Modeling and Characterization of a Self‐Sensing Soft Hydraulic Muscle

Soft artificial muscles are increasingly important for applications in haptics, endoscopic robotics, and humanoid systems. Unlike rigid actuators, which follow deterministic motion paths, soft actuators adaptively deform under external forces, posing significant challenges for accurate control. Conventional approaches rely on external sensors to improve performance, but these add weight, cost, and integration complexity. This article introduces the self‐sensing soft hydraulic muscle (SSHM), a novel actuator that intrinsically provides real‐time feedback of both length and force without the need for external sensors. A robust theoretical model is developed to characterize SSHM behavior and accurately estimate actuation parameters. Experimental validation shows high precision, with root mean square errors of 0.3 N for force and 0.69 mm for length. The study further examines the influence of working fluids on SSHM performance. Among the tested liquids, propylene glycol is identified as optimal, doubling durability compared to water while minimizing hysteresis (0.86%). To demonstrate practical utility, SSHM is integrated into a 3D‐printed humanoid elbow joint, achieving an RMSE of 3.19° for joint angle estimation and 0.45 N for external force sensing. These results highlight SSHM as a compact, adaptive, and scalable platform, advancing soft robotics by unifying actuation and sensing in a single structure.