergonomic Design 2010" database for Japanese individuals to inform the design of the dummy fingers and actuator dimensions. No explicit training datasets were used, as the focus was on experimental evaluation.
Key Insights and Ideas
- Modularity is Key for Individualization: The modular design, with its interchangeable rigid connectors, offers a straightforward way to adapt the actuator system to varying finger lengths and joint configurations. This is a significant advantage over conventional designs that are fixed in length.
- Size Matching Matters Dramatically: Both modular and conventional actuators performed significantly better when their size matched the dummy finger's size. Mismatched pairs resulted in reduced ROM and torque, highlighting the critical need for personalized sizing.
- Modular Actuators Excel in Performance: The modular actuators consistently demonstrated superior bending angles and tip forces compared to their conventional counterparts, even when tested independently. This is attributed to the rigid connectors, which facilitate more efficient force transmission and limit unwanted elastic deformation.
- Stiffness Adaptation is Crucial: While ROM decreased with increasing joint stiffness for both actuator types, the modular actuators showed a greater ability to provide assistive ROM even at higher stiffness levels, making them more suitable for patients with joint contractures.
- Safety Implications of Mismatch: The analysis of force equilibrium revealed that actuator-finger mismatch can lead to off-axis forces (Facx), potentially causing undue stress and injury to finger joints over time.
Key Results
- Bending Performance: Modular actuators achieved a maximal bending angle of 92.5° (ROM) compared to 98.0° for conventional actuators when tested independently. However, the modular actuators delivered a significantly higher tip force (13.2 N vs. 6.5 N at 200 kPa).
- Joint ROM and Torque (Medium Finger): When mounted on a medium-sized dummy finger, modular actuators achieved higher joint ROMs (e.g., 99.6° for PIP) and torques (0.17 Nm for all joints) compared to conventional actuators (e.g., 74.6° for PIP, 0.09 Nm for DIP/PIP/MCP).
- Size Matching Impact: Ratios (RROM and Rtorque) indicated that matched actuator-finger pairs performed substantially better than mismatched pairs. Modular actuators showed higher RROM and Rtorque values, suggesting a greater benefit from precise size matching.
- Stiffness Impact: For a medium-sized dummy finger with mid-level stiffness, modular actuators achieved over 70° of ROM, which is considered sufficient for rehabilitation. The ROM decreased with increasing stiffness for both types, but modular actuators maintained better performance.
Limitations and Future Work
- Limited Parameter Variation: The study used only three levels for finger size and stiffness. Future work should explore finer gradations of these parameters to provide more generalized findings.
- Simplified Dummy Finger: The current dummy finger is optimized for measurement and lacks realistic anatomical details. Developing more anatomically accurate dummy fingers would allow for more nuanced investigations into individual differences.
- Theoretical Analysis: The authors suggest developing Finite Element Analysis (FEA) models to better understand the complex interactions between finger-soft actuator complexes and the effects of these interactions. This would aid in extrapolating findings to a wider range of soft actuators.
Summary updated