Modern assistive robotic control demands parallel advancements in wearable, noninvasive sensing. Electrical impedance myography (EIM)-a noninvasive, wearable technology for dynamically measuring tissue electrical impedance properties-is a promising but poorly established solution for quantifying muscle force and fatigue, given that prior approaches have relied on black-box algorithms when attempting to relate EIM to muscle physiology or have extrapolated relationships between EIM and muscle force from singular, isolated movements. Here, we elucidate direct relationships between EIM and muscle kinematics and kinetics during functional movement using biomechanical principles and traditional measures of joint torque, muscle architecture, and activation. Dual-frequency EIM measurements of the medial gastrocnemius of 10 participants were collected and compared against dynamometer, B-mode ultrasound, and electromyography measurements. Participants performed multiple categories of dynamic movements spanning constrained isometric, self-selected isometric, concentric, and eccentric contractions at multiple force levels, joint angles, and angular velocities. EIM correlated to joint torque only for specific conditions, whereas EIM correlated with muscle length change dynamics across all movements. Mixed-effects modeling and sequential feature analysis showed fascicle length and muscle activation as primary drivers of EIM variance. Last, dual-frequency EIM combined with principal components analysis reliably captured changes in fascicle length and activation during walking. These findings validate EIM as a robust, real-time biomechanical tool for sensing muscle kinematics and neuromuscular activity during functional movement, expanding its use in assistive robotic control and injury prevention.
Physiological motion in dynamic luminal organs, such as respiration-induced lung displacement, compromises the stability of tool-tissue interaction during telerobotic endoluminal imaging and interventions. However, existing devices for stabilizing the distal end of a catheter or endoscope face critical limitations: Rigid shape-locking mechanisms risk pressure-induced trauma, electrical hazard, and thermal injury, whereas balloon-based designs obstruct luminal patency. Moreover, they lack sufficient adjustable-diameter range for anatomies such as the bronchial tree. In this work, we present a reconfigurable helical brace that overcomes these challenges. Constructed of rotationally stacking tubular modules of asymmetric stiffness, the helical brace achieved up to almost 10 times diameter tunability (3.0 to 29.3 millimeters), enabling adaptive anchoring across diverse bronchial diameters. The established brace in a dynamic bronchus phantom reduced respiration-induced catheter displacement by about 96.0% while preserving more than 97.4% luminal flow conductance. Ex vivo and in vivo bronchial studies confirmed safety under prolonged radial loading, with no notable surface trauma or deep tissue injury observed. We demonstrate a helically braced telerobotic manipulator that can operate with submillimeter control accuracy in dynamic environments, enabling targeted transbronchial needle interventions in live porcine models, including drug delivery, bronchus puncture, and fluid aspiration. In addition, in vivo studies on porcine lungs showed that the brace-aided telerobotic catheter enhances the stability of optical coherence tomography imaging, effectively achieving respiratory motion-free conditions. This patency-preserving brace enhances stabilization without compromising physiological function and safety, ensuring accurate telerobotic operations in dynamic luminal environments.
Miniature wireless magnetic robots (MWMRs) hold immense potential for minimally invasive interventions. Real-time tracking of MWMRs is essential for precise in vivo navigation but remains challenging. Tracking MWMRs via their emitted magnetic signals offers a viable solution, yet it has been limited to centimeter-scale robots because isolating weak MWMR signals from overwhelming actuation-induced interference becomes increasingly difficult with miniaturization. Here, we present an artificial intelligence (AI)-driven tracking system to overcome these challenges, enabling precise localization of millimeter-scale MWMRs in the presence of actuation fields. Our approach integrates a sensor array design with a transformer to accurately model and filter the interference, followed by a spatiotemporal transformer that interprets the extracted MWMR signals over a short time window to corresponding three-dimensional coordinates. Our system has been validated in a range of in vitro and in vivo experiments, demonstrating its applicability to closed-loop navigation of MWMRs in unshielded clinical settings.
Assisting underwater locomotion with exoskeletons could reduce energy and air consumption during scuba diving.
Reinforcement learning offers a powerful framework for automatically generating control policies for robot swarms, yet its application to physical multirobot systems has been limited by the simulation-to-reality gap. Despite substantial advances in multiagent reinforcement learning, systematic demonstrations of gradient-based sim-to-real transfer to physical robot swarms remain scarce. Most work remains confined to simulation, and the few real-robot demonstrations lack evaluation across multiple runs and tasks. We show that this barrier can be overcome by embedding structural constraints into policy learning. We introduce attention-based counterfactual policy gradients for behavior selection in robot swarms (SwarmACB), a multiagent actor-critic method in which robots arbitrate among predefined behavior modules—reactive sensorimotor routines such as phototaxis or aggregation—rather than learning low-level motor commands. This architectural choice reduces the variance of learned solutions, which we argue mitigates overfitting to simulation artifacts. Through systematic evaluation on five collective missions using 20 physical e-puck robots (250 real-robot runs), we demonstrate that SwarmACB transfers successfully from simulation to reality, substantially outperforming continuous-action baselines. By progressively constraining policy architecture, perception, and model capacity, we show that smaller, structured policies transfer more reliably than large, unconstrained ones. These findings challenge the assumption that scale is necessary for successful transfer, suggesting instead that principled structural design enables reinforcement learning to produce control software for robot swarms that is both learned automatically and directly deployable in the physical world.
