Proprioceptive Sensing in Soft Robotic Arms Using Soft Strain Sensors

Soft robots actively interact with their environment, exhibiting large distributed deformations due to their structural compliance. These features make them ideal for applications in wearable robotics, medical devices, and assistive technologies. However, a major challenge in their development is integrating sensory systems that do not constrain their movements while enabling accurate estimation of the end-effector position. This work focuses on addressing this limitation in the context of a soft robotic arm. Commercial piezoresistive textile Electrolycra was identified as a suitable candidate for strain sensing. Following electromechanical characterization, Electrolycra sensors were integrated into the soft arm. Using a derived calibration curve and the constant curvature approximation, the system's performance in position tracking was evaluated. The sensing system and estimation method successfully tracked the end-effector's position with errors of 13.1 mm, 13.7 mm, and 4.8 mm in the x, y, and z coordinates, respectively, within a global reference frame during workspace operation. These results demonstrate the potential of Electrolycra-based sensing for soft robotic applications and open avenues for further development, particularly in closed-loop control systems.Clinical Relevance - A sensorized soft robot with enhanced closed-loop control can significantly impact the clinical field by enabling precise and adaptive interactions with patients, improving safety, efficacy, and personalization in rehabilitation, assistive technologies, and motor assessment.