Soft fluidic robots are widely used in various fields due to their advantages such as diversified design, large deformation, simple fabrication, and low cost. However, their development is hindered by issues like poor fluidic power sources, susceptibility to damage, and a lack of intelligent self-protection. Most existing fluidic robots rely on external, bulky pumps, limiting their mobility. While miniaturized rigid pumps are an option, they restrict design and performance. Innovative approaches like gas-liquid phase change or hydrogen peroxide decomposition offer untethered operation, but often suffer from slow actuation speed, reduced force, or poor controllability. Electrohydrodynamic (EHD) technology, manipulating dielectric liquids via electric fields, presents a promising solution. EHD pumps offer advantages such as lightweight, easy embedment, and quiet operation, but existing designs often show poor actuation performance, limiting their use in soft fluidic robots. Damage repair is another significant challenge. Existing self-healing research primarily focuses on elastomer self-healing, which is slow and only effective for small-area damage. Liquid-based self-healing has been explored but suffers from poor adhesion, limited stretchability, and long healing times. There is a need for self-protection mechanisms – the ability of a robot to detect damage and actively take measures for rapid self-healing. While some self-sensing and damage detecting soft devices exist, a fully integrated self-sensing, self-detecting, and self-heating system for soft fluidic robots remains to be achieved. This paper addresses these challenges by presenting a design for self-protection soft fluidic robots inspired by the human body’s self-healing mechanisms.

The existing literature extensively covers soft fluidic robots, highlighting their diverse applications and fabrication methods. However, there's a significant gap in addressing the issues of robust power sources and self-healing capabilities for these robots. Research on EHD pumps shows promise for miniaturization and integration, but current designs often lack the speed and force needed for effective actuation in complex tasks. The self-healing mechanisms explored in the literature mainly focus on material-level self-healing, often insufficient for large-area damage repair in soft robots. The need for untethered operation has led to various energy sources being explored, each with its own drawbacks. The incorporation of intelligent self-protection features through sensing and autonomous response mechanisms is still an underdeveloped area in soft robotics, with most existing solutions only partially addressing aspects of self-sensing or self-repair. This work reviews the literature demonstrating the challenges involved in achieving high-performance untethered actuation and rapid self-healing for soft robots and how the authors aim to address these significant technological shortcomings.

The researchers designed self-protection soft fluidic robots that integrate soft electrohydrodynamic (EHD) pumps, actuators, healing electrofluids, and electronic skins (E-skins). The design process involved four steps: 1. **Design of actuators and E-skins:** Actuators were 3D printed using conductive and non-conductive TPU. E-skins, based on liquid metal, were either coated or embedded into the actuators for sensing and heating. 2. **Design of soft EHD pumps:** A novel conical-array-porous-plate electrode pair design was employed and fabricated via multi-material 3D printing to enhance EHD flow and performance. This improved on previous designs that used cylindrical electrodes. The high manufacturing accuracy of 3D printing also allows the integration of multiple electrode pairs in series or parallel. 3. **Finite element simulation analysis:** Abaqus software was used to model and optimize the actuators and pumps based on material properties obtained from uniaxial tensile tests using Ecoflex 00-30 and 00-20. A hyperelastic incompressible Yeoh material model was employed for accurate representation of the materials behavior. 4. **Fabrication of robots:** Actuators, pumps, and E-skins were assembled, and chambers were filled with healing electrofluids. The resulting robots were designed to perform bending, twisting, and contracting motions. A WiFi module was integrated for wireless control via a smartphone app. The healing electrofluid was synthesized by mixing linalyl acetate functional liquid, methyltracetoxysilane, and dibutyltindilaurate. This combination enabled rapid large-area self-healing of silicone rubber and resulted in a self-healed film with excellent stretchability and adhesion. The E-skin was designed to enable both self-sensing and self-heating. The resistance change in the liquid metal E-skin, resulting from the robot’s deformation, was used for self-sensing and detecting damage. A self-judgment model was developed based on a large dataset of sensing data for both healthy and damaged robots. For self-heating, the E-skin’s temperature could be stabilized at ~160 °C, accelerating the curing of the healing electrofluid. Actuation tests were performed using a custom LabVIEW program, a data acquisition board, and a high-voltage amplifier. A PID closed-loop control system was also implemented using a camera for visual feedback, demonstrating the robot's controllability. To enhance functionality, multiple electrode pairs or actuators were combined. Examples included a high-load capacity robot and a soft gripper capable of catching a falling object. A mechanical sieve that was faster than traditional fluidic robots was also demonstrated, highlighting the potential for various applications.

This research yielded several key findings: 1. **High-Performance Soft EHD Pump:** The researchers developed a novel soft EHD pump design using a conical-array-porous-plate electrode pair fabricated using multi-material 3D printing. This resulted in significantly faster actuation speeds (<0.25 s) compared to existing EHD pumps, enabling high-speed robot actuation. The pump demonstrated a high output force exceeding 0.8N and an actuation stroke of 14 mm, exceeding prior state-of-the-art benchmarks. 2. **Rapid Large-Area Self-Healing:** A new healing electrofluid was synthesized, comprising linalyl acetate functional liquid, methyltracetoxysilane, and dibutyltindilaurate. This electrofluid, upon exposure to air, forms a self-healed film exhibiting excellent stretchability (>1200%) and strong adhesion to different silicone rubbers. This fluid effectively repairs large-area damage (damage where the fracture surface is not in contact), a significant improvement over existing self-healing methods. The curing time of the electrofluid was significantly reduced at elevated temperatures. 3. **Intelligent Self-Protection Behaviors:** The liquid-metal-based E-skin provides self-sensing capabilities, allowing the robot to monitor its state. A self-judgment model was developed to identify damage, prompting the robot to autonomously switch to self-heating mode to accelerate self-healing. This self-protection mechanism includes self-sensing (self-detecting), self-judgment, self-heating, and rapid self-healing, mimicking human self-repair mechanisms. 4. **Enhanced Functionality through Component Combination:** The modular design of the robots allows for the combination of electrodes and actuators to enhance functionality. Examples include a robot with increased load-bearing capacity by integrating multiple electrode pairs and a soft gripper with high-speed actuation capable of catching a falling object, and a mechanical sieve showcasing faster operation compared to prior fluidic robots. The ability to individually control multiple actuators demonstrates the potential of this system for complex tasks. 5. **Demonstrated Applications:** The capabilities of the robots were demonstrated through successful experimentation with bending, twisting, and contracting motions; a soft gripper, a mechanical sieve and a tactile thermal system.

The findings presented in this paper significantly advance the field of soft robotics by directly addressing critical limitations of existing soft fluidic robots. The development of the high-performance soft EHD pump provides a robust and highly efficient power source for untethered operation, overcoming the limitations of slower, less controllable alternatives. The innovative healing electrofluid allows for rapid and large-area self-healing, a crucial step towards creating robust and resilient soft robots capable of operating in unpredictable and potentially damaging environments. The integration of the multi-functional E-skin and the implementation of a self-judgment model represent a major leap towards physically intelligent robots capable of autonomous self-maintenance and self-repair. This design strategy not only tackles existing challenges but also offers a framework for creating highly adaptable soft robots suitable for a wider range of applications. The modular design allows for the straightforward incorporation of additional actuators and sensors, opening up exciting possibilities for the future development of more complex and sophisticated soft robotic systems. This work demonstrates the potential of bio-inspired design in achieving significant advancements in robotics.

This research demonstrates a novel approach to developing self-protection soft fluidic robots, successfully addressing the challenges of power source limitations, damage susceptibility, and a lack of intelligent self-protection mechanisms in existing soft robots. The key contributions are the development of high-performance soft EHD pumps, rapid large-area self-healing electrofluids, and a multi-functional E-skin enabling autonomous self-sensing, self-judgment, self-heating, and self-healing. This design strategy allows for enhanced functionality through modular component combination, as demonstrated through various applications. Future work could explore further miniaturization of the system, expanding the range of materials compatible with the healing electrofluid, and developing more sophisticated control algorithms for more complex tasks.

While this study demonstrates significant advancements in soft robotics, certain limitations exist. The self-healing capability is currently limited to a fracture surface distance of less than 5 cm. Larger damages may be challenging to repair fully. The self-judgment model relies on a predefined threshold; further refinement of the model might improve its accuracy and adaptability. The power consumption, while acceptable for the demonstrated applications, could be further optimized for longer operational times. The current study mainly focuses on silicone rubber; extending the compatibility of the healing electrofluid to a broader range of soft materials would enhance its versatility.