Analysis of Forces Exerted by Shoulder and Elbow Fabric-based Pneumatic Actuators for Pediatric Exosuits

TL;DR

This work tackles safe and effective force generation in soft fabric-based pneumatic actuators for pediatric upper-extremity exosuits by using infant-scale rigs to quantify ROM and body contact forces under varied anchor points and locked joint angles. It compares a shoulder single-cell actuator and an elbow 10-cell bellow actuator, recording load-cell and encoder data while controlling input pressure and actuator positioning. The results show that shoulder ROM is maximized with anchor point S1 and lower peak forces, while the elbow achieves maximal ROM with a symmetric E2 configuration; joint-angle constraints and nonlinear, hysteretic force-pressure relationships emerge across configurations. These findings enable co-optimization of actuator anchoring and control policies to balance functionality and wearability, with future work focusing on dynamics-based modeling and controller integration.

Abstract

To enhance pediatric exosuit design, it is crucial to assess the actuator-generated forces. This work evaluates the contact forces exerted by soft fabric-based pneumatic actuators in an upper extremity pediatric exosuit. Two actuators were examined: a single-cell bidirectional actuator for shoulder abduction/adduction and a bellow-type actuator for elbow extension/flexion. Experiments assessed the impact of actuator anchoring points and the adjacent joint's angle on exerted forces and actuated joint range of motion (ROM). These were measured via load cells and encoders integrated into a custom infant-scale engineered apparatus with two degrees of freedom (two revolute joints). For the shoulder actuator, results show that anchoring it further from the shoulder joint center while the elbow is flexed at $90^\circ$ yields the highest ROM while minimizing the peak force exerted on the body. For the elbow actuator, anchoring it symmetrically while the shoulder joint is at $0^\circ$ optimizes actuator performance. These findings contribute a key step toward co-optimizing the considered exosuit design for functionality and wearability.

Figures (10)

• Figure 1: Experimental setup for force profiling of a fabric-based pneumatic (A) shoulder and (B) elbow actuator. The setup allows various actuator anchoring points and non-actuated joint angle settings (locked in place).
• Figure 2: (A) Schematic diagram of shoulder actuator anchoring: Two anchoring points (S1 at two-thirds and S2 at one-half of the upper arm length from the proximal end) and four fixed elbow angles ($0^{\circ}$, $30^{\circ}$, $60^{\circ}$, and $90^{\circ}$). (B) Schematic diagram of elbow actuator anchoring: Twelve configurations with three anchoring points (E1, E2, E3) and four fixed shoulder angles ($0^{\circ}$, $30^{\circ}$, $60^{\circ}$, and $90^{\circ}$). E1: Actuator cells distributed 6:4 (UA:FA), with attachment points at two-thirds (UA) and one-half (FA) of their lengths. E2: Cells distributed 1:1, with both attachment points at one-half of their lengths. E3: Cells distributed 4:6, with attachment points at one-half (UA) and four-fifths (FA) of their lengths. Actuators shown in thick (red) curves.
• Figure 3: (A) Changes in shoulder joint angle over time in the two shoulder actuator anchorings for different locked elbow angles. Shaded areas represent one standard deviation. (Best viewed in color.) (B) Boxplot comparing the shoulder joint angle between the S1 and S2 configurations.
• Figure 4: Temporal evolution of the forces exerted by the shoulder actuator on the torso (Top) and UA (Bottom) for the two shoulder actuator anchorings and different locked elbow angles.
• Figure 5: Boxplots of the exerted peak force for the shoulder actuator on the torso and UA. Results show significant performance differences owing to different actuator configurations and non-actuated locked elbow angles.