Insect Inspired Intelligence Lab (III)

Aerial–aquatic robots possess a distinctive ability to operate in air, on water, and underwater, requiring designs that function across distinct physical regimes. This review highlights insect-inspired legged and wing-driven designs, identifying core design principles that enable propulsion systems to function across air and water. Biological and robotic strategies for multimodal locomotion are examined, followed by methods for air–water transitions. In air, flapping wings generate lift and thrust through unsteady aerodynamic forces. On water surfaces, locomotion relies on surface tension support and sculling strokes. Underwater, movement is dominated by drag-based rowing or lift-based flapping. Water entry and exit involve static and dynamic strategies, coordinated with posture and stroke timing. Four key design considerations for multimodal systems are identified: leg design and coordination, body posture control, flapping mechanics, and actuation. Unified legged systems for surface and underwater motion, along with posture strategies that stabilize transitions, represent promising avenues. For winged systems, a deeper understanding of stroke plane and body orientation could enable simpler adaptable control. Challenges in actuation and autonomy also remain central to building aerial–aquatic robots. Leveraging bioinspiration offers a path for microrobots to realize their potential for tasks, such as monitoring, disaster response, and search-and-rescue.

A broad class of animals rely on touch sensation for perception. Among insects, the American cockroach (Periplaneta americana) uses a pair of soft antennae with distributed sensors to touch its environment to guide decision making. During touch, forces on the antenna can activate thousands of mechanosensors. To understand the content of this sensory information, it is critical to understand how antenna mechanics shape the transmission of contact forces. Here, we investigated the structure and mechanics of antennae of the American cockroach at the individual annulus level through experiments, modeling, imaging with micro-computed tomography (micro-CT), 3D reconstruction of antenna morphology and finite element modeling (FEM). Our data and model predictions revealed that the antenna flagellum bends according to a kinematic chain model, with annuli connected by softer joints. Whereas the middle region of the antenna consistently fractured under cyclic bending, the tip region remained intact, revealing mechanical specialization along the antenna. Micro-CT imaging revealed an invagination of the exocuticle at annulus intersections of the tip. To test the hypothesis that this structure could enhance flexibility and robustness, we used FEM and confirmed that the invagination allows for larger bending without structural failure (buckling). Applying FEM to a morphologically accurate kinematic chain model of the flagellum revealed the relationship between the local strain at the location of campaniform sensilla, predicting the information available for antenna proprioception. Taken together, these findings reveal biomechanical adaptations of insect antennae and provide a critical step toward a mechanistic understanding of touch sensation in a touch specialist.

Insect-inspired antenna dynamics enable physically intelligent robotic tactile sensing with improved efficiency and classification accuracy.

Soft robotic sensors today struggle to interpret complex tactile scenes without incurring significant computational costs. Inspired by insect antennae—soft, distributed sensors that efficiently process tactile information through physical intelligence—we investigated whether mechanical design and active touch sensing strategies could enhance robotic tactile feature perception. We hypothesized that insect-inspired antenna dynamics, specifically stiffness gradients and active touch speed, could simplify tactile classification. Using bioinspired computational and robophysical models of cockroach antennae, we introduce the notion of tactile tensors—spatiotemporal representations of tactile stimuli shaped by contact location, feature type, and active touch speed. Our analyses show that cockroach-inspired antenna mechanics and active touch speeds significantly improve feature classification accuracy compared to conventional sensors by increasing tactile data sparsity and dispersion. Through sim-to-real transfer, these principles were successfully demonstrated on a miniature distributed soft robotic antenna, validating their effectiveness in real-world robotic systems. Unlike robotic vision systems—which also use distributed sensing but cannot leverage mechanical gradients and contact dynamics—our approach achieves efficient sensing through physically intelligent, adaptive mechanics. Our work for the first time demonstrates how active movement of a mechanically tuned soft, distributed tactile sensor enhances robotic perception. Taken together, this work presents a biologically grounded framework for tactile sensor design that reduces computational load and enhances adaptability.

Laser ablation is emerging as an effective method for fabricating and modifying polymeric membranes. However, the interaction between laser energy and polymers is intricate, being governed by photothermal, photochemical, and photophysical processes. With the added complexity of porous structures, it is challenging to achieve controlled ablation of membranes. Despite many recent demonstrations, a systematic experimental study of laser–porous membrane interactions is still lacking. Here we report detailed studies of the ablation of two commonly used membranes, polyvinylidene fluoride (PVDF) and polyether sulfone (PES), using a femtosecond UV laser. The ablation rates were quantified under varying laser parameters (pulse number and fluence) and membrane characteristics (chemistry, pore size, and porosity). Under single-pulse ablation, both PVDF and PES membranes displayed pore size independent ablation rates at low fluences. At high fluences, the membranes displayed reduced ablation rates but did not exhibit the shielding effect that dominated dense polymer films. Similar trends were observed for multi-pulse ablation of the membranes. Interestingly, the multi-pulse/single-pulse ablation depth ratios showed a power-law-like dependence on the pulse number, which is analogous to the equation describing “incubation effect” observed for analyzing ablation thresholds. By leveraging these findings, we demonstrated laser ablation as a controlled in-plane sectioning methodology for characterizing the pore structures of an asymmetric PES membrane at varying depths. The results provide critical insights into the interaction between femtosecond UV laser and porous membranes, which is important for the continuous development of laser ablation as a method for membrane fabrication and characterization.

The American cockroach (Periplaneta americana) uses its soft antennae to guide decision making by extracting rich tactile information from tens of thousands of distributed mechanosensors. Although tactile sensors enable robust, autonomous perception and navigation in natural systems, replicating these capabilities in insect-scale robots remains challenging due to stringent size, weight, and power constraints that limit existing sensor technologies. To overcome these limitations, we introduce CITRAS (Cockroach Inspired Tactile Robotic Antenna Sensor), a bioinspired, multi-segmented, compliant laminate sensor with embedded capacitive angle sensors. CITRAS is compact (73.7×15.6×2.1 mm), lightweight (491 mg), and low-power (32 mW), enabling seamless integration with miniature robotic platforms. The segmented compliant structure passively bends in response to environmental stimuli, achieving accurate hinge angle measurements with maximum errors of just 0.79 degree (quasistatic bending) and 3.58 degree (dynamic bending). Experimental evaluations demonstrate CITRAS’ multifunctional tactile perception capabilities: predicting base-to-tip distances with 7.75 % error, estimating environmental gap widths with 6.73 % error, and distinguishing surface textures through differential sensor response. The future integration of this bioinspired tactile antenna in insect-scale robots addresses critical sensing gaps, promising enhanced autonomous exploration, obstacle avoidance, and environmental mapping in complex, confined environments.

Interfacing bioelectronic devices with plants can enable nature-based sensing networks through transduction of self-generated electrical signals from living plant nodes into real-time environmental data. A key challenge in developing such platforms is to create stable, biocompatible electrodes that provide sufficient adhesion to plant tissue with minimal impedance drift. In this report, an adhesive gel electrode featuring an inkjet-printed poly(3,4ethylenedioxythiophene) polystyrene sulfonate (PEDOT:PSS) trace on a poly(vinyl alcohol) (PVA) hydrogel substrate with a methyl cellulose adhesive is presented. When attached to the inside of a Venus flytrap (Dionaea muscipula) lobe these biocompatible printed conformable electrodes demonstrate long-term mechanical stability and reliable continuous recording of action potentials. Compared to commercial Ag/AgCl electrodes, the developed bioelectrodes provide higher sensitivity, faster signal dynamics, and comparable signal-to-noise ratios. The adhesive gel electrodes maintain low impedance over extended periods, which is critical for long-term monitoring bioelectronic signals. Low impedance and robust connection to the plant tissue enabled the tracking of diurnal changes. This stability enabled real-time monitoring of a Venus flytrap’s response to environmental factors such as temperature variations and insect activity. Additionally, the bioelectrodes were integrated with low-cost ESP32 microcontroller-based electronics, enabling wireless plant-to-plant communication over long distances. This system demonstrates a biologically integrated platform for developing nature-integrated networks that utilize living plants as sensor nodes.

We aim to improve the adhesion capabilities of electroadhesive pads on rough surfaces by using geometry-driven compliance to increase effective contact area. We present a kirigami-based approach for enhancing compliance through an exploration of geometric features cut into an adhesive disk. We experimentally test a range of geometries, comparing shear adhesion strength to understand structure–function relationships in our chosen parameter space. Our findings indicate that introducing cuts to form serpentine paths in a disk results in longer effective lengths and enhanced compliance, thus requiring less energy to deform into a rough surface. Leveraging this insight and associated scaling analysis, we conclude that serpentine-like features arranged in a radially symmetric wedge configuration achieve high levels of adhesion, even on rough surfaces, enabling robust adhesion relative to featureless electroadhesive disks.

Thermal bubble-driven micro-pumps are an upcoming micro-actuator technology that can be directly integrated into micro/mesofluidic channels, have no moving parts, and leverage existing mass production fabrication approaches. These micro-pumps consist of a high-power micro-resistor that boils fluid in microseconds to create a high-pressure vapor bubble which performs mechanical work. As such, these micro-pumps hold great promise for micro/mesofluidic systems such as lab-on-a-chip technologies. However, to date, no current work has studied the interaction of these micro-pumps with biofluids such as blood and protein-rich fluids. In this study, the effects of organic fouling due to egg albumin and bovine whole blood are characterized using stroboscopic high-speed imaging and a custom deep learning neural network based on transfer learning of RESNET-18. It was found that the growth of a fouling film inhibited vapor bubble formation. A new metric to quantify the extent of fouling was proposed using the decrease in vapor bubble area as a function of the number of micro-pump firing events. Fouling due to egg albumin and bovine whole blood was found to significantly degrade pump performance as well as the lifetime of thermal bubble-driven micro-pumps to less than 10⁴ firings, which may necessitate the use of protective thin film coatings to prevent the buildup of a fouling layer.

Thermal bubble-driven micro-pumps are an upcoming micro-actuator technology that can be directly integrated into micro/mesofluidic channels, have no moving parts, and leverage existing mass production fabrication approaches. As such, these micro-pumps hold great promise for micro/mesofluidic systems such as lab-on-a-chip technologies. To date, thermal bubble-driven micro-pumps have been realized discretely at the micro-scale (10’s of μm) and meso-scale (100’s of μm) which result in flow rates on the order of pL/pulse to nL/pulse respectively. However, no current work has studied how pump performance scales as a function of pump area. In this study, a systematic scaling analysis from micro to meso-scale (10-1000 μm) of thermal bubble-driven micro-pumps is performed to develop reduced parameter one-dimensional (1D) models with pump area dependency. We present, for the first time, an empirical relationship between bubble strength and resistor area across 4 orders of magnitude that generalizes the prevailing reduced parameter 1D model. Namely, q_o (A)=(1.253 x 10^-12)A [kg·m/s] which was found to fit the data with an R value of 0.96. Previously, experimental data were required to estimate the bubble strength of a particular micro/mesofluidic channel with a thermal bubble-driven micro-pump of a given area. In this work, the developed empirical relationship estimates bubble strength as a function of resistor area thus eliminating the need for experimental data to perform a first-order analysis of thermal bubble-driven micro-pumps. We envision such reduced parameter 1D models as an important first-order design tool for micro/mesofluidic designers to predict the size of the pump area needed for a desired flow rate.

Nitinol is a smart material that can be used as an actuator, a sensor, or a structural element, and has the potential to significantly enhance the capabilities of microrobots. Femtosecond laser technology can be used to process nitinol while avoiding heat-affected zones (HAZ), thus retaining superelastic properties. In this work, we manufacture living hinges of arbitrary cross sections from nitinol using a femtosecond laser micromachining process. We first determined the laser cutting parameters, 4.1Jcm⁻² fluence with 5 passes for 5μm ablation, by varying laser power level and number of passes. Next, we modeled the hinges using an analytical model as well as creating an Abaqus finite element method, and showed the accuracy of the models by comparing them to the torque produced by eight different hinges, four with a rectangular cross section and four with an elliptic notch cross section. Finally, we manufactured a prototype miniature device to illustrate the usefulness of these nitinol hinges: a piezoelectric actuated robotic wing mechanism, and we characterized its performance.

Whether walking, running, slithering, or flying, organisms display a remarkable ability to move through complex and uncertain environments. In particular, animals have evolved to cope with a host of uncertainties—both of internal and external origin—to maintain adequate performance in an ever-changing world. In this review, we present mathematical methods in engineering to highlight emerging principles of robust and adaptive control of organismal locomotion. Specifically, by drawing on the mathematical framework of control theory, we decompose the robust and adaptive hierarchical structure of locomotor control. We show how this decomposition along the robust–adaptive axis provides testable hypotheses to classify behavioral outcomes to perturbations. With a focus on studies in non-human animals, we contextualize recent findings along the robust–adaptive axis by emphasizing two broad classes of behaviors: (1) compensation to appendage loss and (2) image stabilization and fixation. Next, we attempt to map robust and adaptive control of locomotion across some animal groups and existing bio-inspired robots. Finally, we highlight exciting future directions and interdisciplinary collaborations that are needed to unravel principles of robust and adaptive locomotion.

Miniature-legged robots are constrained by their onboard computation and control, thus motivating the need for simple, first-principles-based geometric models that connect periodic actuation or gaits (a universal robot control paradigm) to the induced average locomotion. In this paper, we develop a modulable two-beat gait design framework for sprawled planar quadrupedal systems under the no-slip using tools from geometric mechanics. We reduce standard two-beat gaits into unique subgaits in mutually exclusive shape subspaces. Subgaits are characterized by a locomotive stance phase when limbs are in ground contact and a non-locomotive, instantaneous swing phase where the limbs are reset without contact. During the stance phase, the contacting limbs form a four-bar mechanism. To analyze the ensuing locomotion, we develop the following tools: (a) a vector field to generate nonslip actuation, (b) the kinematics of a four-bar mechanism as a local connection, and (c) stratified panels that combine the kinematics and constrained actuation to encode the net change in the system’s position generated by a stance-swing subgait cycle. Decoupled subgaits are then designed independently using flows on the shape-change basis and are combined with appropriate phasing to produce a two-beat gait. Further, we introduce “scaling” and “sliding” control inputs to continuously modulate the global trajectories of the quadrupedal system in gait time through which we demonstrate cycle-average speed, direction, and steering control using the control inputs. Thus, this framework has the potential to create uncomplicated open-loop gait plans or gain schedules for robots with limited resources, bringing them closer to achieving autonomous control.

Legged robot control has improved in recent years with the rise of deep reinforcement learning, however, much of the underlying neural mechanisms remain difficult to interpret. Our aim is to leverage bio-inspired methods from computational neuroscience to better understand the neural activity of robust robot locomotion controllers. Similar to past work, we observe that terrain-based curriculum learning improves agent stability. We study the biomechanical responses and neural activity within our neural network controller by simultaneously pairing physical disturbances with targeted neural ablations. We identify an agile hip reflex that enables the robot to regain its balance and recover from lateral perturbations. Model gradients are employed to quantify the relative degree that various sensory feedback channels drive this reflexive behavior. We also find recurrent dynamics are implicated in robust behavior, and utilize sampling-based ablation methods to identify these key neurons. Our framework combines model-based and sampling-based methods for drawing causal relationships between neural network activity and robust embodied robot behavior.

Most biosensing techniques require complex processing steps that generate prolonged workflows and introduce potential points of error. Here, we report an acoustic pipette to purify and label biomarkers in 70 minutes. A key aspect of this technology is the use of functional negative acoustic contrast particles (fNACPs), which display biorecognition motifs for the specific capture of biomarkers from whole blood. Because of their large size and compressibility, the fNACPs robustly trap along the pressure antinodes of a standing wave and separate from blood components in under 60 seconds with >99% efficiency. fNACPs are subsequently fluorescently labeled in the pipette and are analyzed by both a custom, portable fluorimeter and flow cytometer. We demonstrate the detection of anti-ovalbumin antibodies from blood at picomolar levels (35 to 60 pM) with integrated controls showing minimal nonspecific adsorption. Overall, this system offers a simple and versatile approach for the rapid, sensitive, and specific capture of biomolecules.

Nitinol is a smart material that can be used as an actuator, a sensor, or a structural element, and has the potential to significantly enhance the capabilities of microrobots. Femtosecond laser technology can be used to process nitinol while avoiding heat-affected zones (HAZ), thus retaining superelastic properties. In this work, we manufacture living hinges of arbitrary cross sections from nitinol using a femtosecond laser micromachining process. We first determined the laser cutting parameters, 4.1Jcm⁻² fluence with 5 passes for 5μm ablation, by varying laser power level and number of passes. Next, we modeled the hinges using an analytical model as well as creating an Abaqus finite element method, and showed the accuracy of the models by comparing them to the torque produced by eight different hinges, four with a rectangular cross section and four with an elliptic notch cross section. Finally, we manufactured a prototype miniature device to illustrate the usefulness of these nitinol hinges: a piezoelectric actuated robotic wing mechanism, and we characterized its performance.

While animals swim, crawl, walk, and fly with apparent ease, building robots capable of robust locomotion remains a significant challenge. In this review, we draw attention to mechanosensation-the sensing of mechanical forces generated within and outside the body-as a key sense that enables robust locomotion in animals. We discuss differences between mechanosensation in animals and current robots with respect to (1) the encoding properties and distribution of mechanosensors and (2) the integration and regulation of mechanosensory feedback. We argue that robotics would benefit greatly from a detailed understanding of these aspects in animals. To that end, we highlight promising experimental and engineering approaches to study mechanosensation, emphasizing the mutual benefits for biologists and engineers that emerge from moving forward together.

Miniature robots provide unprecedented access to confined environments and show potential for applications such as search-and-rescue and high-value asset inspection. The capability of body deformation further enhances the reachability of these small robots in cluttered terrains similar to those of insects and soft arthropods. Motivated by this concept, compliant legged articulated robotic insect (CLARI), an insect-scale 2.59 g quadrupedal robot capable of body deformation, is presented. The robot, currently, with tethered electrical connections for power and control is manufactured using laminate fabrication and assembled using origami pop-up techniques. To enable locomotion in multiple shape configurations, a novel body architecture comprising modular, actuated leg mechanisms, is designed. CLARI has eight independently actuated degrees of freedom driven by custom piezoelectric actuators, making it mechanically dextrous. Herein, open-loop robot locomotion at multiple stride frequencies (1–10 Hz) is characterized using multiple gaits (trot, walk, etc.) in three different fixed body shapes (long, symmetric, wide) and the robot’s capabilities are illustrated. Finally, preliminary results of CLARI locomoting with a compliant body in open terrain and through a laterally constrained gap, a novel capability for legged robots, is demonstrated. These results represent the first step toward achieving effective cluttered terrain navigation with adaptable compliant robots in real-world environments.

Discrete and periodic contact switching is a key characteristic of steady-state legged locomotion. This letter introduces a framework for modeling and analyzing this contact-switching behavior through the framework of geometric mechanics on a toy robot model that can make continuous limb swings and discrete contact switches. The kinematics of this model form a hybrid shape-space and by extending the generalized Stokes’ theorem to compute discrete curvature functions called stratified panels, we determine average locomotion generated by gaits spanning multiple contact modes. Using this tool, we also demonstrate the ability to optimize gaits based on the system’s locomotion constraints and perform gait reduction on a complex gait spanning multiple contact modes to highlight the method’s scalability to multilegged systems.

Thermal bubble-driven micro-pumps are an upcoming micro-actuator technology that can be directly integrated into micro/mesofluidic channels, have no moving parts, and leverage existing mass production fabrication approaches. As such, these micro-pumps hold great promise for micro/mesofluidic systems such as lab-on-a-chip technologies. To date, thermal bubble-driven micro-pumps have been realized discretely at the micro-scale (10’s of μm) and meso-scale (100’s of μm) which result in flow rates on the order of pL/pulse to nL/pulse respectively. However, no current work has studied how pump performance scales as a function of pump area. In this study, a systematic scaling analysis from micro to meso-scale (10-1000 μm) of thermal bubble-driven micro-pumps is performed to develop reduced parameter one-dimensional (1D) models with pump area dependency. We present, for the first time, an empirical relationship between bubble strength and resistor area across 4 orders of magnitude that generalizes the prevailing reduced parameter 1D model. Namely, q_o (A)=(1.253 x 10^-12)A [kg·m/s] which was found to fit the data with an R value of 0.96. Previously, experimental data were required to estimate the bubble strength of a particular micro/mesofluidic channel with a thermal bubble-driven micro-pump of a given area. In this work, the developed empirical relationship estimates bubble strength as a function of resistor area thus eliminating the need for experimental data to perform a first-order analysis of thermal bubble-driven micro-pumps. We envision such reduced parameter 1D models as an important first-order design tool for micro/mesofluidic designers to predict the size of the pump area needed for a desired flow rate.

Soft compliant microrobots have the potential to deliver significant societal impact when deployed in applications such as search and rescue. In this research we present mCLARI, a body compliant quadrupedal microrobot of 20mm neutral body length and 0.97g, improving on its larger predecessor, CLARI. This robot has four independently actuated leg modules with 2 degrees of freedom, each driven by piezoelectric actuators. The legs are interconnected in a closed kinematic chain via passive body joints, enabling passive body compliance for shape adaptation to external constraints. Despite scaling its larger predecessor down to 60 % in length and 38% in mass, mCLARI maintains 80% of the actuation power to achieve high agility. Additionally, we demonstrate the new capability of passively shape-morphing mCLARI – omnidirectional laterally confined locomotion – and experimentally quantify its running performance achieving a new unconstrained top speed of ~3 bodylengths/s (60 mms-1). Leveraging passive body compliance, mCLARI can navigate through narrow spaces with a body compression ratio of up to 1.5 × the neutral body shape.

Recurrent neural network-based reinforcement learning systems are capable of complex motor control tasks such as locomotion and manipulation, however, much of their underlying mechanisms still remain difficult to interpret. Our aim is to leverage computational neuroscience methodologies to understanding the population-level activity of robust robot locomotion controllers. Our investigation begins by analyzing topological structure, discovering that fragile controllers have a higher number of fixed points with unstable directions, resulting in poorer balance when instructed to stand in place. Next, we analyze the forced response of the system by applying targeted neural perturbations along directions of dominant population-level activity. We find evidence that recurrent state dynamics are structured and low-dimensional during walking, which aligns with primate studies. Additionally, when recurrent states are perturbed to zero, fragile agents continue to walk, which is indicative of a stronger reliance on sensory input and weaker recurrence.

In motor neuroscience, artificial recurrent neural networks models often complement animal studies. However, most modeling efforts are limited to data-fitting, and the few that examine virtual embodied agents in a reinforcement learning context, do not draw direct comparisons to their biological counterparts. Our study addressing this gap, by uncovering structured neural activity of a virtual robot performing legged locomotion that directly support experimental findings of primate walking and cycling. We find that embodied agents trained to walk exhibit smooth dynamics that avoid tangling — or opposing neural trajectories in neighboring neural space — a core principle in computational neuroscience. Specifically, across a wide suite of gaits, the agent displays neural trajectories in the recurrent layers are less tangled than those in the input-driven actuation layers. To better interpret the neural separation of these elliptical-shaped trajectories, we identify speed axes that maximizes variance of mean activity across different forward, lateral, and rotational speed conditions.

Several species of male fireflies display a dazzling ability to synchronize their flash timings when interacting in mating swarms. When individuals are kept in isolated tents, similar synchronization can be induced via exposure to a fixed-frequency LED near the natural frequency of the individual firefly. This is a closed-loop interaction in which the firefly responds to the artificial light signal, but the light signal does not respond to the firefly. Here we explore trials of the first-ever open-loop communication system between fireflies and LEDs, where the behavior of the firefly and the LED both influence each other. Using applied computer vision, we enable a Raspberry Pi connected to a camera and a LED to communicate with captive fireflies via mathematical models of synchronization. We present details of the implementation and preliminary results from a pilot field season introducing the system to two species of fireflies. Finally, we explore applications of such a system in the context of perturbing natural fireflies and inducing behavioral changes in swarms.

Thermal bubble-driven micro-pumps are an upcoming actuation technology that can be directly integrated into micro/mesofluidic channels to displace fluid without any moving parts. These pumps consist of high power micro-resistors, which we term thermal micro-pump (TMP) resistors, that locally boil fluid at the resistor surface in microseconds creating a vapor bubble to perform mechanical work. Conventional fabrication approaches of thermal bubble-driven micro-pumps and associated microfluidics have utilized semiconductor micro-fabrication techniques requiring expensive tooling with long turn around times on the order of weeks to months. In this study, we present a low-cost approach to rapidly fabricate and test thermal bubble-driven micro-pumps with associated microfluidics utilizing commercial substrates (indium tin oxide, ITO, and fluorine doped tin oxide, FTO, coated glass) and tooling (laser cutter). The presented fabrication approach greatly reduces the turn around time from weeks/months for conventional micro-fabrication to a matter of hours/days allowing acceleration of thermal bubble-driven micro-pump research and development (R&D) learning cycles.

To survive during colony reproduction, bees create dense clusters of thousands of suspended individuals. How does this swarm, which is orders of magnitude larger than the size of an individual, maintain mechanical stability? We hypothesize that the internal structure in the bulk of the swarm, about which there is little prior information, plays a key role in mechanical stability. Here, we provide the first-ever 3D reconstructions of the positions of the bees in the bulk of the swarm using x-ray computed tomography. We find that the mass of bees in a layer decreases with distance from the attachment surface. By quantifying the distribution of bees within swarms varying in size (made up of 4000–10,000 bees), we find that the same power law governs the smallest and largest swarms, with the weight supported by each layer scaling with the mass of each layer to the ≈1.5 power. This arrangement ensures that each layer exerts the same fraction of its total strength, and on average a bee supports a lower weight than its maximum grip strength. This illustrates the extension of the scaling law relating weight to strength of single organisms to the weight distribution within a superorganism made up of thousands of individuals.

The distributed compliance of soft-legged robots enables them to explore complex environments using novel gaits with high levels of safety and adaptability. However, this makes designing, controlling, and motion planning for such systems often challenging due to a large number of unactuated/underactuated degrees of freedom. To address this issue, current solutions often include constraining their motion to a few predefined modes by design and typically require complicated data-driven system identification and controller design techniques. In this work, we present two geometric motion planning frameworks for an autonomous, discretely soft, closed kinematic chain articulated, insect-sized quadrupedal robot with electroadhesive feet – CLARI (Compliant Legged Articulated Robotic Insect). Inspired by geometric mechanics-based gait generation for serpentine robots, we develop ‘gait controllers’ for in-plane, `shape-shifting locomotion of CLARI. The ability to deliberately and precisely control the robot-ground interaction allows us to implement sequential ‘electroadhered’ strides of leg-pairs rendering the locomotion problem kinematic with exact solutions for robot configurations, and global trajectories. We expect this approach to generalize to n-linked robots with m (<n) legs.

The western honey bee (Apis mellifera) is a domesticated pollinator famous for living in highly social colonies. In the spring, thousands of worker bees and a queen self-assemble into a swarm that hangs from a tree branch for several days while worker bees scout for a new hive location. How bees are arranged within the swarm to distribute forces and optimize internal conditions is, so far, not well understood, since the majority of bees are hidden within the swarm. We reconstruct the non-isotropic arrangement of worker bees inside swarms made up of 3000 – 10000 bees using x-ray computed tomography. We find that the structure of the swarm is non-homogeneous. Some bees form stationary layers near the attachment board and scaffold-like chains throughout the swarm. The cross-sectional area of the swarm and packing density of bees within it distributes the weight such that the top layer of bees supports the most weight without overloading individual bees. The remaining bees use the chains as pathways to walk around the swarm, potentially to feed the queen or communicate with one another. This internal structure allows the swarm to protect the queen and adapt to the changing environment, and leaves scout bees free to investigate potential new hives.

Animals such as mice, cockroaches and spiders have the remarkable ability to manoeuvre through challenging cluttered natural terrain. These abilities have been inspiration for legged robotic systems. Recent research indicates that body reorientation along pathways of minimal energy is a key factor influencing such locomotion. We propose to extend this idea with hypothesizing body compliance of soft bodied animals and robots as an alternate yet equally effective strategy to squeeze through cluttered obstacles. We have developed an insect-scale, origami-based quadrupedal robot, CLARI. The robot has body compliance through exoskeletal morphology. Using a carbon fibre multilayer laminate laser micromachining technique and piezoelectric actuators, four independently two degree of freedom legs are controlled. Additional electro adhesive feet allow for controlled motion across non-level surfaces. With the compliance of the exoskeleton CLARI can passively conform to its environment. We aim to verify our hypothesis by constructing a series of insect scale robots with varying body compliance and experimentally determining the preference for body deformation vs reorientation in a given cluttered environment for identical feedforward control commands.

Bioinspired insect sized robots can not only emulate the remarkable locomotory capabilities of their animal counterparts, but can also serve as effective ”at scale” robo-physical models for discovering novel biomechanical principles and validating their underlying physical mechanisms. The limited applicability of readily available off-the-shelf technologies – which enable sensing, control and power autonomy at this scale – is a major factor limiting progress of insect scale robo-physical models and experiments. In this work, we have devised modular and scalable solutions for our insect scale robo-physical model CLARI (compliant legged articulated robot insect). We demonstrate long term untethered operation through wireless energy harvesting and distributed control using custom high voltage smart piezoactuator driver modules. With these developments, we hope to use CLARI to test hypotheses about terradynamic streamlining and energy landscape optimization for locomotion in cluttered terrain, while contributing technologies that foster accelerated progress in the interdisciplinary fields of physics, biology and engineering through open source designs and documentation.

Flies maintain flight even after losing half of a single wing, a feat with no current analogue in robotics. Compensation is achieved through changes in the motion of the intact and damaged wing. However, the impact of wing damage on performance, and the neuromechanical strategies used to compensate for wing damage, are not well understood. By studying intact and injured flies inside a virtual reality arena, we quantified the impact of wing damage on gaze stabilization. Using system identification techniques, we found that wing damage subtly reduced flight performance. By combining flight data with a robophysical model, we discovered that damage compensation is driven by both active and passive mechanics. The robophysical model and flies exhibited a passive increase in wing amplitude and flapping frequency with increasing wing damage. However, flies decreased the amplitude of the intact wing and shifted the abdomen towards the intact wing, suggesting active control. Using control theory, we show that compensation to wing damage is achieved by active changes in internal gains that trade off stability and performance. Studying the response of flying insects to wing injury can inform the design of flapping-wing robots that maintain function following damage.

This material is based upon work supported by the Air Force Office of Scientific Research under Award Number FA9550-20-1-0084 to JMM.

The transport and distribution of particles in the human vascular network play several major roles in physiological phenomena in health and disease. Tracking particles in vasculature in vivo remains a methodological challenge, while in silico modeling involves key assumptions and limitations regarding particle size and flow environment. Here we discuss the development of a physiologically realistic in vitro benchtop flow-loop to track the transport and distribution of particles across human vasculature. Specifically, we illustrate a study on tracking representative embolic particles through a 3D printed phantom of the human common carotid bifurcation and summarize the trends in resulting particle distribution. We will discuss observations on anatomically representative particles driven at different flow speeds and with different temporal wave forms through: (a) idealized phantom models; and (b) anatomically accurate arterial phantom models; and demonstrate how these factors affect the distribution of embolic particles through the common carotid artery. The in vitro design and results are compared and contrasted against an in silico model for embolus transport across the same phantom models, to assess the functionality and accuracy of our approach.

Physical injury often impairs mobility, which can have dire consequences for survival in animals. Revealing mechanisms of robust biological intelligence to prevent system failure can provide critical insights into how complex brains generate adaptive movement and inspiration to design fault-tolerant robots. For flying animals, physical injury to a wing can have severe consequences, as flight is inherently unstable. Using a virtual reality flight arena, we studied how flying fruit flies compensate for damage to one wing. By combining experimental and mathematical methods, we show that flies compensate for wing damage by corrective wing movement modulated by closed-loop sensing and robust mechanics. Injured flies actively increase damping and, in doing so, modestly decrease flight performance but fly as stably as uninjured flies. Quantifying responses to injury can uncover the flexibility and robustness of biological systems while informing the development of bio-inspired fault-tolerant strategies.

Flies compensate for wing injury by intelligent control and benefit from a robust mechanical design.

The impressive locomotion and manipulation capabilities of spiders have led to a host of bioinspired robotic designs aiming to reproduce their functionalities; however, current actuation mechanisms are deficient in either speed, force output, displacement, or efficiency. Here—using inspiration from the hydraulic mechanism used in spider legs—soft-actuated joints are developed that use electrostatic forces to locally pressurize a hydraulic fluid, and cause flexion of a segmented structure. The result is a lightweight, low-profile articulating mechanism capable of fast operation, high forces, and large displacement; these devices are termed spider-inspired electrohydraulic soft-actuated (SES) joints. SES joints with rotation angles up to 70°, blocked torques up to 70 mN m, and specific torques up to 21 N m kg⁻¹ are demonstrated. SES joints demonstrate high speed operation, with measured roll-off frequencies up to 24 Hz and specific power as high as 230 W kg⁻¹—similar to human muscle. The versatility of these devices is illustrated by combining SES joints to create a bidirectional joint, an artificial limb with independently addressable joints, and a compliant gripper. The lightweight, low-profile design, and high performance of these devices, makes them well-suited toward the development of articulating robotic systems that can rapidly maneuver.

The remarkable ability of animals, such as mice, cockroaches and spiders, to manoeuvre through challenging cluttered natural terrain has been a primary inspiration for legged robots. Recent research indicates that body reorientation along pathways of minimal energy is a key factor influencing such locomotion. We propose to extend this idea by hypothesizing that soft bodied animals and robots could employ an alternate yet equally effective strategy relying on their distributed body compliance to squeeze through cluttered obstacles. To demonstrate the same, we have developed a palm-scale, origami-based hexapedal robot, Compliant Legged Articulated Robotic Insect (CLARI) with compliant exoskeletal morphology. The robot, fabricated using a multilayer laminate laser micromachining technique, has six independently actuated two degree of freedom legs. Using this strategy, CLARI is able to passively conform to its environment and move through both horizontally and vertically confined spaces and requires only simple leg mechanics. We aim to verify our hypothesis by constructing a series of robots with varying body compliance and experimentally determining the preference for body deformation vs reorientation in a given cluttered environment for identical feedforward control commands.

Sub-centimeter robotic systems have traditionally used piezoelectric bending actuators to actuate their various degrees of freedom due to their high-bandwidth and ease of control. However, they require high-voltage power electronics to operate, complex transmission mechanisms to amplify motion and thus, have a reduced force output limiting their ready implementation for actuating high-dimensional autonomous systems such as the limb of a legged robot to replicate animal-like dexterity. Voice-coil architecture based electromagnetic motors serve as attractive alternatives. Our origami-inspired, unidirectional, multi-scale fabrication technique leverages carbon fiber laminate and flexure based Sarrus linkage to function as both the housing and transmission unit, and when integrated into a single robot leg mechanism displays long-stroke high-force output and low-voltage operation. Preliminary modeling and experimental results show that our prototype (1.4cm in dimension), using a 3.175mm cubical N52 magnet produces about 50 times its body weight as static Lorenz force on a 41AWG, 46-turn, 4.75 mm diameter tightly wound coil with a constant input current of 0.1 A. Using impedance control, we propose to demonstrate various single-legged high-frequency hopping behaviors.

Climbing robots have applications like search and rescue, surveillance, and inspection. To execute such tasks and successfully navigate their complex environments, inclined, vertical, and inverted climbing is required. To address this challenge, we have developed Microelectroadhesive Treaded Robot (μETR), an origami-based, lightweight, small sized autonomous system. μETR’s key innovations include a seamless electroadhesive tread design utilizing screen printing and custom laser micromaching that adheres it to conductive surfaces, in addition to consisting of a drive alignment system, a tensioner/voltage transferer, crowned wheels, and a rigid chassis. Preliminary experiments suggest that the robot can climb 30 degree slopes, tug loads up to 7 grams, and statically adhere to angles up to approximately 70 degrees. We are currently integrating a custom, lightweight powerboard to make it untethered, as well as demonstrate vertical and inverted climbing on conductive surfaces. In the near future, an interdigitated tread design will be explored to experiment with non-conductive surfaces.

Rather than depending on material composition to primarily dictate performance metrics, metamaterials can leverage geometry to achieve specific properties of interest. For example, reconfigurable metamaterials have enabled programmable shape transformations, tunable mechanical properties, and energy absorption. While several methods exist to fabricate such structures, they often place severe restrictions on manufacturing materials, or require significant manual assembly. Moreover, these arrays are typically composed of unit cells that are either macro-scale or micro-scale in dimension. Here, the fabrication gap is bridged, and laminate manufacturing is used to develop a method for designing reconfigurable metamaterials at the millimeter-scale, that is compatible with a wide range of materials, and that requires minimal manual assembly. In addition to showing the versatility of this fabrication method, how the use of laminate manufacturing affects the behavior of these multi-component arrays is also characterized. To this end, a numerical model that captures the deformations exhibited by the structures is developed, and an analytic model that predicts the strain of the structure under compressive stress is built. Overall, this approach can be leveraged to develop millimeter-scale metamaterials for applications that require reconfigurable materials, such as in the design of tunable acoustics, photonic waveguides, and electromagnetic devices.

Fabrication of high-performance actuators is often a bottle-neck in the creation of microrobots. In this work, we present a new approach to piezoelectric bimorph actuator fabrication that simplifies the process (in part by reducing the number of materials used) without sacrificing actuator performance. We fabricate PZT bimorph actuators with two active layers, as well as bimorphs with four active layers that are under 2 mm wide – these are more challenging to fabricate than the two-layer bimorphs and thus present a useful test case for our new process. The blocked force and free deflection of the actuators agrees well with expected performance, with energy densities of approximately 1.2 J/kg at electric fields of 1.67 V/µm. We expect this process to enable rapid fabrication of piezoelectric actuators with more layers and at smaller scales than was previously attainable.

The ability to efficiently and precisely manipulate objects in inaccessible environments is becoming an essential requirement for many applications of mobile robots, particularly at small sizes. Here, we propose and implement a mobile micromanipulation solution using a piezoelectric microgripper integrated into a dexterous robot, HAMR (the Harvard Ambulatory MicroRobot), that has a size of approximately 4.5 cm by 4 cm by 2.3 cm and a maximum payload of approximately 3 g. Our 100 mg miniature gripper is composed of recurve piezoelectric actuators that produce parallel jaw motions (stroke of 236 ± 32 μm at 200 V) while providing high gripping forces (blocked force of 0.575 N at 200 V), making it effective for micromanipulation applications with tiny objects. Using this gripper, we successfully demonstrated a grasping and lifting task with an object of 1.3 g and thickness of 250 μm at an operating voltage of 100 V. Finally, by taking advantage of the locomotion capabilities of HAMR, we demonstrate mobile manipulation by changing the position and orientation of small objects weighing up to 2.8 g controlled by the movement of the robot. We expect that the addition of this novel manipulation capability will increase the effectiveness of such miniature robots for accomplishing real-world tasks.

Here we present HAMR-Jr, a 22.5mm, 320mg quadrupedal microrobot. With eight independently actuated degrees of freedom, HAMR-Jr is, to our knowledge, the most mechanically dexterous legged robot at its scale and is capable of high-speed locomotion (13.91bodylengthss-1) at a variety of stride frequencies (1-200Hz) using multiple gaits. We achieved this using a design and fabrication process that is flexible, allowing scaling with minimum changes to our workflow. We further characterized HAMR-Jr’s open-loop locomotion and compared it with the larger scale HAMR-VI microrobot to demonstrate the effectiveness of scaling laws in predicting running performance.

The effects of scaling on locomotion performance has fascinated biologists, physicists and engineers alike. In particular, roboticists have exploited dynamic scaling to build systems at sizes varying from a tens of centimeters to several meters. Thanks to recent advances in manufacturing, insect-scale robotics is currently an exciting research direction and holds the promise of tremendous impact in areas of search-and-rescue and high-value asset inspection. However, most robots at this scale have simplified morphology and can only demonstrate basic mobility. Here, we present the newly designed HAMR-Jr, a 22.5mm, 320mg quadrupedal microrobot. With eight independently actuated degrees of freedom, HAMR-Jr is, to our knowledge, the most mechanically dexterous legged robot at its scale and is capable of high-speed locomotion (13.91 bodylengths/s) at a variety of stride frequency (1-200Hz) using multiple gaits. We achieved this using a design and fabrication process that is flexible, allowing us to exploit and implement the physics of scaling with minimum changes to our workflow. We further characterized HAMR-Jr’s open-loop locomotion and compared it with the larger scale HAMR-VI microrobot to demonstrate the effectiveness of scaling laws in predicting running performance.

Limitations in actuation, sensing, and computation have forced small legged robots to rely on carefully tuned, mechanically mediated leg trajectories for effective locomotion. Recent advances in manufacturing, however, have enabled in such robots the ability for operation at multiple stride frequencies using multi-degree-of-freedom leg trajectories. Proprioceptive sensing and control is key to extending the capabilities of these robots to a broad range of operating conditions. In this work, we use concomitant sensing for piezoelectric actuation with a computationally efficient framework for estimation and control of leg trajectories on a quadrupedal microrobot. We demonstrate accurate position estimation (<16 root-mean-square error) and control (<16 root-mean-square tracking error) during locomotion across a wide range of stride frequencies (10 Hz–50 Hz). This capability enables the exploration of two bioinspired parametric leg trajectories designed to reduce leg slip and increase locomotion performance (e.g. speed, cost-of-transport (COT), etc). Using this approach, we demonstrate high performance locomotion at stride frequencies (10 Hz–30 Hz) where the robot’s natural dynamics result in poor open-loop locomotion. Furthermore, we validate the biological hypotheses that inspired the trajectories and identify regions of highly dynamic locomotion, low COT (3.33), and minimal leg slippage (<10%).

The ability to climb greatly increases the reachable workspace of terrestrial robots, improving their utility for inspection and exploration tasks. This is particularly desirable for small (millimeter-scale) legged robots operating in confined environments. This paper presents a 1.48-gram and 4.5-centimeter-long tethered quadrupedal microrobot, the Harvard Ambulatory MicroRobot with Electroadhesion (HAMR-E). The design of HAMR-E enables precise leg motions and voltage-controlled electroadhesion for repeatable and reliable climbing of inverted and vertical surfaces. The innovations that enable this behavior are an integrated leg structure with electroadhesive pads and passive alignment ankles and a parametric tripedal crawling gait. At a relatively low adhesion voltage of 250 volts, HAMR-E achieves speeds up to 1.2 (4.6) millimeters per second and can ambulate for a maximum of 215 (162) steps during vertical (inverted) locomotion. Furthermore, HAMR-E still retains the ability for high-speed locomotion at 140 millimeters per second on horizontal surfaces. As a demonstration of its potential for industrial applications, such as in situ inspection of high-value assets, we show that HAMR-E is capable of achieving open-loop, inverted locomotion inside a curved portion of a commercial jet engine.

Planning locomotion trajectories for legged microrobots is challenging because of their complex morphology, high frequency passive dynamics, and discontinuous contact interactions with their environment. Consequently, such research is often driven by time-consuming experimental methods. As an alternative, we present a framework for systematically modeling, planning, and controlling legged microrobots. We develop a three dimensional dynamic model of a 1.5g quadrupedal microrobot with complexity (e.g., number of degrees of freedom) similar to larger-scale legged robots. We then adapt a recently developed variational contact-implicit trajectory optimization method to generate feasible whole-body locomotion plans for this microrobot, and we demonstrate that these plans can be tracked with simple joint-space controllers. We plan and execute periodic gaits at multiple stride frequencies and on various surfaces. These gaits achieve high per-cycle velocities, including a maximum of 10.87mm/cycle, which is 15% faster than previously measured velocities for this microrobot. Furthermore, we plan and execute a vertical jump of 9.96mm, which is 78% of the microrobot’s center-of-mass height. To the best of our knowledge, this is the first end-to-end demonstration of planning and tracking whole-body dynamic locomotion on a millimeter-scale legged microrobot.

Sensor fabrication for microrobots is challenging due to their small size and low mass. As a potential solution, we present a technique for estimating the velocity of piezoelectric bending bimorph actuators, a popular choice for driving such microscale devices, that requires simple electronics and no additional mechanical components. Our approach relies on the insight that motion of the actuators causes varying strains on the surface on the piezoelectric material, which via the direct piezoelectric effect, results in a current proportional to the actuator velocity. We propose that the actuator be electrically approximated as a parallel combination of a frequency and voltage dependent resistor and capacitor, and a velocity proportional current source. We develop an experimental procedure to measure these quantities, and are able to experimentally determine the actuator tip velocity to within 10% accuracy over a range of voltages (25–200 V) and frequencies (1–2000 Hz, well beyond actuator resonance). We successfully apply this sensing methodology to two microrobots, the RoboBee and the Harvard Ambulatory MicroRobot (HAMR), to estimate the wing and limb motion respectively. We further use sensor feedback to close the loop on HAMR’s leg phase and obtain desired leg trajectories near transmission resonance. The proposed sensor methodology is generic and can be applied to piezoelectric actuators of different geometries and configurations for uses in microrobotic applications.

Exceptional performance is often considered to be elegant and free of ‘errors’ or missteps. During the most extreme escape behaviours, neural control can approach or exceed its operating limits in response time and bandwidth. Here we show that small, rapid running cockroaches with robust exoskeletons select head-on collisions with obstacles to maintain the fastest escape speeds possible to transition up a vertical wall. Instead of avoidance, animals use their passive body shape and compliance to negotiate challenging environments. Cockroaches running at over 1 m or 50 body lengths per second transition from the floor to a vertical wall within 75 ms by using their head like an automobile bumper, mechanically mediating the manoeuvre. Inspired by the animal’s behaviour, we demonstrate a passive, high-speed, mechanically mediated vertical transitions with a small, palm-sized legged robot. By creating a collision model for animal and human materials, we suggest a size dependence favouring mechanical mediation below 1 kg that we term the ‘Haldane limit’. Relying on the mechanical control offered by soft exoskeletons represents a paradigm shift for understanding the control of small animals and the next generation of running, climbing and flying robots where the use of the body can off-load the demand for rapid sensing and actuation.

Performance metrics such as speed, cost of transport, and stability are the driving factors behind gait selection in legged locomotion. To help understand the effect of gait on the performance and dynamics of small-scale ambulation, we explore four quadrupedal gaits over a wide range of stride frequencies on a 1.43 g, biologically-inspired microrobot, the Harvard Ambulatory MicroRobot (HAMR). Despite its small size, HAMR can precisely control leg frequency, phasing, and trajectory, making it an exceptional platform for gait studies at scales relevant to insect locomotion. The natural frequencies of the body dynamics are used to identify frequency regimes where the choice of gait has varying influence on speed and cost of transport (CoT). To further quantify these effects, two new metrics, ineffective stance and stride correlation, are leveraged to capture effects of foot slippage and observed footfall patterns on locomotion performance. At stride frequencies near body resonant modes, gait is found to drastically alter speed and CoT. When running well above these stride frequencies we find a gait-agnostic shift towards energy characteristics that support ‘kinematic running’, which is defined as a gait with a Froude number greater than one with energy profiles more similar to walking than running. This kinematic running is rapid (8.5 body lengths per second), efficient (CoT  =  9.4), different from widely observed SLIP templates of running, and has the potential to simplify design and control for insect-scale runners.

The integration of information from dynamic sensory structures operating on a moving body is a challenge for locomoting animals and engineers seeking to design agile robots. As a tactile sensor is a physical linkage mediating mechanical interactions between body and environment, mechanical tuning of the sensor is critical for effective control. We determined the open-loop dynamics of a tactile sensor, specifically the antenna of the American cockroach, Periplaneta americana, an animal that escapes predators by using its antennae during rapid closed-loop tactilely mediated course control. Geometrical measurements and static bending experiments revealed an exponentially decreasing flexural stiffness (EI) from base to tip. Quasi-static experiments with a physical model support the hypothesis that a proximodistally decreasing EI can simplify control by increasing preview distance and allowing effective mapping to a putative control variable – body-to-wall distance – compared with an antenna with constant EI. We measured the free response at the tip of the antenna following step deflections and determined that the antenna rapidly damps large deflections: over 90% of the perturbation is rejected within the first cycle, corresponding to almost one stride period during high-speed running (~50 ms). An impulse-like perturbation near the tip revealed dynamics that were characteristic of an inelastic collision, keeping the antenna in contact with an object after impact. We contend that proximodistally decreasing stiffness, high damping and inelasticity simplify control during high-speed tactile tasks by increasing preview distance, providing a one-dimensional map between antennal bending and body-to-wall distance, and increasing the reliability of tactile information.

This paper presents the design and development of a flexure-based microgripper, accompanied with a real-time, vision-based force sensing system to handle objects of various sizes ranging from 100 µm to 1 mm. A simulation-based design methodology is adopted to develop an initial microgripper design, which is then optimized using theoretical modeling. The final prototype developed generated a large stroke length of over 500 µm with high-deflection magnification (ratio of the end deflection (output) to the input deflection) of 3.52. A spring system has been incorporated into the microgripper for easy measurement and control of the gripping forces. The web-camera-based visual system enables real-time force measurement with a resolution of 2.37 mN and can be operated in both manual and automatic modes to control the applied forces. The system has successfully demonstrated gripping of a variety of micro-objects including a 100 µm human hair and a 1 mm steel rod with forces as small as 43 mN and 159 mN, respectively.