Table of contents

Underwater robot designs inspired by the behavior, physiology, and anatomy of fishes can provide enhanced maneuverability, stealth, and energy efficiency. Over the last two decades, robotics researchers have developed and reported a large variety of fish-inspired robot designs. The purpose of this review is to report different types of fish-inspired robot designs based upon their intended locomotion patterns. We present a detailed comparison of various design features like sensing, actuation, autonomy, waterproofing, and morphological structure of fish-inspired robots reported in the past decade. We believe that by studying the existing robots, future designers will be able to create new designs by adopting features from the successful robots. The review also summarizes the open research issues that need to be taken up for the further advancement of the field and also for the deployment of fish-inspired robots in practice.

Synthetic composite materials that mimic the structure and composition of mammalian tooth enamel were prepared by mixing fluoroapatite rods (diameter 2–3 μm, thickness about 0.5 μm) and methylmethacrylate (MMA), followed by polymerization either during or immediately after ultracentrifugation, using either a tertiary amine/radical initiator for polymerization at room temperature or a radical initiator for thermal polymerization. This led to mineral-rich composites (mineral content between 50 and 75 wt%). To enhance the mechanical stability and the interaction between fluoroapatite and polymer matrix, small amounts of differently functionalized MMA monomers were added to the co-monomer mixture. Another approach was the coating of the fluoroapatite rods with silica and the polymerization in the presence of a siloxane-functionalized MMA monomer. The hardness of the composites was about 0.2–0.4 GPa as determined by Vickers indentation tests, about 2 times higher than the polymer matrix alone. The composites had a good resistance against acids (60 min at pH 3, 37 °C).

Designing bio-inspired, multifunctional organic–inorganic composite materials is one of the most popular current research objectives. Due to the high complexity of biocomposite structures found in nacre and bone, for example, a one-pot scalable and versatile synthesis approach addressing structural key features of biominerals and affording bio-inspired, multifunctional organic–inorganic composites with advanced physical properties is highly challenging. This article reviews recent progress in synthesizing organic–inorganic composite materials via various self-assembly techniques and in this context highlights a recently developed bio-inspired synthesis concept for the fabrication of hierarchically structured, organic–inorganic composite materials. This one-step self-organization concept based on simultaneous liquid crystal formation of anisotropic inorganic nanoparticles and a functional liquid crystalline polymer turned out to be simple, fast, scalable and versatile, leading to various (multi-)functional composite materials, which exhibit hierarchical structuring over several length scales. Consequently, this synthesis approach is relevant for further progress and scientific breakthrough in the research field of bio-inspired and biomimetic materials.

Tree frogs have become an object of interest in biomimetics due to their ability to cling to wet and slippery surfaces. In this study, we have investigated the adhesion and friction behavior of toe pads of White's tree frog (Litoria caerulea) using atomic force microscopy (AFM) in an aqueous medium. Facilitating special types of AFM probes with radii of ∼400 nm and ∼13 μm, we were able to sense the frictional response without damaging the delicate nanopillar structures of the epithelial cells. While we observed no significant adhesion between both types of probes and toe pads in wet conditions, frictional forces under such conditions were very pronounced and friction coefficients amounted between 0.3 and 1.1 for the sliding friction between probes and the epithelial cell surfaces.

Recent experimental data revealed a stiffening of aged cortical bone tissue, which could not be explained by common multiscale elastic material models. We explain this data by incorporating the role of mineral fusion via a new hierarchical modeling approach exploiting the asymptotic (periodic) homogenization (AH) technique for three-dimensional linear elastic composites. We quantify for the first time the stiffening that is obtained by considering a fused mineral structure in a softer matrix in comparison with a composite having non-fused cubic mineral inclusions. We integrate the AH approach in the Eshelby-based hierarchical mineralized turkey leg tendon model (Tiburtius et al 2014 Biomech. Model. Mechanobiol. 13 1003–23), which can be considered as a base for musculoskeletal mineralized tissue modeling. We model the finest scale compartments, i.e. the extrafibrillar space and the mineralized collagen fibril, by replacing the self-consistent scheme with our AH approach. This way, we perform a parametric analysis at increasing mineral volume fraction, by varying the amount of mineral that is fusing in the axial and transverse tissue directions in both compartments. Our effective stiffness results are in good agreement with those reported for aged human radius and support the argument that the axial stiffening in aged bone tissue is caused by the formation of a continuous mineral foam. Moreover, the proposed theoretical and computational approach supports the design of biomimetic materials which require an overall composite stiffening without increasing the amount of the reinforcing material.

Bio-inspired Laponite (clay)−liquid crystal (LC) polymer composite materials with high clay fractions (>80%) and a high level of orientation of the clay platelets, i.e. with structural features similar to the ones found in natural nacre, have been shown to exhibit a promising behavior in the context of reduced oxygen transmission. Key characteristics of these bio-inspired composite materials are their high inorganic content, high level of exfoliation and orientation of the clay platelets, and the use of a LC polymer forming the organic matrix in between the Laponite particles. Each single feature may be beneficial to increase the materials gas barrier property rendering this composite a promising system with advantageous barrier capacities. In this detailed study, Laponite/LC polymer composite coatings with different clay loadings were investigated regarding their oxygen transmission rate. The obtained gas barrier performance was linked to the quality, respective Laponite content and the underlying composite micro- and nanostructure of the coatings. Most efficient oxygen barrier properties were observed for composite coatings with 83% Laponite loading that exhibit a structure similar to sheet-like nacre. Further on, advantageous mechanical properties of these Laponite/LC polymer composites reported previously give rise to a multifunctional composite system.

To enhance today's artificial flow sensing capabilities in aerial and underwater robotics, future robots could be equipped with a large number of miniaturized sensors distributed over the surface to provide high resolution measurement of the surrounding fluid flow. In this work we show a linear array of closely separated bio-inspired micro-electro-mechanical flow sensors whose sensing mechanism is based on a piezoresistive strain-gauge along a stress-driven cantilever beam, mimicking the biological superficial neuromasts found in the lateral line organ of fishes. Aiming to improve state-of-the-art flow sensing capability in autonomously flying and swimming robots, our artificial lateral line system was designed and developed to feature multi-parameter freestream flow measurements which provide information about (1) local flow velocities as measured by the signal amplitudes from the individual cantilevers as well as (2) propagation velocity, (3) linear forward/backward direction along the cantilever beam orientation and (4) periodicity of pulses or pulse trains determined by cross-correlating sensor signals. A real-time capable cross-correlation procedure was developed which makes it possible to extract freestream flow direction and velocity information from flow fluctuations. The computed flow velocities deviate from a commercial system by 0.09 m s⁻¹ at 0.5 m s⁻¹ and 0.15 m s⁻¹ at 1.0 m s⁻¹ flow velocity for a sampling rate of 240 Hz and a sensor distance of 38 mm. Although experiments were performed in air, the presented flow sensing system can be applied to underwater vehicles as well, once the sensors are embedded in a waterproof micro-electro-mechanical systems package.

An amoeboid unicellular organism, a plasmodium of the true slime mold Physarum polycephalum, exhibits complex spatiotemporal oscillatory dynamics and sophisticated information processing capabilities while deforming its amorphous body. We previously devised an 'amoeba-based computer (ABC),' that implemented optical feedback control to lead this amoeboid organism to search for a solution to the traveling salesman problem (TSP). In the ABC, the shortest TSP route (the optimal solution) is represented by the shape of the organism in which the body area (nutrient absorption) is maximized while the risk of being exposed to aversive light stimuli is minimized. The shortness of the TSP route found by ABC, therefore, serves as a quantitative measure of the optimality of the decision made by the organism. However, it remains unclear how the decision-making ability of the organism originates from the oscillatory dynamics of the organism. We investigated the number of coexisting traveling waves in the spatiotemporal patterns of the oscillatory dynamics of the organism. We show that a shorter TSP route can be found when the organism exhibits a lower number of traveling waves. The results imply that the oscillatory dynamics are highly coordinated throughout the global body. Based on the results, we discuss the fact that the decision-making ability of the organism can be enhanced not by uncorrelated random fluctuations, but by its highly coordinated oscillatory dynamics.

The intrinsic muscular properties of biological muscles are the main source of stabilization during locomotion, and superior biological performance is obtained with low energy costs. Man-made actuators struggle to reach the same energy efficiency seen in biological muscles. Here, we compare muscle properties within a one-dimensional and a two-segmented hopping leg. Different force–length–velocity relations (constant, linear, and Hill) were adopted for these two proposed models, and the stable maximum hopping heights from both cases were used to estimate the cost of hopping. We then performed a fine-grained analysis during landing and takeoff of the best performing cases, and concluded that the force–velocity Hill-type model is, at maximum hopping height, the most efficient for both linear and segmented models. While hopping at the same height the force–velocity Hill-type relation outperformed the linear relation as well. Finally, knee angles between 60° and 90° presented a lower energy expenditure than other morphologies for both Hill-type and constant relations during maximum hopping height. This work compares different muscular properties in terms of energy efficiency within different geometries, and these results can be applied to decrease energy costs of current actuators and robots during locomotion.

The dynamic aeroelastic effects on wings modified with bio-inspired leading-edge (LE) tubercles are examined in this study. We adopt a state-space aeroelastic model via the coupling of unsteady vortex-lattice method and a composite beam to evaluate stability margins as a result of LE tubercles on a generic wing. The unsteady aerodynamics and spanwise mass variations due to LE tubercles have counteracting effects on stability margins with the former having dominant influence. When coupled, flutter speed is observed to be 5% higher, and this is accompanied by close to 6% decrease in reduced frequencies as an indication of lower structural stiffness requirements for wings with LE tubercles. Both tubercle amplitude and wavelength have similar influences over the change in flutter speeds, and such modifications to the LE would have minimal effect on stability margins when concentrated inboard of the wing. Lastly, when used in sweptback wings, LE tubercles are observed to have smaller impacts on stability margins as the sweep angle is increased.

Load-bearing reinforcing elements in a continuous matrix allow for improved mechanical properties and can reduce the weight of structural composites. As the mechanical performance of composite systems are heavily affected by the interfacial properties, tailoring the interactions between matrices and reinforcing elements is a crucial problem. Recently, several studies using bio-inspired model systems suggested that interfacial mechanical interlocking is an efficient mechanism for energy dissipation in platelet-reinforced composites. While cheap and effective solutions are available at the macroscale, the modification of surface topography in micron-sized reinforcing elements still represents a challenging task. Here, we report a simple method to create nanoasperities with tailored sizes and densities on the surface of alumina platelets and investigate their micromechanical effect on the energy dissipation mechanisms of nacre-like materials. Composites reinforced with roughened platelets exhibit improved mechanical properties for both organic ductile epoxy and inorganic brittle cement matrices. Mechanical interlocking increases the modulus of toughness (area under the stress–strain curve) by 110% and 56% in epoxy and cement matrices, respectively, as compared to those reinforced with flat platelets. This interlocking mechanism can potentially lead to a significant reduction in the weight of mechanical components while retaining the structural performance required in the application field.

An improved quasi-steady aerodynamic model for flapping wings in hover has been developed. The purpose of this model is to yield rapid predictions of lift generation and efficiency during the design phase of flapping wing micro air vehicles. While most existing models are tailored for a specific flow condition, the present model is applicable over a wider range of Reynolds number and Rossby number. The effects of wing aspect ratio and taper ratio are also considered. The model was validated by comparing against numerical simulations and experimental measurements. Wings with different geometries undergoing distinct kinematics at varying flow conditions were tested during validation. Generally, model predictions of mean force coefficients were within 10% of numerical simulation results, while the deviations in power coefficients could be up to 15%. The deviation is partly due to the model not taking into consideration the initial shedding of the leading-edge vortex and wing-wake interaction which are difficult to account under quasi-steady assumption. The accuracy of this model is comparable to other models in literature, which had to be specifically designed or tuned to a narrow range of operation. In contrast, the present model has the advantage of being applicable over a wider range of flow conditions without prior tuning or calibration, which makes it a useful tool for preliminary performance evaluations.

This study investigated the effect of the serial connection of two pumping chambers on transport of liquid with increased viscosity. A serially connected valveless piezoelectric micropump was fabricated inspired by the liquid-feeding strategy of a female mosquito drinking liquid with a wide range of viscosities, from nectar to blood. The performance of the micropump was investigated by varying the viscosity of working liquid. Results showed that the optimal phase difference between the two chambers was 180° out-of-phase for all viscosity conditions. The two chambers operating at 180° out-of-phase exhibited higher pumping performance compared with the sum of each single chamber solely actuated, when viscosity increased. The flow patterns in the micropump showed that the rectification efficiency improved with the increase in viscosity. Results indicated that the serially connected valveless piezoelectric micropump is more robust to the increase of viscosity than a single-chamber piezoelectric micropump. This study would be helpful in the design of microfluidic devices for transporting liquids with a wide range of viscosities.

The emission of biosonar pulses in horseshoe bats (family Rhinolophidae) differs from technical sonar in that it has dynamic features at the interface to the free field. When the horseshoe bats emit their biosonar pulses through the nostrils, the walls of a horn-shaped baffle (anterior leaf) are in motion while diffracting the outgoing ultrasonic wave packets. Here, biomimetic reproductions of the dynamic emission shapes of horseshoe bats have been studied for their ability to impose time-variant signatures onto the outgoing pulses. It was found that an elliptical sound outlet with rotating baffles that were attached along the direction of the major axis can be well suited for this purpose. Most importantly, concave baffle shapes were found to produce strongly time-dependent devices characteristics that could reach a root-mean-square-difference between beampatterns of almost 6 dB within a rotation angle of 10°. In contrast to this, a straight baffle shape needs to be rotated over 60° for a similar result. When continuously rotated in synchrony with the emitted pulses, the concave biomimetic baffles produced time-variant device characteristics that depended jointly on direction, frequency, and time. Since such device properties are so easily produced, it appears well worthwhile to explore their use in engineering.

The droplet interface bilayer (DIB) was recently used to show that a 5 nm thick lipid membrane placed near a vibrating synthetic hair could transduce hair motion into electrical current. Herein, we study for the first time mechanoelectrical transduction of hair motion using multi-membrane DIB arrays formed with more than 2 droplets connected in series, and we introduce a transduction model to investigate how airflow across the hair generates current in a membrane-based hair cell. Measurements of sensing currents across every membrane in serial chains of up to 5 connected droplets demonstrate that perturbation of a single hair creates vibrations that propagate across several droplets, allowing for membranes that are not directly attached to the hair to still transduce its motion. Membranes positioned closest to the hair generate the largest currents, while those farther away produce less current due to energy loss from fluid damping. Inserting multiple hairs of different lengths into different droplets in the array yields sensing currents that exhibit multiple characteristic frequencies in addition to location specific current intensities, features that can be used to spatially localize mechanical perturbations. We also develop a transduction model that provides an order-of-magnitude approximation of the sensing current generated by a membrane in response to airflow across the hair. This model provides physical insights into how membrane-based materials can be used for sensing mechanical stimuli—just like nature does.

In this paper a novel flexible joint is proposed for robotic fish pectoral fins, which enables a swimming behavior emulating the fin motions of many aquatic animals. In particular, the pectoral fin operates primarily in the rowing mode, while undergoing passive feathering during the recovery stroke to reduce hydrodynamic drag on the fin. The latter enables effective locomotion even with symmetric base actuation during power and recovery strokes. A dynamic model is developed to facilitate the understanding and design of the joint, where blade element theory is used to calculate the hydrodynamic forces on the pectoral fins, and the joint is modeled as a paired torsion spring and damper. Experimental results on a robotic fish prototype are presented to illustrate the effectiveness of the joint mechanism, validate the proposed model, and indicate the utility of the proposed model for the optimal design of joint depth and stiffness in achieving the trade-off between swimming speed and mechanical efficiency.

This paper presents a biomimetic turtle flipper actuator consisting of a shape memory alloy composite structure for implementation in a turtle-inspired autonomous underwater vehicle. Based on the analysis of the Chelonia mydas, the flipper actuator was divided into three segments containing a scaffold structure fabricated using a 3D printer. According to the filament stacking sequence of the scaffold structure in the actuator, different actuating motions can be realized and three different types of scaffold structures were proposed to replicate the motion of the different segments of the flipper of the Chelonia mydas. This flipper actuator can mimic the continuous deformation of the forelimb of Chelonia mydas which could not be realized in previous motor based robot. This actuator can also produce two distinct motions that correspond to the two different swimming gaits of the Chelonia mydas, which are the routine and vigorous swimming gaits, by changing the applied current sequence of the SMA wires embedded in the flipper actuator. The generated thrust and the swimming efficiency in each swimming gait of the flipper actuator were measured and the results show that the vigorous gait has a higher thrust but a relatively lower swimming efficiency than the routine gait. The flipper actuator was implemented in a biomimetic turtle robot, and its average swimming speed in the routine and vigorous gaits were measured with the vigorous gait being capable of reaching a maximum speed of 11.5 mm s⁻¹.

The unique architecture of iridescent Morpho butterfly scales is known to exhibit different optical responses to various vapours. However, the mechanism behind this phenomenon is not fully quantitatively understood. This work reports on process developments in the micro-fabrication of a Morpho-inspired photonic structure in atomic layer deposited (ALD) materials in order to investigate the vapour optical sensitivity of such artificial nanostructures. By developing recipes for dry and wet etching of ALD oxides, we micro-fabricated two structures: one combining Al₂O₃ and TiO₂, and the other combining Al₂O₃ and HfO₂. For the first time, we report the optical response of such ALD Morpho-like structures measured under a controlled flow of either ethanol or isopropyl alcohol (IPA) vapour. In spite of the small magnitude of the effect, the results show a selective vapour response (depending on the materials used).

In robotics, there is a need for small scale, compliant actuators for use in medical applications or minimally invasive environmental monitoring. Biohybrid devices offer one solution to this need by using muscle cells to actuate compliant scaffolds. Such devices typically use biocompatible synthetic polymers as compliant scaffolds, which require additional processing steps to promote cellular alignment and attachment. Instead, electrocompacted and aligned collagen (ELAC) can be used as a completely organic scaffold, requiring no additional processing steps, with alignment being innately promoted by the topography. Locomotive living machines have been fabricated in this study using ELAC scaffolds. Devices have been produced using either primary cardiomyocytes or primary skeletal muscle cells isolated from chick embryos as actuators. When tested under the same conditions, skeletal muscle cell powered devices were approximately an order of magnitude faster, having a mean velocity of 77.6 ± 86.4 μm min^–1, compared to 9.34 ± 6.69 μm min^–1 for cardiomyocyte powered devices. In conclusion, completely organic living machines have been fabricated using electrocompacted collagen skeletons, and it was found that skeletal muscle powered devices were significantly faster than cardiomyocyte powered devices.

The human hand's versatility allows for robust and flexible grasping. To obtain such efficiency, many robotic hands include human biomechanical features such as fingers having their two last joints mechanically coupled. Although such coupling enables human-like grasping, controlling the inverse kinematics of such mechanical systems is challenging. Here we propose a cortical model for fine motor control of a humanoid finger, having its two last joints coupled, that learns the inverse kinematics of the effector. This neural model functionally mimics the population vector coding as well as sensorimotor prediction processes of the brain's motor/premotor and parietal regions, respectively. After learning, this neural architecture could both overtly (actual execution) and covertly (mental execution or motor imagery) perform accurate, robust and flexible finger movements while reproducing the main human finger kinematic states. This work contributes to developing neuro-mimetic controllers for dexterous humanoid robotic/prosthetic upper-extremities, and has the potential to promote human–robot interactions.

The study investigates the aerodynamic performance and the relation between wing deformation and unsteady force generation of a flapping-wing micro air vehicle in hovering flight configuration. Different experiments were performed where fluid forces were acquired with a force sensor, while the three-dimensional wing deformation was measured with a stereo-vision system. In these measurements, time-resolved power consumption and flapping-wing kinematics were also obtained under both in-air and in-vacuum conditions. Comparison of the results for different flapping frequencies reveals different wing kinematics and deformation characteristics. The high flapping frequency case produces higher forces throughout the complete flapping cycle. Moreover, a phase difference occurs in the variation of the forces, such that the low flapping frequency case precedes the high frequency case. A similar phase lag is observed in the temporal evolution of the wing deformation characteristics, suggesting that there is a direct link between the two phenomena. A considerable camber formation occurs during stroke reversals, which is mainly determined by the stiffener orientation. The wing with the thinner surface membrane displays very similar characteristics to the baseline wing, which implies the dominance of the stiffeners in terms of providing rigidity to the wing. Wing span has a significant effect on the aerodynamic efficiency such that increasing the span length by 4 cm results in a 6% enhancement in the cycle-averaged X-force to power consumption ratio compared to the standard DelFly II wings with a span length of 28 cm.