LM-Jelly: Liquid Metal Enabled Biomimetic Robotic Jellyfish

Abstract

Jellyfish have attracted worldwide attention owing to their fantastic moving styles, which also inspired development in soft robotics to meet the demands of underwater surveillance. In this study, a soft robotic jellyfish integrated liquid metal coil, and magnetic field is proposed for the first time to mimic the soft rowing propulsion of oblate jellyfish. The soft robotic jellyfish is actuated by the entirely soft electromagnetic actuators that enabled the gentle motion. Through conceptual experiments and computational fluid dynamics simulations, we systematically interpreted the mechanism of this robotic jellyfish and various factors to dominate its movement behaviors, which involve vortex formation and ascending modalities. Besides, underwater monitoring and bio-friendliness of robotic jellyfish were also demonstrated to illustrate its potential application scenarios and gentle motion characteristics. This study will help to broaden the vistas for liquid metal enabled bionic robotics in a wide range of underwater applications.

Introduction

In the long-term natural evolution, underwater creatures have developed their own unique propulsion methods, the most common of which is swing propulsion of fish.¹ However, jellyfish, as a typical representative of underwater mollusks, whose body is small and light, but with graceful, smooth, and flexible movement, provides a new perspective for imitating propulsion modes of aquatic animals. Their gentle expansion and contraction movements are fascinating and attractive, which have been gradually studied by many scholars.^2–5

Up to now, it is widely accepted that jellyfish's graceful propulsion mode can be divided into two distinct types by their wake patterns. One is rowing propulsion of oblate jellyfish with isolated and symmetric vortex rings wake, and the other is jet propulsion of prolate jellyfish whose wake consists of vortex followed by trailing jet of water.^6–8 Admittedly, jet propulsion always exhibits higher accelerations,⁹ but the rowing propulsion shows relatively low power consumption and high propulsive efficiencies.⁸

Because of these excellent characteristics of rowing jellyfish, oblate jellyfish-inspired robots¹⁰ have been promoted to improve traditional propeller-driven underwater robots with problems of large size, heavy weight, high cost, unfavorable noise, and hazard to underwater biological environment.¹¹ Therefore, the rowing propulsion is especially promoted.

Robotic jellyfish “Aquejelly” of German Festo company is an outstanding representative of the earliest bionic jellyfish.¹² They are driven by air cylinder, and eight tentacles are designed to help them glide gracefully through water. “Cyro” inspired by Cyanea capillata also has eight robotic arms,¹³ which is large and moving fast, but driven by linear direct current motors. Moreover, to avoid the usage of rigid moving parts, the flexible intelligent materials are explored to make bionic jellyfish, such as shape memory alloys,^14,15 dielectric elastomers,^16,17 magnetic composite elastomers,^18,19 ionic polymer metal composites,^15,20 pneumatic²¹ and hydraulic actuators,²² and so on.

Although the above materials have great actuation performance, a lot of robots are more like ephyra^23,24; their skeleton-like paddles are more similar to a folding umbrella than a real bell-shaped jellyfish. In addition, sometimes specific underwater environments are required.²⁵ Consequently, different driving approaches still need to be discovered.

In contrast with the above, we conceived that liquid metal electromagnetic actuators have unique advantages in imitating the soft propulsion of jellyfish. Liquid metal with perfect fluidity and superb conductivity can be arbitrarily changed in shape,²⁶ which could be widely used in various fields requiring high compliance of materials such as skin electronics.^27,28 At the same time, it can also be made into a liquid coil easily,²⁹ sealed in elastomers as a flexible actuator. Such an actuator has good electromagnetic responsibility and exhibits great driving performance,^30,31 which is conducive to developing a bionic robotic jellyfish. Most importantly, the biosafety of liquid metal renders soft electromagnetic actuators possibility to interact with marine life.^28,32–34

In this article, we proposed and demonstrated a low-power consumption and small design cost flexible robotic jellyfish driven by soft liquid metal-enabled electromagnetic actuator (called LM-Jelly) to imitate the rowing propulsion. The robotic jellyfish makes full use of the buoyancy of its head to integrate permanent magnets, which allows its motion no longer limited by external magnetic field. When applied an alternating current signal on the liquid metal-enabled electromagnetic actuator, the membrane will contract and expand under the action of the Lorenz force and the resulting hydrodynamics behaviors between the membrane and water actuate robotic jellyfish to swim upwards. Liquid metal adopted here owns both high conductivity and fluidity, which implies that the moving parts of the robot are fully flexible and harmless to underwater creatures.

We conducted a series of experiments to explore the factors that affect the movement of robotic jellyfish, including the amplitude, frequency, and duty of input signal, and found the optimal parameters for our LM-Jelly. We also simulated the motion of LM-Jelly to analyze the reasons for the ascending movement. Finally, we demonstrate application of LM-Jelly underwater monitoring and present that its gentle motion has produced almost no effect on biological activities. This research has broadened the application potential of such small bionic robotic jellyfish in underwater detection.

Materials and Methods

Design of LM-Jelly

An aequorea jellyfish without trailing jet is illustrated in Figure 1a, the rowing gaits of which are presented through the continuous alternation of contraction and expansion. Considering that there is no pulse jet, the volume change of the jellyfish bell during motion is ignored. Besides, the dominant interaction between jellyfish and fluid environments when jellyfish swims is at the disc-like bottom. Inspired by the jellyfish, we can simplify their movement as the rowing behavior of a thin film, as depicted in Figure 1b. The movement mechanism of LM-Jelly is mainly based on the flexible membrane flapping, which generates a counter-push force, allowing such jellyfish swim up in water lively.

FIG. 1.

The prototype and working mechanism of LM-Jelly. (a) An aequorea jellyfish without trailing jet in its wake patterns, photo courtesy of Pexels GmbH/Pia. (b) Schematic diagrams of aequorea jellyfish expansion and contraction movement and its simplified model. (c) Three-dimensional view of LM-Jelly and working principle when subjected to a current supply. (d) Exploded view of LM-Jelly. (e) Expansion and contraction of actual LM-Jelly.

To drive the flapping behavior of LM-Jelly, a soft electromagnetic actuator consisting of permanent magnet and liquid metal coil embedded in umbrella-shaped Ecoflex membrane is designed, and its working principle is in enlarged view of Figure 1c. The alternate currents are applied to the liquid metal coil to adjust the continuous alternation of contraction and expansion. In particular, when the liquid metal coil loaded a clockwise current, the umbrella membrane of LM-Jelly overcomes gravity and turns into flat under the Lorentz force during the expansion stage. On the contrary, when applied on a counter clockwise current, the umbrella membrane suffers a downward force and that is contraction stage.

The prototype structure of LM-Jelly is showcased in Figure 1d. As seen from the exploded view, LM-Jelly mainly consists of the jellyfish head and umbrella membrane. The umbrella membrane is linked together with jellyfish head by connecting nails. Importantly, their centers are fixed on the same axis to confine the relative motions of their centers. Herein, the jellyfish head, which is composed of acrylic hemisphere with a thin decoration Ecoflex film, permanent magnet, and sealing base three-dimensional (3D) printed with photosensitive resin, is designed for providing a shelter for permanent magnet and adjusting the buoyancy. To install the permanent magnet without weakening the magnetic field, the sealing base is designed with thickness of 2 mm; the fabrication process is shown in detail in Supplementary Data.

What is more, it is of great importance to balance the gravity and buoyance of LM-Jelly to maintain its levitating posture. It would sink quickly and could not swim up when its density is too high. When density is low, it would float and be incapable of sinking. Besides, the center of buoyancy should stay above its center of mass so that the LM-Jelly can swim steadily. At this stage, there are a few methods with practical trials ever tried to adjust its density, including weight adding method, bladder method, and temporary drainage method, respectively. Specific explanations and their characteristics are presented in Supplementary Table S1. Considering the convenience of operations, temporary drainage method is chosen to alter the density of LM-Jelly to be slightly larger compared with water in subsequent experiments. The actual LM-Jelly is illustrated in Figure 1e.

The core component of LM-Jelly lies in its umbrella membrane. For manufacturing convenience, the membrane is designed flat, and the liquid metal coil is designed as a spiral shape with the consideration of the arrangement of the electrodes and the distribution of the film's center of gravity, as shown in Figure 2a. The fabrication process and design dimensions of the umbrella membrane are shown in Figure 2b and c, respectively. The reserved area in the center is for connecting nail installation.

FIG. 2.

Fabrication process of umbrella membrane. (a) The real umbrella membrane. (b) Main steps of the umbrella membrane fabrication: Molding of the silicone elastomer, bonding to a sheet of elastomer and injection of liquid metal. (c) The design dimensions of umbrella membrane. (d) The front and (e) cross-sectional views of the umbrella membrane under the microscope.

To prevent the liquid metal channel from being damaged during injection, the distance between adjacent arcs is almost twice the width of channel. Before assembling the umbrella membrane into LM-Jelly, liquid metal coil is checked for its conductivity through a multimeter. The resistance of liquid metal coil is about 0.8 Ω and the resistance of which exhibited small change that can be ignored in the process of its deformation, as shown in Supplementary Figure S3. Moreover, its integrity is observed under microscope. As seen from microscope images in Figure 2d and e, the liquid metal coil is filled completely and evenly in the Ecoflex channel, ensuring the reliability during its movement.

Characteristics of actuating performance

Tensile and bending tests are performed to qualitatively compare the difference when deforming between liquid and solid metal. The size of samples and experimental equipment are shown in the Supplementary Data and Supplementary Figure S2. Similar to the umbrella membrane, the samples are designed of Ecoflex 00-30 blocks with liquid metal channels. The samples without liquid metal channels at room temperature are used as control groups, and Ecoflex 00-30 blocks embedded with liquid metal and solid copper wires at room temperature are tested.

The devices and samples are presented in Supplementary Figure S2. The results in Supplementary Figure S2c and d demonstrated that the addition of liquid metal has almost no effect on the performance of the entire Ecoflex 00-30 blocks, whether it is bending or stretching. The subtle differences in the illustrations are negligible, which were caused by the slight slip between Ecoflex and the fixture. However, the addition of the copper wires is bone support for soft materials and makes the Ecoflex blocks harder to deform. The current tests on fish responses behavior with LM-Jelly were approved by the Institutional Animal Care and Use Committee (IACUC) of Tsinghua University, Beijing, China under contract No. SYXK(Jing)2014-0024.

Results and Discussion

Characterization of the actuating performance

After LM-Jelly was assembled, a series of experiments are conducted to test its performance of motion. The rising speed of LM-Jelly head and the amplitude of robotic LM-Jelly are used to describe the performance of motion. The key factors that affect the motion are parameters of input voltage signal, which include magnitude, duty cycle, and frequency.

The displacement of rowing robotic jellyfish is shown in Figure 3a under an optimal input signal with amplitude of 7.5 V, duty cycle of 70%, and frequency of 0.8 Hz. From this displacement overtime curve, it can be concluded that the speed of each cycle is steady and its average speed maintained at 6 mm/s. LM-Jelly moves slightly downwards when the reverse voltage (expansion) is applied and upwards immediately when the forward voltage (contraction) is on. Naturally, when there is no electrical supply, the membrane sags, and LM-Jelly drops under gravity (Supplementary Movie S1).

FIG. 3.

Characterizing locomotion with respect to different operating parameters. (a) Displacement of LM-Jelly overtime curve, the solid line represents the input voltage signal. The illustration shows the position and state of LM-Jelly at the beginning of each cycle. (b) The influence of the applied voltage. The illustration is schematic diagram of angle calculation. (c) The states of contraction and expansion of the robot under different applied voltage. (d) The influence of the duty cycle. (e) The influence of the frequency. The illustrations are expansion angle graphs of actual LM-Jelly.

The first aspect that should be considered to improve the velocity of LM-Jelly is the magnitude of voltage supply. As illustrated in Figure 3b, the input electrical power increases with the amplitude of input voltage increasing, which results in faster velocity when the duty of the input signal is fixed to 50% and the frequency is fixed to 0.8 Hz. Moreover, the start voltage of LM-Jelly is 2.5 V, namely, only when the input voltage is >2.5 V, LM-Jelly could overcome the gravity and swim upwards. But when the applied voltage exceeds 9 V, increasing the voltage has rare effect on its movement except for increasing the generation of Joule heat.

In addition, the range of motion of LM-Jelly umbrella can be expressed by the difference between the contraction angle and expansion angle (shown as purple shadow in Fig. 3b). The greater the difference is, the larger the deformation of robotic LM-Jelly is (Fig. 3c), which can cause LM-Jelly to swim faster under the same duty and frequency.

Duty cycle tests are performed to investigate the dynamics of the present LM-Jelly. In the duty cycle of 10–90% range, the LM-Jelly reveals the maximum movement speed, namely, 6 mm/s when duty cycle is 70%, and the voltage of input electrical signal is fixed to 7.5 V and the frequency to 0.8 Hz, as shown in Figure 3c. When duty cycle is less than 70%, the contraction time is much less than expansion time. Consequently, LM-Jelly umbrella membrane could not reach the maximum contraction extent in the contraction process. On the contrary, when duty cycle is too large, umbrella membrane hardly expands; hence, the velocity becomes lower than the peak.

The frequency of applied voltage is of the same importance for LM-Jelly velocity. When the applied voltage was controlled at 7.5 V, and duty cycle at 70%, frequency range from 0.8 to 1.7 Hz obtains relatively high and similar speed of robotic jellyfish, namely around 6 mm/s. Since the duty cycle is fixed to 70%, the contraction time is larger than expansion whatever the frequency is, and the contraction angle is maintained at a relatively high level like purple line in Figure 3d.

When the frequency is lower than 0.8 Hz, the umbrella membrane has enough time to expand to flat state (left expansion illustration in Fig. 3e). However, it takes too much time to finish one cycle, which would cause low average velocity of LM-Jelly. But when frequency is much higher, one cycle period is too short to fully expand, like the right expansion illustration of Figure 3e, so the velocity declines gradually. Therefore, only balancing the time to expand and cycle, LM-Jelly can swim at a high speed. Combined with the analysis of the influence of the above parameters on the movement, it is found that only when the membrane can fully contract and relax in a short time, can a fast movement be achieved. That is to say, the larger the current, the faster the LM-Jelly moves at the same frequency. When the current increases, LM-Jelly could move more quickly at faster frequencies.

The test performed with our robotic jellyfish has been compared with other adult oblate jellyfish inspired swimmers (Table 1). Generally, robots with rigid moving parts exhibit faster movement speed because of the stronger driving forces. But the speed of LM-Jelly did not show an order of magnitude difference based on its body length. Moreover, compared with other electromagnetic actuation, its motion is not limited by the external magnetic field.

Table 1.

The Characteristic Comparison of LM-Jelly with Other Adult Rowing Jellyfish-Inspired Swimmers

	Speed (L/s)	Actuation method	Range of motion	Rigid driving part
Jellyfish⁸	0.16–0.66	—	Unlimited	No
Cyro¹³	0.49	Mechanical arm	Unlimited	Yes
Underwater vehicle²⁰	0.026	Ionic polymer metal composite actuators	Unlimited	Yes
FludoJelly²¹	0.73	Soft pneumatic composite actuators	Unlimited	Yes
Mini robot³⁵	—	Electromagnetic actuation	Limited	No
Our work	0.2	Electromagnetic actuation	Unlimited	No

Dynamic analysis of the LM-Jelly

It is the interaction between LM-Jelly and fluid that makes the LM-Jelly to move. When the umbrella membrane is subjected to electromagnetic force and contracts, the vortexes are dynamically formed under the membrane. Then the reaction force of the fluid acting on the LM-Jelly would drive it upward. To better reveal the mechanism of its movement, the process was simulated by commercial software COMSOL Multiphysics 5.5. Herein, a mathematical model was built to simulate the dynamic performance of LM-Jelly. The physical model is schematically illustrated in Figure 4a.

FIG. 4.

Simulations of LM-Jelly motion. (a) Physical model of LM-Jelly. (b) Comparison of experimental and simulation movement results of one cycle. (c) Velocity nephogram of LM-Jelly in one cycle. The last photo is an enlarged view of T = 0.87 s. The arrows indicate the velocity of the fluid. (d) Fluorescent display of LM-Jelly vortex formation and shedding during contraction and expansion.

In this study, the following assumptions are used for calculating the locomotion process of LM-Jelly: (1) The influence of electrode wire on the dynamic performance of LM-Jelly is ignored. (2) The heat generation and diffusion are ignored, owing to the excellent conductivity of liquid metal. (3) The entire device is simplified as an axisymmetric model to improve computational efficiency.

The Multiphysics modeling of LM-Jelly locomotion mainly consists of three physical fields, including magnetic field, solid mechanics, and the turbulent flow. The relevant governing equations are presented in the Supplementary Data. The magnetic field is provided by a sintered NdFeB magnet N33UH with remanent flux density of 1.16 T; the spatial distribution of the magnetic field around the permanent magnet is showcased in Supplementary Figure S6. A current signal with amplitude of 0.62 A, frequency of 0.8 Hz, and duty cycle of 70% is applied. The grid is divided as shown in Supplementary Figure S4.

The solution process is divided into two steps, steady state and transient state. First, the steady state is used to solve the magnetic field. The magnetic field and solid mechanics are one-way coupled by electromagnetic force acting on the coil, which could be expressed by Equation (1). The fluid-structure two-way coupling in multiphysics is used to solve solid mechanics and turbulence in the transient solution process, and the time step is set to 0.01 s. Then use the results of the transient solution to iterate to calculate the steady-state process at the next time node. $F_{V} = J \times B$ (1)

where $J$ is the current density; $B$ is the magnetic flux density.

The velocity nephograms of LM-Jelly are presented in Figure 4c. In the stage of the contraction (0–0.87 s), the forward current is applied to liquid metal coil, while the expansion phase (0.88–1.25 s) is suffered to reverse current. As seen from the Figure 4c, during the contraction process, the umbrella membrane acts like a paddle as LM-Jelly rows, and an inwardly vortex is formed under umbrella membrane margin (Supplementary Movie S2). However, the simulated vortex has some differences with that observed in the actual process like the illustration of Figure 4d (also shown in Supplementary Movie S3).

In the real process, thrust is generated by resistance of fluid, and LM-Jelly rises up. As time goes by, it turns to the expansion period, and the umbrella membrane tends to be flat. Following that, the shed vortex develops (Fig. 4d). The movement of LM-Jelly is hindered and begins to decline due to gravity. As shown in Figure 4b, the simulated and experimental results of the whole motion process in one cycle indicate that LM-Jelly rises in the contraction phase and slightly drops in the expansion phase, which share extremely similar trends. Nevertheless, LM-Jelly in the actual situation can hardly be made completely axisymmetric, which responds to electrical signals a little faster than simulation. Furthermore, the assumption of two-dimensional axisymmetric structure makes it to only move axisymmetrically and the amplitude of motion decreased. The velocity of the fluid is smaller than the real process, and the vortex shedding behind can be hardly observed.

The results of present experiments have drawn a conclusion that the faster movement of LM-Jelly requires greater Lorentz force. Except for that, different parameters of the umbrella membrane can also be the factors that affect the locomotion. The above model was used to discuss the influence of the parameters of the membrane on its motion. As shown in Figure 5, the changes of LM-Jelly motion are obtained with different film thickness d1, film modulus E, coil inner circle radius d2, and coil spacing d3 as variables, respectively (Supplementary Fig. S5).

FIG. 5.

Simulations of LM-Jelly motion with different parameters. (a) The thickness d1 of the umbrella membrane. (b) The modulus E of the membrane. (c) The coil inner circle radius d2. (d) The coil spacing d3.

The increases of the thickness d1 and modulus E of the membrane result in the decrease of its velocity, but the effect is little. This is because when the film is thicker or the elastic modulus is larger, the same deformation requires more force but these parameters do not change the magnitude of Lorentz force on the liquid metal coil. But the layout of the coil could affect the magnitude of its suffering Lorentz force.

The velocity of LM-Jelly does not simply increase or decrease with the changes of coil inner circle radius d2 and coil spacing d3. If the d2 or d3 is too small, the Lorentz force is concentrated on the center, and its deformation is also concentrated on the center, which makes it difficult to drive the surrounding movement. But if d2 or d3 is too large, because the magnetic field strength decreases sharply in space (Supplementary Fig. S6), the Lorentz force will also be much smaller, which causes the speed to fail. The parameters we adopted in our experiments, d2 = 16 mm and d3 = 2.8 mm, could help to achieve a better motion.

Application of underwater detection

Traditional propeller-driven robots often work with high noise and vibration, which is hard to move close to some underwater creatures. To exhibit the security and gentleness of our LM-Jelly using a simple and intuitive way, the robotic jellyfish and propeller-driven robot were, respectively, put in a tank to swim together with small real tropical fishes. First, let the small fishes adapt to surroundings for 5 min. Then the robots were started, as shown in Supplementary Movie S4. Even though the jellyfish is in motion and suffered relatively high current, the small fishes were still swimming quietly and did not show any avoidance behaviors. They can even swim around the side of LM-Jelly, as shown in Figure 6a.

FIG. 6.

Experiments on underwater monitoring of LM-Jelly. (a) Sequential snapshots of actual LM-Jelly swimming and Propeller-driven robot with small tropical fishes. Circles mark the location of the fish. (b) The propeller-driven robot. (c) LM-Jelly model equipped with small Bluetooth surveillance camera. (d) Schematic diagram of underwater detection of LM-Jelly, background images courtesy of Pexels GmbH/Pedro Furtado and Jess Vide.

On the contrary, when the propeller-driven robot started up, the fishes stayed away from the robot as far as possible. When the propeller approached, the fishes escaped immediately (Fig. 6a; also see Supplementary Movie S4). The thruster propellers are presented in Figure 6b. From the two different reactions of fishes to robots, we can infer that robotic jellyfish is more friendly to creatures because of its soft motion.

To observe the underwater world from the perspective of LM-Jelly, LM-Jelly is equipped with a small Bluetooth surveillance camera like Figure 6c. Underwater detection is displayed in fish tank as Supplementary Movie S5, in which a yellow fish model is floating as target. It is worth noting that due to the very gentle movement of our LM-Jelly, the view of camera is highly steady despite the quality of the lens being not high enough. Figure 6d shows future detection application of LM-Jelly in the ocean. Because its morphology is extremely similar to real jellyfish, such artificially made “creature” has excellent concealment under water, which is beneficial to underwater reconnaissance and observation.

So far, the LM-jelly has demonstrated its gentle movement for the fully soft electromagnetic actuator. Although the robotic jellyfish shows non-negligible advantages and great application prospects in an underwater environment, there still remains some problems unsolved, for example, the perfect 3D-rotationally symmetrical motion and the kinematics. In our present work, the bell of real adult jellyfish is simplified as a flat membrane, and the spiral design of liquid metal coil result in that contraction is smaller than bending.

To solve this problem, more scientific coil designs are required. In addition, the speed of the robot and the rowing direction are also remained to be solved. 3D swimming of LM-Jelly is definitely needed for underwater detection, which can be achieved by improving coil distribution or adding direction controlling of magnet in the LM-Jelly head. And to speed up the LM-Jelly, stronger magnet and input current density could be more helpful. In a word, there is still a disparity between the mobility of robotic jellyfish and that of real jellyfish, and more efforts should be put into this liquid metal enabled biomimetic robotic jellyfish in the coming time.

Conclusions

In summary, we proposed and demonstrated the biomimetic LM-Jelly by integrating magnetic field and electromagnetic actuator for the first time, whose movement could be regulated by the current signal input on the liquid metal coil. In magnetic fields, the liquid metal energizing coil is subjected to the Lorentz force, which could drive its umbrella membrane to contract and expand in water. In turn, the reaction force of liquid environment drives LM-Jelly upwards. Through controlling variables, we, respectively, clarified the influence of magnitude, duty cycle, and frequency of input signal on its motion and obtained the optimal input signal parameters.

We found that LM-Jelly could swim upwards at a maximum speed of 6 mm/s when input signal was with amplitude of 7.5 V, duty cycle of 70%, and frequency of 0.8 Hz. Its contraction and expansion movement were also simulated to analyze the formation process of vortex caused by its motion, which was basically consistent with the experimental results.

In addition, the robotic jellyfish for underwater detection was proven, and the stability of its motion is conducive to observing the underwater environment. Due to gentle movement of LM-Jelly, it could get along well with easily frightened underwater creatures, which can be a hard thing for conventional rigid robots. More importantly, LM-Jelly is simple to make and drive with no additional rigid moving parts, and it possesses an entirely soft actuator that realizes the movement. In the near future, how to increase the speed of LM-Jelly and how to achieve directional control remained to be further explored. It is believed that the successful preparation of liquid metal-enabled robotic jellyfish will clearly promote the development of flexible underwater robots, which is also expected to provide a feasible path and new vision for the realization of advanced and practical underwater liquid metal-based flexible robots in the near future.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

Funding Information

This work was supported by National Natural Science Foundation of China under Grant Nos. 91748206, the National Key Research and Development Program of China (No. 2020YFC0122301).

Supplementary Material

References

Marchese

, Onal

, Rus

. Autonomous soft robotic fish capable of escape maneuvers using fluidic elastomer actuators. Soft Robot, 2014; 1:75–87.

Bajcar

, Malačič

, Malej

, et al. Kinematic properties of the jellyfish Aurelia sp. Hydrobiologia, 2008; 616:279–289.

Dabiri

, Colin

, Costello

, et al. Flow patterns generated by oblate medusan jellyfish: field measurements and laboratory analyses. J Exp Biol, 2005; 208:1257–1265.

McHenry

, Jed

. The ontogenetic scaling of hydrodynamics and swimming performance in jellyfish (Aurelia aurita). J Exp Biol, 2003; 206:4125–4137.

Sahin

, Mohseni

, Colin

. The numerical comparison of flow patterns and propulsive performances for the hydromedusae Sarsia tubulosa and Aequorea victoria. J Exp Biol, 2009; 212:2656–2667.

Dabiri

, Colin

, Katija

, et al. A wake-based correlate of swimming performance and foraging behavior in seven co-occurring jellyfish species. J Exp Biol, 2010; 213:1217–1225.

Costello

, Colin

, Dabiri

. Medusan morphospace: phylogenetic constraints, biomechanical solutions, and ecological consequences. Invertebr Biol, 2008; 127:265–290.

Gemmell

, Costello

, Colin

, et al. Passive energy recapture in jellyfish contributes to propulsive advantage over other metazoans. Proc Natl Acad Sci USA, 2013; 110:17904–17909.

Colin

, Costello

. Morphology, swimming performance and propulsive mode of six co-occurring hydromedusae. J Exp Biol, 2002; 205:427.

10.

Bujard

, Giorgio-Serchi

, Weymouth

. A resonant squid-inspired robot unlocks biological propulsive efficiency. Sci Robot, 2021; 6:eabd2971.

11.

McKenna

, Ross

, Wiggins

, et al. Underwater radiated noise from modern commercial ships. J Acoust Soc Am, 2012; 131:92–103.

12.

AquaJellies 2.0. https://www.festo.com/group/en/cms/10227.htm (accessed June 16, 2018).

13.

Villanueva

, Marut

, Michael

, et al. Biomimetic autonomous robot inspired by the Cyanea capillata (Cyro). Bioinspir Biomim, 2013; 8:046005.

14.

Almubarak

, Punnoose

, Maly

, et al. KryptoJelly: a jellyfish robot with confined, adjustable pre-stress, and easily replaceable shape memory alloy NiTi actuators. Smart Mater Struct, 2020; 29:075011.

15.

Bar-Cohen

, Najem

, Leo

. A bio-inspired bell kinematics design of a jellyfish robot using ionic polymer metal composites actuators. EAPAD, 2012; 8340:83401Q.

16.

Cheng

, Li

, Liang

, et al. Untethered soft robotic jellyfish. Smart Mater Struct, 2019; 28:015019.

17.

Christianson

, Bayag

, Li

, et al. Jellyfish-inspired soft robot driven by fluid electrode dielectric organic robotic actuators. Front Robot AI, 2019; 6:00126.

18.

Ren

, Hu

, Dong

, et al. Multi-functional soft-bodied jellyfish-like swimming. Nat Commun, 2019; 10:2703.

19.

, Ze

, Zhang

, et al. Symmetry-breaking actuation mechanism for soft robotics and active metamaterials. ACS Appl Mater Interfaces, 2019; 11:41649–41658.

20.

Najem

, Sarles

, Akle

, et al. Biomimetic jellyfish-inspired underwater vehicle actuated by ionic polymer metal composite actuators. Smart Mater Struct, 2012; 21:094026.

21.

Joshi

, Kulkarni

, Tadesse

. FludoJelly: experimental study on jellyfish-like soft robot enabled by soft pneumatic composite (SPC). Robotics, 2019; 8:56.

22.

Frame

, Lopez

, Curet

, et al. Thrust force characterization of free-swimming soft robotic jellyfish. Bioinspir Biomim, 2018; 13:064001.

23.

Higgins JE

III

, Ford

, Costello

. Transitions in morphology, nematocyst distribution, fluid motions, and prey capture during development of the scyphomedusa Cyanea capillata. Biol Bull, 2008; 214:29–41.

24.

Nawroth

, Lee

, Feinberg

, et al. A tissue-engineered jellyfish with biomimetic propulsion. Nat Biotechnol, 2012; 30:792–797.

25.

, Go

, Ko

, et al. Magnetic actuated pH-responsive hydrogel-based soft micro-robot for targeted drug delivery. Smart Mater Struct, 2016; 25:027001.

26.

Chen

, Wang

H-Z

, Zhao

R-Q

, et al. Liquid metal composites. Matter, 2020; 2:1446–1480.

27.

Wang

, Ren

, Liu

. Liquid metal enabled electrobiology: a new frontier to tackle disease challenges. Micromachines Basel, 2018; 9:360.

28.

Wang

, Guo

, Liu

. Liquid metal based soft robotics: materials, designs, and applications. Adv Mater Technol, 2018; 4:1800549.

29.

Jin

, Park

, Hong

, et al. Stretchable loudspeaker using liquid metal microchannel. Sci Rep, 2015; 5:11695.

30.

Guo

, Sheng

, Gong

, et al. Liquid metal spiral coil enabled soft electromagnetic actuator. Sci China Technol Sci, 2017; 61:516–521.

31.

Mao

, Drack

, Karami-Mosammam

, et al. Soft electromagnetic actuators. Sci Adv, 2020; 6:eabc0251.

32.

Wang

, Zhang

, Guo

, et al. Conformable liquid metal printed epidermal electronics for smart physiological monitoring and simulation treatment. J Micromech Microeng, 2018; 28:034003.

33.

Sun

, Yuan

, Rao

, et al. Amorphous liquid metal electrodes enabled conformable electrochemical therapy of tumors. Biomaterials, 2017; 146:156–167.

34.

Kim

J-H

, Kim

, So

J-H

, et al. Cytotoxicity of gallium-indium liquid metal in an aqueous environment. ACS Appl Mater Interfaces, 2018; 10:17448–17454.

35.

, Na

, Lee

, et al. A jellyfish-like swimming mini-robot actuated by an electromagnetic actuation system. Smart Mater Struct, 2012; 21:057001.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

3.66 MB

4.06 MB

0.89 MB

16.80 MB

11.54 MB

0.79 MB

0.00 MB