Publications

Dexterous locomotion, such as immediate direction change during fast movement or shape reconfiguration to perform diverse tasks, are essential animal survival strategies which have not been achieved in existing soft robots. Here, we present a kind of small-scale dexterous soft robot, consisting of an active dielectric elastomer artificial muscle and reconfigurable chiral-lattice foot, that enables immediate and reversible forward, backward and circular direction changes during fast movement under single voltage input. Our electric-driven soft robot with the structural design can be combined with smart materials to realize multimodal functions via shape reconfigurations under the external stimulus. We experimentally demonstrate that our dexterous soft robots can reach arbitrary points in a plane, form complex trajectories, or lower the height to pass through a narrow tunnel. The proposed structural design and shape reconfigurability may pave the way for next-generation autonomous soft robots with dexterous locomotion.

No abstract is provided for this article.

Enabling robots with contact-rich manipulation remains a pivotal challenge in robot learning, which is substantially hindered by the data collection gap, including its inefficiency and limited sensor setup. While prior work has explored handheld paradigms, their rod-based mechanical structures remain rigid and unintuitive, providing limited tactile feedback and posing challenges for human operators. Motivated by the dexterity and force feedback of human motion, we propose FreeTacMan, a human-centric and robot-free data collection system for accurate and efficient robot manipulation. Concretely, we design a wearable gripper with dual visuo-tactile sensors for data collection, which can be worn by human fingers for intuitive control. A high-precision optical tracking system is introduced to capture end-effector poses while synchronizing visual and tactile feedback simultaneously. We leverage FreeTacMan to collect a large-scale multimodal dataset, comprising over 3000k paired visual-tactile images with end-effector poses, 10k demonstration trajectories across 50 diverse contact-rich manipulation tasks. FreeTacMan achieves multiple improvements in data collection performance compared to prior works, and enables effective policy learning for contact-rich manipulation tasks with self-collected dataset. The full suite of hardware specifications and the dataset will be released to facilitate reproducibility and support research in visuo-tactile manipulation.

Artificial tactile sensing capable of measuring shear and normal forces is crucial for the diverse human‐machine interactions and dexterous robotic manipulations. However, existing soft multi‐axis force sensors usually suffer from limited detectable directions and complex coupling mechanisms, limiting their applications in realistic wearable and robotic systems. Here, a hierarchically‐interlocked, three‐axis soft iontronic sensor is presented with an asymmetrical electrode pattern that leverages the mortise‐and‐tenon structure and ultracapacitive principle to detect omnidirectional shear and normal forces. The designed sensor facilitates the decoupling process and achieves enhanced sensing performances including high accuracy, fast response ability, and mechanical robustness. Prototypical integration and application of the sensor are demonstrated to perform wearable telecontrol of virtual platforms and assist robot gripper through closed‐loop force feedback. A sensing array is further developed to construct a touch panel and identify handwriting based on the continuously measured directions of applied forces. This work may offer a potentially promising solution for the next‐generation intelligent electronic skins requiring multimodal tactile sensing information.

No abstract is provided for this article.

By 3D printing lattice structure with active materials, the structures can exhibit shape and functional changes under external stimulus. However, the programmable shape changes of the 3D printed lattice structures are limited due to the complex geometries, nonlinear behaviors of the active materials, and the diverse external stimuli. In this work, we propose a design framework combining experiments, theoretical modeling, and finite element simulations for the controllable shape changes of the 3D printed horseshoe under thermal stimulus. The theoretical model is based on a phase evolution model that combines the geometrical nonlinearity and the material nonlinearity. Results show that the shapes with positive or negative Poisson's ratio and bending intermediate shapes can be programmed by tuning the geometrical parameters and the temperature distribution. This work provides a method to aid the design of 3D printed functional lattice structures and have potential applications in soft robotics, biomedicine, and energy absorbing fields.

No abstract is provided for this article.

As a frontier of biology, mechanobiology plays an important role in tissue and biomedical engineering. It is a common sense that mechanical cues under extracellular microenvironment affect a lot in regulating the behaviors of cells such as proliferation and gene expression, etc. In such an interdisciplinary field, engineering methods like the pneumatic and motor-driven devices have been employed for years. Nevertheless, such techniques usually rely on complex structures, which cost much but not so easy to control. Dielectric elastomer actuators (DEAs) are well known as a kind of soft actuation technology, and their research prospect in biomechanical field is gradually concerned due to their properties just like large deformation (> 100%) and fast response (< 1 ms). In addition, DEAs are usually optically transparent and can be fabricated into small volume, which make them easy to cooperate with regular microscope to realize real-time dynamic imaging of cells. This paper first reviews the basic components, principle, and evaluation of DEAs and then overview some corresponding applications of DEAs for cellular mechanobiology research. We also provide a comparison between DEA-based bioreactors and current custom-built devices and share some opinions about their potential applications in the future according to widely reported results via other methods.

Tactile feedback is of great significance for amputees to improve the controllability of prosthetic hands and obtain tactile information regarding the interacting objects, which remains a significant challenge for neuroprosthetic hands. In this study, we present a method to restore finger-specific tactile sensations on the projected finger map of a unilateral forearm amputee with a sensory soft neuroprosthetic hand through electrotactile stimulation. On this basis, five soft touch sensors embedded in the fingertips are first adopted to measure the pressure changes of the soft neuroprosthetic hand with the touched objects. The measured pressure information is then accordingly encoded into electrotactile stimulation patterns to trigger an electrical stimulator that outputs programmable electrical pulses on the projected finger map of the amputee. In this manner, the finger tactile sensation can be elicited, which can help the amputee to distinguish the finger press state and discriminate the curvature and hardness of the touched objects. Experimental results show that, based on the different stimulation regions, the amputee subject can instantaneously distinguish the tactile sensation of a single finger or multiple fingers with an accuracy of 98.57% and 91.71%, respectively. By programming the frequencies of the electrical pulses, the amputee subject can successfully discriminate the touching objects with different curvatures and hardnesses with an accuracy of 97.26% and 97.93%, respectively. Finally, we demonstrate that the amputee subject can achieve closed-loop control of the sensory soft neuroprosthetic hand by integrating a myoelectric control interface and electrotactile feedback to achieve multilevel perception.