Recent progress on underwater soft robots: adhesion, grabbing, actuating, and sensing

Abstract

The research on biomimetic robots, especially soft robots with flexible materials as the main structure, is constantly being explored. It integrates multi-disciplinary content, such as bionics, material science, mechatronics engineering, and control theory, and belongs to the cross-disciplinary field related to mechanical bionics and biological manufacturing. With the continuous development of various related disciplines, this area has become a hot research field. Particularly with the development of practical technologies such as 3D printing technology, shape memory alloy, piezoelectric materials, and hydrogels at the present stage, the functions and forms of soft robots are constantly being further developed, and a variety of new soft robots keep emerging. Soft robots, combined with their own materials or structural characteristics of large deformation, have almost unlimited degrees of freedom (DoF) compared with rigid robots, which also provide a more reliable structural basis for soft robots to adapt to the natural environment. Therefore, soft robots will have extremely strong adaptability in some special conditions. As a type of robot made of flexible materials, the changeable pose structure of soft robots is especially suitable for the large application environment of the ocean. Soft robots working underwater can better mimic the movement characteristics of marine life in the hope of achieving more complex underwater tasks. The main focus of this paper is to classify different types of underwater organisms according to their common motion modes, focusing on the achievements of some bionic mechanisms in different functional fields that have imitated various motion modes underwater in recent years (e.g., the underwater sucking glove, the underwater Gripper, and the self-powered soft robot). The development of various task types (e.g., grasping, adhesive, driving or swimming, and sensing functions) and mechanism realization forms of the underwater soft robot are described based on this article.

1 Introduction

In this review, we mainly discuss the common motion patterns of many underwater organisms and focus on recent developments in this field according to the classification of the same motion pattern. In the field of soft adhesion, soft suckers with adhesion capacity, such as octopus-like, remora-like, and clingfish-like adhesion, are highlighted. In the field of soft grippers, this paper focuses on the tentacles of the octopus and the soft grasping claws of mammals. In the field of soft actuators, the oscillating drive of fish, the jet drive of jellyfish, and the soft drive of the underwater crawling or floating motion of starfish and sea spiders are introduced. In the field of soft sensors, the development of flow or pressure sensors that mimic the lateral line of fish and the tactile sensation of skin is described. Finally, the challenges and prospects of soft robots in future underwater applications are summarized. The main structure is shown in Figure 1.

FIGURE 1

2 Soft adhesion

In terms of soft adhesion, inspired by octopus, remora, starfish, and other organisms, scholars have conducted a study focusing on the morphology and kinematics of these organisms and then developed it into a major direction of flexible robots at present. In the field of adhesion, this paper divides adhesion into active and passive. The main way to distinguish the two is that active adhesion needs external energy for adhesion.

2.1 Octopus-like

By studying the structure of octopus suckers (Tramacere et al., 2014a; Tramacere et al., 2013; Tramacere et al., 2012), the understanding of the mechanism of octopus suckers has been improved to some extent. Therefore, based on the development of flexible materials, research on some cup structure (octopus structure) suckers (Tramacere et al., 2014b; Tramacere et al., 2015) is further promoted. Certain research achievements have been made in recent years, whether it is dry-type, wet-type (Qiao et al., 2018; Wang and Hensel, 2021; Wang Yue et al., 2020), or dry and wet dual-purpose (Wang Yue et al., 2019) suckers. Under the development of passive adhesion, some progress has also been made in a series of active adhesion.

Frey et al. (2022) were inspired by octopus suckers and integrated embedded sensors, processors, controllers, and other components to develop a controllable octopus sucker that can freely control surface adhesion. This sucker can be well-attached to some irregular surfaces and can perform certain intelligent control. As shown in Figure 2, when the sensor is close to the specified distance, the suction cup begins to perform vacuum adhesion. When the adhesion surface needs to be removed, the suction cup will be inflated to remove the adhesion.

FIGURE 2

Swift et al. (2020) introduced an active polydimethylsiloxane membrane, which uses pneumatic control and tunable rigidity material to study the adhesion force and used pneumatic positive and negative pressure to cooperatively control the contact stiffness and geometry to perform adhesion and release. The active polydimethylsiloxane film can be extremely extended, and the adhesion and release time is approximately 0.1 s, which is very suitable for applications requiring rapid adhesion and release.

Wang et al. (2021b) introduced a new magnetically driven and energy-saving intelligent sucker based on the elastic energy storage mechanism of an octopus sucker. This sucker has the characteristics of rapid adjustment of adhesion strength and high-speed switching of adhesion state. The main adhesion characteristic of these structures is to manipulate the magnetic particles to imitate the movement of the acetabular roof of octopus suckers.

2.2 Remora-like

Based on the morphology of the remora research (Gamel et al., 2019; Beckert et al., 2015; Flammang and Kenaley, 2017), researchers found that it differs from the mimic fish sucker. The biggest characteristic of this type of adhesion is the presence of lamella and spinule support to achieve adhesion. In addition, the different lamella and spinule geometry makes the shear adhesion significantly different.

In passive adhesion, Lee et al. (2019) made polymer-based adhesives inspired by remora suckers and conducted a series of experiments. As shown in Figure 3, the tensile and shear strengths of the adhesive model in water were 266.8 and 194.2 kPa, respectively. The adhesive maintained excellent adhesion and friction properties after multiple tests.

FIGURE 3

In active adhesion, several researchers (Li et al., 2022; Wang et al., 2017a; Wang et al., 2019b; Wang Siqi et al., 2020; Wang Siqi et al., 2022) proposed an active biomimetic sucker made of multi-material composite based on morphology and kinematics of remora, using laser processing and 3D printing technologies. This type of sucker is made of composite material and can produce higher strength than the weight of the disc. The whole mechanism is similar to the disc of live remora. Studies have shown that suckers combined with hard spines and soft materials can significantly increase the friction force on the disks, which have been tested on the substrates with different roughness.

2.3 Clingfish-like

Earlier studies have been conducted on the adhesion mechanism of clingfish (Green and Louise Barber, 1988), which revealed the microscopic structure and mechanism of adhesion of clingfish from a biological perspective. At the present stage, the focus is more on the study of the adhesion capacity of clingfish, and the evaluation criteria are generally the influence of dirt and surface roughness on the adhesion of clingfish (Ditsche et al., 2014; Wainwright et al., 2013). Based on these studies, scholars (e.g., Ditsche and Adam, 2019; Sandoval et al., 2019) further transformed their understanding of clingfish into artificial bionic suckers and conducted a series of tests. The prototype sucker was developed by Ditsche Petra et al., which could stick to the rock and rough surface, and the strength of the sucker was up to 70 kPa on a rough surface with a particle size of 270 μm.

2.4 Other mimics

In addition, except for some widely known bionic artificial suckers, researchers studied other organisms (e.g., starfish). Sandoval et al. (2022) proposed a directed adhesion sucker. Based on the directed adhesion characteristics of specific materials, they developed a type of sucker that is easy to absorb when loading in one direction and release when loading in the other direction. The characteristic of the sucker is that it can achieve obvious adhesion and detachment without other active control (Ishida et al., 2022). In other words, the sucker produces a strong adhesion capacity at a certain angle and can easily detach when the load is applied in the other direction.

2.5 Challenges and limitations

The main task of adhesion is to solve the problem of how to carry out the connection operation between two objects in various complex wet environments. Furthermore, the evaluation of the strength and reliability of the connection is an area of challenge. This type of surface adhesion can be applied to the docking of the underwater autonomous vehicle, the launching of the underwater autonomous vehicle, and the joint movement of master and slave robots. In conclusion, as shown in Table 1, the limitations and challenges can be summarized as follows.

The main characteristics of the octopus-like suction cup are the wide range of applications (Eroğlu and Ebru Devrim, 2023) and the adjustable pressure. However, the resistance to shear stress needs to be further improved.

The remora-like sucker has added bone spurs compared to the octopus sucker. Therefore, it has better resistance to tangential force in a single direction. However, its structure is more complex, its overall structure is non-homogeneous, and the manufacturing cost is relatively high.

The clingfish-like sucker has a larger adsorption edge than the octopus-like sucker, which enables the sucker to obtain a certain good adsorption capacity with a larger surface roughness.

Furthermore, other mimic structures can make a good effort in the adhesive area (Soltannia and Sameoto, 2014), such as gecko-inspired (Soltannia and Sameoto, 2014) and sea star-inspired.

3 Soft gripper

As a very popular application direction of flexible mechanisms, the development of soft grippers underwater is also very promising. Particularly in grasping some vulnerable marine specimens, the soft gripper can better play its own advantages of flexible materials. In addition, the flexible gripper can be combined with the existing rigid gripper to further improve the grabbing range of targets. At the present stage, mainstream claw design inspiration comes from the construction of imitations of the human hand and other structures, such as the construction of octopus tentacles. The movement of the gripper is generally performed using the deformation of shape memory alloy (SMA) and/or soft cavity to grasp the target object. The general structure of these claws includes three parts: wrist, finger, and grasping surface. According to the different swing positions of the fingers, the gripper can be subdivided into parallel gripper, intermeshed claw, and cage claw.

The classification standard of the gripper is arranged according to the analysis of Mazzeo et al. (2022), in which the parallel gripper refers to the gripper whose fingers keep parallel when opening and closing. The intermeshed claw is a claw with three or more fingers with the same end, rotating around the end, and the edges of the fingers have certain contact. The cage claw refers to the combination of multiple fingers to envelope a certain area.

3.1 Basic part introduction

With regard to wrist research, Gong et al. (2018a) designed a cylindrical soft arm structure with a corrugated texture and drove it pneumatically. They conducted underwater research under laboratory conditions and quantitatively analyzed the open-loop kinematic model performance of this model. The position error caused by the open-loop model control was tested without visual and sensor feedback.

In addition, Kurumaya et al. (2018) designed a cylindrical, flexible wrist joint structure that uses a pneumatic drive under normal atmospheric conditions but uses a hydraulic drive for functional testing under a high static pressure environment (equivalent to at least 2,300 m water depth pressure).

In addition to the cylindrical wrist, Shen et al. (2021), inspired by origami technology, developed a wrist with driving and sensing functions, which is controlled by fluid, as shown in Figure 4. The origami process throughout the wrist enables excellent high-load output and robust and accurate sensing performance without the need for external cladding of traditional sensors, greatly simplifies the fabrication process, and enables the wrist to be applied to complex interactive tasks in harsh environments.

FIGURE 4

Sun et al. (2020) studied the bending degree of soft, curved fingers driven by the typical bellows tube under different pressure conditions and concluded that the external water pressure could make the actuator bend more under different external pressures (1 atm∼15 MPa). In other words, the increase in the bending angle is positively correlated with the environmental pressure.

For the study on claw surface, Jiang et al. (2021) proposed seven finger surfaces designed by observing the claw shape of Boston lobster and made them with soft silicone. This new soft silicone claw surface has accumulated more than 14,000 grasp attempts on land (71.4%) and underwater (28.6%). The experiments first selected the optimal design through land experiments. Subsequently, the underwater capability was further tested. It was verified that the bionic finger surface design could increase the grasping success rate by up to 18.2%. In summary, a certain number of studies have been performed on the wrist joint, fingers, and finger surface type of claw. Different types of wrists, fingers, and contact surfaces will produce different claw forms and application effects.

In this chapter, some of the following shaped claws are taken as an example, classified by wrist joints, fingers, and contact surfaces to discuss their respective application fields and the testing environment, as shown in Figure 5.

FIGURE 5

3.2 Cage claws

For example, in the aspect of cage claws, Capalbo Enrico et al. (2022) used rigid claw wrist and cage-shape fluid, flexible fingers to squeeze the internal pad to form some envelope surfaces that fit the grasping object in order to achieve the fine holding function of parallel claw and envelope function of cage claw. In addition, its inner liner (the finger surface) uses an antagonistic hydraulic cylinder to match the volume loss of the liner, and this mechanism has undergone a certain amount of underwater testing in the laboratory.

Earlier studies (Gong et al., 2019; Gong et al., 2018b; Gong et al., 2021) have investigated a 3-DoF soft mechanical arm for fragile grasping in shallow water space using a flexible pneumatic wrist with a reverse bending structure, a cage-type flexible pneumatic gripper, and a silicone gripper surface, as shown in Figure 6. The team first studied the trajectory and working space of the gripper and then conducted real-time closed-loop pickup and placement experiments of the manipulator using a binocular and hand-held camera. The experimental results showed that the claw could successfully collect eight sea thorns and one sea cucumber within 20 min.

FIGURE 6

Wu et al. (2022) introduced a cage-type gripper that is combined with a flexible hydraulic claw wrist and a flexible hydraulic claw finger. Its touching surface is a silicone sucker with adhesive ability. The design of the claw was inspired by Stauroteuthis syrtensis. The feature of the gripper is very distinct. When the gripper contacts the underwater target object, the local suction port on the array sucker is blocked, which changes the flow rate in the array suction port to detect whether an object is grasped, as shown in Figure 7. The gripper is ideal for situations where cameras cannot be used, such as murky underwater environments. In addition, the adhesive and grasping of objects (such as turtles) were tested in the underwater environment of the laboratory.

FIGURE 7

Licht et al. (2016) and Licht et al. (2017) developed a more abstract cage claw, in which the holding wrist is rigid, but the gripping finger and finger surface are composed of a latex balloon filled with a mixture of fresh water and glass beads (approximately 200 µm in diameter). The claw mainly relies on the deformation of the latex balloon to grip a specific object, which is equivalent to an infinite number of fingers for cage grasping. Experiments conducted at sea prove that the gripper can be used for deep-sea grasping missions of more than 100 atm.

3.3 Intermeshed claws

In terms of the intermeshed claw, Galloway et al. (2016) and Vogt et al. (2018) designed a gripper with a rigid claw wrist, a hydraulically flexible intermeshed claw finger, and a sponge claw surface. The fingers were typical corrugated claw fingers and were equipped with two sets of replaceable and disassembled fingers, which were convenient for twisting or intermeshing grasping. The final test was mainly conducted on site sampling of specimens, such as coral.

Sinatra et al. (2019) proposed an ultra-flexible soft robot actuator that mainly consists of a rigid claw wrist, a hydraulically driven ultra-flexible intermeshed claw finger, and a nanomaterial finger surface. The grasping test was conducted in the laboratory and aquarium. The main content of the test was to grasp delicate gelatinous marine biological specimens such as jellyfish under the sea. A series of assessments were performed to test the collection area of the claw and the robustness of the claw against external forces.

3.4 Parallel claws

In the study of parallel claws, Shen et al. (2020) proposed a design and control method for the underwater robotic arm, whose main structures include a hydraulic wrist, parallel claws combined with rigid and soft materials, and a soft finger face, as shown in Figure 8. The main feature of the manipulator is that it uses a rigid frame and a soft driver as the finger of the claw, which has formatted a soft manipulator prototype with 4-DoF and 15 N load. The gripper conducted an underwater test in the laboratory.

FIGURE 8

3.5 Challenges and limitations

At present, the underwater application of comparison grippers with different structures is mainly for sample collection, including but not limited to the sampling of jellyfish, Echini, and coral samples. For different grippers, the development direction is limited by the application field. As shown in Table 2, the comparison is as follows.

Compared with other claw shapes, the parallel claw is better for the precision grasp of small objects and more suitable for capturing steady or low movement targets. In other words, the moving object is unsuitable for a parallel claw to grasp, for example, catching a moving jellyfish. However, the parallel claw is inevitably an important component for delicate grasping. The further challenge is to improve the ability to grasp and precision in a complex environment.

As a regional capturing gripper, the cage claw is limited by its own envelope structure, and its grasp performance for small objects is not as fine as that of the parallel claw (Herrero-Pérez and Martínez-Barberá, 2022; Songlin et al., 2017). However, its regional capture ability for small marine creatures (e.g., moving fish) has shown unique advantages. The next step of this type of claw is improving the ability of the grasp area and the efficiency of grasping.

Similar to a claw, the intermeshed claw has a certain grasp area and a certain grasp accuracy that has a unique advantage, such as grasping for stick-like objects. Its intermeshed claw fingers make it easier to grasp and hold the grasped object. Furthermore, their fingers are not likely to interfere. Although the intermeshed claw has the advantages of cage and parallel claws, the design of the structure highly depends on the grasping object. When the grasping object is unknown, the ability to grasp declines sharply. The problem has become a hot spot in intermeshed claw research.

4 Soft actuator

Among the diversity of marine life, animals walk in different ways. Like most fish in the ocean, they swing by bending their bodies to provide a driving force in the water (e.g., swing locomotion). The octopus and jellyfish are driven by pressure differences (e.g., jet drive). Additionally, starfish and sea spiders like to walk underwater or on the surface. Among them, swing locomotion drives are divided into BCF locomotion (e.g., most fish and water snakes) and MPF locomotion (e.g., manta rays) (Raj and Thakur, 2016).

4.1 Locomotion

4.1.1 BCF locomotion

As the fish that people have been in contact with and understood for the longest time, research on swing locomotion has also made great progress (Wolf et al., 2020). In the research direction of BCF locomotion, some scholars (Berg et al., 2022) modeled the thunniform swimming mode and introduced an open-source soft robotic fish that achieves a maximum speed of 0.85 m per second using a propulsion system that can adjust the shape of the sine wave to achieve a higher oscillation frequency, as shown in Figure 9.

FIGURE 9

Katzschmann et al. (2018) and Katzschmann et al. (2016) used water as the transmission fluid to drive the swing of the soft fish tail through the water circulation in the internal waterway channel. In order to achieve the purpose of swimming, the main energy supply mode was using a gear pump to promote the water pressure. In addition, a camera, an underwater wireless transceiver, and pitching fins are added to the propulsion frame to control the pitching and swimming movements of the whole fish.

Marchese et al. (2014) focused on the fast escape response of soft fish. Under the premise of having all the subsystems of the traditional robot (power, drive, processing, and control), the experiment of simulating the escape of bionic fish was performed. The kinematics and controllability of the robot during the simulated escape response were analyzed and compared with those of the biological fish. In conclusion, the soft robot has an input–output relationship similar to that of the biological fish during the escape response.

Shintake et al. (2018) developed a fish robot with a length of 150 mm, a thickness of 0.75 mm, and a weight of 4.4 g using dielectric elastomer actuators (DEAs) to mimic the swing of fish swimming. The robotic fish swims by body, BCF locomotion, or both and is made using laminated silicone layers and two counter-configured DEAs. Only two DEAs should be driven to produce waves, thus achieving a fish-like motion, as shown in Figure 10. The design was guided by a mathematical model based on Euler–Bernoulli beam theory and considered the heterogeneous geometry of the robot and the hydrodynamic effect of water. By comparing the modeling results of the robot fish with the experimental results, it is concluded that the measured peak frequency of the thrust generated by the robot is similar to the natural frequency calculated by the model. At 0.75 Hz, the peak swimming speed of the robot is 37.2 mm/s (0.25 body length/s), which swims much like a real fish.

FIGURE 10

Zhang et al. (2018) used an electromagnetic drive to build ostraciiform robots. A two-segment caudal fin design was used, which is a substantial improvement over the existing single-segment caudal fin. The proposed two-segment fins help reduce drag force, improve resonant frequency, and achieve a wide range of controllable thrust. Six types of caudal fins were compared in the overall experiment, more than 1,200 tests were designed, the resonance frequency was nearly three times higher than that of the single-segment homogeneous caudal fin, and the thrust was 12% higher. It is also proved that the peak frequency can be selected by customizing the tail fin according to the application.

Zhu et al. (2019) introduced a platform (Tunabot) by mimicking Thunnus albacares and Scomber scombrus to study the high-frequency swimming performance of fish. The kinematics, speed, and force of the body were measured to quantify swimming performance and study the flow field generated by the tail while increasing the tail beating frequency. The Tunabot (255 mm long) can achieve a maximum tail-clap frequency of 15 Hz. The swimming speed is equivalent to four body lengths per second and has a range of 9.1 km if it swims at a speed of 0.4 m/s. Although the swimming speed is 1.0 m/s, the range is 4.2 km (assuming a 10 Wh battery).

4.1.2 MPF locomotion

The aforementioned study describes how swaying from BCF locomotion can provide a certain amount of power to swim, and this is the swaying structure adopted by most fish. However, in addition to BCF locomotion, MPF locomotion can drive organisms to swim. For example, Li et al. (2021) developed a soft animatronic fish with a fully integrated onboard power supply and remote control system. Driven entirely by a soft electroactive structure composed of dielectric elastomers and ionic conducting hydrogels, the fish swam at a speed of 6.4 cm/s (0.69 body length per second) and behaved stably over a wide range of temperatures. Because of its transparency, it can be used for stealth navigation.

Similarly, Li et al. (2017), inspired by the structure of deep-sea snails and fish, developed a tethered soft robot for deep-sea exploration. Its onboard power, control, and driving electronics are all integrated into the silicone matrix to avoid the influence of pressure. The dielectric elastomer material used in the robot’s flapping wings was carefully designed to enable the robot to launch successfully during field tests at 10,900 m deep in the Mariana Trench and to swim freely at 3,224 m deep in the South China Sea. The research work highlights the potential to design soft and lightweight equipment for extreme conditions.

4.2 JetDrive

Cuttlefish and jellyfish use jets to exert a reaction force on themselves to swim through the water, which means that in addition to some creatures that swim through water using oscillations, some creatures swim through water using the jet drive. At the same time, this inflow mode has inspired many scholars to study jet flow, such as Bi and Zhu (2021), who studied the influence of nozzle geometry on system performance during a single discharge at a low Reynolds number. The force decomposition algorithm was used to decouple thrust into three parts: normal stress at the exit, jet flux, and time derivative of internal fluid momentum. The paper concluded that the stroke ratio and deformation time history were fixed. In this case, the total thrust generated by the system was mainly determined by the nozzle outlet size.

Based on the open-source platform (OpenFOAM), Zhang et al. (2020) adopted the CFD method to solve the efflux fluid and conducted theoretical research, mainly studying vortex ring evolution and propulsion efficiency characteristics. In the first part, the pulse process of different strokes was simulated and compared with the standard experiment. In the second part, a reasonable open-water property calculation method was proposed to study the vortex ring thruster (VRT). The thrust coefficient and propulsive efficiency were obtained. The results showed that the high efficiency of VRT was close to 1 at a higher propulsion rate. In addition, the efficiency of four asymmetric velocity schemes and three symmetrical velocity schemes was compared. The final results showed that if the pulsation effect was prominent, a smaller stroke could achieve greater efficiency.

Cheng et al. (2019) described a tethered soft robot jellyfish with high maneuverability, which could imitate the performance of natural jellyfish. The prototype electrode is made of carbon grease sandwiched between two thin dielectric elastomer films. The frame of the material is made of silicone elastomer with six plastic paddles attached to it. The robot jellyfish recorded a maximum swimming speed of 1 cm/s and a peak thrust of 0.00012 N. A finite element simulation was developed to theoretically study the performance of the robot jellyfish. By embedding a compact remote-controlled power source, the robotic jellyfish is autonomous. In this case, the maximum peak velocity is approximately 0.5 cm/s. The working principle of the bionic robot jellyfish can be used for the field of research and guide the design of soft robots and flexible devices.

Inspired by jellyfish creatures, Christianson et al. (2019) connect a ring of rimless DEAs to the non-expandable layer to generate a single structure curved toward the passive side to generate power stroke and effectively restore the original structure when the robot glides. The flexible rimless DEAs can use fluid electrodes to apply a voltage to the film and can achieve efficient movement of the eel robot without the need for a rigid frame. The average speed of the soft robot swimming is 3.2 mm/s. This work demonstrated the feasibility of using DEAs with fluid electrodes for low-power, silent operation in underwater environments.

The jellyfish machines of Frame et al. (2018) are pneumatically driven, with eight pneumatic network tentacle actuators extending outward from their centers. These jellyfish robots can swim and turn in the ocean. Moreover, jellyfish can pass through pores narrower than the nominal diameter of jellyfish. Each of the five jellyfish robots has a different body composition, tentacle actuator, and shore hardness to facilitate the study of these three factors and determine the influence of the tentacle stroke on the measured thrust. All three factors significantly affect average thrust generation, and the half-stroke drive amplitude peaks at a frequency of 0.8 Hz.

Godaba et al. (2016) mainly studied a jellyfish robot with DEA, which has muscle-like characteristics, such as large deformation and high energy density. The deformation and force of the actuator were experimentally tested. The performance of the actuator was analyzed by theoretical simulation, and the results were qualitatively consistent with the experimental results. Preliminary studies have shown that the jellyfish robot, based on dielectric elastomer technology, can move effectively in water. The robot also exhibits fast response and high load capacity (relative to its self-weight)

Wang Tao et al. (2019) studied the ability of pulse-jet underwater soft robots to turn using a steerable nozzle through experiments. The drive of the robot is based on the periodic conversion of slow-charged elastic potential energy into fluid kinetic energy to perform a cyclic pulsed jet. A steerable nozzle is added at the end of the jet pipe of the robot to deflect the jet in order to conduct thrust vector control. The results show that the drive scheme has a higher control ability from static to starting. The turning radius is approximately half the bionic body length, and the optimal nozzle deflection angle is 35°. The most important factor affecting the turning efficiency is the fluid momentum loss caused by nozzle deflection.

Wang Yuzhe et al. (2022) developed a completely transparent soft-bodied jellyfish robot, which is transparent and can move in all directions. The robot is driven by transparent dielectric elastomer drivers (DEAs), a hybrid silver nanowire network, and conductive polymer poly(3,4-ethylenedioxythiophene):poly(styrenesulfonate)/waterborne polyurethane as a compatible electrode. The electrode exhibits a great tensile property, low stiffness, high transmittance, and excellent electrical conductivity under large strain. Therefore, with the very high-bond film as the medium layer and polydimethylsiloxane as the surface coating, the maximum area strain of the highly transparent DEAs based on the hybrid electrode can reach 146%, and the hysteretic loss is only 3%. Driven by these transparent DEAs, the soft jellyfish robot can achieve vertical and horizontal movements in water by imitating the actual pulsating rhythm of Aurelia aurita.

Zhang et al. (2021) used a biomimetic soft-robotic siphon (BSRS) as the propulsion unit, which was composed of a new central flow regulative duct (CFRD) surrounded by three siphon actuation muscles (SAMs). Hydraulic pressurization of SAMs enables thrust vectoring by deflecting the BSRS and flow regulation by proportionally alternating CFRD ports. Flexural deformation and flow regulation experiments were performed using the BSRS prototype. The results show that the bending range of the BSRS is more than 180°, and the current limiting capacity is up to 100%. The robot achieves a burst effect by exceeding a constant flow rate of up to 50%, achieving a huge thrust increase in a very short time. This work demonstrates the feasibility of combining omnidirectional deflection with flow regulation in a soft robot mechanism, paving the way for compact water jet propulsion of underwater vehicles.

4.3 Crawling/floating

In terms of unptderwater crawling, Ishida et al. (2019) developed an underwater crawling robot whose top and legs could be inflated and deflated. The change in body shape affects its hydrodynamic properties. When the robot (2.87 N underwater weight) needs to stay still in the flow, an asymmetrical body is more resistant to body sliding. Body shape had a significant effect on the robot’s ability to walk upstream, with uninflated bodies able to walk in a flow of 0.09 m/s, but larger inflated bodies were pushed backward. The paper demonstrated that such robots can detect changes in velocity using commercial flow sensors and respond by deforming into a hydrodynamically preferable shape.

Inspired by starfish, Patterson et al. (2020) proposed a mobile cordless underwater crawling soft robot (PATRICK). PATRICK consists of five flexible legs driven by 20 SMA wires, providing a wide variety of motion possibilities through its large input space. The experiment shows that the robot can be ordered to move to the target state using the planning mode. These experiments provide examples of closed-loop state space target searching of tethered underwater flexible crawling robots, and some progress has been made in fully autonomous soft mobile robot systems.

Inspired by sea spiders, Tan et al. (2021) proposed a soft robot design method for underwater crawling robots, as well as a rigid-flexible hybrid multi-joint leg design. The joint has a quasi-linear range of motion, and at the same time, the inherent flexibility of the soft actuator is utilized to maintain good passive impact compliance. Figure 11 shows the overall and joint structures of the flexible robot.

FIGURE 11

Huang et al. (2021) introduced a model based on discrete differential geometry to solve the nonlinear deformation problem when the soft material structure interacts with the liquid environment in terms of water walking. An untethered omnidirectional star swimming soft robot is designed by imitating the structure of a starfish. The robot has been tested to be able to move with multiple swimming gaits. The quantitative agreement between experiment and simulation demonstrates the potential application of this numerical model in robot design and model-based control schemes.

Moreover, on this basis, Huang et al. (2022) studied a completely cordless soft frog swimming robot framework based on physical modeling, motion planning, and control. They used the discrete elastic rods (DERs) physics engine to disperse the soft robot into many stretchable and flexible rods. In terms of hardware, an untethered water soft robot is designed to perform a frog-like rowing behavior, as shown in Figure 12.

FIGURE 12

4.4 Challenges and limitations

Compared with the existing turbine drive, the low swing frequency of the flexible drive is very suitable for the observation of marine biological population activities, and the high energy conversion efficiency (Gupta et al., 2019) is very suitable for underwater endurance operations. Although the benefits are enormous, part of the robot is made of flexible materials with a complex underwater environment, which has raised concerns about its stability and durability. These problems also need certain research and development to handle. The related progress is shown in Table 3.

In addition, locomotion showed a strong energy conversion advantage and a high movement speed (within a certain swing frequency). However, there was a certain gap compared to the current research on the movement speed, and the endurance and stability should be discussed in a certain number of experiments.

As a swimming mode with intermittent output, Jet Drive can well imitate the swimming of creatures such as octopuses (Krieg et al., 2015). However, its complex turning structure and energy dissipation are unavoidable. At present, the research is mainly focused on the optimization of the overall mechanism and underwater test. Further work could be the movement affected by the external environment, especially under high pressure. It still needs much time for testing in the deep salty water area.

The crawling and floating movements are easier to realize compared to other neutral buoyancy movements. However, this type of movement sacrifices most of the area underwater so that it can only move in a certain place. Despite this limitation, the robot’s carrying capacity and structure stability have made a space in actuation (Ren and Yu, 2021). Further progress could be achieved through structure optimization and control strategy evaluation (Cianchetti et al., 2015).

5 Soft sensor

As fish use the lateral lines of their bodies to sense water velocity and vibrations, octopuses use suckers and tentacles to sense the presence of objects. Perception, as an important source of information, prompts people to further develop flexible underwater sensors, such as velocity sensors inspired by the lateral line of fish (Jiang et al., 2016; Liu et al., 2016) and skin sensors inspired by animal skin. Sensors combined with the characteristics of the underwater environment have achieved certain development in recent years (Bora et al., 2018; Han et al., 2018).

5.1 Velocity sensors

In the field of hair sensing, Abels et al. (2016) constructed a microelectromechanical system (MEMS) structure to simulate the biological superficial neuromasts found in the lateral organs of fish. Based on the piezoresistive strain principle of the cantilever beam driven along the stress, the local flow velocity was measured to understand the sensitivity to temperature changes and response to changes in the relative flow direction. Furthermore, two cantilever flow sensors with opposite and parallel bending directions (Abels et al., 2019) were designed and compared with a single cantilever hair sensor.

Furthermore, Asadnia et al. (2015) developed an artificial hair cell flow sensor array. The flow sensor adopted polymer hair cells manufactured by stereolithography and mounted on a microdiaphragm with a floating bottom electrode. By installing a polymer tube with pores, external fluid can be directed toward hair cells embedded in the tube. The experimental results show that the vibration of 35 Hz oscillating dipole stimulation has high sensitivity and low detection limits. The flexible array of such sensors is capable of locating using underwater dual stimuli.

Inspired by the artificial lateral system with different conduction mechanisms (Jiang et al., 2019a) in the field of artificial flow sensors, Jiang et al. (2017) developed a cantilever flow sensing element based on polymer materials, formed by lamination of polypropylene and polyvinylidene fluoride (PVDF). The cantilever flow sensing element was integrated into the polydimethylsiloxane (PDMS) pipe, as shown in Figure 13. The sensor has high-pass filtering capability to attenuate low-frequency stimuli, and the pressure gradient detection limit is approximately 11 Pa/m at a frequency of 115 ± 1 Hz. Its structure is flexible and can resist certain interference.

FIGURE 13

5.2 Skin sensors

In terms of underwater flexible contact sensors, He et al. (2022) proposed an adaptive soft sensor whose manufacturing method can be directly used in 3D printing. The whole component is controlled by pneumatic pressure, and the stiffness can be changed within a certain range. The influence of stiffness adjustment on sensing characteristics is studied. The results show that this class of adaptive sensors exhibits good sensitivity, high sensor repeatability, low latency, and good manufacturing repeatability.

Furthermore, Ibarra et al. (2022) used 3D printing technology to print out sensitive, thin, and conductive channels in the soft matrix to construct strain sensors. The effects would also change according to different wiring modes of conduction channels.

Ren et al. (2022) introduced an anti-expansion hydrogel. This hydrogel has high toughness, compressive modulus, ionic conductivity, and, most importantly, anti-expansion properties, with only 9% equilibrium expansion when in water for 30 days. In addition, this research demonstrates some basic application scenarios of this hydrogel as a sensor to better shorten the development cycle of the sensor based on the hydrogel.

5.3 Challenges and limitations

As a very specific structure in soft robots (Ma et al., 2019), sensors can perform much work in the underwater environment, which significantly differs on land (Ma et al., 2020).

Due to the differences in sensing media, compared with sensors in dry air, the development of underwater sensors is more inclined to detect low-frequency signals (Zhai et al., 2021), such as the detection of dipoles and the flow rate of the fluid. In addition, sensors on underwater contact forces and other aspects have been further developed.

Compared with the normal sensor, the underwater sensor is characterized by the differences under different working environments, as shown in Table 4. Therefore, certain requirements have been commanded for waterproof performance, detection frequency, and anti-interference ability (Jiang et al., 2019b; Jiang et al., 2020; Jiang et al., 2019a). At present, underwater sensors are mainly based on vibration sensing mimicked by the lateral line structure of fish and pressure sensing modeled on human skin. At this stage, the development has achieved good results, but further experiments are needed to prove its stability, sustainability, and other performance.

6 Conclusion

Soft robot structures, combined with existing new materials and new control methods (Youssef et al., 2022), compared with traditional rigid robots in an underwater environment, as well as their lightness, dexterity, and various realization methods, have garnered considerable interest among scholars (Fang et al., 2022).

In the aspect of adhesion, soft tissue can be active or passive, and the adhesion is firm. Its excellent adhesion structure and underwater adaptability can be well bonded to the adsorbed surface (Chen et al., 2020). However, some problems remain to be solved in the aspect of the sucker. First, the adhesion performance of an object with an extremely rough surface is not good. Second, the surface with stains such as oil will have a negative impact on the adhesion capacity. Finally, the sucker will have a certain fatigue after repeated deformation (Kang et al., 2021).

In terms of the gripper, the soft grasping mechanism has a good grasping performance for vulnerable samples, and its diversified degrees of freedom are also very attractive (Rus and Tolley, 2015). However, the materials for soft grippers need to be further developed, and the structure and processing technology should be simplified. The control strategy also needs to be further studied and explored in combination with the material and motion characteristics of the specific mechanism (Wang and Cui, 2021).

In terms of the actuator, the underwater driving prototype has the advantages of low noise, making it more suitable for the habitat environment of underwater organisms and incorporating multi-functional composite mechanisms (e.g., transparency and low power advantages), which has attracted several scholars to study (Fu et al., 2021). However, further research is needed on the construction materials and testing of some prototypes in deep water (below 2,000 m), and the driving strategy can be further optimized and developed (Sun et al., 2022).

In terms of the soft sensor, as the most widely used for new materials, the fish-like lateral line and skin-like sensing devices can be used in several underwater applications. However, the properties of the new materials should be studied, including mechanical and chemical properties. In addition, the signal should be collected and quantified. Finally, a self-feedback correction is performed to be formally put into application (Won et al., 2021).

In this paper, several aspects of the development of the underwater soft robot are briefly introduced from the perspective of application. By starting from the common marine organisms, several main motion modes of the underwater soft robot are analyzed (viz., adhesion, grasping, driving, and sensing). The most common adhesion is the morphological biomimetics of clinging animals, such as octopus, remora, and clingfish. In addition, the gripper mainly studies and analyzes the behavior of predation and capture of organisms. In the aspect of actuators, it is briefly introduced that the tail drive of the swinging fish, the jet drive of the jellyfish, and the crawling drive of the starfish spider. Finally, some applications of biomimetic underwater sensors in the field of sensing are briefly introduced. This paper aims to help readers quickly understand the recent progress in different aspects of underwater soft mechanisms in order to guide engineers and technicians to a comprehensive understanding of specific problems in specific fields.

References

• 1

  AbelsC.QualtieriA.De VittorioM.MegillW. M.RizziF. (2016). A bio-inspired real-time capable artificial lateral line system for freestream flow measurements. Bioinspiration Biomimetics11 (3), 035006. 10.1088/1748-3190/11/3/035006

  • CrossRef
  • Google Scholar

• 2

  AbelsC.QualtieriA.LoberT.MariottiA.ChambersL. D.De VittorioM.et al (2019). Bidirectional biomimetic flow sensing with antiparallel and curved artificial hair sensors. Beilstein J. Nanotechnol.10, 32–46. 10.3762/bjnano.10.4

  • CrossRef
  • Google Scholar

• 3

  AsadniaM.Ajay GiriP. K.MiaoJ.WarkianiM. E.MichaelTriantafyllouS. (2015). Artificial fish skin of self-powered micro-electromechanical systems hair cells for sensing hydrodynamic flow phenomena. J. R. Soc. Interface12 (111), 20150322. 10.1098/rsif.2015.0322

  • CrossRef
  • Google Scholar

• 4

  BeckertM.FlammangB. E.NadlerJ. H. (2015). Remora fish suction pad attachment is enhanced by spinule friction. J. Exp. Biol. January, jeb218, 3551–3558. 10.1242/jeb.123893

  • CrossRef
  • Google Scholar

• 5

  BergS. C. v. d.ScharffR. B. N.RusákZ.Wu.J. (2022). OpenFish: Biomimetic design of a soft robotic fish for high speed locomotion. HardwareX12, e00320. 10.1016/j.ohx.2022.e00320

  • CrossRef
  • Google Scholar

• 6

  BiX.ZhuQ. (2021). Effect of nozzle geometry on the performance of pulsed-jet propulsion at low Reynolds number. J. Fluids Struct.107, 103402. 10.1016/j.jfluidstructs.2021.103402

  • CrossRef
  • Google Scholar

• 7

  BillardA.KragicD. (2019). Trends and challenges in robot manipulation. Science364 (6446), eaat8414. 10.1126/science.aat8414

  • CrossRef
  • Google Scholar

• 8

  BoraM.Ajay GiriP. K.MiaoJ.TriantafyllouM. S. (2018). Sensing the flow beneath the fins. Bioinspiration Biomimetics13 (2), 025002. 10.1088/1748-3190/aaa1c2

  • CrossRef
  • Google Scholar

• 9

  Capalbo EnricoC.TomainoD.BrunoF.RizzoD.PhillipsB.et al (2022). A soft robotic gripper with neutrally buoyant jamming pads for gentle yet secure grasping of underwater objects. IEEE J. Ocean. Eng.47 (4), 975–983. 10.1109/JOE.2022.3156746

  • CrossRef
  • Google Scholar

• 10

  ChenY.MengJ.GuZ.WanX.JiangL.WangS. (2020). Bioinspired multiscale wet adhesive surfaces: Structures and controlled adhesion. Adv. Funct. Mater.30 (5), 1905287. 10.1002/adfm.201905287

  • CrossRef
  • Google Scholar

• 11

  ChengT.LiG.LiangY.ZhangM.LiuB.WongT. W.et al (2019). Untethered soft robotic jellyfish. Smart Mater. Struct.28 (1), 015019. 10.1088/1361-665X/aaed4f

  • CrossRef
  • Google Scholar

• 12

  ChristiansonC.BayagC.LiG.JadhavS.GiriA.AgbaC.et al (2019). Jellyfish-inspired soft robot driven by fluid electrode dielectric organic robotic actuators. Front. Robotics AI6, 126. 10.3389/frobt.2019.00126

  • CrossRef
  • Google Scholar

• 13

  CianchettiM.CalistiM.MargheriL.KubaM.LaschiC. (2015). Bioinspired locomotion and grasping in water: The soft eight-arm OCTOPUS robot. Bioinspiration Biomimetics10 (3), 035003. 10.1088/1748-3190/10/3/035003

  • CrossRef
  • Google Scholar

• 14

  DitscheP.AdamS. (2019). Learning from northern clingfish (gobiesox maeandricus): Bioinspired suction cups attach to rough surfaces. Philosophical Trans. R. Soc. B Biol. Sci.374 (1784), 20190204. 10.1098/rstb.2019.0204

  • CrossRef
  • Google Scholar

• 15

  DitscheP.WainwrightD. K.SummersA. P. (2014). Attachment to challenging substrates – fouling, roughness and limits of adhesion in the northern clingfish (gobiesox maeandricus). J. Exp. Biol.217 (14), 2548–2554. 10.1242/jeb.100149

  • CrossRef
  • Google Scholar

• 16

  EroğluM.Ebru DevrimŞ. P. (2023). A simple manufacturing process of the miniaturised octopus-inspired underwater soft robotic grippers. J. Adhesion Sci. Technol.37 (7), 1163–1176. 10.1080/01694243.2022.2043056

  • CrossRef
  • Google Scholar

• 17

  FangJ.ZhuangY.LiuK.ChenZ.ZhouL.KongT.et al (2022). A shift from efficiency to adaptability: Recent progress in biomimetic interactive soft robotics in wet environments. Adv. Sci.9 (8), 2104347. 10.1002/advs.202104347

  • CrossRef
  • Google Scholar

• 18

  FlammangB. E.KenaleyC. P. (2017). Remora cranial vein morphology and its functional implications for attachment. Sci. Rep.7 (1), 5914. 10.1038/s41598-017-06429-z

  • CrossRef
  • Google Scholar

• 19

  FrameJ.LopezN.CuretO.ErikEngebergD. (2018). Thrust force characterization of free-swimming soft robotic jellyfish. Bioinspiration Biomimetics13 (6), 064001. 10.1088/1748-3190/aadcb3

  • CrossRef
  • Google Scholar

• 20

  FreyS. T.Tahidul HaqueA. B. M.RaviT.KrotzE. V.LeeC.Cole Haverkamp MarkvickaB. E. J.et al (2022). Octopus-inspired adhesive skins for intelligent and rapidly switchable underwater adhesion. Sci. Adv.8 (28), eabq1905. 10.1126/sciadv.abq1905

  • CrossRef
  • Google Scholar

• 21

  FuS.WeiF.YinC.YaoL.WangY. (2021). Biomimetic soft micro-swimmers: From actuation mechanisms to applications. Biomed. Microdevices23 (1), 6. 10.1007/s10544-021-00546-3

  • CrossRef
  • Google Scholar

• 22

  GallowayK. C.BeckerK. P.PhillipsB.KirbyJ.LichtS.TchernovD.et al (2016). Soft robotic grippers for biological sampling on deep reefs. Soft Robot.3 (1), 23–33. 10.1089/soro.2015.0019

  • CrossRef
  • Google Scholar

• 23

  GamelK. M.GarnerA. M.FlammangB. E. (2019). Bioinspired remora adhesive disc offers insight into evolution. Bioinspiration Biomimetics14 (5), 056014. 10.1088/1748-3190/ab3895

  • CrossRef
  • Google Scholar

• 24

  GiordanoG.CarlottiM.MazzolaiB. (2021). A perspective on cephalopods mimicry and bioinspired technologies toward proprioceptive autonomous soft robots. Adv. Mater. Technol.6 (12), 2100437. 10.1002/admt.202100437

  • CrossRef
  • Google Scholar

• 25

  GodabaH.LiJ.WangY.ZhuJ. (2016). A soft jellyfish robot driven by a dielectric elastomer actuator. IEEE Robotics Automation Lett.1 (2), 624–631. 10.1109/LRA.2016.2522498

  • CrossRef
  • Google Scholar

• 26

  GongZ.ChenB.LiuJ.FangX.LiuZ.WangT.et al (2019). An opposite-bending-and-extension soft robotic manipulator for delicate grasping in shallow water. Front. Robotics AI6, 26. 10.3389/frobt.2019.00026

  • CrossRef
  • Google Scholar

• 27

  GongZ.ChengJ.ChenX.SunW.FangX.HuK.et al (2018a). A bio-inspired soft robotic arm: Kinematic modeling and hydrodynamic experiments. J. Bionic Eng.15 (2), 204–219. 10.1007/s42235-018-0016-x

  • CrossRef
  • Google Scholar

• 28

  GongZ.ChengJ.HuK.WangT.WenL. (2018b). “An inverse kinematics method of a soft robotic arm with three-dimensional locomotion for underwater manipulation,” in Procceedings of the 2018 IEEE International Conference on Soft Robotics (RoboSoft), Livorno, April 2018 (IEEE), 516–521. 10.1109/ROBOSOFT.2018.8405378

  • CrossRef
  • Google Scholar

• 29

  GongZ.FangX.ChenX.ChengJ.XieZ.LiuJ.et al (2021). A soft manipulator for efficient delicate grasping in shallow water: Modeling, control, and real-world experiments. Int. J. Robotics Res.40 (1), 449–469. 10.1177/0278364920917203

  • CrossRef
  • Google Scholar

• 30

  GreenD. M.Louise BarberD. (1988). The ventral adhesive disc of the clingfish gobiesox maeandricus: Integumental structure and adhesive mechanisms. Can. J. Zoology66 (7), 1610–1619. 10.1139/z88-235

  • CrossRef
  • Google Scholar

• 31

  GulJ. Z.KimSuY.ChoiK. H. (2018). Fully 3D printed multi-material soft bio-inspired whisker sensor for underwater-induced vortex detection. Soft Robot.5 (2), 122–132. 10.1089/soro.2016.0069

  • CrossRef
  • Google Scholar

• 32

  GuptaU.QinL.WangY.GodabaH.ZhuJ. (2019). Soft robots based on dielectric elastomer actuators: A review. A review’. Smart Mater. Struct.28, 103002. 10.1088/1361-665X/ab3a77

  • CrossRef
  • Google Scholar

• 33

  HanZ.LiuL.WangK.SongH.ChenD.WangZ.et al (2018). Artificial hair-like sensors inspired from nature: A review. J. Bionic Eng.15 (3), 409–434. 10.1007/s42235-018-0033-9

  • CrossRef
  • Google Scholar

• 34

  HeL.NicolasH.NanayakkaraT.MaiolinoP. (2022). 3D-Printed soft sensors for adaptive sensing with online and offline tunable stiffness. Soft Robot.9 (6), 0074. 10.1089/soro.2021.0074

  • CrossRef
  • Google Scholar

• 35

  Herrero-PérezD.Martínez-BarberáH. (2022). Soft gripper design and fabrication for underwater grasping. Appl. Sci.12 (21), 10694. 10.3390/app122110694

  • CrossRef
  • Google Scholar

• 36

  HuangX.HuangW.PattersonZ.Zhijian RenM.KhalidJ.CarmelM. (2021). Numerical simulation of an untethered omni-directional star-shaped swimming robot,” in Proceedings of the 2021 IEEE International Conference on Robotics and Automation (ICRA), Xi’an, China (IEEE), 11884–11890. 10.1109/ICRA48506.2021.9561301

  • CrossRef
  • Google Scholar

• 37

  HuangX.PattersonZ. J.SabelhausA. P.HuangW.ChinK.RenZ.et al (2022). Design and closed‐loop motion planning of an untethered swimming soft robot using 2D discrete elastic rods simulations. Adv. Intell. Syst.4 (10): 2200163. 10.1002/aisy.202200163

  • CrossRef
  • Google Scholar

• 38

  IbarraA.Darbois-TexierB.MeloF. (2022). Designing a contact fingertip sensor made using a soft 3D printing technique. Soft Robot.9 (6), 1210–1219. 10.1089/soro.2021.0128

  • CrossRef
  • Google Scholar

• 39

  IshidaM.DrotmanD.ShihB.HermesM.LuharM.MichaelTolleyT. (2019). Morphing structure for changing hydrodynamic characteristics of a soft underwater walking robot. IEEE Robotics Automation Lett.4 (4), 4163–4169. 10.1109/LRA.2019.2931263

  • CrossRef
  • Google Scholar

• 40

  IshidaM.SandovalJ. A.LeeS.HuenS.TolleyM. T. (2022). Locomotion via active suction in a sea star-inspired soft robot. IEEE ROBOTICS AUTOMATION Lett.7 (4), 10304–10311. 10.1109/lra.2022.3191181

  • CrossRef
  • Google Scholar

• 41

  JiangH.HanX.JingY.WanF.SongC. (2021). Rigid–soft interactive design of a lobster-inspired finger surface for enhanced grasping underwater. Front. Robotics AI8787187. 10.3389/frobt.2021.787187

  • CrossRef
  • Google Scholar

• 42

  JiangY.FuJ.ZhangD.ZhaoY. (2016). Investigation on the lateral line systems of two cavefish: Sinocyclocheilus macrophthalmus and S. Microphthalmus (cypriniformes: Cyprinidae). J. Bionic Eng.13 (1), 108–114. 10.1016/S1672-6529(14)60164-5

  • CrossRef
  • Google Scholar

• 43

  JiangY.MaZ.CaoB.GongL.FengL.ZhangD. (2019b). Development of a tactile and slip sensor with a biomimetic structure-enhanced sensing mechanism. J. Bionic Eng.16 (1), 47–55. 10.1007/s42235-019-0005-8

  • CrossRef
  • Google Scholar

• 44

  JiangY.MaZ.FuJ.ZhangD. (2017). Development of a flexible artificial lateral line canal system for hydrodynamic pressure detection. Sensors17 (6), 1220. 10.3390/s17061220

  • CrossRef
  • Google Scholar

• 45

  JiangY.MaZ.ZhangD. (2019a). Flow field perception based on the fish lateral line system. Bioinspiration Biomimetics14 (4), 041001. 10.1088/1748-3190/ab1a8d

  • CrossRef
  • Google Scholar

• 46

  JiangY.ZhaoP.MaZ.ShenD.LiuG.ZhangD. (2020). Enhanced flow sensing with interfacial microstructures. Biosurface Biotribology6 (1), 12–19. 10.1049/bsbt.2019.0043

  • CrossRef
  • Google Scholar

• 47

  KangM.SunK.SeongM.HwangI.JangH.ParkS.et al (2021). Applications of bioinspired reversible dry and wet adhesives: A review. Front. Mech. Eng.7668262. 10.3389/fmech.2021.668262

  • CrossRef
  • Google Scholar

• 48

  KatzschmannR. K.JosephD.MacCurdyR.RusD. (2018). Exploration of underwater life with an acoustically controlled soft robotic fish. Sci. Robotics3 (16), eaar3449. 10.1126/scirobotics.aar3449

  • CrossRef
  • Google Scholar

• 49

  KatzschmannR. K.MarcheseA. D.Rus.D. (2016). “Hydraulic autonomous soft robotic fish for 3D swimming,” in Experimental robotics. Springer tracts in advanced robotics. Editors Ani HsiehM.KhatibOussamaKumarVijay (Cham: Springer International Publishing). 10.1007/978-3-319-23778-7_27

  • CrossRef
  • Google Scholar

• 50

  KriegM.SledgeI.MohseniK. (2015). Design considerations for an underwater soft-robot inspired from marine invertebrates. Bioinspiration Biomimetics10 (6), 065004. 10.1088/1748-3190/10/6/065004

  • CrossRef
  • Google Scholar

• 51

  KurumayaS.PhillipsB. T.BeckerK. P.MichelleH.RosenD. F. G.KevinC. G.et al (2018). A modular soft robotic wrist for underwater manipulation. Soft Robot.5 (4), 399–409. 10.1089/soro.2017.0097

  • CrossRef
  • Google Scholar

• 52

  LeeS. H.SongH. W.KangB. S.KwakM. K. (2019). Remora-inspired reversible adhesive for underwater applications. ACS Appl. Mater. Interfaces11 (50), 47571–47576. 10.1021/acsami.9b16350

  • CrossRef
  • Google Scholar

• 53

  LiG.ChenX.ZhouF.LiangY.XiaoY.CaoX.et al (2021). Self-powered soft robot in the Mariana Trench. Nature591 (7848), 66–71. 10.1038/s41586-020-03153-z

  • CrossRef
  • Google Scholar

• 54

  LiH.LiX.WangB.ShangX.YaoJ. (2023). A fault-tolerant soft swallowing robot capable of grasping delicate structures underwater. IEEE ROBOTICS AUTOMATION Lett.8 (6), 3302–3309. 10.1109/LRA.2023.3266668

  • CrossRef
  • Google Scholar

• 55

  LiL.WangS.ZhangY.SongS.WangC.TanS.et al (2022). Aerial-aquatic robots capable of crossing the air-water boundary and hitchhiking on surfaces. Sci. Robotics7 (66), eabm6695. 10.1126/scirobotics.abm6695

  • CrossRef
  • Google Scholar

• 56

  LiT.LiG.LiangY.ChengT.DaiJ.YangX.et al (2017). Fast-moving soft electronic fish. Sci. Adv.3 (4), e1602045. 10.1126/sciadv.1602045

  • CrossRef
  • Google Scholar

• 57

  LichtS.CollinsE.Ballat-DurandD.Lopes-MendesM. (2016). Universal jamming grippers for deep-sea manipulation, in Procceedings of the IEEEOCEANS 2016 MTS/IEEE Monterey, Monterey, CA, September 2016 (IEEE), 1–5. 10.1109/OCEANS.2016.7761237

  • CrossRef
  • Google Scholar

• 58

  LichtS.CollinsE.MendesM. L.BaxterC. (2017). Stronger at depth: Jamming grippers as deep sea sampling tools. Soft Robot.4 (4), 305–316. 10.1089/soro.2017.0028

  • CrossRef
  • Google Scholar

• 59

  LiuG.WangA.WangX.LiuP. (2016). A review of artificial lateral line in sensor fabrication and bionic applications for robot fish. Appl. Bionics Biomechanics20161–15. 10.1155/2016/4732703

  • CrossRef
  • Google Scholar

• 60

  MaZ.JiangY.WuP.XuY.HuX.GongZ.et al (2019). Constriction canal assisted artificial lateral line system for enhanced hydrodynamic pressure sensing. Bioinspiration Biomimetics14 (6), 066004. 10.1088/1748-3190/ab3d5a

  • CrossRef
  • Google Scholar

• 61

  MaZ.JiangY.DongZ.ZhangD.ZhiwuH.DeyuanZ. (2020). Hydrodynamic perception using an artificial lateral line device with an optimized constriction canal. J. Bionic Eng.17 (5): 909–919. 10.1007/s42235-020-0084-6

  • CrossRef
  • Google Scholar

• 62

  MarcheseA. D.OnalC. D.Rus.D. (2014). Autonomous soft robotic fish capable of escape maneuvers using fluidic elastomer actuators. Soft Robot.1 (1), 75–87. 10.1089/soro.2013.0009

  • CrossRef
  • Google Scholar

• 63

  MazzeoA.AguzziJ.CalistiM.CaneseS.VecchiF.StefanniS.et al (2022). Marine robotics for deep-sea specimen collection: A systematic review of underwater grippers. Sensors22 (2), 648. 10.3390/s22020648

  • CrossRef
  • Google Scholar

• 64

  NegrelloF.StuartH. S.CatalanoM. G. (2020). Hands in the real world. Front. Robotics AI6, 147. 10.3389/frobt.2019.00147

  • CrossRef
  • Google Scholar

• 65

  PattersonZ. J.SabelhausA. P.KeeneC.TessH.CarmelM. (2020). An untethered brittle star-inspired soft robot for closed-loop underwater locomotion. Avaialble at: http://arxiv.org/abs/2003.13529.

  • Google Scholar

• 66

  QiaoS.WangL.HaK. H.LuN. (2018). Suction effects of craters under water. Soft Matter14 (42), 8509–8520. 10.1039/C8SM01601A

  • CrossRef
  • Google Scholar

• 67

  RajA.ThakurA. (2016). Fish-inspired robots: Design, sensing, actuation, and autonomy—A review of research. Bioinspiration Biomimetics11 (3), 031001. 10.1088/1748-3190/11/3/031001

  • CrossRef
  • Google Scholar

• 68

  RenJ.LiuY.WangZ.ChenS.MaY.WeiH.et al (2022). An anti-swellable hydrogel strain sensor for underwater motion detection. Adv. Funct. Mater.32 (13), 2107404. 10.1002/adfm.202107404

  • CrossRef
  • Google Scholar

• 69

  RenK.YuJ. (2021). Research status of bionic amphibious robots: A review. Ocean. Eng.227, 108862. 10.1016/j.oceaneng.2021.108862

  • CrossRef
  • Google Scholar

• 70

  RusD.TolleyM. T. (2015). Design, fabrication and control of soft robots. Nature521 (7553), 467–475. 10.1038/nature14543

  • CrossRef
  • Google Scholar

• 71

  SandovalJ. A.JadhavS.QuanH.DimitriDeheynD.TolleyM. T. (2019). Reversible adhesion to rough surfaces both in and out of water, inspired by the clingfish suction disc. Bioinspiration Biomimetics14 (6), 066016. 10.1088/1748-3190/ab47d1

  • CrossRef
  • Google Scholar

• 72

  SandovalJ. A.IshidaM.JadhavS.HuenS.TolleyM. T. (2022). Tuning the morphology of suction discs to enable directional adhesion for locomotion in wet environments. Soft Robot.9 (6), 1083–1097. 10.1089/soro.2021.0096

  • CrossRef
  • Google Scholar

• 73

  ShenZ.ZhaoY.ZhongH.TangK.ChenY.YinX.et al (2021). Soft origami optical-sensing actuator for underwater manipulation. Front. Robotics AI7, 616128. 10.3389/frobt.2020.616128

  • CrossRef
  • Google Scholar

• 74

  ShenZ.ZhongH.XuE.ZhangR.YipK. C.ChanL. L.et al (2020). An underwater robotic manipulator with soft bladders and compact depth-independent actuation. Soft Robot.7 (5), 535–549. 10.1089/soro.2019.0087

  • CrossRef
  • Google Scholar

• 75

  ShintakeJ.CacuccioloV.SheaH.FloreanoD. (2018). Soft biomimetic fish robot made of dielectric elastomer actuators. Soft Robot.5 (4), 466–474. 10.1089/soro.2017.0062

  • CrossRef
  • Google Scholar

• 76

  SinatraN. R.ClarkB.TeepleD. M. V.Kevin KitP.DavidF. G.RobertJ. W. (2019). Ultragentle manipulation of delicate structures using a soft robotic gripper. Sci. Robotics4 (33): eaax5425. 10.1126/scirobotics.aax5425

  • CrossRef
  • Google Scholar

• 77

  SoltanniaB.SameotoD. (2014). Strong, reversible underwater adhesion via gecko-inspired hydrophobic fibers. ACS Appl. Mater. Interfaces6 (24), 21995–22003. 10.1021/am5075375

  • CrossRef
  • Google Scholar

• 78

  SonglinN.LiuX.HuiJ.MaZ.YinF. (2017). Simulation and experiment study on deformation characteristics of the water hydraulic flexible actuator used for the underwater gripper. IEEE Access8, 191447–191459. 10.1109/ACCESS.2020.3032179

  • CrossRef
  • Google Scholar

• 79

  SubadR. A. S. I.MahmudH. S.ParkK. (2022). Soft multi-directional force sensor for underwater robotic application. Sensors22 (10), 3850. 10.3390/s22103850

  • CrossRef
  • Google Scholar

• 80

  SunB.LiW.WangZ.ZhuY.QuH.GuanX.et al (2022). Recent progress in modeling and control of bio-inspired fish robots. J. Mar. Sci. Eng.10 (6), 773. 10.3390/jmse10060773

  • CrossRef
  • Google Scholar

• 81

  SunE.WangT.ZhuS. (2020). An experimental study of bellows-type fluidic soft bending actuators under external water pressure. Smart Mater. Struct.29 (8), 087005. 10.1088/1361-665X/ab9518

  • CrossRef
  • Google Scholar

• 82

  SwiftM. D.ColeB. H.ChristopherJ. S.DohgyuH.RaymondH. P.KevinT. T.et al (2020). Active membranes on rigidity tunable foundations for programmable, rapidly switchable adhesion. Adv. Mater. Technol.5 (11), 2000676. 10.1002/admt.202000676

  • CrossRef
  • Google Scholar

• 83

  TanQ.ChenY.LiuJ.ZouK.YiJ.LiuS.et al (2021). Underwater crawling robot with hydraulic soft actuators. Front. Robotics AI8, 688697. 10.3389/frobt.2021.688697

  • CrossRef
  • Google Scholar

• 84

  TramacereF.AlexanderK.KleinteichT.GorbS. N.MazzolaiB. (2014b). Structure and mechanical properties of Octopus vulgaris suckers. J. R. Soc. Interface11 (91), 20130816. 10.1098/rsif.2013.0816

  • CrossRef
  • Google Scholar

• 85

  TramacereF.AppelE.MazzolaiB.GorbS. N. (2014a). Hairy suckers: The surface microstructure and its possible functional significance in the Octopus vulgaris sucker. Beilstein J. Nanotechnol.5, 561–565. 10.3762/bjnano.5.66

  • CrossRef
  • Google Scholar

• 86

  TramacereF.BeccaiL.KubaM.GozziA.AngeloB.MazzolaiB. (2013). The morphology and adhesion mechanism of Octopus vulgaris suckers. PLoS ONE8 (6), e65074. 10.1371/journal.pone.0065074

  • CrossRef
  • Google Scholar

• 87

  TramacereF.BeccaiL.MattioliF.SinibaldiE.MazzolaiB. (2012). “Artificial adhesion mechanisms inspired by Octopus suckers,” in Proccedings of the 2012 IEEE International Conference on Robotics and Automation, St Paul, MN, USA, May 2012 (IEEE), 3846–3851. 10.1109/ICRA.2012.6225058

  • CrossRef
  • Google Scholar

• 88

  TramacereF.PugnoN. M.KubaM. J.MazzolaiB. (2015). Unveiling the morphology of the acetabulum in Octopus suckers and its role in attachment. Interface Focus5 (1), 20140050. 10.1098/rsfs.2014.0050

  • CrossRef
  • Google Scholar

• 89

  VogtD. M.KaitlynP. B.BrennanT. P.MoritzA. G.RandiD. R.ShankT. M.et al (2018). Shipboard design and fabrication of custom 3D-printed soft robotic manipulators for the investigation of delicate deep-sea organisms. PLOS ONE13 (8): e0200386. 10.1371/journal.pone.0200386

  • CrossRef
  • Google Scholar

• 90

  WainwrightD. K.KleinteichT.KleinteichA.GorbS. N.SummersA. P. (2013). Stick tight: Suction adhesion on irregular surfaces in the northern clingfish. Biol. Lett.9 (3), 20130234. 10.1098/rsbl.2013.0234

  • CrossRef
  • Google Scholar

• 91

  WangH.HeX.FaresAbu-DakkaJ.KyrkiV.YangC.LiX.et al (2022b). A bidirectional soft biomimetic hand driven by water hydraulic for dexterous underwater grasping. IEEE Robotics Automation Lett.7 (2), 2186–2193. 10.1109/LRA.2022.3143297

  • CrossRef
  • Google Scholar

• 92

  WangQ.FanC.GuiY.ZhangL.ZhangJ.SunL.et al (2021a). Engineered mechanosensors inspired by biological mechanosensilla. Adv. Mater. Technol.6 (11), 2100352. 10.1002/admt.202100352

  • CrossRef
  • Google Scholar

• 93

  WangS.LiL.ChenY.WangY.SunW.XiaoJ.et al (2019b). “A bio-robotic remora disc with attachment and detachment capabilities for reversible underwater hitchhiking,” in Proceedings of the 2019 International Conference on Robotics and Automation (ICRA), Montreal, QC, Canada, May 2019 (IEEE), 4653–4659. 10.1109/ICRA.2019.8793703

  • CrossRef
  • Google Scholar

• 94

  WangS.LiL.SunW.WainwrightD.WangH.ZhaoW.et al (2020b). Detachment of the remora suckerfish disc: Kinematics and a bio-inspired robotic model. Bioinspiration Biomimetics15 (5), 056018. 10.1088/1748-3190/ab9418

  • CrossRef
  • Google Scholar

• 95

  WangS.LiL.ZhaoW.ZhangY.WenL. (2022a). A biomimetic remora disc with tunable, reversible adhesion for surface sliding and skimming. Bioinspiration Biomimetics17 (3), 036001. 10.1088/1748-3190/ac4e7a

  • CrossRef
  • Google Scholar

• 96

  WangS.LuoH.LinghuC.SongJ. (2021b). Elastic energy storage enabled magnetically actuated, Octopus‐inspired smart adhesive. Adv. Funct. Mater.31 (9), 2009217. 10.1002/adfm.202009217

  • CrossRef
  • Google Scholar

• 97

  WangT.LidtkeA. K.Giorgio-SerchiF.WeymouthG. D. (2019c). “Manoeuvring of an aquatic soft robot using thrust-vectoring,” in Proccedings of the 2019 2nd IEEE International Conference on Soft Robotics (RoboSoft), Seoul, Korea (South), April 2019 (IEEE), 186–191. 10.1109/ROBOSOFT.2019.8722732

  • CrossRef
  • Google Scholar

• 98

  WangY.HenselR. (2021). Bioinspired underwater adhesion to rough substrates by cavity collapse of cupped microstructures. Adv. Funct. Mater.31 (31), 2101787. 10.1002/adfm.202101787

  • CrossRef
  • Google Scholar

• 99

  WangY.KangV.ArztE.WalterF.HenselR. (2019a). Strong wet and dry adhesion by cupped microstructures. ACS Appl. Mater. Interfaces11 (29), 26483–26490. 10.1021/acsami.9b07969

  • CrossRef
  • Google Scholar

• 100

  WangY.KangV.WalterF.ArztE.HenselR. (2020a). Switchable underwater adhesion by deformable cupped microstructures. Adv. Mater. Interfaces7 (23), 2001269. 10.1002/admi.202001269

  • CrossRef
  • Google Scholar

• 101

  WangY.YangX.ChenY.WainwrightD. K.KenaleyC. P.GongZ.et al (2017a). A biorobotic adhesive disc for underwater hitchhiking inspired by the remora suckerfish. Sci. Robotics2 (10), eaan8072. 10.1126/scirobotics.aan8072

  • CrossRef
  • Google Scholar

• 102

  WangY.ZhangP.HuangH.ZhuJ. (2022c). Bio-inspired transparent soft jellyfish robot. Soft Robot.10 (3), 590–600. 10.1089/soro.2022.0027

  • CrossRef
  • Google Scholar

• 103

  WangZ.CuiW. (2021). For safe and compliant interaction: An outlook of soft underwater manipulators. Proc. Institution Mech. Eng. Part M J. Eng. Marit. Environ.235 (1), 3–14. 10.1177/1475090220950911

  • CrossRef
  • Google Scholar

• 104

  WolfZ.JusufiA.VogtD. M.LauderG. V. (2020). Fish-like aquatic propulsion studied using a pneumatically-actuated soft-robotic model. Bioinspiration Biomimetics15 (4), 046008. 10.1088/1748-3190/ab8d0f

  • CrossRef
  • Google Scholar

• 105

  WonP.KimK. K.KimH.JungParkJ.HaI.ShinJ.et al (2021). Transparent soft actuators/sensors and camouflage skins for imperceptible soft robotics. Adv. Mater.33 (19), 2002397. 10.1002/adma.202002397

  • CrossRef
  • Google Scholar

• 106

  WuM.ZhengX.LiuR.HouN.AfridiW. H.Hussain AfridiR.et al (2022). Glowing sucker Octopus (Stauroteuthis syrtensis)‐Inspired soft robotic gripper for underwater self‐adaptive grasping and sensing. Adv. Sci.9 (17), 2104382. 10.1002/advs.202104382

  • CrossRef
  • Google Scholar

• 107

  YoussefS. M.SolimanM.SalehM. A.MousaM. A.ElsamantyM.RadwanA. G. (2022). Underwater soft robotics: A review of bioinspiration in design, actuation, modeling, and control. Micromachines13 (1), 110. 10.3390/mi13010110

  • CrossRef
  • Google Scholar

• 108

  ZhaiY.ZhengX.XieG. (2021). Fish lateral line inspired flow sensors and flow-aided control: A review. J. Bionic Eng.18 (2), 264–291. 10.1007/s42235-021-0034-y

  • CrossRef
  • Google Scholar

• 109

  ZhangR.ShenZ.WangZ. (2018). Ostraciiform underwater robot with segmented caudal fin. IEEE Robotics Automation Lett.3 (4), 2902–2909. 10.1109/LRA.2018.2847198

  • CrossRef
  • Google Scholar

• 110

  ZhangR.ShenZ.ZhongH.TanJ.HuY.WangZ. (2021). A cephalopod-inspired soft-robotic siphon for thrust vectoring and flow rate regulation. Soft Robot.8 (4), 416–431. 10.1089/soro.2019.0152

  • CrossRef
  • Google Scholar

• 111

  ZhangX.WangJ.WanD. (2020). CFD investigations of evolution and propulsion of low speed vortex ring. Ocean. Eng.195, 106687. 10.1016/j.oceaneng.2019.106687

  • CrossRef
  • Google Scholar

• 112

  ZhuJ.WhiteC.WainwrightD. K.Di SantoV.LauderG. V.Bart-SmithH. (2019). Tuna robotics: A high-frequency experimental platform exploring the performance space of swimming fishes. Sci. Robotics4 (34), eaax4615. 10.1126/scirobotics.aax4615

  • CrossRef
  • Google Scholar