5 October 2023 in Nature Machine Intelligence
16 April 2025 in IEEE Transactions on Instrumentation and Measurement
Soft actuators driven by shape memory alloy (SMA) wires are capable of undergoing a large deformation and offering a fast response, which allows them to assemble soft robots for diverse operations. However, their further applications, especially in scenarios of force control and compliant operation, are limited due to the lack of force-sensing capability. The main challenges in force sensing of SMA soft actuators include two factors: 1) Their deformable structures and relatively small volumes limit the use of rigid force sensors; 2) It is difficult to derive the loaded external force through mechanical modeling. In this work, the magnitude and the direction angle of a two-dimensional external force were predicted according to the shape information of a soft SMA planar actuator (SPA) using a force-magnitude prediction model and a force-direction prediction model, respectively. Both the two prediction models were trained by a shallow neural network based on the shape information of the SPA collected automatically through experiments. The effectiveness of the two trained models was proved through validation analysis and prediction experiments. In the magnitude range 0-400 mN (nearly four times the SPA’s weight (10.5 g)) and the direction angle range 0-60° of external force, mean absolute prediction errors of 16.56 mN and 2.06° were achieved, respectively. Finally, two application experiments employing the prediction capabilities in the magnitude and the direction angle of external forces were presented. These results proved that SPAs can utilize their deformable structures to achieve both actuation and two-dimensional force perception. This work is expected to improve the sensing capability and extend the force-related application range for SMA soft actuators.
19 September 2025 in IEEE/ASME Transactions on Mechatronics
Shape memory alloy (SMA) driven soft manipulators (SDSMs), characterized by great system integration and lightweight structure, present significant application potential in confined working environments. However, the compact structure poses an obstacle to the integration of effective force-sensing systems, limiting the development of advanced control methods considering the constraints of the tension of the SMAs. In this article, the kinematic and static models of an SDSM were established to estimate the external force only using the self-sensing shape information. Simulations and experimental results validate the feasibility of the models and the force estimation method, demonstrating accurate external load perception under different bending states. Then, a motion-planning algorithm considering static constraints was proposed to solve both the inverse kinematics of the SDSM and the tensions of the SMAs. Finally, a tension-based motion controller of the SDSM was built based on the calculated tensions and state-space equations of the SMAs. A root mean square error of 0.0145 in bending angle tracking and a tip position tracking repeatability of less than 0.4 mm underscore the effectiveness of the proposed motion controller. This work shows significant implications for enhancing the environmental interaction capabilities of SDSMs and will further broaden their application prospects.
27 January 2026 in Research Square
Robotic dexterous operation necessitates a tactile sensing system that possesses capabilities equivalent to those of the human finger, but this benchmark remains unfulfilled by current sensors. Here, we establish a frequency-separated afferent encoding and neuronal multiplexing mechanism that fundamentally breaks from the conventional image-processing paradigm of visuotactile sensors. This mechanism enables, for the first time, human finger-level tactile intelligence by seamlessly integrating the recognition of contact force, surface roughness, and slip, which is implemented in a sensor named HydroPalm. HydroPalm achieves force perception with remarkable accuracy (R² = 0.965), producing force heatmaps strikingly similar to a human finger. Moreover, it demonstrates finger-level acuity in resolving micro-scale textures with a roughness average (Ra) of 0.258 μm, far surpassing existing tactile sensors. Its millisecond-level latency (< 75 ms) allows predictive detection of incipient slip, facilitating human-like reflexive grip adjustments. The pioneering finger-level tactile perception endows the underwater robotic pipe assembly with force perception, texture recognition, and slip prediction. This establishes a foundation for integrating human-level tactile intelligence into robotic dexterous manipulation.