Search

Soft robotic devices require highly stretchable and flexible materials to ensure that they can conform to delicate objects and work safely alongside humans. As a result, new flexible actuation technologies must be developed so that soft robotic platforms can achieve their desired function. Dielectric elastomer actuators are an emerging actuator technology that possesses high energy density in a lightweight package. However, since these actuators utilize a thin, soft elastomer film, they are particularly delicate and depending on their configuration, do not efficiently outcouple their outputted mechanical energy. Rigid frames, metallic springs, and other mechanisms have been used in the past to improve the performance of these actuators in various configurations, but these highly stiff components are not ideal for soft robotic applications. Novel strategies can be implemented to mitigate these issues. In this work, we highlight new strategies to create novel dielectric elastomer actuator configurations with high power density, large strokes, and high force output. First, we discuss the development of a core-free rolled dielectric elastomer. We then discuss a flexible DEA-based fluidic pump that can output high flowrates and withstand large pressures for its small size. Finally, we detail a novel mechanism to control the deformation direction of DEAs, and leverage this mechanism to create a patch-like haptic interface for virtual reality applications.

Inspired by the biology, soft robots have drawn tremendous attention due to its large and continuous deformation, friendly human-machine interaction, large number of degrees of freedom (DOFs), capability of absorbing energy. They have been explored in broad applications ranging from dexterous soft gripper to the novel assistive devices. In the recent decade, numerous soft actuating materials and deformable structures have been developed to construct soft robots, including hydrogels, shape memory polymers (SMPs), dielectric elastomer actuators (DEAs), fluid elastomer actuators (FEAs) and magnetic actuators. However, those materials and structures have well-known limitations such as slow actuation speed, irreversibility, high voltage input and bulky controlling systems. Liquid crystal elastomers (LCEs), as newly emerging soft actuating materials, exhibit large and reversible deformation and versatile actuation modes. Based on the molecular structure, LCE can be viewed as a combination of liquid crystal molecules and polymer networks. When the LCE is heated above the critical temperature, it can generate large deformation because of the nematic-isotropic phase transition. However, in terms of the practical use of LCE, a few challenges exist such as lack of programmable operation and slow responsive speed for LCEs, which need to be addressed.In this dissertation, we first integrate flexible heating wire into LCE tube, forming electrically controlled soft tubular actuator. By selectively applying low electrical voltage, this soft tubular actuator can exhibit multiple actuation modes, such as different directional bending and homogeneous contraction. The LCE soft tubular actuator can also be integrated to construct untethered robot that can execute multiple functionalities. To address the slow responsive speed of LCE based soft actuator, we embed microfluidic channel into LCE, forming vascular LCE soft actuator. Through alternatively injecting hot and cold fluid into its internal fluidic channel, the vascular LCE soft actuator can generate fast actuation as well as recovery. In addition, by introducing the disulfide bonds into the LCE materials, the newly obtained vascular LCE based soft actuator has shown repairability and recyclability. Finally, we use electrospinning technique to fabricate LCE microfiber that can be actuated by NIR light. We demonstrate that the electrospun LCE fiber can be easily integrated to micro-robotic system and machine as artificial muscle fiber.

Adventitious bioinspiration describes the observation of a biological phenomenon and work to understand and replicate it. The dissertation examines how this philosophy can be applied to familiar model organisms in understudied loading modes, or familiar bioinspired concepts to understudied applications. That wood achieves high strength and toughness through its cellular solid structure is well known, but how this macrostructure interacts with its mesostructural anatomy of distinct cell types and cell wall behavior is less understood, especially in non-quasistatic, compressive loading modes. We propose that by studying wood in impact and torsion, novel bioinspiration may be obtained from perhaps the oldest known and most widely used biological material.First, impact testing was performed to understand why only certain species of wood were used to craft striking weapons and fortifications across eras and civilizations. It was hypothesized that these tree species combined unique anatomical features that allowed high energy absorption in low velocity dynamic loading. Scanning electron microscopy (SEM) and micro-computed tomography (μ-CT) revealed that trees with the greatest energy absorption per unit density combined uniform vessel distribution, intermittent rays, interlocking grain growth, and optimal fiber adhesion. In particular, a new hierarchical progressive delamination was observed in which not only tracheids ruptured and tore out, but their helically wound cell walls also unraveled to absorb energy. This contrasted from wood quasi-static behavior in which density was a reliable predictor of mechanical behavior without accounting for these mesostructural features. Next, torsional testing on cholla cacti were conducted to understand their adaptations to twisting induced by high winds. Though tall trees are typically bent by wind loading, shorter, branching plants with high skewness (like cholla) are twisted instead. It was hypothesized that the helical macroporosity winding around their hollow wooden stalk was an adaptation to maximize torsional stiffness while minimizing mass in the resource-poor desert. Novel mesostructural characterization methods of laser- scanning and photogrammetry were used alongside traditional optical microscopy, SEM, and μ-CT to identify mechanisms responsible for torsional resistance. These methods, in combination with finite element analysis (FEA) revealed how cholla meso and macro-porosity and fibril orientation contributed to highly density-efficient mechanical behavior. Selective lignification and macroscopic tubercle pore geometry contributed to density-efficient shear stiffness, while mesoscopic wood fiber straightening, delamination, pore collapse, and fiber pullout provided extrinsic toughening mechanisms. These energy absorbing mechanisms were enabled by the hydrated material level properties. Together, these hierarchical behaviors allowed the cholla to far exceed bamboo and trabecular bone in its ability to combine specific torsional stiffness, strength, and toughness.This dissertation also explores how familiar bioinspired concepts can be applied to understudied applications by designing and testing hierarchically structured jamming devices. In the soft-robotics field, bio-inspiration is often cited, pointing to the animal-like forms created—however, the concept of hierarchical architecture common to biological materials has yet to be applied effectively. It was shown how that by considering the hierarchical structure of the medium (primary level), the organization of jamming media (secondary level), and the organization of jammers (tertiary level) new functionalities not possible with conventional jamming technology could be obtained. At the primary level, optimal compositions of fibrous flakes and grains were identified to improve stiffness and strength per unit weight; fish-inspired ganoid scales were used to create flexible armors. At the secondary level, layers and grains were combined in the tensile and compressive faces of beams to maximize mechanical properties, while ganoid scales of different compositions were layered to create mechanical gradients, among other combinations of jamming media. Finally, at the tertiary level, the isotropy of triaxially woven jammers was demonstrated relative to traditional biaxial jammers; a cylindrical “finger-trap” weave with adjustable radius was shown. The improved mechanical weight-efficiency, anisotropy control, mechanical property gradients, and other features enabled by considering hierarchical design in jamming promise new application spaces for an established field, such as reactive wearable armors. Finally, wood-templated silicon carbides were infiltrated with epoxy to investigate a scalable method of fabricating easily shapable ceramic-polymer composites with both flexural strength (σ Flexure) and fracture toughness (KIC). In situ SEM and μ-CT of fracture toughness samples indicated a lack of interaction between the ceramic and epoxy phases, leading to only marginal improvement over the rule of mixtures for KIC and underperformance for σ Flexure. Preliminary tests of surface oxidized wood-templated SiC showed qualitatively improved crack stability in in situ SEM.

In this article, a novel actuator called armor-based stable force pneumatic artificial muscle (AS-PAM) is presented. AS-PAM has a sealed chamber made of polygonal parts and film, which helps the actuator to be lightweight (∼100 g) and achieve a large contraction ratio (60%). It has an armor and a constraint to guide its motion, which keeps its force output [6.25 N/(cm

The overarching goal of this thesis relies in the exploration of design criteria, such as optimal material layout and geometry, to (1) create and (2) enhance the performance of innovative actuators and sensors, respectively. First, we present a bio-inspired multifunctional active skin, which is a two-dimensional architected material that exhibits local and/or global programmable, rapid, and reversible, out-of-plane surface texture morphing when actuated by in-plane tension. Here, by introducing geometrical imperfection or notches at judiciously chosen locations in a preconceived, auxetic geometry, we can effectively control the directionality of out-of-plane deformations for applications such as camouflage, surface morphing, and soft robotic grippers. Second, an innovative “sensing mesh” capable of resolving both spatial strain magnitudes and directionalities for distributed strain field monitoring is introduced. The sensing mesh leverages the same concept of patterning to impart unique sensing capabilities in piezoresistive nanocomposites. In particular, the use of a grid-like layout with high aspect ratio struts resolves the strain sensing directionality limitations observed in previously reported sensing skins while also enhancing the sensitivity of the piezoresistive graphene-based thin films once coupled with an electrical impedance tomography conductivity mapping technique. Finally, the design considerations explored in this thesis contribute to the development of an inverse design methodology of architected materials in which functionality parameters are first indicated and, then, optimal topologies for maximum effectiveness are indicated.

Structural colors, derived from existing natural creatures, have aroused widespread attention in the materials regulation for different applications. Here, inspired by the color adjusting mechanism of hummingbird, we present a novel shape-memory structural color hydrogel film by introducing shape memory polymers (SMPs) into synthetic inverse opal scaffold structure. The excellent flexibility as well as the inverse opal structure of the hydrogel films imparts them with stable stretchability and brilliant structural colors. Benefiting from the transient structural anisotropy of copolymers, the hybrid films are possessed with shape-morphing behaviors capability. Based on the shape transformations and color responsiveness performance, we have demonstrated diverse structural color actuators with complex shapes for different tasks. Notably, as the photothermal responsive graphene quantum dots were integrated into the hydrogel, the hybrid films could also be endowed with the feature of light-controlled reversible deformation with synchronous structural color variation. These features demonstrate that the presented shape-memory structural color hydrogel film is valuable for soft robotics with multi-functions of sensing, communication and disguise.

Programmable materials have artificially designed physical shapes responding to external stimuli, as well as high design capability and high flexibility. Here, we propose a microfiber-shaped programmable material with an axial pattern of stimuli-responsive (SR) and nonresponsive hydrogels. The SR pre-gel solution was mixed to sodium alginate pre-gel solution for instantaneous gelation with ionic crosslinking and solidified on a nonresponsive hydrogel microfiber with a valve-controlled microfluidic system. A design of microfiber-shaped programmable material (patterned position of SR regions) could be flexibly altered by changing a coded sequence program. We confirmed that the three-dimensional (3D) coil-like structures were self-folded at the patterned SR regions responding to the thermal stimulus and that the chirality of the self-folded 3D coil-like structures depends on the condition of the stimulus to the microfiber. Finally, interaction with objects using the programmable microfiber as a soft actuator was demonstrated. Our microfiber-shaped programmable materials expand possibilities of fiber-based materials in biomimetics and soft robotics fields.

An untethered, soft robot using liquid crystal elastomer (LCE) actuators, onboard power, and wireless Bluetooth control was developed. LCE actuators were thermally triggered using Joule heating and demonstrated an ∼5 N force pull capacity per LCE. A20% repeatable strain was demonstrated over100 cycles with minimal loss of strain at high cycle numbers. The LCE actuators were horizontally oriented to maximize conversion of LCE contraction to overall robot movement. A battery and control board were integrated into the body of the robot, which allowed for Bluetooth control of the LCE on/off cycle. System level programming and design were implemented to offset the slow recovery associated with LCE actuators. The multiple LCE actuator legs were programmed to allow individual control of on/off cycles for each leg. LCE leg actuation was alternated between inner and outer legs to provide horizontal movement with minimized loss of motion during the LCE recovery cycle by actuating one set of legs during the recovery cycle of the other set for a maximum movement speed of 1.27 cm/min. Path control was also demonstrated by turning the robot by actuating two LCE legs on one side of the robot. The robot was able to pull up to 1400 g in ideal frictional conditions, allowing the possibility of payload transport, additional battery storage, or onboard sensors. Additional design considerations are discussed to further improve overall robot speed in the future by combining system and material level design considerations.

Fiber reinforcement is a crucial attribute of soft-bodied muscular hydrostats that have the ability to undergo large deformations and maintain their posture. Helically wound fibers around the cylindrical worm body help control the tube diameter and length. Geometric considerations show that a fiber winding angle of 54.7°, called the magic angle, results in a maximum enclosed volume. Few studies have combined both experimental and theoretical techniques to explore the effects of fiber winding at varied angles on the large deformation mechanics of fiber-reinforced elastomers (FRE). We fabricated FRE materials in transversely isotropic layouts varying from 0° to 90° using a custom filament winding technique and characterized the nonlinear stress-strain relationships using uniaxial and equibiaxial experiments. We used these data within a continuum mechanical framework to propose a novel constitutive model for incompressible FRE materials with embedded extensible fibers. The model includes individual contributions from the matrix and fibers in addition to coupled terms in strain invariants, I1 and I4. The deviatoric stress components show inversion at fiber orientation angles near the magic angle in the FRE composites. These results are useful in soft robotic applications and in the biomechanics of fiber-reinforced tissues such as the myocardium, arteries, and skin.

To improve performance of thermopneumatic soft actuators, which have recently been developed for various industrial applications, we embedded different nanoscale materials into their elastomer bodies. This yields a significant enhancement in the actuator performance via improving the mechanical and thermal properties of the elastomer bodies. In addition, the use of nanoinclusions diminished losses of the working fluid from the actuators by decreasing vapor leaks through the elastomer body and thus improving longevity. Notably, when using different working fluids with low boiling temperatures, the operating temperature range of the actuators can be lowered and widened. The hybrid approach proposed in this study is expected to advance the industrial feasibility of thermopneumatic actuators.

Robotic hands with tactile perception can perform more advanced and safer operations, such as material recognition. Nanowires with high sensitivity, fast response, and low power consumption are suitable for multifunctional flexible tactile sensors to provide the tactile perception of robotic hands. In this work, we designed a multifunctional soft robotic finger with a built-in nanoscale temperature-pressure tactile sensor for material recognition. The flexible multifunctional tactile sensor integrates a nanowire-based temperature sensor and a conductive sponge pressure sensor to measure the temperature change rate and contact pressure simultaneously. The developed nanoscale temperature and conductive sponge pressure sensor can reach a high sensitivity of 1.196%/°C and 13.29%/kPa, respectively. With this multifunctional tactile sensor, the soft finger can quickly recognize four metals within three contact pressure ranges and 13 materials within a high contact pressure range. By combining tactile information and artificial neural networks, the soft finger can recognize the materials precisely with a high recognition accuracy of 92.7 and 95.9%, respectively. This work proves the application potential of the multifunctional soft robot finger in material recognition.

It is desirable to fabricate an antifatigue gel for skin-mimicking sensors on the demand of long-term durability in practical usage. Here, we developed a physically cross-linked eutectogel based on a poly(vinyl alcohol)/poly(acrylic acid) (PVA/PAA) binary polymer skeleton and a deep eutectic solvent (DES). In this eutectogel, uniformly distributed PVA crystalline domains acted as stable physical cross-linkers, and high-density hydrogen bonds possessed great reversibility. Such a polymer network structure was expected to endow this eutectogel with excellent mechanical strength, stretchability, and a self-recovery ability. Specifically, this eutectogel exhibited a superior tensile strength of 2.6 MPa, a fracture strain of 680%, and a fracture toughness of 8.39 MJ m-3. In cyclic stretching/releasing tests with a fixed strain of 100%, this eutectogel could recover its mechanical properties within a 600 s resting time. Based on this self-recoverable eutectogel, a reliable flexible sensor was fabricated, which possessed good sensitivity and stability over a wide strain range (1-300%). More importantly, the flexile sensor was able to maintain a highly repeatable response signal during 1000 consecutive stretching/releasing cycles, showing outstanding long-term durability. Given the excellent sensing performance, this eutectogel has promising potential in wearable electronics, human-machine systems, and soft robotics.

Deformable 3D structures have emerged to revolutionize next-generation flexible electronics. In this study, a large out-of-plane deformable kirigami-based structure integrated with traditional functional materials has been successfully applied to wirelessly sense mechanical vibration and pressure. Unlike spiral inductor coils that lack mechanical stability, the inductor coils supported with polymer kirigami designs, comprising concentric circles with alternately connected hinges among the consecutive layers, offer exceptional mechanical stability. The wireless sensor shows a good linear response (Adj. R2 = 0.99) between the shift in resonant frequency as a function of extension. Moreover, the sensor device exhibits excellent cycling mechanical stability and minimal hysteresis, as confirmed by the experiments performed for over 5 d. An acceleration sensor (0-20 ms-2) with high linearity (Adj. R2 = 0.99) is introduced. Furthermore, a highly sensitive low-pressure sensor is demonstrated wirelessly in real time. Thus, the sensor can wirelessly monitor mechanical vibration and pressure. It can be applied for motion tracking, health monitoring, soft robotics, and deformation detection in battery-free deformable electronic devices.

Skin-mountable capacitive-type strain sensors with great linearity and low hysteresis provide inspiration for the interactions between human and machine. For practicality, high sensing performance, large stretchability, and self-healing are demanded but limited by stretchable electrode and dielectric and interfacial compatibility. Here, we demonstrate an extremely stretchable and self-healing conductor via both hard and soft tactics that combine conductive nanowire assemblies with double dynamic network based on π-π attractions and Ag-S coordination bonds. The obtained conductor outperforms the reported stretchable conductors by delivering an elongation of 3250%, resistance change of 223% at 2000% strain, high durability, and multiresponsive self-healability. Especially, this conductor accommodates large strain of 1500% at extremely knotted and twisted deformations. By sandwiching hydrogel conductors with a newly developed dielectric, ultrahigh stretchability and omni-healability are simultaneously achieved for the first time for a capacitive strain sensor inspired by metal-thiolate coordination chemistry, showing great potentials in wearable electronics and soft robotics.

Magnetic liquid metal droplets (MLMDs) have been proven to be very important in many fields such as flexible electronics and soft robotics. Usually, soft magnetic particles such as nickel (Ni) and iron (Fe) are mixed or suspended into the liquid metal to obtain soft MLMDs (S-LMDs), which can be easily manipulated under the magnetic field due to the favorable deformability and flexibility. In addition, hard magnetic particles such as neodymium iron boron (NdFeB) with a high residual magnetization can also be dispersed into the liquid metal and the hard MLMDs (H-LMDs) become more compact due to the interaction between internal particles induced by remanence. This work reports a kind of H-LMDs with high surface tension, high flexibility and mechanical robustness, whose electrical conductivity and strength are better than the S-LMDs. Under the magnetic field, the H-LMDs have a faster response time (0.58 s) and a larger actuating velocity (4.45 cm/s) than the S-LMDs. Moreover, the H-LMDs show excellent magnetic controllability, good elasticity and favorable mobility, as demonstrated by magnetically actuated locomotion, bounce tests and rolling angle measurements. Finally, the droplets can be further applied in wheel-driven motors and micro-valve switches, which demonstrates their high application potential in robotic manipulation and microfluidic devices.

Hydrogels constructed from natural sources have received increased attention recently, including applications in biomedical fields. They are protein or polysaccharide cross-linked scaffolds that display water retention and are able to recognize host cargos. Their excellent biocompatibility does not always combine with high mechanical strength (up to 136 kPa) and thermostability, making them less useful in biomedical applications. This paper reports biocompatible gelatin hydrogels, double cross-linked via imine and Diels-Alder (DA) dynamic covalent frameworks. They showed integrated advantages of adjustable and durable mechanical strength, good thermal stability, biocompatibility for promoting cell growth and reasonable degradable rate. These hydrogels possess remarkable self-healing property, acid/alkali resistance at 65 °C and good integrity in organic solvents at 130 °C, holding great potential for biomedical applications in the areas such as cartilage regeneration, articular reconstruction or soft robotics.

Soft robots built with active soft materials have been increasingly attractive. Despite tremendous efforts in soft sensors and actuators, it remains extremely challenging to construct intelligent soft materials that simultaneously actuate and sense their own motions, resembling living organisms’ neuromuscular behaviors. This work presents a soft robotic strategy that couples actuation and strain-sensing into a single homogeneous material, composed of an interpenetrating double-network of a nanostructured thermo-responsive hydrogel poly(N-isopropylacrylamide) (PNIPAAm) and a light-absorbing, electrically conductive polymer polypyrrole (PPy). This design grants the material both photo/thermal-responsiveness and piezoresistive-responsiveness, enabling remotely-triggered actuation and local strain-sensing. This self-sensing actuating soft material demonstrated ultra-high stretchability (210%) and large volume shrinkage (70%) rapidly upon irradiation or heating (13%/°C, 6-time faster than conventional PNIPAAm). The significant deswelling of the hydrogel network induces densification of percolation in the PPy network, leading to a drastic conductivity change upon locomotion with a gauge factor of 1.0. The material demonstrated a variety of precise and remotely-driven photo-responsive locomotion such as signal-tracking, bending, weightlifting, object grasping and transporting, while simultaneously monitoring these motions itself via real-time resistance change. The multifunctional sensory actuatable materials may lead to the next-generation soft robots of higher levels of autonomy and complexity with self-diagnostic feedback control.

Materials that can assist with perception and responsivity of an engineered machine are said to promote physical intelligence. Physical intelligence may be important for flexible and soft materials that will be used in applications like soft robotics, wearable computers, and healthcare. These applications require stimuli responsivity, sensing, and actuation that allow a machine to perceive and react to its environment. The development of materials that exhibit some form of physical intelligence has relied on functional polymers and composites that contain these polymers. This review will focus on composites of functional polymers that display physical intelligence by assisting with perception, responsivity, or by off-loading computation. Composites of liquid crystal elastomers, shape-memory polymers, hydrogels, self-healing materials, and transient materials and their functionalities are examined with a viewpoint that considers physical intelligence. Graphic Abstract

Hydrogels with multifunctional properties have attracted considerable interest owing to their wide range of applications in soft robotics, wearable device development, and tissue engineering. In th...

A reliable estimate for bending stiffness is critical to many soft robot models when predicting everything from robot-environment contact to buckling resistance. Current methods for predicting actuator bending stiffness rely on highly accurate knowledge of material characteristics, which are not trivial to obtain for composite actuators. Additionally, current models for fluidic actuators often depend on a pressure-independent bending stiffness despite pressure playing a non-negligible role in bending stiffness behavior. Methods to measure actuator stiffness often require costly instrumentation to measure or perturb the motions and forces required to measure actual bending stiffness. We introduce a simple free-fold test to empirically estimate the bending stiffness of slender soft actuators—pressurized or unpressurized and composite or homogeneous—which requires the measurement of one distance from a single image of a specific robot pose. The resulting model also shows that the change in actuator weight per unit length can be used to determine the dependence of bending stiffness on actuation pressure.

Conventional soft robotic grippers have been developed based on soft pneumatic actuators or using smart soft materials, but they need external pressure sources, are easily deteriorated or used in millimetric-scale, which leads to increase of the independency, instability, and vulnerability. In this study, we propose an EPM–MRE (Electropermanent Magnet–Magnetorheological elastomer) actuation system, which strengthens the independency and stability of soft actuation by non-contact magnetic force. We established its fundamental principle and investigated parameter design through developing a suction cup as a robotic gripper. We prototyped an EPM-MRE actuated suction cup based on magnetic-charge and tensile modeling as well as optimized the structure of EPM by introducing axisymmetric EPM with frustum pole and suction cup by introducing bi-silicone structure to uniformly activate MRE membrane. The activation of EPM by different current pulses and suction force affected by contact shapes and air gaps were tested. Evaluations showed that the suction cup could be activated with 10 ms and generate a maximum suction force of 9.2 N at a steady state with 0 J energy consumption. In conclusion, the EPM-MRE soft actuation system can be used as a robotic gripper.

Magnetorheological (MR) fluids have unique characteristics that make them well-suited for soft robotic actuation systems. Two particularly interesting applications are bistable valves and force-amplified grippers, which only require power to switch states. These design topologies can not only make MR fluid actuation systems more efficient, but can also allow them to be constructed from entirely compliant materials. In this work, we describe new bistable valve mechanism concepts; the first design is constructed using traditional rigid components while the second design is composed of entirely compliant materials. This includes liquid metal coils, compliant magnetic composites, and silicone flexures. The MR valve concept is modeled using magnetic field theory and scaling laws are determined from MR fluid holding pressure experiments. Then, experimental results are used to show that rough tubing surfaces can withhold an average of 5× more pressure than smooth tubing surfaces. Both rigid and compliant valve topologies are then constructed and tested, showing the ability to hold up to 88 kPa and 37 kPa of pressure, respectively. These valves are then implemented for controlling a soft multi-fingered PneuNet gripper and a new force-amplified MR fluid gripper, which has the ability to provide up to 324 $\%$ higher load capacity than purely pressure-driven mechanisms (with the addition of a 25 mm diameter and 12.7 mm thick Nd-Fe-B magnet). Finally, we show how these bistable valves and grippers can be used for grasping and manipulation applications.

Hydrogel actuators displaying programmable shape transformations promise to be core components in future biomedical and soft robotic devices. However, current hydrogel actuators have shortcomings, including poor mechanical properties, slow response, and lack of shape reprogrammability, which limit their practical applications . Existing molecular designs offer limited efficiency in synergistically addressing these issues in a single hydrogel system. Herein, we propose a strategy to develop hydrogel actuators with muscle-mimetic aligned microfibrillar morphology, combining thermoinduced microphase separation and mechanical alignment. The key to our design is the introduction of metal–phenolic complexes, which not only induce irreversible sol–gel transition via the concentrated coordinate ions above lower critical solution temperature (LCST) but also fix the alignment of bundle network due to dynamic network rearrangement. Our design concept is observed to simultaneously achieve excellent mechanical properties (tensile strength ≈ 1.27 MPa, toughness ≈ 2.0 MJ m–3) and ultrafast actuation (40.1% thermal contraction as short as 1 s), which is a long-lasting challenge in the field. In addition, the dynamic hydrogels can be reprogrammed into spiral, helical, and biomimetic actuators. This work opens new opportunities to realize real-world applications for smart hydrogels as soft machines by fundamentally breaking the current property limit.

Stretchable electronics that can elongate elastically as well as flex are crucial to a wide range of emerging technologies, such as wearable medical devices, electronic skin, and soft robotics. Critical to stretchable electronics is their ability to withstand large mechanical strain without failure while retaining their electrical conduction properties, a feat significantly beyond traditional metals and silicon-based semiconductors. Herein, we present a review of the recent advances in stretchable conductive polymer nanocomposites with exceptional stretchability and electrical properties, which have the potential to transform a wide range of applications, including wearable sensors for biophysical signals, stretchable conductors and electrodes, and deformable energy-harvesting and -storage devices. Critical to achieving these stretching properties are the judicious selection and hybridization of nanomaterials, novel microstructure designs, and facile fabrication processes, which are the focus of this Review. To highlight the potentials of conductive nanocomposites, a summary of some recent important applications is presented, including COVID-19 remote monitoring, connected health, electronic skin for augmented intelligence, and soft robotics. Finally, perspectives on future challenges and new research opportunities are also presented and discussed.

Living organisms such as fishes1, cephalopods2 and clams3 are cryptically coloured with a wide range of hues and patterns for camouflage, signalling or energy regulation. Despite extensive efforts to create colour-changing materials and devices4, it is challenging to achieve pixelated structural coloration with broadband spectral shifts in a compact space. Here, we describe pneumatically inflating thin membranes of main-chain chiral nematic liquid crystalline elastomers that have such properties. By taking advantage of the large elasticity anisotropy and Poisson’s ratio (>0.5) of these materials, we geometrically program the size and the layout of the encapsulated air channels to achieve colour shifting from near-infrared to ultraviolet wavelengths with less than 20% equi-biaxial transverse strain. Each channel can be individually controlled as a colour ‘pixel’ to match with surroundings, whether periodically or irregularly patterned. These soft materials may find uses in distinct applications such as cryptography, adaptive optics and soft robotics. Pneumatically actuated membranes made from a chiral nematic liquid crystalline elastomer supported by poly(dimethylsiloxane) layers are assembled into pixelated colour devices, where each individual pixel can be tuned throughout the entire visible spectrum.

Liquid metal represents a highly conductive and inherently deformable conductor for the development of stretchable electronics. The widespread implementations of liquid metal towards functional sensors and circuits are currently hindered by the lack of a facile and scalable patterning approach. In this study, we report a fully solution-based process to generate patterned features of the liquid metal conductor. The entire process is carried out under ambient conditions and is generally compatible with various elastomeric substrates. The as-prepared liquid metal feature exhibits high resolution (100 μm), excellent electrical conductivity (4.15 × 104S cm−1), ultrahigh stretchability (1000% tensile strain), and mechanical durability. The practical suitability is demonstrated by the heterogeneous integration of light-emitting diode (LED) chips with liquid metal interconnects for a stretchable and wearable LED array. The solution-based technique reported here is the enabler for the facile patterning of liquid metal features at low cost, which may find a broad range of applications in emerging fields of epidermal sensors, wearable heaters, advanced prosthetics, and soft robotics.

Thanks to their continuously deformable features, soft bending actuators, including pneumatic and electric actuators, are increasingly used in continuum robots. Further development, however, is limited by the complexity of the multicavity structure in pneumatic actuators and the instability of the electromechanical coupling in dielectric elastomer actuators. In this article, by combining the advantages of the two actuation techniques, a novel hybrid electro-pneumatic bending actuator is proposed that is capable of bending with significantly reduced complexity and improved stability. A timing control method is proposed that enables rapid and significant bending. A low voltage is used for bending guidance with air-pressure concurrently to produce a large bending stroke. The results of experiments indicate that, with increasing air pressure, the bending angle varies in three different stages. These stages are (I) a low-pressure guidance stage, (II) an instability induction stage, and (III) a stable deformation stage. By harnessing the instability at stage II, the driving voltage can be substantially reduced to avoid electric breakdown. By studying the effect of prestretch on bending, the controllability of the bending deformation is improved. The simulation is also used to verify the control method and predict the development path of soft robots.

In nature, hydrostatic, endo- and exo-skeletons are widely observed, and provide essential rigidity and anchoring points for the application of muscular forces. The efficient interface between a hard skeleton and soft muscle in biology is made possible by a complex hierarchy of structures and composite materials, extending from the nano- to the meso-scale. In contrast, artificial constructs which aim to bridge this hard-soft interface are prone to failure due to local discontinuities and concentrations in stress and strain which lead to material ruptures, delamination and tearing. In this article, the concept of a stiffness-graded electroactive material (SGEM) is proposed which emulates the soft-rigid interface in the nature biological systems and provides both electromechanical activity and the smooth stiffness gradient needed to bridge these two extreme states. This is achieved by programming the diffusion of a rigid filler material (polyvinyl chloride) in a liquid plasticizer (diisodecyl adipate). It is shown that the resulting stiffness gradient can match that of biological tissues such as smooth and skeletal muscles, and that the distal rigid region can be drilled and bonded and significant loads can be safely applied. Additionally, the resulting composite shows electroactive capability through graded anodophilic actuation characteristics. This protocol can be extended to numerous morphologies such as vertical or radial gradients depending on the deployment of two precursor ingredients. Finally, example applications including surface morphing and motion generation are demonstrated. The embodied stiffness gradient and electroactivity make this concept suitable for the development of bio-integrating and wearable artificial muscles systems and more effective soft robots.