Biological systems achieve great efficiency with tactile sensing by multiplexing receptors along shared afferent pathways, whereas soft robots, on the other hand, still rely on bulky, one-to-one sensor routing. To introduce this capability to soft-engineered systems, we present a soft optical network that utilizes input-agnostic junctions to multiplex time-of-flight signals. By connecting elastomeric waveguides into a network, we consolidate multiple sensory zones into a single shared afferent channel. We developed a curve-fitting optimization approach to deconvolve overlapping time-of-flight signals. The amplitude and arrival time of those signals convey information about the deformation state of each of the sensor paths in the robot. Testing with a prototype incorporating hard optical junctions demonstrates that the system disambiguates press and stretch deformations across four sensing zones. The prototype network produced 3 distinct signal peaks associated with the different zones. While each zone behaved differently, the system could generally detect changes at 20-gram increments of pressing, and signal peaks 1 and 3 showed sensitivity to stretching. Finally, we discuss an open-face molding approach for fabricating fully soft junctions, and describe its limitations, providing insights to guide the future development of fully soft, monolithic, proprioceptive, and exteroceptive networks.

Multi-scale biomimetic fish propulsion CFD models were established to investigate how formation arrangement affects propulsive efficiency and energy transfer. Inspired by avian V-shaped formations, several representative configurations were designed, and dynamic mesh techniques were used to simulate upstream swimming under incoming-flow conditions. By varying individual arrangement, the analysis was extended from single-fish to collective propulsion, examining how hydrodynamic interactions affect vorticity, force characteristics, power consumption, and efficiency. Asymmetric (G3) and symmetric formations (G4) were compared in terms of pressure distribution, velocity field, and vortex structures, while a larger symmetric formation (G5) was analyzed to assess school-size effects. Results show that the symmetric formation generates a stable energy-transfer pathway, improving overall propulsive efficiency by approximately 25% compared with the asymmetric formation. Increasing school size under symmetric conditions enhances the hydrodynamic advantages of symmetry but introduces additional energy.

Soft robotic grippers are well suited to handling delicate and irregular objects, yet many remain difficult to deploy in practice due to high hardware overhead (pumps, valves, motors) and continuous energy demand during holding. This study presents a sustainable multi-stable soft gripper that decouples activation from holding and reduces manufacturing waste through mandrel-based, support-free multi-material fused filament fabrication (FFF). The gripper is a thin-walled cylindrical shell patterned with a circumferential auxetic kirigami lattice composed of bistable unit cells. Sequential snap-through transitions among the graded cell groups give rise to multistable behaviour at the gripper level, while the fully closed and fully deployed states serve as the functional grasping and object-acquisition states, respectively. A magnetically responsive distal tip, allows non-contact triggering to switch the structure into an expanded, truncated-cone open state for object acquisition, followed by snap-back to the closed state for passive grasping. Mechanical characterizations quantified activation and holding performance for two gripper sizes (36 mm and 58 mm initial diameters). The small gripper, with 25g weight and up to 48 payload-to-weight ratios, reliably activated across pulling speeds from 10 to 1385 mm·min[-1], showing repeatable multi-peak snap-through behaviour and consistent activation displacement of ~20 mm, with peak activation forces of approximately 10-12 N under six-magnet loading. Object pull-out tests demonstrated stable passive holding, with maximum pull-out forces ranging from ~3-5 N for 40-48 mm circular objects (small gripper) and up to ~17 N for non-circular objects (large gripper), with higher forces often arising from passive geometric interlocking. Finite element simulations using a Mooney-Rivlin hyperelastic TPE model reproduced the snap-through sequence and instability locations, supporting interpretation of the activation mechanics. Robotic demonstrations on a 3-DoF arm further validated practical pick-and-place capability without onboard actuation or sensing.

Adaptive gait control under external perturbations requires coordinated modulation of both local muscle activation and global muscular organisation. To capture these multi-level adaptations, this study integrates co-contraction share analysis, which characterises distributed muscle contributions beyond conventional antagonist-pair co-activation indices, with Principal Component Analysis (PCA) as an analytical framework for examining Electromyographic (EMG) responses to frontal-plane angular perturbations during overground walking. EMG signals are recorded from 10 subjects on distal and proximal lower-limb muscles during the Pre-Perturbation Stride (PPS), Perturbation Stride (PS), and Subsequent Stride (SS) under four perturbation conditions: Midstance Ipsilateral (MI), Midstance Contralateral (MC), Touchdown Ipsilateral (TI), and Touchdown Contralateral (TC). Co-contraction share is quantified using summed values, peak values, and peak locations across defined gait phases, while PCA is used to characterise synergistic activation patterns. Perturbation-related modulation is observed across all strides with effect magnitudes ranging from moderate to very large. PPS generally shows more stable activation patterns. PS exhibits the strongest perturbation-induced changes, whereas SS shows either partial recovery or persistence of altered activation depending on muscle group and perturbation condition. Distal muscles demonstrate the largest and most sustained effects, particularly in co-contraction magnitude and timing, indicating a dominant role in immediate stabilisation and post-perturbation recovery. Proximal muscles exhibit earlier, but generally smaller adjustments, consistent with whole-body coordination. PCA reveals fewer significant changes than co-contraction share analysis, possibly reflecting both preservation of global coordination structure and lower sensitivity to localised perturbation effects. These findings support a hierarchically organisation of perturbation responses and demonstrate the value of combining co-contraction share analysis with PCA for characterising muscular adaptation during perturbated gait. The identified control strategies may provide useful design principles for adaptive locomotion control to perturbation in bio-inspired legged robotic and wearable assistive technologies.

Aerodynamic performance in flying animals can be controlled not only by changes in wing shape and size, but also by changes in wing motion. In flapping wings, the two wings interact aerodynamically, where the interaction depends on wingbeat phase, wing proximity, and wing attitude. Although observations of free-flying animals show that wings are relatively close during forward flight, how these interactions influence force production remain unclear. Here we used a robotic flapping wing and quantitative flow measurements to test the effect of wing interactions while varying mean wing elevation and pitch timing during upstroke-downstroke transitions. We show that both force magnitude and direction depend strongly on these parameters and their interaction. The fitted response models showed that high wing position combined with early pitching maximized vertical force, whereas low-to-mid wing position with synchronized-to-late pitching increased thrust. Efficiency metrics were highest near mid-to-high wing position combined with late pitching. In addition, we found that transition phases strongly affected thrust generation and produced substantial body pitch torques. These findings demonstrate that small kinematic adjustments can markedly alter aerodynamic performance and be used for tailless flight control. This offers mechanistic explanations for observed animal wing motions and novel strategies for controlling flapping drones.

Gecko-inspired fibrillar adhesives are promising for impact-contact applications, yet the mechanism underlying adhesion degradation after a single impact event remains unclear. Existing studies have primarily focused on micropillar buckling as a possible degradation mechanism, whereas whether additional degradation pathways exist during impact remains unresolved. To address this issue, impact and quasi-static adhesion experiments were conducted on four fibrillar adhesive units (FAUs) with identical micropillar adhesive layers but different backing layers. Comparative analysis of mechanical responses and interfacial images indicates that, for the present short-pillar FAUs, tensile loading during post-impact rebound induces local interfacial fracture, which serves as the primary degradation pathway. Micropillar buckling alone cannot explain the observed degradation. Backing-layer-related compliance and dissipation suppress rebound and reduce the driving force for interfacial fracture, thereby mitigating adhesion degradation. A reduced-order rebound-stage model based on fracture mechanics qualitatively reproduces the adhesion-degradation trend and backing-dependent response, further supporting the fracture-based interpretation. Finally, transient buckling and fracture-activation analyses reveal that the applicability of this interpretation depends jointly on micropillar aspect ratio and intrinsic adhesion energy, with buckling emerging as a potential competing mechanism as the aspect ratio increases.

Soft growing robots achieve extension through material eversion and exhibit remarkable compliance and adaptability in confined environments. However, most existing soft growing robots still rely on tethered laboratory pneumatic supplies, onboard electric compressors, or pre-charged gas reservoirs for actuation. In contrast, chemical reaction--based actuation for soft growing robots remains largely unexplored. Inspired by the internally driven tip-growth process observed in biological systems such as pollen tubes, this work presents a chemically driven soft growing robot in which gas generated on demand from liquid reactants replaces a conventional compressed-gas supply. To quantitatively characterize {the relationship between chemical gas generation and eversion-based growth, a gas-generation--pressure--growth model is developed by linking supply-dependent gas generation, pressure evolution, growth-induced cavity expansion, and eversion mechanics. The model elucidates the relationship between gas generation and robot elongation, enabling predictable regulation of growth velocity and startup time through reactant concentration and liquid-feed-rate control. Experimental results demonstrate efficient and stable growth performance. The measured growth velocity, internal pressure, and startup time agree closely with theoretical predictions, validating both the proposed actuation strategy and the developed model within the tested operating conditions. The presented framework establishes a biologically inspired approach to internally actuated soft growing robots and provides a foundation for the future development of more self-contained growing robotic systems, while the present study focuses specifically on chemically generated pneumatic actuation and growth dynamics.

An experimentally informed model of gliding barn-owl flight with prey carriage and passive wing load alleviation is developed to examine trim, static stability, and dynamic response. Baseline trim predictions agree with prior owl data and are shown to be more consistent with a minimum-sink condition than with minimum drag coefficient or maximum lift-to-drag ratio. A generalized static-stability analysis demonstrates that vertical center-of-gravity offset and drag-induced moments, including those associated with prey carriage, contribute directly to the pitching-moment balance, so that the pitching-moment slope is not determined by fore-aft center-of-gravity position alone. Prey carriage is found to drive the generalized static margin further into the unstable region and to shorten the time to double amplitude of the dominant mode. In the dynamic model, passive wing load alleviation attenuates the dominant unstable mode and shifts its character, as measured by unit-invariant modal participation factors, from a fast short-period-like response to a slower, more phugoid-like response as compliance increases. With passive alleviation present, modest proportional wing-pitch feedback is sufficient to place the system poles in the stable left-half plane. These results indicate that load carriage and passive compliance can reshape the stability properties of gliding flight, reducing the control required to maintain steady trajectories without requiring a classically stable configuration.

Flapping-wing micro aerial vehicles (FWMAVs) offer distinct advantages, including high thrust-to-weight ratios, low acoustic signatures, and superior maneuverability. Despite these benefits, the operational versatility of FWMAVs is strictly limited by the lack of robust takeoff mechanisms adaptable to diverse terrains. The inherent unsteady aerodynamics and low payload capacity of these vehicles make integrating such systems particularly challenging. This paper addresses this limitation by developing a spring-loaded, jumping-based takeoff mechanism fabricated from PLA. The viability was validated through indoor tethered experiments that characterized the mechanism's launch-phase performance, including a peak vertical interface force of 1005.5 N. For the given FWMAV configuration and the tested launch-angle range, the best measured compromise between apex height and forward velocity was provided. The integrated tests demonstrated flight-compatible ballistic launch conditions; sustained free flight and the subsequent transition to powered flapping flight were not evaluated in this study. This study offers critical insights into the dynamics of ground-to-air transitions for bio-inspired robots, demonstrating that a jump-assisted strategy can overcome the energy constraints associated with vertical liftoff. Ultimately, this system provides a viable ground-launch stage that establishes the necessary kinematic initial conditions for subsequent transition to flapping flight, significantly expanding the mission profile and environmental adaptability of FWMAVs.

This study presents a bio-inspired open knee joint for biofidelic robots with an anatomically correct range of motion. Bio-inspired joints contribute to robot performance and understanding living organisms by providing dynamic motion data as the real-world physical twin. The proposed joint structure features the integration of the open joint and ligamentous constraint. In particular, the spherical envelope of the condyles contributes to the rotation other than pitch rotation. By applying a bio-inspired ligament arrangement, the developed knee joint performed pitch and yaw motion. The relationship between the ligament length and joint orientation was derived by forward kinematics and the ligament wrapping. The arrangement of the bio-inspired collateral ligaments was determined with the iterative calculation method considering the object wrapping, which predicted the yaw range of motion. Verification of the proposed approach includes the evaluation of the range of motion on the robotic knee joint equipped with the motor-driven tendons. The robot validation tests include the rotational motions around the pitch and yaw axes. The tests showed that the proposed joint structure can be a relevant approach for replicating the range of motion of a human knee joint. Future work includes the motion control of the joint system and integration into a real robot.

Birds routinely solve high-speed, three-dimensional collision-avoidance problems that remain a central challenge for autonomous micro-aerial vehicles (MAVs). While pairwise avoidance is well studied in budgerigars, comparatively little is known about how avoidance kinematics change when more than two birds must simultaneously resolve a potential collision-the regime in which distributed control becomes attractive for MAV swarms. Using an openly available stereo-camera dataset (1v1,2v2,3v3 trials), we extracted nine kinematic indices per encounter and tested for group-size effects using dual statistical frameworks supplemented by stratified, subsampling-robustness and noise-floor sensitivity analyses. The two most robust findings are (i) that the optic-flow time-to-contact marginis qualitatively conserved across group sizes (the time-course is preserved while the absolute margins compress with group size), and (ii) that the minimum inter-bird separation does not shrink with group size and in fact grows mildly. By contrast, evidence that birds in the 3v3 condition react later, modulate forward speed less, or fly more curved paths is conditionally supported at best: the reaction-distance effect replicates in only 19.5% of subsampled iterations, and the path-complexity effects are significant only after trial-level aggregation. The agility metrics (peak lateral acceleration and minimum turn radius) are heavily confounded with tracking noise; a majority of 3v3 encounters fall below the empirical noise floor and these values should not be used as an engineering target. These observations are consistent with-but do not strongly mandate-a bioinspired design hypothesis in which a fixed-based visual policy is paired with a group-size- adaptive reaction threshold. A proof-of-concept multi-agent simulation of this controller resolves pairwise head-on encounters reliably (100% collision-free) but degrades sharply with group size and density, indicating that a purely reactive single-threat policy is insufficient for swarm-scale operation and that an explicit multi-threat coordination mechanism is required.

Humanoid musculoskeletal robots driven by McKibben muscles offer inherent compliance and safe physical interaction, making them ideal for human-centered environments. However, developing the capacity for learning and generalizing reaching movements to new spatial targets remains a complex mission due to the sophisticated coordination required between perception, decision-making, and control mechanisms. Inspired by the human motor system, we propose a brain-inspired framework for motion learning and generalization on a musculoskeletal robotic arm, integrating two key neural principles: motor primitive encoding and cerebellum-inspired predictive correction. We first learn reusable motor primitives from sparse demonstrations using a multilayer perceptron optimized via iterative imitation learning. To generalize to novel targets, we propose a cerebellum-inspired analytic motor primitive composition (CAMPC) method that integrates a differentiable forward model and muscle-synergy-aware regularization to compute optimal blending weights. Experiments show that our CAMPC method reduces mean positioning error to 2.45 cm in simulation and generalizes to hardware using human demonstrations without any reinforcement learning, thereby avoiding hazardous online policy exploration on fragile pneumatic actuators. On the physical robot, CAMPC attains the lowest mean error of 3.03 cm, outperforming other baselines (paired-tests,). This demonstrates that our brain-inspired method enables humanoid musculoskeletal robots to generalize to untrained static reaching targets with high accuracy and data efficiency within a constrained desktop workspace.

Cheetahs are characterized by large spinal flexion and extension during high-speed running, yet the dynamical role of the phase relationship between spinal motion and limb support remains unclear. We aimed to clarify how this phase relationship affects running performance, focusing on the effect of asymmetric spinal stiffness. Using a simple sagittal planar cheetah model with asymmetric stiffness of the spine undergoing extension or flexion, we numerically searched for periodic bounding solutions over a range of stiffness parameters and compared their ground reaction forces (GRFs), horizontal velocities, and stability. We obtained both cheetah-like solutions, in which the spine extends after hindlimb liftoff and flexes after forelimb liftoff, and non-cheetah-like solutions, in which the spine flexes after hindlimb liftoff and extends after forelimb liftoff. Under asymmetric spinal stiffness, cheetah-like solutions reduced GRFs while maintaining horizontal velocity more effectively than non-cheetah-like solutions. The phase relationship between spinal motion and stance timing is a key determinant of high-speed running performance. These findings provide a dynamical understanding of cheetah locomotion and suggest design principles for spined legged robots.

Spinning samaras provide a promising biological basis for small wind turbine rotors, but the engineering translation of their three-dimensional morphology and the mechanisms governing performance losses below and above the low-tip-speed-ratio (TSR) optimum remain insufficiently understood. In this study, a small horizontal-axis wind turbine rotor was developed from aµCT-informed engineering reconstruction of an Acer buergerianum samara. Its aerodynamic performance was evaluated using transient URANS simulations with the SSTmodel and a sliding mesh, supported by grid and time-step sensitivity studies and validation against NREL Phase VI torque measurements. Atand, increasing the blade number from two to five raisedfrom 0.2445 to 0.3767, whereas a sixth blade provided only a further 1.01% increase. Among the sampled pitch angles,produced the highestof approximately 0.388. Forand, a refined sweep identified a near-flat peak overand a high-efficiency interval defined byof. For the same configuration,atincreased from 0.3698 to 0.4021 asincreased fromto. Mechanistic analysis showed that the peak is sustained by a continuous positive power-contribution band over the mid-to-outer span. At low TSR, suction-side separation, near-wake recirculation, and axial-velocity deficit restrict the effective work-producing region; at high TSR, viscous resisting torque largely cancels the pressure-driven torque. Together, these findings explain the low-TSR peak and provide mechanism-based guidance for blade-number selection and TSR control of the present samara-derived rotor.

Animals with a slender body such as eels progress by propagating body waves; this swimming mode is called anguilliform. Though anguilliform swimming has been studied extensively, its optimality and gait selection principles remain unknown. Here, through model-based optimal gait analysis, we propose that three dimensionless characteristic parameters-the stride length (SL), muscle mechanical efficiency and environmental force anisotropy-shape the gait, and they respectively determine a whole or a fraction of body undulation, a more C-shaped or S-shaped body wave, and the overall body undulation amplitude and swim direction. The body(muscle)-fluid properties and the intended swim speed influence the gait via these three. Observed biological gait features, including those of young animals and the counterintuitive ones, are shown to conform, qualitatively, to these principles, suggesting that their gaits may be explained by optimality with respect to the total of fluid and muscle damping power losses, with the latter occupying a small percentage. In addition, the prevailing feature of caudally increased body curvature may be a result of an adjustment to the gait due to the tapered body geometry and/or oncoming fluid vortices. Biological gaits could thus be understood as a composition of aforementioned gait features. Also explained by optimal gaits are the negative correlation between the Strouhal number and the SL, and the wider range of the Strouhal number of anguilliform swimming (0.31-0.93) than that of carangiform swimming (0.2-0.4). The neural signals for muscle activation are also calculated from the optimal gaits to predict control strategies underlying the observed gaits. Our results benefit both the understanding of biological locomotion and the gait planning for bio-inspired underwater vehicles.