Keywords: Soft Sensors and Actuators, Soft Robot Materials and Design, Biomimetics
Abstract: Liquid crystal elastomers (LCE) have emerged as a fascinating candidate for soft actuation. The shape memory polymer undergoes a thermo-responsive nematic to isotropic transition, capable of producing excellent reversible axial actuation at accessible temperatures. However, liquid crystal elastomer actuators have been limited in their force capabilities, which are lower than other hallmark soft actuators. Herein, we introduce variable stiffness particles, in the form of low-melting point alloy Field's metal, into an LCE matrix to induce variable stiffness properties and enhanced actuation stresses. We find that when we include low volume concentrations of FM (<10 vol.%) into LCE, we achieve a 5X increase in actuation stress. Furthermore, we find that incorporating around 30 vol.% FM in LCE (yielding a FM-LCE composite) imparts excellent variable stiffness properties on the LCE. In a cold state FM-LCE acts similar to a metal with high stiffness, but when the FM is melted, the composite mechanically behaves like the surrounding LCE matrix. Introducing more functionalities in the form of variable moduli and enhanced actuation stresses in LCE will enable them to be more applicable to a variety of disciplines.

Keywords: Grippers and Other End-Effectors
Abstract: Soft robotics is an emerging technology in which engineers create flexible devices for use in a variety of applications. In order to advance the wide adoption of soft robots, ensuring their trustworthiness is essential; if soft robots are not trusted, they will not be used to their full potential. In order to demonstrate trustworthiness, a specification needs to be formulated to define what is trustworthy. However, even for soft robotic grippers, which is one of the most mature areas in soft robotics, the soft robotics community has so far given very little attention to formulating specifications. In this work, we discuss the importance of developing specifications during development of soft robotic systems, and present an extensive example specification for a soft gripper for pick-and-place tasks for grocery items. The proposed specification covers both functional and non-functional requirements, such as reliability, safety, adaptability, predictability, ethics, and regulations. We also highlight the need to promote verifiability as a first-class objective in the design of a soft gripper.

Keywords: Modeling, Control, and Learning for Soft Robots
Abstract: Knowing the state of a robot is critical for many problems, such as feedback control. For continuum robots, state estimation is an incredible challenge. First, the motion of a continuum robot involves many kinematic states, including poses, strains, and velocities. Second, all these states are infinite-dimensional due to the robot's flexible property. It has remained unclear whether these infinite-dimensional states are observable at all using existing sensing techniques. Recently, we presented a solution to this challenge. It was a mechanics-based dynamic state estimation algorithm, called a Cosserat theoretic boundary observer, which could recover all the infinite-dimensional robot states by only measuring the velocity twist of the tip. In this work, we generalize the algorithm to incorporate tip pose measurements for more tuning freedom. We also validate this algorithm offline using experimental data recorded from a tendon-driven continuum robot. We feed the recorded tendon force and tip measurements into a numerical solver of the Cosserat rod model based on our robot. It is observed that, even with purposely deviated initialization, the state estimates by our algorithm quickly converge to the recorded ground truth states and closely follow the robot's actual motion.

Keywords: Modeling, Control, and Learning for Soft Robots, Learning and Adaptive Systems, Tendon/Wire Mechanism
Abstract: Controlling soft mobile robots that perform limbless locomotion is costly to develop due to the need to consider friction and the complexity of movement mechanics. There are methods using reinforcement learning (RL) to create controllers for complex soft caterpillar robots. However, these often involve learning through simulation, and model inaccuracies can lead to reduced controller performance upon deployment. In this paper, we created a soft caterpillar robot driven by tendons with two motors and trained a controller using RL. By training with real soft robots without using simulations, we created a learning model that works effectively even with soft robots' complex dynamics, without performance degradation upon deployment. Using a model-based learning algorithm enabled quick policy learning, even with real robots that typically require time-consuming sampling. The learning model we developed could achieve locomotion in forward tasks after about one hour of training. After training, the actual robot was capable of moving at approximately 36.7 mm/s. To the best of our knowledge, this is the first instance of learning locomotion for a soft mobile robot's crawling using only a real robot.

Keywords: Biologically-Inspired Robots, Soft Robot Materials and Design, Biomimetics
Abstract: Pipelines, vital for fluid transport, pose an important yet challenging inspection task, particularly in small, flexible biological systems, that robots have yet to master. In this study, we explored the development of an innovative robot inspired by the ovipositor of parasitic wasps to navigate and inspect pipelines. The robot features a flexible locomotion system that adapts to different tube sizes and shapes through a mechanical inflation technique. The flexible locomotion system employs a reciprocating motion, in which groups of three sliders extend and retract in a cyclic fashion. In a proof-of-principle experiment, the robot locomotion efficiency demonstrated positive linear correlation (r=0.6434) with the diameter ratio (ratio of robot diameter to tube diameter). The robot showcased a remarkable ability to traverse tubes of different sizes, shapes and payloads with an average of (70%) locomotion efficiency across all testing conditions, at varying diameter ratios (0.7 ∼ 1.5). Furthermore, the mechanical inflation mechanism displayed substantial load-carrying capacity, producing considerable holding force of (13 N), equivalent to carrying a payload of (≈5.8 Kg) inclusive the robot weight. This novel soft robotic system shows promise for inspection and navigation within tubular confined spaces, particularly in scenarios requiring adaptability to different tube shapes, sizes, and load-carrying capacities. This novel design serves as a foundation for a new class of pipeline inspection robots that exhibit versatility across various pipeline environments, potentially including biological systems.

Keywords: Legged Robots, Sensor-based Control, Soft Robot Materials and Design
Abstract: In this study, a chair-type asymmetric tripedal low-rigidity robot was designed based on the three-legged chair character in the movie “Suzume” and its gait was generated. Its body structure consists of three legs that are asymmetric to the body, so it cannot be easily balanced. In addition, the actuator is a servo motor that can only feed-forward rotational angle commands and the sensor can only sense the robot’s posture quaternion. In such an asymmetric and imperfect body structure, we analyzed how gait is generated in walking and stand-up motions by generating gaits with two different methods: a method using linear completion to connect the postures necessary for the gait discovered through trial and error using the actual robot, and a method using the gait generated by reinforcement learning in the simulator and reflecting it to the actual robot. Both methods were able to generate gait that realized walking and stand-up motions, and interesting gait patterns were observed, which differed depending on the method, and were confirmed on the actual robot. Our code and demonstration videos are available here: https://github.com/shin0805/ChairTypeAsymmetricalTripedalRo bot.git

Keywords: Human-Centered Robotics, Biologically-Inspired Robots, Soft Robot Materials and Design
Abstract: Exploring the intersection of soft robotics and well-being, this research aims to alleviate stress and create restorative environments. Against the backdrop of rising mental health concerns, we introduce a novel soft robotic flower based on an inflatable membrane mechanism. By replicating the gentle movements of a natural lotus flower, this innovation aims to induce calming and meditative interactions. Our design leverages its simplicity on the use of soft materials, providing a nuanced alternative to earlier prototypes. Through a comprehensive evaluation, we demonstrate that this soft robotic flower significantly reduces stress and enhances well-being, setting the stage for a new era of therapeutic human-robot interaction.

Keywords: Soft Robot Applications, Hydraulic/Pneumatic Actuators, Prosthetics and Exoskeletons
Abstract: Bladder cancer and urinary dysfunctions represent significant challenges to human health and patients’ quality of life. Radical cystectomy is the primary treatment for muscle-invasive bladder cancer, but the associated bladder reconstruction surgery is unable to replicate the voiding capabilities of the native organ. In this scenario, the development of an implantable artificial bladder with controlled voiding capabilities has the potential to overcome the limitations of current clinical solutions. This study explores the synergy between an artificial bladder and an active artificial detrusor constituted by two soft actuators integrated on a soft sleeve. In particular, hydraulic balloon actuators have been selected to enhance the voiding performances of the artificial bladder while preserving softness for smooth and minimally-invasive implantation. Experimental results demonstrate that the actuators can reach a displacement of 31.0 ± 0.7 mm at 40 mL inlet volume and generate output forces of 26.4 ± 3.4 N, 6.8 ± 1.4 N, and 1.5 ± 0.4 N at 25%, 50%, and 75% of the maximum displacement, respectively. The soft detrusor enables promising voiding capabilities of the artificial bladder in terms of Post-Voiding Residual (PVR) volume, Voiding Efficiency (VE), and Voiding Time (VT), reaching values equal to 2.2 ± 2.0 mL, 98.9 ± 1.0 %, and 83.9 ± 4.7 s, respectively.

Keywords: Grippers and Other End-Effectors, Soft Robot Applications, Soft Robot Materials and Design
Abstract: Underwater soft grippers exhibit potential for applications such as monitoring, research, and object retrieval. However, existing underwater gripping techniques frequently cause disturbances to ecosystems. In response to this challenge, we present a novel underwater gripping framework comprising a lightweight gripper affixed to a specialized submarine pod deployable via drone. This approach minimizes water disturbance and enables efficient navigation to target areas, enhancing overall mission effectiveness. The pod allows for underwater motion and is characterized by four degrees of freedom. It is provided with a custom buoyancy system, two water pumps for differential thrust and two for pitching. The system allows for buoyancy adjustments up to a depth of 6 meters, as well as motion in the plane. The 3-fingered gripper is manufactured out of silicone and was successfully tested on objects with different shapes and sizes, demonstrating a maximum pulling force of up to 8 N when underwater. The reliability of the submarine pod was tested in a water tank by tracking its attitude and energy consumption during grasping maneuvers. The system also accomplished a successful mission in a lake, where it was deployed on a hexacopter. Overall, the integration of this system expands the operational capabilities of underwater grasping, makes grasping missions more efficient and easy to automate, as well as causing less disturbance to the water ecosystem.

Keywords: Soft Sensors and Actuators, Force and Tactile Sensing
Abstract: Soft robots can interact safely with humans and delicate objects, conform to complex shapes, and operate in unstructured environments, making them appealing for a wide variety of applications.

However, they are often limited in terms of loading and sensing capabilities. In this paper, we propose a soft actuator integrating a spine designed to improve its overall strength, thus providing a larger bending angle when the actuator is subjected to loads. The spine is stiffer than the main actuator body, thus reducing the yield of the structure when loads are applied.

Furthermore, the spine is designed to be mechanically interfaced with a tactile sensor, allowing the deformation of the actuator body to be measured. Experimental results show that the use of the spine improves both the loading and sensing performance compared to a spineless actuator. An additional analysis is also performed to find a proper material to manufacture the spine that provides the best tradeoff between sensing and loading performance.

Keywords: Tendon/Wire Mechanism, Compliant Joint/Mechanism
Abstract: In recent years, there has been much research on soft robots with variable stiffness mechanisms. There are various methods for achieving switching stiffness, one of which is the wire method. This consists of multiple segments with through holes, and a wire passed through the center. This method has an advantage in that the stiffness can be increased to the point where the segment material is about to break, and the time required for stiffness switching is very short. However, there have been few studies on methods to shorten the distance between segments, and the size of the whole body is problematic. In this study, we propose a novel bead shape for use in wire-driven variable stiffness mechanisms. This bead is shorter than the distance between beads, and less than half the bead radius, which is the limit of the conventional bead distance. Various theoretical bead design models are developed, and a prototype of the proposed mechanism is fabricated. The usefulness of the proposed bead shape is demonstrated experimentally by measuring the radius of curvature and holding torque, including comparisons with conventional beads. Finally, the characteristics revealed by the experiments are discussed, and future works and applications under development are presented.

Keywords: Soft Sensors and Actuators, Soft Robot Materials and Design, Soft Robot Applications
Abstract: In this study, we developed a new soft actuator "MORI-A" using buckling of 3D printed structure, which has the　characteristics of soft modular robot: (1) reconfigurable, (2) capable of various deformations, and (3) adaptable to the environment. This module can realize bending, Shrinking, and shearing. However, while this module combines convenience and a variety of deformation patterns, it incorporates only a silicone rubber contraction element for actuation, and is not capable of deformation through pressure-induced expansion. We therefore extended the MORI-A module to develop an actuator that can generate diametrically opposite deformation patterns under both pressure and vacuum. We were able to realize this principle using a single air channel and uniform hardness silicone rubber, while maintaining the simplicity of this module. Deformation characteristics were investigated during contraction, expansion, and bending. In addition, multiple modules were connected to change the direction of deformation and reproduce S-shaped bending.

Keywords: Compliant Joint/Mechanism, Additive Manufacturing, Legged Robots
Abstract: Laminate structures with nonlinear stiffness and controllable bending properties have applications as appendages for soft robot locomotion. In this paper, we develop a laminate-based structure for soft-robotics called Jagged Anisotropic Mechanically Jamming (JAMJam) sheets. JAMJam sheets are comprised of two flexible layers with rigid anisotropic structures 3D printed onto the flexible surface. The rigid elements between the flexible sheets are displaced longitudinally as the sheets are bent and when the elements make contact the bending stiffness increases and the resulting curvature can be controlled. In this work we study a simple example of a JAMJam sheet consisting of controllable bending curvature which is anisotropic along the forward and backward directions. We characterize the JAMJam behavior through force-displacement measurements and we implement the JAMJam sheets as appendages in a swimming robot. We compare the JAMJam mechanics under representative limb motions for a swimming robot and compare two types of appendages. Lastly, a robot utilizing the appendages is designed for shallow-water environments, which can both walk while semi-submerged at up to 9 cm/s and swim at 7.2 cm/s. The robot is capable of locomotion in both turf and sandy laboratory conditions, and can traverse over obstacles in its path.

Keywords: Soft Robot Applications, Haptics and Haptic Interfaces, Soft Sensors and Actuators
Abstract: Regular user interface screens can display dense and detailed information to human users but miss out on providing somatosensory stimuli that take full advantage of human spatial cognition. Therefore, the development of new haptic displays can strengthen human-machine communication by augmenting visual communication with tactile stimulation needed to transform information from digital to spatial/physical environments. Shape-changing interfaces, such as pin arrays and robotic surfaces, are one method for providing this spatial dimension of feedback; however, these displays are often either limited in maximum extension or require bulky mechanical components. In this paper, we present a compact pneumatically actuated soft growing pin for inflatable haptic interfaces. Each pin consists of a rigid, air-tight chamber, an inflatable fabric pin, and a passive spring-actuated reel mechanism. The device behavior was experimentally characterized, showing extension to 18.5 cm with relatively low pressure input (1.75 psi, 12.01 kPa), and the behavior was compared to the mathematical model of soft growing robots. The results showed that the extension of the soft pin can be accurately modeled and controlled using pressure as input. Finally, we demonstrate the feasibility of implementing individually actuated soft growing pins to create inflatable haptic surfaces.

Keywords: Soft Sensors and Actuators, Soft Robot Materials and Design, Soft Robot Applications
Abstract: Soft robotic systems are promising for a wide range of applications from locomotion to manipulation. In particular, fluidic dielectric elastomer actuators, such as Hydraulically Amplified Self-healing ELectrostatic (HASEL) actuators, demonstrate several compelling properties when compared to other soft robotic actuators such as lower power consumption, faster response times, and self-sensing capabilities. Current HASEL actuator research has thoroughly characterized linear HASEL actuators, but there is a lack of geometric characterization for bending HASEL actuators. This paper addresses this gap through the characterization of two important design parameters that affect out-of-plane bending of planar HASEL actuator designs. In particular, a sinusoidal wave pattern is parameterized by the period length and the minimum channel width. The ratio of the period length to channel width is shown to be a good predictor for the curvature of the HASEL actuators when bending out-of-plane. The experimentally derived relationship is then used to demonstrate different grasp types for various objects.

Keywords: Modeling, Control, and Learning for Soft Robots, Soft Robot Materials and Design, Soft Sensors and Actuators
Abstract: Soft manipulators have experienced a significant development over the past few years thanks to their versatility, compliance, and safety. Soft manipulators made of hyperelastic materials present a modeling challenge due to their combined geometric and material nonlinearity, alongside their leveraged compliance. This paper considers a thick cylindrical pneumatic actuator in the context of a soft parallel robot. The input pressure and the external axial force to which it is subjected are modeled. The analytical solutions for these models are then derived using the Yeoh strain energy density function. The advantage of using the Yeoh material model is guaranteeing the existence of an analytical solution regardless of the material used for the fabrication of the cylindrical actuator. The accuracy of the models is evaluated using Finite Element Analysis, and a sensitivity analysis is also carried out to test their robustness. Finally, experimental results are provided and the importance of the material characterization is highlighted. The purpose of this work is to lay the foundation for future studies in modeling a soft parallel robot with three soft thick cylindrical “legs".

Keywords: Modeling, Control, and Learning for Soft Robots, Physical Human-Robot Interaction, Prosthetics and Exoskeletons
Abstract: This paper presents a morphological learning framework to predict/estimate the contact forces between a human finger and a pneumatically actuated soft robotic digit by harvesting the reservoir computing power of the robot's soft bodies. During the actuation, the soft robot and human finger physically interact at multiple contact points; to predict the contact forces, a physical reservoir framework was established. The soft actuator segment was considered as the physical reservoir integrated into a linear readout. The deformation of the soft-bodied was visually tracked using an image processing code during actuation based on distributed attached markers while the force sensors measured the contact forces. The linear readout adds up the relative displacements to estimate the interaction forces. The readout weights were trained using a linear regression based on the desired output forces. On the other hand, a theoretical lumped model of the soft actuator segment, analogous to the markers, is developed using a network of mass-spring-damper. The model was simulated with experimentally measured input pressure force acting on the mass points while the spring length was calculated and added up using the linear readout. The physical reservoirs were trained for three individual soft actuators and exploited for estimating the contact forces. The effectiveness of the learning framework was experimentally and theoretically demonstrated, where it was able to estimate the actual contact forces between the soft robotic digit and the anthropomorphic finger model within an acceptable range of errors.

Keywords: Sensor-based Control, Soft Sensors and Actuators, Learning and Adaptive Systems
Abstract: The extraction of tactile information from an environment depends on the action, sensor design, and the environment itself. This makes identifying an optimal action for a given task, i.e. classification, challenging and typically requires extensive data-rich experiments. We propose utilizing Mutual Information (MI), a rapid-to-evaluate information content metric that considers the shared information between two random variables as a means of comparing say{tactile images} from different action-sensor-environment pairings. We propose this as a heuristic for selecting actions that offer the highest reliability or the greatest ability to classify different environments. As a demosntration, MI was used to search for an optimal action to distinguish two environments using Bayesian Optimization.The use of MI could guide the data-efficient evaluation and optimization of action selection and multi-modal sensor design.

Keywords: Additive Manufacturing, Soft Robot Applications, Soft Robot Materials and Design
Abstract: In this work, a novel, monolithic fabrication method is proposed and developed for laminated pneumatically actuated pouch motors. Combining the concepts of lamination, 3D printing and thermal bonding, the proposed process is capable of making low profile, pneumatic soft robots driven by pouch actuators. Two main procedures, the mask heat pressing and modified 3d printing were introduced and tested, preferred manufacturing parameters were obtained by experiments. The fabrication, due to its simplicity and flexibility, enables fast prototyping of pneumatic actuation systems, which could potentially contribute to future soft robotic implementations.

Keywords: Soft Robot Materials and Design, Simulation and Animation, Grippers and Other End-Effectors
Abstract: Computational design can excite the full potential of soft robotics, but it has the drawback of being highly nonlinear in terms of material, structure, and contact. To date, enthusiastic research interests have been demonstrated for individual soft fingers, but the frame design space (how each soft finger is assembled) remains largely unexplored. Computational design remains challenging for the finger-based soft gripper to grip across multiple geometrically distinct object types successfully. Including the design space for the gripper frame can bring huge difficulties for conventional optimization algorithms and fitness calculation methods due to the exponential growth of design space. This work proposes an automated computational design optimization framework that generates gripper diversity to individually grasp geometrically distinct object types based on a quality-diversity approach. This work first discusses a significantly large design space (28 design parameters) for a finger-based soft gripper, including the rarely-explored design space of finger arrangement. Then, a contact-based Finite Element Modelling (FEM) is proposed in SOFA to output high-fidelity grasping data for fitness evaluation and feature measurements. Finally, diverse gripper designs are obtained from the framework while considering features such as the volume and workspace of grippers. This work bridges the gap of computationally exploring the vast design space of finger-based soft grippers while grasping large geometrically distinct object types with a simple control scheme.

Keywords: Soft Robot Applications, Soft Robot Materials and Design, Biologically-Inspired Robots
Abstract: Soft robotics have been proposed for space applications due to their inherent compliance and novel locomotion methods. They provide promising solutions to current challenges in space environments; however, extreme temperature, radiation and vacuum conditions prevent the usage of traditional soft materials. We present a design methodology focused on exploring the use of flexible metallic modules as the basis for a modular soft robotic limb applicable to space. We generalise this limb into a family of locomotion platforms to enable the development and evaluation of novel forms of locomotion specifically targeted to soft robotics in space environments. Preliminary validation of the system's functionality in extreme temperatures is conducted by operating a cryogenically cooled soft robotic limb. Finally, we present our methods for evaluating soft robotic locomotion in analogue space environments using a soft robotic triped, a member of the family of proposed locomotion platforms. This research contributes to the ongoing exploration of soft robotics usage in space, offering a framework for the comparison and evaluation of novel locomotion strategies.

Keywords: Modeling, Control, and Learning for Soft Robots, Soft Robot Applications, Soft Robot Materials and Design
Abstract: This paper introduces a Cosserat rod based mathematical model for modeling a self-controllable variable curvature soft continuum robot. This soft continuum robot has a hollow inner channel and was developed with the ability to perform variable curvature utilizing a growing spine. The growing spine is able to grow and retract while modifies its stiffness through milli-size particle (glass bubble) granular jamming. This soft continuum robot can then perform continuous curvature variation, unlike previous approaches whose curvature variation is discrete and depends on the number of locking mechanisms or manual configurations. The robot poses an emergent modeling problem due to the variable stiffness growing spine which is addressed in this paper. We investigate the property of growing spine stiffness and incorporate it into the Cosserat rod model by implementing a combined stiffness approach. We conduct experiments with the soft continuum robot in various configurations and compared the results with our developed mathematical model. The results show that the mathematical model based on the adapted Cosserat rod matches the experimental results with only a 3.3% error with respect to the length of the soft continuum robot.

Keywords: Soft Robot Materials and Design, Compliant Joint/Mechanism, Tendon/Wire Mechanism
Abstract: Soft and continuum robots have garnered great interest in recent years due to their ability to reconfigure, navigate complex environments, and enhance safety during unplanned collisions. To combine the advantages (e.g. higher payloads, simpler kinematics) of rigid robots with these desirable attributes, we present here a manipulator which is able to switch back and forth between being a rigid-link manipulator and a soft continuum robot. Manipulator stiffness is altered using phase changes of a low melting point alloy. The robot can switch between stiff and flexible states using thermoelectric heat pumps which enable local heating and cooling of both robot sections and robot joints. We evaluate the expanded reach afforded by the soft state, the overall workspace, and the increase in payload capacity enabled by the stiff state.

Keywords: Modeling, Control, and Learning for Soft Robots, Soft Sensors and Actuators
Abstract: Soft actuators, distinguished by their complex non-linear behavior, are difficult to model analytically and cumbersome to prototype. Finite element (FE) models allow for more efficient behavioral prediction, but often require onerous setup, especially for large systems. We present a physics-informed neural network model formed by combining a low fidelity analytical model and input-convex neural networks to learn an underlying energy potential for the actuator from experimental and finite element simulation data. In doing this, the neural network can provide sufficiently accurate predictions about systems made up of multiple units, essentially scaling the model from a single unit to an assembly of many. To test this concept, we compare predictions of the deformation of a 5-actuator system from an FE model and from the physics-informed neural network. The neural network, which provides a prediction similar in accuracy to the FE equivalent, can more easily be adjusted to execute systems of greater quantities of units without drastic increases in computational consumption. In this way, we can scale our predictive understanding with adequate accuracy without compounding resources.

Keywords: Soft Sensors and Actuators, Soft Robot Materials and Design
Abstract: Tape spring mechanisms are useful in soft robotics for their ability to extend long distances and exhibit both high strength and high flexibility. In this work we introduce a new tape-spring based actuator concept that uses internal fluidic actuation to achieve: 1) stiffness control of tape-spring beams, and 2) shape modulation of a length-constrained tape-spring actuator. This concept relies on the ability of sealed tape-springs to form a rolling pressure seal when the mechanism is bent, and thus through differential pressure control across the bend we can reconfigure the actuator shape. We first describe the fabrication methods for sealed tape-spring actuators and then characterize their mechanical and hydraulic properties. Three-point bend tests of open and hydraulically sealed tape-spring beams enable us to determine how fluid pressure influences mechanical stiffness. We measured the sealing characteristics of the rolling bend seal by measuring mass-flow through the bend at different input pressures. These characterization experiments allowed us to build a length-constrained actuator which can change stiffness, and actuate through differential pressure control. Fluidic control of rolling-seal tape spring actuators presents new opportunities for using tape-spring mechanisms in situations where electromagnetic actuation may be unfavorable or infeasable.

Keywords: Haptics and Haptic Interfaces, Soft Robot Applications, Soft Robot Materials and Design
Abstract: Haptic research is the study of touch. Surprisingly, both quality and availability of haptic displays are still considered a limiting factor to the impact of the field of haptic research today. Some of these limitations can potentially be overcome by soft, high-resolution displays based on soft robotics principles. We demonstrate a large area haptic display based on an array of soft robotic actuators called pneumatic unit cells (PUCs) that were previously studied as individual, standalone units. We address two underlying practical challenges in the design and control of such an array: reliable and efficient fabrication of a large number of PUCs and reliable and reversible interconnection of the soft actuator layer to a rigid manifold. The device was developed for a specific haptic experiment where the PUCs are actuated in groups that form concentric circles, to simulate areas of small to large size. The design concept can be adapted to address different haptics and perception questions, and may be applied as a practical soft interface in human-machine interaction.

Keywords: Soft Sensors and Actuators, Force and Tactile Sensing, Prosthetics and Exoskeletons
Abstract: With the rise of soft robotics, a wide range of soft sensorised skins have been proposed and developed. One particular group, fluidic electronic skins, have been shown to be promising but, so far, have been limited to small scales. Here, we present a large-scale (35,440 mm^2) sensory skin that uses an embedded channel system filled with saline and which is aimed to work as a prosthetic liner. Mechanical forces on the skin translate into channel deformations which can be measured by changes in impedance through external electrodes. We developed a novel fabrication process based on methods commonly used in prosthodontics and jewellery making to overcome challenges related to the larger size of the skin. We tested two machine learning techniques, i.e., artificial neural networks and random forests, to learn the mapping of the impedance changes to the location of the physical interaction. The results provide new insights on how to improve the design, in particular, in how to improve the channel structure to increase sensory performance.

Keywords: Modeling, Control, and Learning for Soft Robots, Flexible Robotics, Contact Modeling
Abstract: Soft material robotic systems offer inherent safety and flexibility due to their low material stiffness. Therefore, soft material robots are prone to operate in unknown environments and fulfill tasks that involve and even exploit contact with the environment. Moving to the application of soft robots, incorporating validated contact models in modeling frameworks can be crucial for simulation tasks in, e.g. design optimization, motion planning or control. Cosserat rod models have proven themselves not only to be accurate but also computationally efficient for slender soft continuum robots (SCRs). However, only recently the topic of contact modeling has been introduced to Cosserat rod frameworks for SCRs. In this paper, for the first time we present and analyze an approach to include contact modeling in a widely used shooting-based Cosserat rod implementation. Evaluation against detailed finite element (FE) simulations indicate comparable accuracy, while the computational time remains a small fraction. Simulated data for the considered contact scenarios reveal a consistent level of agreement to experimental data, with minor discrepancies. The results are a promising basis on which further contact investigations can build.

Keywords: Multifingered Hands, Grippers and Other End-Effectors, Dexterous Manipulation
Abstract: Dexterous robotic manipulation in unstructured environments can aid in everyday tasks such as cleaning and caretaking. Anthropomorphic robotic hands are highly dexterous and theoretically well-suited for working in human domains, but their complex designs and dynamics often make them difficult to control. By contrast, parallel-jaw grippers are easy to control and are used extensively in industrial applications, but they lack the dexterity for various kinds of grasps and in-hand manipulations. In this work, we present DELTAHANDS, a synergistic dexterous hand framework with Delta robots. The DELTAHANDS are soft, easy to reconfigure, simple to manufacture with low-cost off-the-shelf materials, and possess high degrees of freedom that can be easily controlled. DELTAHANDS' dexterity can be adjusted for different applications by leveraging actuation synergies, which can further reduce the control complexity, overall cost, and energy consumption. We characterize the Delta robots' kinematics accuracy, force profiles, and workspace range to assist with hand design. Finally, we evaluate the versatility of DELTAHANDS by grasping a diverse set of objects and by using teleoperation to complete three dexterous manipulation tasks: cloth folding, cap opening, and cable arrangement. We open-source our hand framework at https://sites.google.com/view/deltahands/.

Keywords: Modeling, Control, and Learning for Soft Robots, Hardware-Software Integration in Robotics, Reinforcement Learning
Abstract: Polymer-based soft robots are difficult to characterize due to their non-linear nature. This difficulty is compounded by multiple additional degrees of movement freedom which adds complexity to any control strategy proposed. The following work proposes and demonstrates a modular framework to test, debug and characterize soft robots using the robot operating system (ROS), to enable modeless deep reinforcement learning control strategies through hardware-in-the-loop system training. The framework is demonstrated using an actor-critic algorithm to learn a locomotion policy for a two-actuator pneu-net soft robot with integrated resistive flex sensors. The resulting locomotion strategy achieves a 270% increase in velocity and over a 400% increase in distance traveled over 1000 steps when compared to a random control strategy implemented on the same system.

Keywords: Soft Sensors and Actuators, Force and Tactile Sensing
Abstract: Most braille-reading robotic sensors employ a discrete letter-by-letter reading strategy, despite the higher potential speeds of a biomimetic sliding approach. We propose a complete pipeline for continuous braille reading: frames are dynamically collected with a vision-based tactile sensor; an autoencoder removes motion-blurring artefacts; a lightweight YOLO v8 model classifies the braille characters; and a data-driven consolidation stage minimizes errors in the predicted string. We demonstrate a state-of-the-art speed of 315 words per minute at 87.5 percent accuracy, more than twice the speed of human braille reading. Whilst demonstrated on braille, this biomimetic sliding approach can be further employed for richer dynamic spatial and temporal detection of surface textures, and we consider the challenges which must be addressed in its development.

Keywords: Mapping, Autonomous Vehicle Navigation, Object Detection, Segmentation and Categorization
Abstract: Topological maps are a common framework for enabling autonomous robotic navigation. To be effective for robotic exploration the maps must be able to be generated quickly and compact enough to store on lightweight hardware. Here we propose a novel 3D topological map called Sphere-Graph which has adaptive edge lengths, can be quickly generated, and can be used to semantically identify hallways and rooms to produce a compact representation of complex environments. We give examples of the Sphere-Graph representation of large 3D urban and cave environments.

Keywords: Modeling, Control, and Learning for Soft Robots, Optimization and Optimal Control, Underactuated Robots
Abstract: Continuum soft robots are nonlinear mechanical systems with theoretically infinite degrees of freedom (DoFs) that exhibit complex behaviors. Achieving motor intelligence under dynamic conditions necessitates the development of control-oriented reduced-order models (ROMs), which employ as few DoFs as possible while still accurately capturing the core characteristics of the theoretically infinite-dimensional dynamics. However, there is no quantitative way to measure if the ROM of a soft robot has succeeded in this task. In other fields, like structural dynamics or flexible link robotics, linear normal modes are routinely used to this end. Yet, this theory is not applicable to soft robots due to their nonlinearities.

In this work, we propose to use the recent nonlinear extension in modal theory -called eigenmanifolds- as a means to evaluate control-oriented models for soft robots and compare them. To achieve this, we propose three similarity metrics relying on the projection of the nonlinear modes of the system into a task space of interest. We use this approach to compare quantitatively, for the first time, ROMs of increasing order generated under the piecewise constant curvature (PCC) hypothesis with a high-dimensional finite element (FE)-like model of a soft arm. Results show that by increasing the order of the discretization, the eigenmanifolds of the PCC model converge to those of the FE model.

Keywords: Soft Robot Applications, Hydraulic/Pneumatic Actuators, Modeling, Control, and Learning for Soft Robots
Abstract: In this paper, we present a modular pressure control system called PneuDrive that can be used for large-scale, pneumatically-actuated soft robots. The design is particularly suited for situations which require distributed pressure control and high flow rates. Up to four embedded pressure control modules can be daisy-chained together as peripherals on a robust RS-485 bus, enabling closed-loop control of up to 16 valves with pressures ranging from 0-100 psig (0-689 kPa) over distances of more than 10 meters. The system is configured as a C++ ROS node by default. However, independent of ROS, we provide a Python interface with a scripting API for added flexibility. We demonstrate our implementation of PneuDrive through various trajectory tracking experiments for a three-joint, continuum soft robot with 12 different pressure inputs. Finally, we present a modeling toolkit with implementations of three dynamic actuation models, all suitable for real-time simulation and control. We demonstrate the use of this toolkit in customizing each model with real-world data and evaluating the performance of each model. The results serve as a reference guide for choosing between several actuation models in a principled manner. A video summarizing our results can be found here: url{https://bit.ly/3QkrEqO}.

Keywords: Learning and Adaptive Systems, Modeling, Control, and Learning for Soft Robots
Abstract: Soft robots are appealing in a wide variety of tasks thanks to their inherent advantages in safety, compliance, and adaptability. However, accurate modelling and control of soft robots are still significantly challenging. This paper proposes a control scheme implementing an online learning strategy. The architecture is composed of (i) a data-driven model generating a feedforward signal, and (ii) a feedback controller. The latter has two roles. Firstly, it corrects the action of the feedforward controller when the tracking error increases. Secondly, it generated a learning signal to train the data-driven model, allowing for online adaptation of the feedforward signal with respect to changes in the dynamic of the system.

Experimental results show that the proposed method provides better performance compared with a PID controller when applied to a trajectory following task. Furthermore, our controller is shown to be capable of online adaptation to sudden changes in the dynamics of the soft robot due to a variable payload.

Keywords: Modeling, Control, and Learning for Soft Robots, Soft Sensors and Actuators
Abstract: Development of soft actuators often considers the actuator as a unit with a single force and positional output, but the compliance of soft actuators means this is a narrow view of the design space. Recent work has begun to extended single actuator design to produce set trajectories or movements, though often as design of integrated systems. In this work, we propose an actuator design space in this direction, a generalization of serial pneumatic artificial muscles (sPAMs) which allows variable actuation along the length of a compliant backbone. This result is enabled by an evolution of the classical Pleated Pneumatic Artificial Muscle (PPAM) model, expanding the range of actuator geometries for which the force-strain behavior can be predicted. As such, muscle units which are partially inactive or asymmetrically constrained can be predicted and combined in a single serial PAM. Our model enables an extended design space for serial PAMs, which is applied to actuating an inflated beam structure to achieve pre-programmed compound curvatures.

Keywords: Optimization and Optimal Control, Modeling, Control, and Learning for Soft Robots, Soft Robot Applications
Abstract: Over the past decades, trajectory optimization (TO) has become an effective solution for solving complex mo- tion generation problems in robotics, ranging from autonomous driving to humanoids. Yet, TO methods remain limited to robots with tens of degrees of freedom (DoFs), limiting their usage in soft robotics, where robots often depict hundreds of DoFs in general. In this work, we introduce a generic method to perform trajectory optimization based on continuum mechanics to describe the behavior of soft robots. The core idea is to condense the dynamics of the soft robot in the constraint space in order to obtain a reduced dynamics formulation, which can then be plugged into numerical TO methods. In particular, we show that these condensed dynamics can be easily coupled with differential dynamic programming methods for solving TO problems involving soft robots. This method is evaluated on three different soft robots with different geometries and actuation.

Keywords: Modeling, Control, and Learning for Soft Robots, Optimization and Optimal Control, Compliant Joint/Mechanism
Abstract: This paper details a reliable control method for highly nonlinear dynamical systems such as soft robots. We call this method model evolutionary gain-based predictive control or MEGa-PC. The method uses an evolutionary algorithm to optimize a set of controller gains via model predictive control. We demonstrate the performance of MEGa-PC in simulation for a single-link inverted pendulum and a three-link inverted pendulum, and on physical hardware for a three-joint continuum soft robot arm with six degrees of freedom. MEGa-PC is compared to prior work that used Nonlinear Evolutionary Model Predictive Control or NEMPC. The new method performs similarly to NEMPC in terms of accumulated cost over the entire trajectory, however, MEGa-PC generalizes better to real-world applications where safety is paramount, the dynamic model is uncertain, the system has significant latency, and where the previous sampling-based method (NEMPC) resulted in significant steady-state error due to model inaccuracy.

Keywords: Hydraulic/Pneumatic Actuators, Soft Robot Applications, Soft Robot Materials and Design
Abstract: Inflated-beam soft robots, such as tip-everting vine robots, can control curvature by contracting one beam side via pneumatic actuation. This work develops a general finite element modeling approach to characterize their bending. The model is validated across four pneumatic actuator types (series, compression, embedded, and fabric pneumatic artificial muscles), and can be extended to other designs. These actuators employ two bending mechanisms: geometry-based contraction and material-based contraction. The model accounts for intricate nonlinear effects of buckling and anisotropy. Experimental validation includes three working pressures (10, 20, and 30 kPa) for each actuator type. Geometry-based contraction yields significant deformation (92.1% accuracy) once the buckling pattern forms, reducing slightly to 80.7% accuracy at lower pressures due to stress singularities during buckling. Material-based contraction achieves smaller bending angles but remains at least 96.7% accurate. The open source models available at http://www.vinerobots.org support designing inflated-beam robots like tip-everting vine robots, contributing to waste reduction by optimizing designs based on material properties and stress distribution for effective bending and stress management.

Keywords: Surgical Robotics: Steerable Catheters/Needles, Hydraulic/Pneumatic Actuators, Soft Robot Materials and Design
Abstract: Soft robotics have propelled advancements in medical applications such as endovascular interventions through the development of soft steerable catheters. Despite their potential, existing catheters often require intricate manufacturing or complex control systems due to their inherent hysteresis and nonlinear material properties. This study introduces a novel catheter system featuring a 3D-printed hollow tip with two pneumatic bending units, facilitating rotation about two distinct axes without the need for additional fabrication steps. Finite element analysis simulations were used to optimize the catheter design while minimizing its diameter to 6.4 mm. Furthermore, an external robotic control system was integrated to perform physical experiments to assess the bending capabilities of the catheter tip. After calibration, the system exhibited proficient shape control, achieving a wide bending range from -47 degrees to 169 degrees under the integrated control system, and effectively conforming to 'C', 'S', and 'J' configurations with 0.53 degrees closed-loop accuracy. The catheter prototype demonstrated low hysteresis and high repeatability. The entire catheter system presents a pragmatic and cost-effective approach for the rapid prototyping and development of pneumatic steerable catheters. This novel system is ideally suited for testing preliminary concepts and educational applications in endovascular interventions.

Keywords: Education Robotics, Soft Robot Materials and Design, Biologically-Inspired Robots
Abstract: Designing physically intelligent soft robots is more than actuator fabrication; shaping a responsive material for autonomous task execution, i.e., in materia programming, needs a broad skill set, including a good command of crafts. “Thinking with hands” during manual material shaping is often performed unconsciously in a pre-prototyping phase without deserved attention. Yet, such ideation can be enabling in creating a starting point for successive optimization and modeling. We suggest a workshop, aimed at the highest accessibility, on physical intelligence and in materia coding. The workshop specifically focuses on rapid ideation of robot embodiments, with minimal burden of fabrication complexity and even need for specific skills such as programming languages. The employed technique of layering wood veneer as the active material and waterproof tape as the passive material into bioinspired bilayer actuators can be mastered even by preschoolers in ten minutes, giving vast opportunities to code intelligent behavior and accomplish situated tasks via intuitive manual shaping (cutting and combining) the material. Wood, as the central material in the workshop, creates immediate personal engagement with participants and fosters creativity. Bilayer-based embodiments showcase material intelligence via morphological changes in response to humidity and give immediate, visual, and tactile feedback. When situated in an environment and exposed to humidity, the bilayer-based robot starts the execution of a task defined by the encoded shape of its layers and layer arrangement, exemplifying physical intelligence. The wood bilayer robotics workshop has been tested more than five times, confirming accessibility to a broad demographic. Contextualizing apparently simple crafting steps fosters not only public outreach but also encourages explorations in the morphological space for enriched robot embodiments.

Keywords: Soft Robot Applications, Soft Robot Materials and Design
Abstract: Peristalsis has inspired the creation of diverse robotic systems in the study of ailments affecting gastrointestinal tract function, and the development of physical simulators for this vital series of organs. SoftTract is a novel prototype pneumatic design using overlapping concentric pneumatic constrictors to increase the effectiveness of artificial peristalsis. SoftTract exploits an all-soft body with in-extensible mesh reinforcement to move and manipulate an artificial bolus. This actuation concept is demonstrated in a prototype SoftTract which shows improved performance and the benefit of increased media control within the device. The overlapping of chambers reduces the potential for media to become stuck in boundary regions between sequential units and reduces the complexity of control. The design further replicates the compliance of the natural tissues through the use of an all-soft composition. Experimental trials demonstrate a maximum average bolus transport speed of 5.1 mm/s with a sealing pressure of 0.25 bar. This pressure is crucial in preventing debris regurgitation, surpassing the performance of the human oesophagus. The pliable SoftTract prototype circumvents the limitations of rigid materials, thus advancing the ultimate objective of realizing an artificial oesophagus replacement.

Keywords: Soft Robot Applications, Soft Robot Materials and Design, Grasping
Abstract: One of the trendsetting themes in soft robotics has been the goal of developing the ultimate universal soft robotic gripper. One that is capable of manipulating items of various shapes, sizes, thicknesses, textures, and weights. All the while still being lightweight and scalable in order to adapt to use cases. In this work, we report a soft gripper that enables delicate and precise grasps of fragile, deformable, and flexible objects but also excels in lifting heavy objects of up to 1617x its own body weight. The principle behind the soft gripper is based on extending the capabilities of electroadhesion soft grippers through the enhancement principles found in metamaterial adhesion cut and patterning. This design amplifies the adhesion and grasping payload in one direction while reducing the adhesion capabilities in the other direction. This counteracts the residual forces during peeling (a common problem with electroadhesive grippers), thus increasing its speed of release. In essence, we are able to tune the maximum strength and peeling speed, beyond the capabilities of previous electroadhesive grippers. We study the capabilities of the system through a wide range of experiments with single and multiple-fingered peel tests. We also demonstrate its modular and adaptive capabilities in the real-world with a two-finger gripper, by performing grasping tests of up to 5 different multi-surfaced objects.

Keywords: Wearable Robots, Tendon/Wire Mechanism, Modeling, Control, and Learning for Soft Robots
Abstract: Cable-driven exosuits are gaining attention in the upper limb assistive scenario thanks to their portability, comfort, and low cost. In order to guarantee an effective and transparent use of the device, the interaction force between the arm of the user and the supporting cable needs to be carefully tracked. In this work, we proposed and identified a Bowden cable transmission model and we implemented three different direct force controllers: SL) sensorless model-based controller, PI) proportional-integral controller, PIM) combining the PI and the model-based terms. The three control strategies were tested on a motorized mannequin and results show that the lowest tracking error when supporting the mannequin arm with 50% gravity compensation was achieved by the PIM strategy (RMSE = 6.9 N). The SL strategy also achieved a good performance (RMSE = 8.8 N) while minimizing oscillations around the reference force. The PI controller achieved the worst performances both in terms of RMSE (18.2 N) and oscillating behaviour. This work outlined two possible strategies for two different use-cases: 1) a sensor-based strategy (PIM) when the use-case requires high tracking precision, 2) a sensorless strategy (SL) when the use-case requires a more robust, simple, and cost-effective device.

Keywords: Soft Robot Materials and Design, Soft Sensors and Actuators, Soft Robot Applications
Abstract: Dielectrophoretic elastomers actuators are a new class of actuator capable of achieving out-of-plane deformation by the stimulation of external electric fields, providing an alternative approach for soft electric actuators. Pneumatic pumps based on dielectrophoretic elastomer actuators have been presented with a simple structure, facile fabrication, and low cost. Here, we demonstrate an improved version of pneumatic pump with a stand-alone soft materials structure and optimized air chamber design. The component configuration, actuation mechanism, and working performance are investigated, which reveals the enhanced output and durability. By using flexible building blocks, the pump can be bent and twisted, representing a step forward in the use of dielectrophoretic elastomers in soft robotic applications.

Keywords: Soft Robot Materials and Design, Grasping, Soft Robot Applications
Abstract: Suction cups and closed bellows actuators are commonly used in industrial pick and place tasks. They can be combined and actuated by a single negative pressure line, providing a simple coupling between suction and actuation. However, the capabilities of this -- seemingly very simple -- system go beyond just being able to grasp and retract an object. By varying the morphology of the suction cup and by tuning airflow between the two serially connected elements, retraction of the bellows changes and functional embodied behaviours emerge. This work explores the novel concept of flexible Sucking & Retracting Bellows (SuReBellows) comprised of a suction cup and a long-stroke bellows actuator, and how they can be tuned for morphologically intelligent behaviours. Our primary investigation explores how the retraction of the bellows can be tuned to enable ``exploration'' on the surface of an unknown object in order to make a good vacuum seal contact for grasping. Several suction cup designs were tested in grasping a general object set, by varying the compliance and airflow between the bellows and the suction cup. The results show that SuReBellows can explore the surface of objects and find grasp points just by cycling one pressure signal for actuation, with the search pattern and adaptability encoded in its morphology. The grasping force of the system was then investigated. Finally, demonstrations of interesting applications such as bin picking and grasping in confined environments are shown.

Keywords: Soft Robot Materials and Design, Soft Robot Applications, Biologically-Inspired Robots
Abstract: Punctures and surface damage can be catastrophic for soft robotic systems. Inspired by the self-sealing deformations of plants and the anisotropic distribution of collagen in the human skin's dermis, we present a dual-layer skin with controllable wound closure. Anisotropic elastomer channels are used to tune the directionality of the skin's mechanical stiffness, whilst an outermost hydrogel layer grows and shrinks as humidity is varied, generating stresses which close and open the wounds.

We demonstrate and characterize the core principles of both mechanisms, before exploring the effects of composite parameters and morphologies. Repeatable and precise damage is inflicted using a CNC scalpel blade. Finally, we demonstrate the use of these mechanisms to seal scalpel cuts on an artificial hand, immediately restoring electrical channels broken by the wound.

Keywords: Soft Robot Materials and Design, Soft Sensors and Actuators
Abstract: Manipulating airflow is important for controlling pneumatically actuated soft robots, however, current switching techniques suffer from leakage under high pressure (>200 kPa) or require a complex fabrication process. We propose a new method for reliably and repeatably cutting off airflow by harnessing pre-loaded torsional forces applied to our tubing. The switching distance and hysteresis of our pre-twisted tubing are programmable by varying the tube length and the twisting angle. Our experiments demonstrate the use of pre-twisted tubing to implement CMOS equivalent fluidic switches configured as NOT-, AND-, and OR-gates, and a distance sensor for feedback control for the oscillation of a PneuNet. Our approach of pre-loading tubes with a torsional force allows for simplicity, integrated functionality, and the capability of manipulating high-pressure, fluidic signals mainly at the cost of tubing.

Keywords: Deep Learning in Robotics and Automation, Soft Robot Applications, Soft Robot Materials and Design
Abstract: Locomotion tracking of a soft water robot that is propelled and steerable by dielectric elastomer actuators (DEAs), is proposed and demonstrated using a hybrid machine learning model. Measuring at 6 cm x 17 cm x 5 cm and weighing 28 g, this soft water robot can achieve a speed of 3 cm/s, equivalent to 50% of its body length per second (BL/s), while consuming approximately 1.0 W of power. Its buoyancy is enabled by the hexagonal honeycomb base made up of silicone. The DEA actuators are constructed from conductive carbon grease as the electrodes on the elastomer. The steerable soft water robot can complete a full 90° turn within 18 seconds. A comprehensive finite element analysis (FEA) is conducted on the actuators and the soft water robot to provide an in-depth study of the soft robot’s performance including actuation force, travel distance, actuation angle, and steering disposition. For precise locomotion tracking, a novel hybrid machine learning model is developed, comprising two multi-layer perceptron branches in a single in-house model. Two additional bypass layers are added for steering behavior classification. The model attains a real-time accuracy of 98% for both the water robot’s displacement and steering behavior prediction.

Keywords: Simulation and Animation, Computational Geometry, Soft Robot Materials and Design
Abstract: Shape morphing can be achieved using thin, planar sheets with patterned strain-limiting constraints to direct differential growth into desired shapes. Our previous work produced such shape-shifting sheets by patterning the surface of an inflatable planar sheet with variable stiffness fiber constraints placed in patterns optimized by a multi-objective evolutionary algorithm. As a pipeline, we generated two specialized fiber patterns, {P_i; P_j} to produce shapes {S_i; S_j}, which we then layered onto the same sheet. However, this layering approach produced translation inefficiencies in hardware including fiber redundancies. To reduce fiber crowding and increase shape fidelity, we propose that fibers identified as similar in both P_i and P_j can be implemented as a single fiber belonging to both patterns. Extending our previous work, herein we implement a post-optimization fiber merging protocol. Applying the protocol to sheets patterned with fibers targeting two shape pairs, cylinder/sphere and simple saddle/monkey saddle, we demonstrate that fiber merging is a useful strategy to reduce the total number of fibers required for shape-matching, thus reducing sheet bulk and weight. By measuring the transfer error between the actual hardware shapes and target (programmed)shapes, we learn that fiber merging is increasingly useful with increasing fiber pattern complexity as measured by fiber count. The transfer error was decreased for one or both shapes for all fiber mergers implemented in the simple saddle/monkey saddle shape-pair, as compared to the original fiber patterns {P_i; P_j}.

Keywords: Manipulation Planning, Simulation and Animation, Soft Robot Applications
Abstract: Multisection continuum arms are bio-inspired manipulators that combine compliance, payload, dexterity, and safety to serve as co-robots in human-robot collaborative domains. Their hyper redundancy and complex kinematics, however, pose many challenges when performing path planning, especially in dynamic environments. In this paper, we present a W-Space based Rapidly Exploring Random Trees * path planner for multisection continuum arm robots in dynamic environments. The proposed planner improves the existing state-of-art planners in terms of computation time and the success rate, while removing the need for offline computation. On average, the computation time of our approach is below 2 seconds, and its average success rate is around 70%. The computation time of the proposed planner significantly improves that of the state-of-the-art planner by roughly a factor of 20, making the former suitable for real-time applications. Moreover, for application domains where the obstacle motion is not very predictable (e.g., human obstacles), the proposed planner significantly improves the success rate of state-of-the-art planners by nearly 50%. Lastly, we demonstrate the feasibility of several generated trajectories by replicating the motion on a physical prototype arm.

Keywords: Grippers and Other End-Effectors, Soft Robot Applications
Abstract: Aerial grasping using flying robots necessitates gripper designs that can accommodate position/orientation errors to grasp different objects. This study investigates two bistable gripper designs that can automatically close upon contacting an object. Both grippers leverage pre-stressed spring steel bands (PSSB), with one having a single PSSB and the other having two PSSBs in a cross-shape. Both are equipped with a single motor to drive a cable system routed through flexible joints on the bands to ensure consistent opening. The same cable system, together with a friction adjustment mechanism, also ensures sequential closing of the gripper. We conduct experiments to evaluate and compare the gripping performance of both designs. On average, the single-band gripper can grasp within 0.14 s and open in 3 s, while the cross-band gripper takes 0.24 and 4 s, respectively. Both designs can still grasp even when a cylinder makes off-center contact with the PSSB, up to a displacement of 50 mm. Further grasping tests demonstrate both grippers can grasp cylinders with different diameters that approach the gripper with different orientations. Finally, we demonstrate that both grippers can grasp real-world objects (e.g., water bottles, video game controllers, etc.). Our grippers, with their compact, lightweight design and passive grasping capabilities, hold the potential to advance aerial grasping for diverse applications.

Keywords: Wearable Robots, Soft Sensors and Actuators, Medical Robots and Systems
Abstract: Despite recent advances in wearable technology, interfacing movement assistance devices with the human body remains challenging. We present a stretchable pneumatic sleeve that can anchor an exosuit actuator to the human arm with a low displacement of the actuator’s mounting point relative to the body during operation. Our sleeve has the potential to serve as an adaptable attachment mechanism for exosuits, since it can adjust its pressure to only compress the arm as much as needed to transmit the applied exosuit forces without a large displacement. We discuss the design of our sleeve, which is made of fabric pneumatic artificial muscle (fPAM) actuators formed into bands. We quantify the performance of nine fPAM bands of various lengths and widths, as well as three sleeves (an fPAM sleeve, a series pouch motor (SPM) sleeve as in previous literature, and an off the shelf hook and loop sleeve), through the measurement of the compressing force as a function of pressure and the localized pulling force that can be resisted as a function of both pressure and mounting point displacement. Our experimental results show that fPAM bands with smaller resting length and/or larger resting width produce higher forces. Also, when inflated, an fPAM sleeve that has equivalent dimensions to the SPM sleeve while fully stretched has similar performance to the SPM sleeve. While inflated, both pneumatic sleeves decrease the mounting point displacement compared to the hook and loop sleeve. Compared to the SPM sleeve, the fPAM sleeve is able to hold larger internal pressure before bursting, increasing its possible force range. Also, when not inflated, the fPAM sleeve resists the pulling force well, indicating its ability to provide anchoring when not actuated.

Keywords: Soft Robot Applications, Soft Robot Materials and Design, Compliant Joint/Mechanism
Abstract: In-pipe locomotion is a great proving ground for the environmental flexibilities of soft robots. The 2023 Robosoft conference locomotion competition captured many of these challenges and opportunities as a series of in-pipe obstacles for soft robots to navigate. Locomotion through pipes of different sizes, and surfaces, and navigating T-junctions was demonstrated. This paper presents the design of the competition-winning robot, using inchworm-inspired bellows actuated locomotion to move forward via anisotropic friction from the robot body design. The robot can traverse pipe diameters of 5cm to 10cm and can navigate T-junctions via a compliant guide rod that can be rotated to choose heading direction and cross over gaps in the pipeline. This paper also shows further development of the competition robot to generalise for more pipe designs by having the setae be a compliant four-bar linkage mechanism. The final design of the robot only requires one onboard motor to rotate the semi-passive guide rod for navigation and one pneumatic line for the bellows actuator, with the compliant seta design taking up minimal volume in the pipe compared to other wall-mounting inchworm solutions.

Keywords: Modeling, Control, and Learning for Soft Robots
Abstract: Actively tuning mechanical properties in soft robots is now feasible due to advancements in soft actuation technologies. In soft manipulators, these novel actuators can be distributed over the robot body to allow greater control over its large number of degrees of freedom and to stabilize local deformations against a range of disturbances. In this paper, we present a hierarchical policy for stiffness control for such a class of soft manipulators. The stiffness changes induce desired deformations in each segment, thereby influencing the manipulator's end-effector position. The algorithm can be run as an online controller to influence the manipulator's stable states -- as we demonstrate in simulation -- or offline as a design algorithm to optimize stiffness distributions -- as we showcase in a hardware demonstration. Our proposed hierarchical control scheme is agnostic to the stiffness actuation method and can extend to other soft manipulators with nonuniform stiffness distributions.

Keywords: Modeling, Control, and Learning for Soft Robots, Underactuated Robots, Optimization and Optimal Control
Abstract: Soft robotic manipulators offer operational advantage due to their compliant and deformable structures. However, their inherently nonlinear dynamics presents substantial challenges. Traditional analytical methods often depend on simplifying assumptions, while learning-based techniques can be computationally demanding and limit the control policies to existing data. This paper introduces a novel approach to soft robotic control, leveraging state-of-the-art policy gradient methods within parallelizable synthetic environments learned from data. We also propose a safety oriented actuation space exploration protocol via cascaded updates and weighted randomness. Specifically, our recurrent forward dynamics model is learned by generating a training dataset from a physically safe mean reverting random walk in actuation space to explore the partially-observed state-space. We demonstrate a reinforcement learning approach towards closed-loop control through state-of-the-art actor-critic methods, which efficiently learn high-performance behaviour over long horizons. This approach removes the need for any knowledge regarding the robot's operations /capabilities and sets the stage for a comprehensive benchmarking tool in soft robotics control.

Keywords: Soft Sensors and Actuators, Tendon/Wire Mechanism, Medical Robots and Systems
Abstract: The kinematic monitoring of human joint movements is beneficial in various applications, including biomechanics, rehabilitation, and wearable robotics. For consistent monitoring of human motions in different situations, a portable monitoring system with a compact form factor would be highly useful. However, most existing devices are susceptible to external disturbances or require a tedious and repetitive calibration process. In this paper, we propose a flexible and compact joint angle detection mechanism that allows for the estimation of the full configuration of the target joint, as well as the measurement of a flexion angle. The proposed approach incorporates a simple geometry, whereby the configuration of a two-dimensional rigid body can be determined by the lengths of three wires. A stand-alone device for the analysis of knee flexion based on the proposed mechanism is demonstrated and experimentally evaluated considering the in-plane assumption and the out-of-plane errors. The proposed sensor system estimates the flexion angle with root-mean-square errors of 4.8º, 5.1º, and 4.7º for an in-plane, a translated, and a rotated configurations, respectively.

Keywords: Biologically-Inspired Robots, Soft Robot Materials and Design, Soft Robot Applications
Abstract: Fish commonly combine pectoral fin motion with tail and body motion to create a swimming gait. The rowing pectoral fin motion is characterized by distinct power and recovery strokes where the soft fin bends or rotates significantly to change the hydrodynamic profile of the appendage between the two strokes. In this work, we take inspiration from fish to create an underwater robot that uses a rowing gait to create net forward thrust. We present a mechanical linkage using two cams on a single drive axis to drive both the yawing and rolling rotations of the fin with a single actuated degree of freedom. We also describe the kinematics and geometric constraints of the linkage and how they relate to properties of the cams. We show that a soft fin can create asymmetric thrust in conjunction with the cam-follower transmission and that a fin can produce large spikes in negative force while still creating net forward work throughout the gait cycle.

Keywords: Soft Sensors and Actuators, Force and Tactile Sensing, Tendon/Wire Mechanism
Abstract: Cable-driven force transmission is one of the most widely used actuation mechanisms in robotics, finding applications in robotic arms, grippers, and wearable robots. However, measuring cable tension directly is challenging due to limited space and difficulties in attaching sensors to flexible cables. In this paper, we propose a single-axis load cell designed for direct integration into a high-strength woven cable, enabling tension measurement. A small portion of the cable becomes slightly stretchable by infusing an uncured polymer into its weaving pattern, and a liquid-metal thin-film soft sensor is affixed externally. As the cable experiences axial strain, the thin- film soft sensor stretches and changes its electrical resistance, providing real-time tension information. To mitigate the hysteresis and the nonlinearity caused by the viscoelastic effect of the cable and tendon materials, a metal spring casing surrounds the sensor, making the behavior of the sensor governed by that of the spring. The proposed sensor is experimentally characterized, and the result shows a linearity in the range of 5 N to 100 N with negligible hysteresis. It is tested with a twisted-string actuator to control the axial force generated by the actuator through a closed-loop feedback system as an application.

Keywords: Modeling, Control, and Learning for Soft Robots, Software, Middleware and Programming Environments, Optimization and Optimal Control
Abstract: Aggressive and accurate control of complex dynamical systems, such as soft robots, is especially challenging due to the difficulty of obtaining an accurate and tractable model for real-time control. Learned dynamic models are incredibly useful because they do not require derivation of an analytical model, they can represent complex, nonlinear behavior directly from data, and they can be evaluated quickly on graphics-processing units (GPUs). In this paper, we present an open-source Python library to further current research in model-based control of soft robot systems. Our library for Modeling of Learned Dynamics (MoLDy), is designed to generate learned forward models of complex systems through a data-driven approach to hyperparameter optimization and learned model training. Included in the MoLDy library, we present an open-source version of NEMPC (Nonlinear Evolutionary Model Predictive Control), a previously published control algorithm validated on soft robots. We demonstrate the ability of MoLDy and NEMPC to accurately perform model-based control on a physical pneumatic continuum joint. We also present a benchmarking study on the effect of the loss metric used in model training on control performance. The results of this paper serve to guide other researchers in creating learned dynamic models of novel systems and using them in closed-loop control tasks.

Keywords: Additive Manufacturing, Surgical Robotics: Steerable Catheters/Needles, Micro/Nano Robots
Abstract: Patent Ductus Arteriosus (PDA) is a heart condition in which the ductus arteriosus—a blood vessel connecting the pulmonary artery to the aorta in a fetus—fails to undergo closure after birth. A PDA can be an important factor in neonates born with severe congenital heart disease (CHD) or born prematurely. With the advent of new intravascular stent technologies, treatments based on ductus arteriosus stenting can now be completed in many cases; however, difficulties remain in accessing the ductus arteriosus in small babies successfully using current guidewire-catheter systems. Recent developments for soft robotic endovascular instruments that leverage control schemes hold distinctive potential for addressing these access challenges, but such technologies are not yet at the sizes required for navigating neonatal vasculature safely and efficiently. In an effort to meet this clinical need, this work presents an approach for 3D printing 1.5 French (Fr) soft robotic guidewires that transition from straight to “S”-shaped configurations under the application of fluidic (e.g., pneumatic or hydraulic) loading. Two distinct dual-opposing segmented soft actuators, including a symmetric and asymmetric system design (both with heights of 2.5 mm), were 3D printed onto 1.1 Fr capillaries in 35–60 minutes via “Two-Photon Direct Laser Writing (DLW)”. Experimental results revealed that both designs not only withstood pressures of up to 550 kPa, but also exhibited increased opposing bending deformations—corresponding to decreased radii of curvature—with increasing pressure. In combination, this study serves as a critical foundation for next-generation fluidically actuated soft robotic guidewire-catheter systems for PDA interventions.

Keywords: Soft Robot Materials and Design, Biomimetics, Soft Robot Applications
Abstract: Biologically inspired soft anguilliform swimming robots show great promise in underwater exploration. Their soft bodies promise to reduce the chance of harming humans or wildlife with which they come into contact, and also to reduce their chance of becoming stuck in complex environments. Furthermore, the efficiency of anguilliform swimming may enable long duration operation. However, the design and fabrication of soft anguilliform swimming robots remain challenging. Here we present a design concept for a modular soft anguilliform robot. To address the challenge of consistently fabricating modules for this design, we also present a new fabrication method that combines injection molding and lost-core molding for the consistent fabrication of bi-directional fluidic elastomer actuator modules. We evaluate the consistency of the fabrication method through visual inspection, weight measurements, bending angle measurements and the measurement of force output of the actuator. This work represents a step towards an autonomous eel-inspired soft robot, as well as a new fabrication approach that may enable a number of other new soft robotic systems.

Keywords: Soft Sensors and Actuators, Modeling, Control, and Learning for Soft Robots, Soft Robot Materials and Design
Abstract: Shape reconstruction of deformable actuators by means of intrinsic sensors is highly crucial for an accurate control of soft robots. Given the unstructured nature of the intrinsic sensors employed, a common approach is to leverage recurrent neural networks to estimate the position of a number of points along the main axis of the actuator. The shape is, then, reconstructed by fitting a kinematic model to the estimated points on the actuator. This paper proposes an alternative method in which a parameterized curve is chosen to model the deformation of the actuator. Feedback from the intrinsic sensors are utilized to directly estimate the model parameters by means of a neural network. The performance of the proposed approach was tested on a setup with an array of soft strain sensors attached to a tendon-driven actuator. The experiments were configured so that the actuator interacts with an external environment that applies a variable load on the deforming body, inducing a significant variation in the curvature of the backbone of the actuator. The proposed approach achieved an average estimation error of 1.16 mm in the tip position (0.6% of the actuator length) and 1.2 degrees in the tip orientation (less than 1% of the maximum tip orientation).

Keywords: Hydraulic/Pneumatic Actuators, Medical Robots and Systems, Soft Robot Applications
Abstract: Despite its multiple advantages over classical procedures, interventional magnetic resonance imaging (iMRI) remains a rarely used method in percutaneous tumor biobsy and ablation of the liver. Mainly due to the difficult and special environment in the MRI scanner, automation of the process through the use of robots remains a challenge. Soft robotic systems, made of soft materials and actuated by pneumatic pressure, offer a way to assist interventionalists in their daily work in iMRI. In this paper, we present a soft robotic actuator that leverages switchable interlocking strain-limiting structures for bending and position locking. Comb-like structures in separate chambers are used to create the actuator's bending motion through force-transmitting form-closure.

Keywords: Soft Robot Materials and Design, Hydraulic/Pneumatic Actuators, Grippers and Other End-Effectors
Abstract: For robots to be useful for real-world applications, they should be safe around humans, deployable and passively compressible to a fraction of their size, and low-cost for practical use. A soft robotic arm potentially meets these requirements. The space efficiency of the soft robotic arm can be greatly enhanced by making it retractable. Here, we introduce an inflatable robotic arm capable of both deployment and retraction by rolling out from a small package and rolling in. We have conducted experiments characterizing the performance of the bellows actuator in torque, angle, and pressure. We also demonstrate that the rolling mechanism, based on the bilayer film, occupies minimal space when stowed, but can be deployed to access a large workspace. In addition, we demonstrate the feasibility of the inflatable robotic arm through the pick-and-place function to clear objects on the floor.

Keywords: Modeling, Control, and Learning for Soft Robots, Hydraulic/Pneumatic Actuators, Soft Sensors and Actuators
Abstract: We introduce a self-contained pneumatic actuator capable of 1.43 times stiffness gain from 1332 N/m to 1913 N/m without needing an external air source or valve. The design incorporates an air chamber bellows and a spring bellows, connected and sealed. Stiffness modulation is achieved by altering the air chamber volume. We present an approach for computing the volume, pressurized force, and stiffness of a single bellows component, as well as methods for composing single bellows models to predict the change in stiffness of the dual bellows actuator as a function of air chamber compression. We detail the fabrication of the actuator and verify the models on the fabricated prototype. This actuator holds promise for future integration in tunable stiffness robots demanding high strength and adaptability in dynamic scenarios.

Keywords: Biologically-Inspired Robots
Abstract: Transformation between locomotion modes in multimodal robots poses significant design challenges to meet the growing demand for exploration in various environments. Design constraints involved for an aerial-aquatic robot must account for air and water maneuverability efficiently while minimizing the amount of energy expenditure. Recently, flying fish (Exocoetidae) have gained increasing attention for their ability to switch locomotion modes. Flying fish wings are elastic membranes with no muscles present; instead, fin rays lend support to the flexible structure. A bioinspired deployable wing is presented here based on observations of the flying fish. Specifically, we propose a lightweight and self-stiffening foldable membrane. Self-stiffening is achieved by leveraging spring origami bistability design principles allowing for multistable operations including deployment (~1 s), and collapse for flying, and swimming modes, respectively. Additive manufacturing was utilized to fabricate the wing weighing 7.5% of the target 500 g multimodal robot. Additionally, structural and aerodynamic experiments designed for gliding speeds of up to 8.5 m/s revealed that the wing remained deployed and avoided undesired collapse. The same wind tunnel experiments revealed a higher lift coefficient, and glide ratio for the proposed wing, compared to a flat wing, and rigid wing, respectively, owing to the multifunctional membrane architecture.

Keywords: Modeling, Control, and Learning for Soft Robots, Grippers and Other End-Effectors, Soft Sensors and Actuators
Abstract: Knowing the gripping force being applied to an object is important for improving the quality of the grip, as well as preventing surface damage or breakage of fragile objects. In the case of soft grippers, however, an attaching or embedding of force/pressure sensors can compromise their adaptability or constrain their design in scenarios involving significant deformation/deployment. In this paper, we present a vision-based neural network(OriGripNet) that estimates gripping force by combining RGB image data with key parameters extracted from the physical features of a soft gripper. Real-world force data was collected using a reconfigurable test object with an embedded load cell while image data was collected by an RGB camera mounted on the wrist of a robotic arm. In addition, joint position information of the pneumatically driven origami gripper extracted from the images, and applied pressure were used for training of OriGripNet. OriGripNet showed a mean average error(MAE) of 0.0636N when tested for untrained objects, yet some values exhibited errors exceeding 20%. Nevertheless, the results show that pressure, joint position, and image information have their own strengths in force estimation, contact estimation, and that they have a synergistic effect on the performance when combined.

Keywords: Mechanism Design, Prosthetics and Exoskeletons
Abstract: Most existing anthropomorphic robotic fingers are either too stiff to offer compliance, or too soft to provide postural stability. Yet human subjects tend to stiffen their finger when producing fingertip forces and lower their joint stiffness when grasping objects. Variable joint stiffness is therefore required to offer compliance and postural stability to the finger when interacting with its environment. We therefore propose the novel design of a robotic anthropomorphic finger capable of variable stiffness by making use of the embodied intelligence design principle through multifunctionality of the hardware parts. The ligaments of the finger are not only used to connect the phalanges together, but also to provide local variable stiffness at the finger joints through the use of miniature McKibben pneumatic artificial muscles. This novel design can therefore offer compliance at lower stiffness levels and postural stability and a higher applied force at higher stiffness levels while keeping the finger look and movement anthropomorphic and its control quite basic. The developed anthropomorphic finger with variable stiffness was tested by interacting with a flat surface. The finger presented a significantly higher stiffness (6.7*10^{-3} +/- 3*10^{-3} Nm/rad) when the stiffening system was used than when the finger was purely used in compliance mode, without stiffness adaptation (4.8*10^{-3} +/- 0.7*10^{-3} Nm/rad).

Keywords: Modeling, Control, and Learning for Soft Robots, Learning and Adaptive Systems, Grasping
Abstract: Dexterous manipulation, often facilitated by multi-fingered robotic hands, holds solid impact for real-world applications. Soft robotic hands, due to their compliant nature, offer flexibility and adaptability during object grasping and manipulation. Yet, benefits comes with challenges, particularly in the control development for finger coordination. Reinforcement Learning (RL) can be employed to train object-specific in-hand manipulation policies, but limiting adaptability and generalizability. We introduce a Continual Policy Distillation (CPD) framework to acquire a versatile controller for in-hand manipulation, to rotate different objects in shape and size within a four-fingered soft gripper. The framework leverages Policy Distillation (PD) to transfer knowledge from expert policies to a continually evolving student policy network. Exemplar-based rehearsal methods are then integrated to mitigate catastrophic forgetting and enhance generalization. The performance of the CPD framework over various replay strategies demonstrates its effectiveness in consolidating knowledge from multiple experts and achieving versatile and adaptive behaviors for in-hand manipulation tasks.

Keywords: Grippers and Other End-Effectors, Grasping, Biologically-Inspired Robots
Abstract: The increasing demand for food production coupled with a shortage of labor in the agricultural sector has underscored the critical need for capable autonomous grasping solutions. This presents a formidable research challenge due to the unstructured and delicate nature of agricultural environments. Soft grippers, with their hyperelastic material properties, offer an ideal solution for handling delicate harvests without causing damage, all while obviating the need for precise force control. Current robotic end-effectors employed in harvesting tasks predominantly consist of finger-based grippers, which are effective for picking objects one at a time but fall short when dealing with crops that grow in clusters. The drive for automation in agriculture necessitates productivity levels that can at least match human capabilities like handling multiple items in a single grasp. In response to this challenge, this study presents the design and fabrication of a scooping suction-based gripper, drawing inspiration from the shape and feeding mechanism observed in common starfish. This innovative gripper shows potential for cluster grasping applications, thereby enhancing productivity in agricultural harvesting processes. The efficacy of the gripper design elements was validated through finite element analysis, force testing, and grasping experiments using representative objects.

Keywords: Tendon/Wire Mechanism, Soft Robot Materials and Design, Underactuated Robots
Abstract: The use of intrinsically compliant tensegrity structures in manipulation systems is an attractive research topic. In this paper a 3D compliant robotic arm based on a stacked tensegrity structure consisting of x-shaped rigid members is considered. The rigid members are interconnected by a net of prestressed, tensioned members with pronounced intrinsic elasticity and by inelastic tensioned members. The system’s motion is achieved by length-change of the inelastic tensioned members. The operating principle of the system is discussed with the help of kinematic considerations and verified by experiments.

Keywords: Compliant Joint/Mechanism, Soft Robot Materials and Design, Additive Manufacturing
Abstract: This paper introduces a novel compliant mechanism combining lightweight and energy dissipation for aerial physical interaction. Weighting 400g at take-off, the mechanism is actuated in the forward body direction, enabling precise position control for force interaction and various other aerial manipulation tasks. The robotic arm, structured as a closed-loop kinematic chain, employs two deported servomotors. Each joint is actuated with a single tendon for active motion control in compression of the arm at the end-effector. Its elasto-mechanical design reduces weight and provides flexibility, allowing passive-compliant interactions without impacting the motors' integrity. Notably, the arm's damping can be adjusted based on the proposed inner frictional bulges. Experimental applications showcase the aerial system performance in both free-flight and physical interaction. The presented work may open safer applications for Micro Aerial Vehicle (MAV) in real environments subject to perturbations during interaction.

Keywords: Soft Robot Materials and Design, Grippers and Other End-Effectors, Software, Middleware and Programming Environments
Abstract: The design of soft robots’ deformable bodies is complex, partly due to the trade-off between softness for motion and stiffness for force generation. Self-contacts in soft structures can be used to address this problem but have been marginally investigated. In parallel, parametric designs and computational optimization tools constitute an important trend paving the way toward shareable and reproducible results. In this paper, we study the potential of self-contacts in the design of soft grippers using a multi-objective design optimization environment. This open-source environment is targeted toward grasping tasks and is used both for design and model calibration. Soft fingers with improved grasping quality and energy are obtained, taking into account different friction coefficients at the contacts and different shapes of the objects to grasp. They are experimentally validated in terms of mechanical behavior and grasping performances.

Keywords: Force and Tactile Sensing, Biomimetics, Human Factors and Human-in-the-Loop
Abstract: The development of human-like skills in robots benefits from understanding how humans perform tasks. Leveraging tactile or haptic training information to develop robot controller training is still an uncultivated land according to our current knowledge. Furthermore, to benefit from the advantages of state-of-the-art learning models for fine-grained manipulation of thumb and forefinger, we must learn from the tactile interactions in addition to the kinematics. We propose a wearable soft sleeve for a human index finger and thumb which includes fluidic pressure sensors that provide tactile information across these two fingers from several `taxels' or sensory receptors. By combining this with a means of capturing the location and form of contact, through the transfer of ink, we can create unique data sets which provide pressure and contact information. We captured the tactile interactions performed by humans to perform fine task execution such as turning pages. This was analyzed and interpreted to gain insights into human manipulation, and can also serve as a source of training data to guide a compliant robot hand to perform the same task. We consider this experiment as an effort to discover a solution for fine manipulation tasks based on limited tactile data while retaining dexterity and a way to demonstrate and realise fine-grained human-like skills.