Carmel Majidi

ABSTRACT Chronic disease arises slowly through complex physiological factors that require continuous monitoring and treatment. Compliant bioelectronics with soft, stretchable, and tissue‐conformal designs allow for in vivo continuous monitoring through electrophysiological, mechanical, and biochemical modalities, while minimizing tissue irritation and immune system response. In this review, we discuss recent progress in material development, device design, and versatile sensing modalities that allow stable long‐term operation in dynamic biological environments. We focus on chronic electrophysiological systems for neural and cardiac recording, mechanical sensing systems for strain and pressure, and electrochemical sensors for molecular biomarkers. In addition, we examine self‐powered bioelectronics based on piezoelectric and triboelectric energy conversion mechanisms that eliminate the requirement for external batteries, whereas multimodal and closed‐loop systems combine sensing with therapeutic feedback. We consider key parameters like material biocompatibility, device flexibility, and long‐term stability that are required for chronic monitoring to maintain stable signal quality over long time periods relevant for clinical recordings. These technologies for compliant bioelectronics enable early disease diagnosis, personalized treatment and continuous intervention for patients, narrowing the gap between laboratory study and routine clinical applications.

Millions of people around the world need assistance with feeding. Robotic feeding systems offer the potential to enhance autonomy and quality of life for individuals with impairments and reduce caregiver workload. However, their widespread adoption has been limited by technical challenges such as estimating bite timing, the appropriate moment for the robot to transfer food to a user’s mouth. In this work, we introduce WAFFLE: Wearable Approach For Feeding with LEarned Bite Timing, a system that accurately predicts bite timing by leveraging wearable sensor data to be highly reactive to natural user cues such as head movements, chewing, and talking. We train a supervised regression model on bite timing data from 14 participants and incorporate a user-adjustable assertiveness threshold to convert predictions into proceed or stop commands. In a study with 15 participants without motor impairments with the Obi feeding robot, WAFFLE performs statistically on par with or better than baseline methods across measures of feeling of control, robot understanding, and workload, and is preferred by the majority of participants for both individual and social dining. We further demonstrate WAFFLE’s generalizability in a study with 2 participants with motor impairments in their home environments using a Kinova 7DOF robot. Our findings support WAFFLE’s effectiveness in enabling natural, reactive bite timing that generalizes across users, robot hardware, robot positioning, feeding trajectories, foods, and both individual and social dining contexts. Videos are located at https://sites.google.com/view/bitetiming/.

ABSTRACT 3‐D skin surface reconstruction offers promise for objective and quantitative dermatological assessment, but no portable, high‐resolution device exists that has been validated and used for depth reconstruction across various body locations. We present a compact 3‐D skin reconstruction probe based on GelSight tactile imaging with a custom elastic gel and a learning‐based reconstruction algorithm for micron‐level wrinkle height estimation. Our probe, integrated with a load cell for force sensing for consistent contact, achieves a mean absolute error of 12.55 on wrinkle‐like test objects. In a study with 15 participants without skin disorders, we provide the first validated wrinkle depth metrics across multiple body regions. We further demonstrate statistically significant reductions in wrinkle height at three locations following over‐the‐counter moisturizer application. Our work offers a validated tool for clinical and cosmetic skin analysis, with potential applications in diagnosis, treatment monitoring, and skincare efficacy evaluation.

Software-defined radios and cognitive radios can switch their operating frequency across a wide range. Although analog and digital circuits enable cross-frequency operation of much of the RF frontend, there is a lack of practical antennas that are frequency-reconfigurable over such a wide range. Prior rigid antennas are either narrow-band with high radiation efficiency, or wide-band with low radiation efficiency. This study presents a soft, stretchable, compact, omnidirectional liquid metal composite antenna system to enable high radiation efficiency when reconfiguring to different frequencies across a wide band. On the hardware front, an extendable multi-branch stretchable liquid metal based antenna geometry is optimized that preserves high efficiency when switching across a wide range of frequencies. On the software front, a control algorithm is incorporated that ensures that the antenna automatically configures to the optimal received signal strength. The antenna design enables a strain limit of 33% for two independent actuatable branches, further enabling a switchable operating frequency range across 1-5 GHz. Experimental characterization demonstrates that PASTA reduces the radiation loss by 10-20 dB over the operating frequency range compared to state-of-the-art rigid and flexible antennas.

Printed Electronics In a collaboration between Carnegie Mellon, Gonzaga, and Coimbra Universities, Carmel Majidi and co-workers introduce in their Research Article (10.1002/adfm.202519702), a multi-material soft composite ink and direct ink write manufacturing process capable of producing highly functional, multilayer flexible and stretchable wearables designed to digitally augment human sensing and vision.

There has been tremendous progress in soft printable electronics with liquid metal (LM) for creating elastic, stretchable, and thin-film circuitry. However, such innovation risks outpacing sustainability. As the world moves toward softer, thinner, and more integrated electronics, these advancements may accelerate the exponential rise of electronic waste-projected to surpass 74 million tons by 2030 and reach ≈120 million tons by 2050. Therefore, in developing LM-based electronics, it is critical to address the urgent need for eco-friendly recycling strategies. This review proposes a paradigm shift toward recyclable, repairable, renewable, and resilient ("4R") soft electronics enabled by LM composites. Gallium-based alloys and their eutectics, such as Galinstan and EGaIn, offer a rare convergence of high conductivity, fluidic deformability, and recyclability-ideally suited for stretchable circuits, soft robotics, and wearables. Yet, their promise remains incomplete without scalable, eco-friendly recovery methods. Advanced extraction and recycling approaches are examined, including mechanical-chemical and physicochemical methods using deep eutectic solvents (DESs) and ionic liquids for selective recovery with minimal environmental impact. Emphasis is placed on materials selection, substrate compatibility, and integration strategies that enable circular lifecycles. LM composites offer a path to redefine electronics-not only for performance but also for durability, self-healing, and sustainability.

Wireless research has advanced in utilizing channel diversity and beamforming for more efficient communication, sensing, and harvesting ambient energy. We demonstrate our wireless robotic platform that utilizes radio-frequency beamforming for robot actuation. The platform delivers a maximum of 60 watts of power accurately towards soft robotic actuators by efficient frequency-aware beamforming. We also engineer soft actuators to absorb microwaves of specific frequencies to enable selective actuation. In this demonstration, we show a simplified version of our system that achieves frequency-selective actuation of two actuators to enable simple robot locomotion.

Abstract Skin‐interfacing electrodes are central to health monitoring, rehabilitation, stimulation, gaming, and AR/VR. Next‐generation wearables demand high signal quality, conformability, and long‐term comfort, yet existing gel and dry electrodes fall short. It is presented microprinted biphasic soft electrodes whose 3D microstructure and embedded liquid‐metal droplets enlarge true skin contact and reduce contact impedance to 4.7 kΩ—a 14.1‐fold reduction relative to Ag/AgCl (66.1 kΩ). The electrodes enable forehead EEG detection of the Berger effect, previously impractical with conventional electrodes, and deliver robust SNR during prolonged wear and motion. A simple model links infill geometry to effective surface area, guiding optimization and validated in n = 10 subjects. These versatile electrodes enable high‐fidelity monitoring of muscle, heart, and brain signals for healthcare and advanced human‐machine interfaces (HMIs), establishing a transformative paradigm for wearable bioelectronics in healthcare, neuroprosthetics, and advanced HMIs.

Millions of people around the world need assistance with feeding. Robotic feeding systems offer the potential to enhance autonomy and quality of life for individuals with impairments and reduce caregiver workload. However, their widespread adoption has been limited by technical challenges such as estimating bite timing, the appropriate moment for the robot to transfer food to a user's mouth. In this work, we introduce WAFFLE: Wearable Approach For Feeding with LEarned bite timing, a system that accurately predicts bite timing by leveraging wearable sensor data to be highly reactive to natural user cues such as head movements, chewing, and talking. We train a supervised regression model on bite timing data from 14 participants and incorporate a user-adjustable assertiveness threshold to convert predictions into proceed or stop commands. In a study with 15 participants without motor impairments with the Obi feeding robot, WAFFLE performs statistically on par with or better than baseline methods across measures of feeling of control, robot understanding, and workload, and is preferred by the majority of participants for both individual and social dining. We further demonstrate WAFFLE's generalizability in a study with 2 participants with motor impairments in their home environments using a Kinova 7DOF robot. Our findings support WAFFLE's effectiveness in enabling natural, reactive bite timing that generalizes across users, robot hardware, robot positioning, feeding trajectories, foods, and both individual and social dining contexts.

Abstract Augmenting human sensing with wearable electronics has the potential to revolutionize how to interact with the environment, enabling “superhuman” abilities to incorporate new data streams into the daily experience. However, existing soft conductive inks and printing methods for soft electronics microfabrication are limited in their ability to produce the complex multi‐layer circuit architectures required for supporting CMOS cameras and other multiplexed sensing capabilities. Here, this challenge is addressed by introducing a soft‐matter conductive ink that is compatible with direct ink write (DIW) printing for creating complex, sticker‐like, multilayer circuits for augmented human sensing and vision. The ink is composed of fluidic, elastomeric, and nanomaterial phases that enable unique combinations of DIW‐compatible rheology, post‐print elasticity, and electrical conductivity that are tailored for rapid printing of intrinsically soft, multi‐layered digital circuits. The use of electrical overpasses obviates the need for complicated vias, enabling sophisticated circuits to be fabricated in a simple, rapid process with no manual steps. This fabrication approach is validated with three functional demonstrations: a wearable pressure and temperature sensor, a high‐resolution pressure sensing array, and a wearable CMOS camera circuit that livestreams a high‐resolution (1024 × 768) and low‐latency (<0.2 ms) video feed to an augmented reality headset.

Sensor arrays for soft and stretchable electronics typically require a large number of electrical leads that scale with the number of sensing nodes within an array. Whereas rigid electronics can incorporate transistors and complex CMOS architectures for sensor multiplexing, soft material circuits are largely limited in how individual sensing nodes can be addressed. To overcome this challenge, we introduce an approach that utilizes a variable-frequency notch filter to encapsulate resistive or capacitive soft-material and rigid sensors for use in the detection of phenomena including material strain, applied force, and temperature. This design builds on the rich and growing body of research using capacitive and resistive soft material sensors, providing a customizable framework on which to build distributed sensing networks. In particular, we investigate how incorporating strain-sensitive components into a notch filter enables tuning of the sensor’s sensitivity, independent of the particular capacitor or resistor design, allowing sensor responses to be isolated and allow network reconfiguration without calibration. Our approach offers a straightforward means of connecting multiple sensors, even of different types, in a common sensing network with minimal wiring and interface requirements. The active filters were designed to fit in a small area and require minimal rigid components to facilitate incorporation into highly stretchable or bendable soft systems.

Three-dimensional (3-D) skin surface reconstruction offers promise for objective and quantitative dermatological assessment, but no portable, high-resolution device exists that has been validated and used for depth reconstruction across various body locations. We present a compact 3-D skin reconstruction probe based on GelSight tactile imaging with a custom elastic gel and a learning-based reconstruction algorithm for micron-level wrinkle height estimation. Our probe, integrated into a handheld probe with force sensing for consistent contact, achieves a mean absolute error of 12.55 micron on wrinkle-like test objects. In a study with 15 participants without skin disorders, we provide the first validated wrinkle depth metrics across multiple body regions. We further demonstrate statistically significant reductions in wrinkle height at three locations following over-the-counter moisturizer application. Our work offers a validated tool for clinical and cosmetic skin analysis, with potential applications in diagnosis, treatment monitoring, and skincare efficacy evaluation.

Abstract We report that a dielectric polymer chain, constrained at both ends, sharply collapses when exposed to a high electric field. The chain collapse is driven by nonlocal dipolar interactions and anisotropic polarization of monomers, a characteristic of real polymers that prior theories were unable to incorporate. Once collapsed, a large number of chain monomers accumulate at the center location between the chain ends, locally increasing the electric field and polarization by orders of magnitude. The chain collapse is sensitive to the orientation of the applied electric field and chain stretch. Our findings not only offer new ways for rapid actuation and sensing but also provide a pathway to discover the critical physics behind instabilities and electrical breakdown in dielectric polymers.

Abstract This work presents a soft, flexible, and skin-adhering electrode for a wearable bio-signal acquisition device. This dry micropatterned electrode is made with an ink composed of styrene-isoprene-styrene block co-polymer elastomer embedded with silver microflakes. Its applicability for surface bioelectronic signal acquisition is examined and compared to clinical grade gel electrodes. These sensors show a skin-electrode interface impedance of less than 30 kOhm at 10 kHz, which is comparable to standard gel electrodes. Inspired by the setal stalks that gecko lizards use to adhere to surfaces, these electrodes are patterned with vertically aligned micro-pillars that, through contact splitting, allow the electrode to adhere to human skin. These micro-structures have also been designed to make the electrodes robust for acquiring bio-potentials even in the presence of contaminating particles or other foreign objects at the interface between the skin and electrode. Together, these characteristics make the soft, micro-structured electrodes well-suited for wearable bio-signal acquisition devices that can be re-used or applied to skin when it is covered with dirt or other contaminants.

Soft and miniaturized robots possess the capability to operate inside narrow, confined environments. However, powering soft robots inside these environments with on-board batteries or wired connections to external power supplies can significantly restrain their mobility. Similarly, wireless actuation approaches are constrained by near-field actuation, line-of-sight operation, or indiscriminate actuation of many actuators. To provide higher mobility for wireless soft robot to operate inside non-line-of-sight scenarios, we present a radio-frequency system that introduces frequency-selective actuation of liquid crystal elastomer actuators. We create liquid crystalline elastomer actuators with a low actuation temperature and embed them with conductive traces that resonate and heat by selected frequencies of radio-frequency excitation in the 2.40 GHz range. We further develop a wireless actuation platform that infers the wireless channel and beamforms towards the actuator to achieve efficient beamforming. Demonstrations show our system is capable of selectively actuating different actuators while the robot is in motion and obstructed by occlusions.

Auxetic Intelligence In article number 2502222, Il-Kwon Oh and co-workers present a glove-type auxetic wearable haptic interface (GAWH), employing auxetic meta-structures integrated with shape memory alloy wires. This design delivers multimodal tactile feedback–static pressure, dynamic vibration, and variable stiffness–to all finger joints. The GAWH exhibits adaptive fitting with embodied mechanical intelligence, conforming comfortably to diverse finger sizes and motions. Its capabilities enable intuitive grasping interactions in virtual environments and facilitate enhanced human-robot interactions through nuanced haptic-assisted teleoperation.

Haptic surfaces can convey substantial tactile information, benefiting diverse applications such as robotic teleoperation, medical simulation, virtual reality (VR), and education. A critical challenge in haptic device technology is generating haptic sensations associated with mechanical compliance. Employing soft materials for adaptable interfaces that dynamically alternate between soft and hard tactile feedback further complicates this challenge. In this study, we address stiffness-controlled feedback through the design and fabrication of magnetically actuatable micro-structures using a UV-curable magnetic elastomer. These micro-structures, covered by a thin magnetic elastomer layer, form a pad that functions as an encounter-type haptic display when magnetically stimulated. User evaluations demonstrated clear differentiation in perceived softness, confirming the effectiveness of the designed haptic interface.

Abstract The global prevalence of hypertension affects billions of individuals, yet only 21% of those impacted manage their condition. Although commercial non‐clinical blood pressure monitoring devices are available, they are bulky and inconvenient, so they are incapable of continuous and long‐term monitoring. Consequently, these devices are not appropriate for tracking dynamic health changes influenced by individual behaviors and lifestyles, which can lead to severe complications. Therefore, continuous blood pressure monitoring in everyday environments is essential for the prevention and treatment of blood pressure‐related issues. Here, a continuous blood pressure monitoring system is introduced that features a soft and compact form factor, leveraging the inherent metallic conductivity, stretchability, and biocompatibility of laser‐sintered liquid metal conductors. The liquid metal conductors capture both mechanical and electrical signals originating from the heart due to their high sensitivity to detect blood pulses and fluidity‐driven conformability, offering continuous blood pressure signals. Moreover, the continuous blood pressure is measured non‐invasively without additional components such as a cuff. As proof, cardiac activity is continuously monitored before and after exercise, including the recovery phase. The innovative system is anticipated to address the urgent needs of billions of individuals suffering from cardiovascular issues.

Soft robotics offers the opportunity to create dexterous machines that can safely handle delicate objects. Grippers made from deformable actuators and compliant materials can deform around the objects with which they come in contact. The continuum mechanics of flexible manipulators can be leveraged for safe manipulation tasks such as twisting and grasping during manufacturing. However, to achieve this goal, contact sensing and controls for manipulators in these soft systems still remain a challenge in the field. This paper demonstrates a shape-memory alloy actuated soft gripper, with each finger able to bend about multiple axes. This enables the soft gripper to perform twisting tasks and handle various and fragile objects. Using capacitive bend sensors, we also demonstrate that the measured impedance of motion can be used as a proxy for contact, greatly increasing performance in a delicate manipulation task

Matching the swimming efficiency and agility of fish has remained an elusive goal in underwater robotics. Such locomotion capabilities rely on complex vortex interactions between the robot's body and the surrounding fluid. However, simulating these dynamics, which are governed by coupled ordinary and partial differential equations, is significantly more difficult than the multi-body dynamics of classical rigid robotic systems. We present a differentiable framework for simulating strongly coupled fluid-robot multiphysics as a unified optimization problem. The coupled manipulator and incompressible Navier-Stokes equations are derived together from a single Lagrangian using the principle of least action. We employ discrete variational mechanics to derive a stable, well-conditioned, and physically accurate scheme for jointly simulating articulated bodies and the surrounding fluid. We leverage the implicit function theorem to compute derivatives of the fully coupled dynamics. Using this simulator and its gradients, we realize undulating swimming gaits and optimize a highly dynamic C-start escape maneuver for a bioinspired eel robot. We validate both gaits on physical hardware, demonstrating successful sim-to-real transfer. Simulation code, hardware data, and schematics for the eel robot can be found here: https://unified-fluid-robot-multiphysics.github.io/

Abstract Multimaterial 3D printing enables the integration of materials with vastly different mechanical properties. Yet, in practice, existing multimaterial 3D printing methods are often constrained in the range of achievable properties within a single print, necessitating continued reliance on manual assembly for several applications such as soft robotics. Various material systems are developed to address this limitation by incorporating novel chemistries and/or modifying process parameters. Complementing these advancements, a thiol‐ene‐based photopolymer resin is introduced for multimaterial 3D printing that broadens the achievable range of elastic modulus and hardness, spanning five orders of magnitude and two full Shore hardness scales, respectively. It is defined by two extreme materials with disparate cross‐linking densities, but compatible photopolymerization mechanisms, providing access to intermediate properties upon blending. Moreover, continuous gradients are demonstrated through in situ material mixing and controlled diffusion, even at sub‐filament scales. The versatile resin is compatible with direct ink writing, vat photopolymerization, and potentially other processes like material jetting, opening opportunities in soft electronics, personal protective equipment, and biomedical implants.

Abstract Soft robots, made from flexible materials, offer excellent shock absorption and recovery but remain vulnerable to cuts and punctures. To address this limitation, research on self-healing materials has gained significant attention. This paper systematically classifies key components of self-healing soft robots, focusing on self-healing polymers and damage-detection sensors while analyzing recent research trends. A fully functional self-healing soft robot requires an integrated system where self-healing materials, damage detection, and autonomous recovery mechanisms work seamlessly together. This paper categorizes self-healing polymers into intrinsic and extrinsic mechanisms and classifies self-healing sensors based on their damage detection methods—conductive, capacitive, optical, and pneumatic. These sensors are crucial in assessing damage and optimizing the healing process. Enhancing reliability, stability, adaptability, durability, healing speed, and autonomy is essential for practical implementation. Achieving these require advanced sensor characterization, nonlinear modeling, autonomous control, and the integration of self-heating and energy-harvesting technologies. This paper advances self-healing soft robot development toward real-world applications by emphasizing an integrated design approach.

Ensuring safety via safety filters in real-world robotics presents significant challenges, particularly when the system dynamics is complex or unavailable. To handle this issue, learning-based safety filters recently gained popularity, which can be classified as model-based and model-free methods. Existing model-based approaches requires various assumptions on system model (e.g., control-affine), which limits their application in complex systems, and existing model-free approaches need substantial modifications to standard RL algorithms and lack versatility. This paper proposes a simple, plugin-and-play, and effective model-free safety filter learning framework. We introduce a novel reward formulation and use Q-learning to learn Q-value functions to safeguard arbitrary task specific nominal policies via filtering out their potentially unsafe actions. Due to its model-free nature and simplicity, our framework can be seamlessly integrated with various RL algorithms. We validate the proposed approach through simulations on double integrator and Dubin's car systems and demonstrate its effectiveness in real-world experiments with a soft robotic limb.

Nuanced and multidimensional finger based haptic feedback is of particular interest for virtual reality (VR) and augmented reality (AR) applications because it can well recreate real-life sensations. However, existing haptic approaches are typically limited by rigid and bulky equipment that significantly degrades wearability and the overall user experience. Here, we present a new class of lightweight and flexible finger haptic systems powered by shape memory alloy (SMA) actuators specifically designed to provide highly controllable and nuanced cutaneous feedback. Composed of a soft 3D-printed flexible finger cap structure and elastic cover, the device’s material architecture is optimized to allow for shape conformity to human fingers, enhancing wearability and user comfort. Furthermore, it enables eleven distinctive motions that allow for a diverse variety of haptic sensations crucial to advanced VR and AR applications.

Intelligent digital apparel, which integrates electronic functionalities into clothing, represents the future of healthcare and ubiquitous control in wearable devices. Realizing such apparel necessitates developing meter-scale conductive fibers with high toughness, conductivity, stable conductance under deformation, and mechanical durability. In this study, we present a heterostructure printing method capable of producing meter-scale (~50 m) biphasic conductive fibers that meet these criteria. Our approach involves encapsulating deformable liquid metal particles (LMPs) within a functionalized thermoplastic polyurethane matrix. This encapsulation induces in situ assembly of LMPs during fiber formation, creating a heterostructure that seamlessly integrates the matrix's durability with the LMPs' superior electrical performance. Unlike rigid conductive materials, deformable LMPs offer stretchability and toughness with a low gauge factor. Through precise twisting using an engineered annealing machine, multiple fiber strands are transformed into robust, electrically stable meter-scale electrodes. This advancement enhances their practicality in various intelligent digital apparel applications, such as stretchable displays, wearable healthcare systems, and digital controls.

Abstract Soft dielectric elastomers that can exhibit extremely large deformations under the action of an electric field enable applications such as soft robotics, biomedical devices, and energy harvesting among others. A key impediment in the use of dielectric elastomers is failure through instability mechanisms or dielectric breakdown. In this work, by using a group theory-based approach, we provide a closed-form solution to the bifurcation problem of a paradigmatical elastomer actuator and discover an interesting result: at a critical electric field, the elastomer becomes impervious to Treloar–Kearsley instability. This limit is reached prior to the typical dielectric breakdown threshold. Our results thus establish a regime of electrical and mechanical loads where the dielectric elastomer is invulnerable to all common failure modes.

High Resolution Image Download MS PowerPoint Slide Combining droplets of liquid metal (LM) with nanomaterials often introduces synergistic thermal or electrical properties that are not found in the constituent materials alone. However, in these existing systems, LM droplets maintain a statistically uniform dispersion and are not capable of self-assembly or aggregation. These composites are limited by their need for high volume fractions of LM (>60 vol %) to achieve high thermal properties, introducing LM leaking as a drawback for thermal management and wearable electronic applications. In this work, we show that coating nanoscale droplets of eutectic gallium–indium (EGaIn) LM with small volume fractions of Ti 3 C 2 T x MXenes (0.25 vol %) results in a unique LM morphology in which droplets self-assemble to form semisolid aggregates. This is accomplished by wrapping MXene sheets around individual LM droplets to create “sticky” particles that form self-assembled aggregates when mixed with a silicone oil. By introducing aggregation as a design parameter in soft LM composites, the thermal and electric resistance of the composite is shown to change dramatically. In contrast to silicone-based composites containing LM droplets or MXene nanosheets alone, these MXene-LM-silicone-based composites exhibit an exponential increase in thermal and electrical conductivity with decreasing interfacial thickness with significantly lower LM volume fractions (25 vol %) while avoiding LM rupture and bleed-out. This could enable more effective composites, reducing the amount of filler material required for thermal interface materials (TIM) and printed electronics.

Most walking organisms tend to have relatively light limbs and heavy bodies in order to facilitate rapid limb motion. However, the limbs of brittle stars (Class Ophiuroidea) are primarily comprised of dense skeletal elements, with potentially much higher mass and density compared to the body disk. To date, little is understood about how the relatively unique distribution of mass in these animals influences their locomotion. In this work, we use a brittle star inspired soft robot and computational modeling to examine how the distribution of mass and density in brittle stars affects their movement. The soft robot is fully untethered, powered using embedded shape memory alloy actuators, and designed based on the morphology of a natural brittle star. Computational simulations of the brittle star model are performed in a differentiable robotics physics engine in conjunction with an iterative linear quadratic regulator to explore the relationship between different mass distributions and their optimal gaits. The results from both methods indicate that there are robust physical advantages to having the majority of the mass concentrated in the limbs for brittle star-like locomotion, providing insight into the physical forces at play.

Abstract The versatility of fingers stems from their ability to interact with surroundings by taking poses, performing various grasping modes, and making dynamic bending motions. However, conventional glove‐type wearable interfaces often fail to fully engage the structural diversity and dynamic motions of fingers, providing tactile feedback to limited areas and restricting the full utilization of finger versatility for rich interactions. Here, a glove‐type auxetic wearable haptic (GAWH) interface that employs an auxetic meta‐design to offer embodied mechanical intelligence is introduced. This unique design ensures the positioning of haptic‐effective regions on all finger joints with high conformability, despite the variations in finger sizes among individuals. The GAWH delivers multimodal tactile feedback including static pressure, dynamic vibration, and variable stiffness with high spatial resolution to all joints. This enables users to interact with objects through diverse grasping modes and enhanced perception. Consequently, the GAWH provides real‐time meta‐linkage to virtual reality, allowing users to perform grasping tasks intuitively. Additionally, the GAWH facilitates interaction between humans and robots by enabling users to understand how robots feel during haptic‐assisted teleoperation. Overall, this work establishes a foundation for wearable haptic interfaces that can simultaneously enhance versatility, adaptability, and diversity, leveraging the full potential of the versatile fingers.

Physically assistive robots in home environments can enhance the autonomy of individuals with impairments, allowing them to regain the ability to conduct self-care and household tasks. Individuals with physical limitations may find existing interfaces challenging to use, highlighting the need for novel interfaces that can effectively support them. In this work, we present insights on the design and evaluation of an active control wearable interface named HAT, Head-Worn Assistive Teleoperation. To tackle challenges in user workload while using such interfaces, we propose and evaluate a shared control algorithm named Driver Assistance. Finally, we introduce the concept of passive control, in which wearable interfaces detect implicit human signals to inform and guide robotic actions during caregiving tasks, with the aim of reducing user workload while potentially preserving the feeling of control.

Abstract Compliant mechanisms with reconfigurable degrees of freedom are gaining attention in the development of kinesthetic haptic devices, robotic systems, and mechanical metamaterials. However, available devices exhibit limited programmability and form-customizability, restricting their versatility. To address this gap, we propose a metastructure concept featuring reconfigurable motional freedom and tunable stiffness, adaptable to various form factors and applications. These devices incorporate passive flexures and actively stiffness-changing rods to modify kinematic freedom. A rational design pipeline informs the flexures’ topological arrangements, geometric parameters, and control signals based on targeted mobilities, enabling the creation of unitary joints with up to six degrees of freedom. Our demonstrative application examples include a wrist device that has an effective stiffness of 0.370 Nm/deg (unlocked state, 5% displacement) to 2.278 Nm/deg (locked state, 1% displacement) to enable dynamic joint mobility control, a haptic thimble device (2.27-52.815 Nmm −1 at 1% displacement) that mimics the sensation of touching physical materials ranging from soft gel to metal surfaces, and a wearable device composed of multiple joints tailored for the arm and hand to augment haptic experiences or facilitate muscle training. We believe the presented method can help democratize compliant metastructures development and expand their versatility for broader contexts.

Dielectric elastomers have significant potential for new technologies, ranging from soft robots to biomedical devices, driven by their ability to display complex shape changes in response to electrical stimulus. However, an important shortcoming of current realizations is that large voltages are required for useful actuation strains. This work proposes, and demonstrates through theory and numerical simulations, a strategy to achieve large and controlled actuation by exploiting the electromechanical analogue of the Treloar-Kearsley (TK) instability. The key idea is to use the fact that the TK instability is a symmetry-breaking bifurcation, which implies the existence of a symmetry-driven constant-energy region in the energy landscape. This provides for nonlinear soft modes with large deformations that can be accessed with very small external stimulus, which is achieved here by applying a small in-plane electric field. First, the bifurcation and postbifurcation behavior of the electromechanical TK instability are established theoretically in the idealized setting of uniform deformation and electric field. Next, building on this, a finite-element analysis of a realistic geometry with patterned top and bottom electrodes is applied to demonstrate large and soft shape changes driven by small voltage differences across the electrodes.

holes, offering advantages beyond conventional PCB technologies. This approach allows users to insulate or connect top and bottom circuits as needed, even when the circuits overlap or intersect. The inherent properties of liquid metal facilitate pressure-induced sintering, working in synergy with textiles to provide users with the ability to dynamically alter circuits. This unique feature enables real-time customization, allowing for the addition, removal, or replacement of circuits through straightforward cutting and stitching processes. Demonstrating these characteristics, we showcase diverse applications, including a wristband with a replaceable LED indicator circuit, a reversible teddy bear cloth with two distinct functions, and a customizable DIY heating glove. This double-sided textile circuit that is patterned with pressure-controlled drawing offers new possibilities for multifunctional wearable electronics, bridging the gap between traditional PCBs and flexible smart textiles.

Abstract Electrically conductive hydrogels are a promising class of materials for soft electronics and robotics that mimic the mechanics of natural biological tissue. However, these materials are typically derived from petrochemical sources and their production typically involves hazardous solvents and monomers that limit accessibility and environmental compatibility. This study introduces a biomaterial hydrogel composite in which a percolating network of silver microflakes is suspended in a natural, gelatin‐based matrix. The composite is primarily composed of inexpensive, food‐safe ingredients and fabrication is achieved using accessible consumer‐grade equipment. The resulting material system is mechanically soft, stretchable up to 470% strain, and highly conductive up to 3.1 × 10 3 S cm −1 , with properties that can be tailored based on material composition and processing conditions. In addition to experimental characterization of its material properties, this conductive gelatin composite is shown to be applicable for a variety of uses cases in soft matter circuitry and bioelectronics.

Ensuring safety via safety filters in real-world robotics presents significant challenges, particularly when the system dynamics is complex or unavailable. To handle this issue, learning-based safety filters recently gained popularity, which can be classified as model-based and model-free methods. Existing model-based approaches requires various assumptions on system model (e.g., control-affine), which limits their application in complex systems, and existing model-free approaches need substantial modifications to standard RL algorithms and lack versatility. This paper proposes a simple, plugin-and-play, and effective model-free safety filter learning framework. We introduce a novel reward formulation and use Q-learning to learn Q-value functions to safeguard arbitrary task specific nominal policies via filtering out their potentially unsafe actions. The threshold used in the filtering process is supported by our theoretical analysis. Due to its model-free nature and simplicity, our framework can be seamlessly integrated with various RL algorithms. We validate the proposed approach through simulations on double integrator and Dubin's car systems and demonstrate its effectiveness in real-world experiments with a soft robotic limb.

Physically assistive robots present an opportunity to significantly increase the well-being and independence of individuals with motor impairments or other forms of disability who are unable to complete activities of daily living. Speech interfaces, especially ones that utilize Large Language Models (LLMs), can enable individuals to effectively and naturally communicate high-level commands and nuanced preferences to robots. Frameworks for integrating LLMs as interfaces to robots for high level task planning and code generation have been proposed, but fail to incorporate human-centric considerations which are essential while developing assistive interfaces. In this work, we present a framework for incorporating LLMs as speech interfaces for physically assistive robots, constructed iteratively with 3 stages of testing involving a feeding robot, culminating in an evaluation with 11 older adults at an independent living facility. We use both quantitative and qualitative data from the final study to validate our framework and additionally provide design guidelines for using LLMs as speech interfaces for assistive robots. Videos, code, and supporting files are located on our project website1

Physically assistive robots present an opportunity to significantly increase the well-being and independence of individuals with motor impairments or other forms of disability who are unable to complete activities of daily living (ADLs). Speech interfaces, especially ones that utilize Large Language Models (LLMs), can enable individuals to effectively and naturally communicate high-level commands and nuanced preferences to robots. In this work, we demonstrate an LLM-based speech interface for a commercially available assistive feeding robot. Our system is based on an iteratively designed framework, from the paper “VoicePilot: Harnessing LLMs as Speech Interfaces for Physically Assistive Robots,” that incorporates human-centric elements for integrating LLMs as interfaces for robots. It has been evaluated through a user study with 11 older adults at an independent living facility. Videos are located on our project website1