ArticlePDF Available

Bioinspiration and biomimetics in marine robotics: a review on current applications and future trends

March 2024

Authors:

Amrita Vishwa Vidyapeetham

Over the past few years, the research community has witnessed a burgeoning interest in biomimetics, particularly within the marine sector. The study of biomimicry as a revolutionary remedy for numerous commercial and research-based marine businesses has been spurred by the difficulties presented by the harsh maritime environment. Biomimetic marine robots are at the forefront of this innovation by imitating various structures and behaviours of marine life and utilizing the evolutionary advantages and adaptations these marine organisms have developed over millennia to thrive in harsh conditions. This thorough examination explores current developments and research efforts in biomimetic marine robots based on their propulsion mechanisms. By examining these biomimetic designs, the review aims to solve the mysteries buried in the natural world and provide vital information for marine improvements. In addition to illuminating the complexities of these bio-inspired mechanisms, the investigation helps to steer future research directions and possible obstacles, spurring additional advancements in the field of biomimetic marine robotics. Considering the revolutionary potential of using nature's inventiveness to navigate and thrive in one of the most challenging environments on Earth, the conclusion of the current review urges a multidisciplinary approach by integrating robotics and biology. The field of biomimetic marine robotics not only represents a paradigm shift in our relationship with the oceans, it also opens previously unimaginable possibilities for sustainable exploration and use of marine resources by understanding and imitating nature's solutions.

Content uploaded by Karthik V Shankar

Content may be subject to copyright.

Bioinspir. Biomim. 19 (2024) 031002 https://doi.org/10.1088/1748-3190/ad3265

RECEIVED

1 December 2023

REVISED

6 February 2024

ACC EPT ED FOR PUB LICATI ON

11 March 2024

PUBLISHED

2 April 2024

TOPICAL REVIEW

Bioinspiration and biomimetics in marine robotics: a review on

current applications and future trends

Amal Prakash1, Arjun R Nair1, H Arunav1, Rthuraj P R1, V M Akhil2, Charbel Tawk3

and Karthik V Shankar1,4,∗

1Department of Mechanical Engineering, Amrita Vishwa Vidyapeetham, Amritapuri, India

2School of Interdisciplinary Research, Indian Institute of Technology, Delhi, India

3Department of Industrial and Mechanical Engineering, School of Engineering, Lebanese American University, Byblos, Lebanon

4Centre for Flexible Electronics and Advanced Materials, Amrita Vishwa Vidyapeetham, Amritapuri, India

∗Author to whom any correspondence should be addressed.

E-mail: karthikvs@am.amrita.edu and vmakhil@iitd.ac.in

Keywords: biomimetic aquatic robot, marine organisms, propulsion mechanism, fish locomotion

Abstract

Over the past few years, the research community has witnessed a burgeoning interest in

biomimetics, particularly within the marine sector. The study of biomimicry as a revolutionary

remedy for numerous commercial and research-based marine businesses has been spurred by the

difficulties presented by the harsh maritime environment. Biomimetic marine robots are at the

forefront of this innovation by imitating various structures and behaviors of marine life and

utilizing the evolutionary advantages and adaptations these marine organisms have developed over

millennia to thrive in harsh conditions. This thorough examination explores current developments

and research efforts in biomimetic marine robots based on their propulsion mechanisms. By

examining these biomimetic designs, the review aims to solve the mysteries buried in the natural

world and provide vital information for marine improvements. In addition to illuminating the

complexities of these bio-inspired mechanisms, the investigation helps to steer future research

directions and possible obstacles, spurring additional advancements in the field of biomimetic

marine robotics. Considering the revolutionary potential of using nature’s inventiveness to

navigate and thrive in one of the most challenging environments on Earth, the current review’s

conclusion urges a multidisciplinary approach by integrating robotics and biology. The field of

biomimetic marine robotics not only represents a paradigm shift in our relationship with the

oceans, but it also opens previously unimaginable possibilities for sustainable exploration and use

of marine resources by understanding and imitating nature’s solutions.

1. Introduction

The Earth’s vast oceans, covering 71% of its sur-

face area, harbor many essential resources, includ-

ing water, minerals, oil, and diverse living organ-

isms, crucial to modern society. Despite this import-

ance, most of the ocean remains largely unexplored

and inaccessible throughout human history [1,2]

Consequently, the efficient and sustainable explor-

ation and exploitation of marine resources have

become one of the leading research interests within

the scientific community. Underwater robots, with

their wide range of applications and immense poten-

tial value, emerge as essential components of marine

development and underwater activities. Researchers

have developed novel marine robots for specialized

missions, including autonomous underwater vehicles

(AUVs) and remotely operated vehicles. The mari-

time robotics market is expected to reach USD

6.74 billion by 2025, indicating significant growth

soon [3].

Marine robotics finds applications across a diverse

range of sectors, each with unique challenges and

opportunities [4]:

•Science and environment: through breakthroughs

in energy provision, perception, navigation, com-

munication, control, and autonomy, marine

robotics substantially contributes to scientific

endeavors, especially in oceanography, geological

sampling, and seafloor mapping. They are also

useful for long-term environmental monitoring

Bioinspir. Biomim. 19 (2024) 031002 A Prakash et al

and are excellent at resolving pollution and

oil spills.

•Military: the military sector leverages marine

robotics for underwater surveillance and stealth

operations.

•Ocean mining and oil industry: in the oil and

mining sector, marine robots play a critical role in

assisting operations such as offshore oil drilling,

undersea structure building and maintenance,

resource assessment, and ocean survey.

•Other applications: besides these main applic-

ations, marine robotics is used in underwater

communication, nuclear power plant inspection,

ship hull inspection, internal tank inspection, and

power cable installation and inspection. They can

also be used as underwater fishery rangers, provid-

ing entertainment like underwater tours. These

uses highlight marine robotics’ adaptability to vari-

ous industries.

The various challenges posed by the harsh mari-

time environment have spurred the development of

marine robotics. With life in the ocean for 3.7 bil-

lion years, evolutionary advantages and adaptations

of various marine species offer valuable insights for

innovation and have become essential in shaping

marine technologies. Natural selection has honed the

mechanical systems in fish, making them highly effi-

cient in their habitats. Harnessing these adaptations

inspires innovative designs for improving the effi-

ciency and interaction of man-made systems with the

aquatic environment.

Humanity has always drawn inspiration from

nature, as seen by the development of early tools and

modern technologies. The foundation for translating

biological solutions into technical applications is that

biological structures are well-suited for their specific

purposes and can provide interesting applications [5].

Bio-inspired fish robots outperform conventional

underwater vehicles thanks to great efficiency, agility,

low noise, and little fluid disruption [6]. The soph-

isticated solutions in biological systems have out-

performed the most advanced manmade underwater

sensors in maneuvering through challenging under-

water situations. The merging of biology and techno-

logy shows promise for solving the unique challenges

of the maritime environment through the application

of biomimetic marine robotics.

The current review commences with explor-

ing various classifications for underwater robots,

subsequently delving into a comprehensive analysis

of marine organisms categorized by their respect-

ive propulsive mechanisms. It then proceeds to

provide an overview of recent research endeavors

and provides instances of robotic systems that har-

ness or emulate each of these propulsion methods.

It then proceeds with a few examples of marine

bio-inspired sensing technologies. The following dis-

cussion focuses on current research directions and

obstacles in marine biomimetic robotics, providing a

comprehensive overview of this cutting-edge area.

1.1. Terminologies

The diversity of terminologies related to biomim-

icry can lead to confusion due to varying interpret-

ations, and therefore, it is crucial to establish a com-

mon understanding [7]. The current review adheres

to the standard definitions outlined in ISO Standard

18458:2015 [8], ensuring consistency and clarity in

the use of terminology. By adopting these standard-

ized definitions, the review aims to mitigate discrep-

ancies arising from cultural factors, professional spe-

cializations, and individual perspectives, providing a

solid foundation for communication and compre-

hension in biomimicry. Some of the important defin-

itions adopted in the current review are as follows:

•Bioinspiration-creative approach based on the

observation of biological systems.

•Biological system-coherent group of observable

elements originating from the living world span-

ning from nanoscale to macroscale.

•Biomimicry/Biomimetism-philosophy and inter-

disciplinary design approaches use nature to meet

sustainable development challenges.

•Biomimetics-interdisciplinary cooperation of bio-

logy and technology or other fields of innovation

to solve practical problems through the function

analysis of biological systems, their abstraction into

models, and the transfer into and application of

these models to the solution.

•Bionics-technical discipline that seeks to replicate,

increase, or replace biological functions by their

electronic and/or mechanical equivalents.

•Sustainability/Sustainable development that sat-

isfies the present requirements without risking that

future generations will not be able to satisfy them.

2. Classification of different types of

bioinspired and biomimetic marine

robots

The current review identifies several distinct categor-

ization frameworks in classifying biomimetic mar-

ine robotic systems. The first classification criterion

scrutinizes the application domains of these robotic

systems within the marine sector, encompassing sci-

entific research, geological sampling, environmental

monitoring, military applications, ocean mining, and

the oil industry. This delineation is crucial for under-

standing these biomimetic systems’ diverse roles in

addressing marine challenges [4]. The second classi-

fication criterion involves categorization based on the

phylum of the organism that serves as the inspiration

for the robotic system. Leveraging the extensive World

repository of marine organism data, this classifica-

tion provides insights into the taxonomic origins of

Bioinspir. Biomim. 19 (2024) 031002 A Prakash et al

Figure 1. Approximate geographic coverage of the larger Regional Species within the World Register of Marine Species.

Reproduced from [9]. CC BY 4.0.

biomimetic designs [9]. The approximate geographic

coverage of the larger regional species within the

WoRMS database is shown in figure 1. However, the

primary focus of this review centers on the third

classification criterion, which classifies biomimetic

marine robotic systems based on the type of marine

locomotion they emulate [10], illustrated in figure 2.

The subsequent subsection offers an in-depth ana-

lysis of this classification, unraveling the intricacies of

locomotion-inspired designs and their implications

for advancing marine robotic technology.

2.1. Classification based on locomotion

When examining biological locomotion in marine

species, a wide range of locomotive techniques offer

an abundance of analytical criteria to consider. As

a phenomenon, marine locomotion is systematically

categorized into three principal classes: fin oscilla-

tion, fin undulation, and jet propulsion. Each clas-

sification encompasses various locomotive mechan-

isms, contributing to the comprehensive understand-

ing of animal movement [10]. Various fins utilized by

these locomotive mechanisms are shown in figure 3.

The coming sub-sections will look briefly into these

various classes of locomotion.

2.1.1. Fin oscillation

The nomenclature used to categorize fin oscilla-

tion is derived from extensive research conducted

by Sfakiotakis et al [11], which offers a thorough

description of various fish locomotion types. This

classification is the broadest of the ones that have

been established, encompassing a wide range of bio-

logical systems that offer various options in size,

speed, endurance, and turning skills. The propulsion

actuation systems within this classification show sig-

nificant variety depending on the sub-classification

under examination. This variation allows for many

missions, including close-quarter maneuvering, burst

swimming, and endurance. The sub-classifications

encompass caudal fin swimmers, pectoral fin swim-

mers, dorsal & anal fin propelled species, and finally,

the category of combination oscillation representing

a combination of multiple fin types. In addition to

highlighting the variety of biological locomotion, this

variation within the fin oscillation categorization also

lays the groundwork for understanding the various

functional adaptations aquatic creatures display in

their search for effective movement.

2.1.1.1. Caudal fin—anguilliform, thunniform,

subcarangiform, and carangiform

The category of caudal fin Swimmers often denoted as

body caudal fin (BCF) swimmers, is distinguished by

the generation of thrust primarily achieved through

the undulatory movement of the body, particularly

by the tail as shown in figures 4–6. This method

of propulsion is characteristic of faster swimmers or

those possessing high endurance. Notably, a substan-

tial proportion of these fish species exhibit predatory

Bioinspir. Biomim. 19 (2024) 031002 A Prakash et al

Figure 2. Classification Based on the type of marine locomotion the system emulates.

behavior [12]. Moreover, many caudal fin swimmers

are also known to be migratory animals, emphasizing

their adaptability and long-distance swimming cap-

abilities. Within the BCF classification, distinct sub-

types are discerned, each with its unique mode of

locomotion. A detailed description and examples of

these subtypes are given in table 1.

2.1.1.2. Pectoral fin-labriform

Labriform swimming, characterized by pectoral fin

locomotion, is tailored for slower speed and agile

swimming [16] and is exemplified by species such as

wrasses and parrotfish, which are shown in figure 7,

which efficiently navigate close quarters within reef

ecosystems. Labriform swimmers may occasionally

integrate the use of the caudal fin, particularly dur-

ing burst swimming or when their pectoral muscles

are fatigued, they exhibit limited endurance when

relying solely on their pectoral fins for propulsion.

Structurally, labriform swimmers possess pectoral

fins with rigid fin rays within the fin membrane,

enabling effective swimming [17]. This swimming

style encompasses two distinct modes: flapping, char-

acterized by upstroke and downstroke with the

Bioinspir. Biomim. 19 (2024) 031002 A Prakash et al

Figure 3. Various fins are that are utilized for locomotion. Reprinted from [10], Copyright (2018), with permission from Elsevier.

Table 1. Classification of caudal fin swimmers.

Sub types Description Examples

Anguilliform •Largest body oscillation for caudal fin oscillation.

•Characterized by a sinuous and serpentine motion.

•Extremely flexible with small turning radii.

•Efficient migratory swimmers: European Eels are efficient

migratory swimmers [13].

•Found in reefs, highly flexible due to a large number of

vertebrae [14].

•Extremely efficient propulsion, sparing energy for

migratory swimming.

European Eel, Giant Moray, Sea

Snake

Figure 4. Sea Snake. Reprinted

from [10], Copyright (2018),

with permission from Elsevier.

Subcarangiform

and carangiform

•Highly similar, distinguished by initiation point along the

body.

•Subcarangiform uses more head movement, and

oscillation closer to the head.

•Carangiform uses one-third of the posterior body for

oscillation.

•Studied due to accessibility and efficiency as power thrust

fish [15].

•Variety in sizes, weights, and speeds

•Less flexible than anguilliform.

Subcarangiform: Chinook

salmon, lake sturgeon

Carangiform: bonefish, giant

trevally

Figure 5. Chinook Salmon.

Reprinted from [10], Copyright

(2018), with permission from

Elsevier.

Thunniform •Limited body oscillation to the last quarter.

•Rigid tail with powerful oscillations.

•Very streamlined bodies for efficiency.

•Extremely efficient with sustained top speed.

•Some are warm-blooded, providing advantages in

endurance and power [12].

Yellowfin tuna, Bottlenose

dolphin

Figure 6. Yellowfin tuna.

Reprinted from [10], Copyright

(2018), with permission from

Elsevier.

leading edge of the pectoral fin and suitable for

close-quarter maneuvering, and rowing, employed

for faster speeds [10,11]. Combining these modes

allows for swifter speeds, exemplifying the versatility

of labriform swimming across diverse fish species in

adapting to distinct locomotive demands.

Bioinspir. Biomim. 19 (2024) 031002 A Prakash et al

2.1.1.3. Dorsal and anal fin- tetraodontiform

The dorsal & anal Fin locomotion category, exem-

plified by species such as the sharptail mola, slender

sunfish, and ocean sunfish shown in figure 8, is not-

ably underrepresented in the literature. Species in

this class use two huge paddle fins—one dorsal and

one anal—for propulsion and are thought to have

the least efficient design among the fishes described,

despite their astounding size [11]. The rigid ribbed

fins allow non-traditional oscillation motions to con-

trol depth and direction. However, because of their

poor yaw and pitch control, these animals have dif-

ficulty navigating in close quarters [10]. The distinct

characteristics of the tetraodontiform class, contrib-

ute to their unique ecological niche within the aquatic

ecosystem.

2.1.1.4. Dorsal/anal, caudal, and pectoral

fin—ostraciiform

Fish that mostly live in reef and close-quarters habit-

ats and can regulate their mobility via pectoral and

dorsal/anal fin oscillations are classified as ostracii-

forms. A distinctive and stiff body form resembling

a box characterizes this class; species like the Cowfish

and Boxfish, depicted in figure 9, are prime examples.

The unique anatomical configuration allows for pre-

cise fin oscillations, which enhances maneuverab-

ility in tight places. Despite their agility, ostracii-

form fishes are limited in their speed because of

their box-like body structure and relatively ineffi-

cient fin actuation [18]. Like labriform swimming,

ostraciiform species move by flapping their pec-

toral fins vertically and horizontally. Moreover, the

anal and dorsal fins execute coordinated flapping

in phase to give stability and forward propulsion.

Although ostraciiform fishes rarely use their little

caudal fin for burst propulsion, it does not happen

very often.

2.1.2. Fin undulation

The Fin Undulation locomotion class is distinguished

by using individual rib stimulation in fins to pro-

duce undulatory waves, which let aquatic anim-

als generate thrust and maintain stability and con-

trol. Interestingly, certain members of this class

can manipulate the direction of the undulating

wave, which allows for null-speed turning or swim-

ming both forward and backward [11,19]. Various

swimmers, including ambush predators and high-

endurance animals, are included in this class. The fin

undulation class has a more diverse membership than

the fin oscillation class, although it is smaller [10].

The sub-classifications encompass rajiforms, gymno-

tiforms, amiiforms, balistiform, and diodontiform,

presenting many locomotion techniques within the

fin undulation class.

2.1.2.1. Rajiforms

Fish with rajiform locomotion are distinguished by

their characteristic undulations of extended pectoral

fins. This unusual adaptation allows fish to move

through the water by modulating a single control sur-

face, the pectoral disc, by undulating rays [20]. This

class comprises mostly animals like skates and rays,

whose bodies are made of cartilage, which gives them

incredible flexibility. Various examples of rajiform

swimmers are shown in figure 10. Swimming capab-

ilities vary within the rajiform category, with manta

and cownose rays efficiently cruising due to their

rigid fin structures, enabling gliding, while species

like the common stingray and electric rays, with flex-

ible fins, exhibit excellent maneuverability in close-

quarter environments [10].

2.1.2.2. Amiiform

Amiiform locomotion is distinctive for its utilization

of an undulating dorsal fin, a characteristic feature

observed in select species such as the African aba and

the bowfin, shown in figure 11. When in amiiform

mode, these fish swim by flexing their long dorsal

fins, and keeping their bodies straight [11]. The gym-

narchus niloticus is a prime example of this class, with

an extended dorsal fin that spans most of its body

length and tapers to a posterior point. The dorsal fin’s

locomotor waves exhibit variable amplitude and can

go either direction, especially when the animal turns

or brakes [21]. Amiiforms are not very fast, but they

are very agile, able to move both forward and back-

ward by changing the direction of the wave motion

produced by the dorsal fin. The dorsal fin structure

comprises a flexible membrane surrounding tightly

packed fin ribs or rays. The muscles on either side

of the membrane contract to cause intricate wrink-

ling. This intricate undulatory motion grants amii-

form animals tight turning and multi-directional

capabilities, enabling effective navigation in various

directions [22].

2.1.2.3. Gymnotiform

Gymnotiform locomotion represents the inverted

counterpart of amiiform mode, where propulsion is

achieved through undulations of a long anal fin, a

characteristic feature observed in species such as gym-

notus carapo [21]. Various examples of gymnotiform

swimmers are shown in figure 12. In this mode, fishes

lacking dorsal and caudal fins rely on the undulat-

ing ventral anal fin for propulsion, akin to amiiform

animals [23]. Unlike engaging in extensive migratory

swims, gymnotiform species focus on hunting within

their preferred zones to minimize vulnerability to lar-

ger predators. The undulating fin at the body’s pos-

terior provides forward and backward movements,

coupled with the body’s bend, enabling these fishes

to execute complex maneuvering [24]. The angle of

Bioinspir. Biomim. 19 (2024) 031002 A Prakash et al

Figure 7. (a) Banded Wrasse. (b) Schelegel’s Parrotfish. Reprinted from [10], Copyright (2018), with permission from Elsevier.

Figure 8. Ocean sunfish. Reprinted from [10],

Figure 9. Boxfish. Reprinted from [10], Copyright (2018),

with permission from Elsevier.

the undulatory fin to the vertical axis grants a higher

degree of freedom in movement, allowing gymnoti-

form species, such as the ghost knifefish, to achieve

various directional changes by adjusting undulation

waves and the fin’s angle. Pectoral fins serve as con-

trol surfaces for pitch and roll control, enhancing the

overall maneuverability of gymnotiform fishes.

2.1.2.4. Balistiform

The balistiform mode of locomotion, observed in

fishes like the Picasso Triggerfish, which is shown

in figure 13, involves propulsion through synchron-

ized movements of both the dorsal and anal fins,

collectively referred to as medial pair fin (MPF)

undulation locomotion. This mode offers two altern-

ative forms: undulatory or oscillatory movements

of the median fins, analogous to the anguilliform

and carangiform modes of body flexure, respectively

[23]. Balistiform fishes primarily utilize their MPFs

for undulation propulsion, occasionally employing

the caudal fin when the muscles controlling the fin

ribs become fatigued, enhancing endurance [25].

Balistiform fishes, typically characterized by modest

size, are not known for extensive migrations; they

are commonly found and studied in shallow reef

environments.

2.1.2.5. Diodontiform

The diodontiform mode of locomotion, exempli-

fied by species like the porcupine fish or pufferfish

shown in figure 14, is characterized by the genera-

tion of propulsive thrust through propagating undu-

lations in large pectoral fins. In this mode, undula-

tions are induced through the flapping of the fins,

with up to two full wavelengths visible across them

[26]. Diodontiforms use broad, undulating pectoral

fins for slow movements, which are complemen-

ted by oscillating dorsal and anal fins that provide

a push for faster speeds [27]. Diodontiform swim-

ming is primarily found in shallow-water reef habitats

and is not adapted for migratory purposes. Another

example in this category is the Bluegill, which uses its

pectoral fin undulation for swimming at low speeds

and its caudal fin at greater rates [28].

2.1.3. Jet propulsion

Pulsatile jet propulsion, a prevalent swimming mode

across diverse aquatic taxa ranging from chordates

to cnidarians, encompasses a spectrum of sophistic-

ated adaptations that significantly amplify propulsion

capabilities in biological systems compared to their

mechanical counterparts. Key features contributing

to this enhanced efficiency include flexible propulsive

structures, variable jet orifice diameter, vortex inter-

actions, and energy capture during the refill phase

[29]. Categorically, jet propulsion manifests in four

Bioinspir. Biomim. 19 (2024) 031002 A Prakash et al

Figure 10. (a)Cownose ray, (b)common torpedo, (c)manta ray. Reprinted from [10], Copyright (2018), with permission from

Elsevier.

Figure 11. (a) African Aba and (b)Bowfin. Reprinted from [10], Copyright (2018), with permission from Elsevier.

Figure 12. (a) Black knifefish, (b) electric eel. Reprinted from [10], Copyright (2018), with permission from Elsevier.

Figure 13. Picasso triggerfish. Reprinted from [10],

distinct classes, each characterized by specific organ-

isms and propulsion mechanisms and they are Bell

constriction, mantle constriction, shell compression

for swift and agile jet propulsion through the rapid

closure of their shells, and a combination of jet and

mantel fin undulation.

2.1.3.1. Bell constriction

Bell constriction locomotion is a kind of propulsion

unique to jellyfish. It is defined by the coordinated

contraction and relaxation of subumbrellar muscles

inside the bell, which causes the bell’s volume to con-

strict and cause water to be ejected. The effective-

ness of this motion is closely linked to ocean cur-

rents. A schematic representation of an oblate-shaped

Jellyfish and a prolate-shaped Jellyfish is shown

in figure 15, which exhibit diverse swimming pat-

terns, engaging in two primary modes: jet propulsion

and rowing propulsion [30]. Prolate-shaped jelly-

fish, characterized by a bullet-shaped bell, employ

jet propulsion through higher-frequency bell con-

tractions at an increased energy cost. In contrast,

oblate-shaped jellyfish with flat-shaped bells opt for

rowing propulsion characterized by slower but more

voluminous bell contractions, achieving greater water

displacement. Circumferential constrictions, while

not universal, enable oblate jellyfish to achieve full

bell deformation. Rowing propulsion, more preval-

ent in biological Jellyfish, is associated with energy

Bioinspir. Biomim. 19 (2024) 031002 A Prakash et al

Figure 14. Common pufferfish. Reprinted from [10], Copyright (2018), with permission from Elsevier.

Figure 15. Parts of (a) oblate-shaped jellyfish and (b) prolate-shaped jellyfish. Reprinted from [10], Copyright (2018), with

permission from Elsevier.

efficiency, leveraging ocean currents for extended

journeys [10].

2.1.3.2. Mantel constriction

The octopus is a prime example of mantel constric-

tion, a unique type of propulsion in which move-

ment is accomplished by creating a siphon and

then filling the mantle cavity with water. The main

method by which octopuses move is by combining

siphon propulsion with limb movement, particularly

the walking gait made possible by their arms [31].

Figure 16 shows an image of an octopus denoting

its various parts. Mantel constriction serves diverse

purposes for octopuses, including hunting, defense,

and achieving rapid swimming capabilities. Defensive

burst swimming, marked by the swift expulsion of

water, enables octopus to elude predators efficiently.

In instances of threat, octopus may deploy a defens-

ive ink release during burst swimming as a distraction

tactic [32].

2.1.3.3. Shell compression

Mollusks with a basic structure like the Atlantic

bay scallop are good examples of shell compres-

sion locomotion, which involves the dynamic open-

ing and shutting of the two shell halves to pro-

duce a characteristic clapping-action movement, as

shown in figure 17. This intricate propulsion mech-

anism involves water absorption into the abductor

muscle of scallops, followed by the forceful expulsion

Figure 16. Image of an octopus denoting its various parts.

Reprinted from [10], Copyright (2018), with permission

from Elsevier.

of water between the shells, resulting in the char-

acteristic clapping motion. The primary purpose of

shell compression in scallops lies in its utility as an

escape mechanism from predators [10]. However, it

is imperative to acknowledge that shell compression

is not an endurance-oriented propulsion method as

the energy expenditure associated with this mechan-

ism is significant.

2.1.3.4. Undulating mantel fin and mantel constriction

Biological systems that use a dynamic combination

of mantel constriction and undulating mantel fin for

propulsion, such as the squid and cuttlefish depicted

Bioinspir. Biomim. 19 (2024) 031002 A Prakash et al

Figure 17. Parts of mollusks (Atlantic Bay Scallop). Reprinted from [10], Copyright (2018), with permission from Elsevier.

Figure 18. (a) Squid and (b) cuttlefish. Reprinted from

[34], Copyright (2008), with permission from Elsevier.

in figure 18, demonstrate an amazing degree of plasti-

city in underwater locomotion. This adaptable tactic

allows for multiple swimming modes, with the undu-

lation fin-to-jet propulsion ratio changing accord-

ing to needs. Notably, these cephalopods can navigate

both forward and backward, executing rapid changes

in swimming directions through the synergistic use

of undulating fin and mantel constriction [33]. While

this propulsion approach offers meticulous control at

slower speeds, it endows cuttlefish and Squids with

unparalleled agility and remarkable speed. Certain

squid species, for instance, can generate propulsive

forces potent enough to launch them out of the water.

3. Bioinspired and biomimetic marine

robotic systems

Bioinspired fish robots are superior to traditional

underwater vehicles, boasting heightened efficiency,

superior maneuverability, minimal noise, and neg-

ligible fluid disturbance. Integrating nature-inspired

principles, encompassing the study of fish bion-

ics and robotic fish design, holds pivotal signific-

ance for developing next-generation submersibles. A

noticeable surge in research activity within the bio-

mimetics or biomimicry field, particularly evident

in recent years, underscores the increasing interest

and dedication of the research community toward

understanding and emulating natural phenomena.

This heightened focus on biomimetics, with a not-

able impact in the marine sector, presents a com-

pelling solution to the unique challenges posed by

the extreme conditions of the maritime environment.

These conditions, affecting commercial and research-

based marine industries, necessitate innovative solu-

tions, and biomimetic marine robots emerge as a cru-

cial response. Researchers and engineers take a signi-

ficant stride in marine innovations by tapping into the

evolutionary advantages and adaptations that various

marine organisms have developed to thrive in harsh

environments.

While developing biomimetic/bioinspired robots,

the choice of the material plays an integral part.

Generally, these materials can be categorized into

three main types: rigid, soft, or hybrid. Stiff materi-

als like steel, aluminum, or ABS plastic are commonly

used to build rigid robots. Typically, hydraulic sys-

tems or electric motors are used to drive them. Such

robots are essential in industry assembly lines because

of their sturdy construction, which enables them to

perform jobs demanding accuracy, rapid speed, and

significant force. However, their inflexibility makes

them less adaptable in chaotic, crowded settings or

Bioinspir. Biomim. 19 (2024) 031002 A Prakash et al

close to people [35]. As an illustration, many mar-

ine lives move in complex and elegant ways, pro-

pelled by soft structures that lack hard parts and are

mostly made of flexible materials that allow them

to flex while in use, soft robots are modeled after

these natural models. Their morphological altera-

tions are profound and range from somewhat flex-

ible to quite squishy. The primary objective of soft

robotics is to create machines that emulate the adapt-

ability and capabilities found in animals. This flex-

ibility and softness make soft robots ideal for tasks

in cluttered, unstructured environments and more

closely mimic the movements of living organisms

[36]. Using soft materials in robot design is instru-

mental in bridging the gap between traditional rigid

robots and the intricacies of natural locomotion.

In the upcoming subsection, we classify existing

biomimetic/bioinspired AUVs and marine/aquatic

robotic systems. This classification is based on the

specific type of marine locomotion they emulate.

It provides a structured framework to comprehend

the diverse bioinspired robotic systems designed for

underwater exploration and tasks.

3.1. System based on anguilliform

The domain of Anguilliform AUVs presents an

intriguing convergence of bioinspired design and

advanced robotics, showcasing distinct advantages

over traditional underwater vehicles. These AUVs,

characterized by their serpentine motion and signi-

ficant body undulation, use biomimicry principles

to navigate underwater environments efficiently.

Typically sized around 1 m, these AUVs demonstrate

versatility in speed, with some designs incorporat-

ing depth control mechanisms [10]. For instance, the

miniature autonomous robot (MAR) was designed

for underwater exploration and maneuvering, show-

casing an endurance range of 2–5 h. MAR utilizes a

helical fin system, closely mirroring the efficiency of

aquatic animal movements, enabling accurate steer-

ing and diving through its pectoral fins and an actu-

ated tail. Furthermore, the scalability of MAR for

swarm or cluster deployment enhances its applicabil-

ity in confined spaces where traditional thrusters may

face limitations [37]. A detailed description of MAR

and a few other robotic systems based on anguilliform

is given in table 2.

Emulating anguilliform locomotion extends bey-

ond traditional rigid robotics to soft robotics, intro-

ducing flexibility that enables intricate movements

and adaptability in cluttered or unstructured envir-

onments. Soft eel robots featuring soft pneumatic

actuators deform during use, showcasing flexib-

ility and adaptability. These robots can be util-

ized for delicate tasks with considerable efficiency;

for example, an eel-inspired robot could achieve a

straight-swimming speed of 10.5 cm s−1, demon-

strating a cost of transport (COT) of 19.21 in optimal

conditions [38]. The hydrodynamic models of these

soft robots are designed to correlate thrust force

with the beat frequency of their undulating bod-

ies, optimizing conditions for efficient underwater

movement.

Another research looks at how caudal fin shape

affects robots that resemble snakes and eels and iden-

tifies optimal swimming parameters of the fin since

these robots swim most efficiently and quickly at cer-

tain caudal fin angles. Various caudal fin angles used

in this investigation are shown in figure 19. The study

found that stronger and more organized vortex rings

are produced by the ideal design, defined by an 85◦

leading edge angle and a 120◦convex trailing edge

angle. This enhances swimming efficiency and velo-

city. The wake structure created by the snake-like

robot is shown by numerical simulations to be made

up of two parallel, disconnected verse van Karman

vortex rings. This ideal caudal fin design reveals the

efficiency and agility of aquatic movement, similar

to the common caudal fin designs seen in slender-

bodied animal swimmers in biology [39].

Deep reinforcement learning (DRL) represents a

paradigm shift in controlling underactuated robotic

eels. The online control method based on DRL

enables the robotic eel to learn to swim through

200 epochs of training autonomously, each consisting

of 500 steps. The methodology effectively addresses

complex control challenges, particularly in scenarios

difficult to model conventionally. The online control

method demonstrates the longest forward distance

and superior overall performance compared to off-

line control methods. Its adaptability and efficiency

in energy use make it a promising avenue for future

robotic fish applications, including those in environ-

ments with disturbances [40].

The synthesis of bioinspired design and advanced

robotics in anguilliform AUVs unveils a spectrum of

underwater exploration and tasks possibilities. These

advancements, from helical fins to soft robotic adapt-

ations and optimal caudal fin geometry, underscore

the potential for innovative solutions in underwa-

ter robotics. Integrating DRL further propels the

autonomy and adaptability of underactuated robotic

eels, contributing to the evolving landscape of biomi-

metic marine robotics.

3.2. System based on thunniform

Thunniform-based underwater systems, notable for

their limited body undulation to the last quarter, rigid

tails with powerful oscillations, streamlined bodies

for efficiency, and impressive top speeds reaching up

to 2.1 m s−1, represent a class of underwater vehicles

Bioinspir. Biomim. 19 (2024) 031002 A Prakash et al

Table 2. Examples of systems based on anguilliform.

Investigator Description System Parameters

Struebig et al [37]•MAR for underwater exploration with

helical fin structure.

•Inner/outer loop steering, actuated tail,

raspberry Pi, Wi-Fi control.

•Precision via pitch, roll, off-axis forces;

neutral buoyancy, separate pectoral

fins for submerged swimming;

adaptable tail, swarm scalability.

•Length: 108 cm

•Speed: 0.47 m s−1

•Power consumption reduction: up to

50%

•Overall efficiency: 0.176 N/W

•Thrust: 4.7 N

•Power: 26.6 W

•Weight: 7500 g

Nguyen and Ho [38]•Soft eel robot mimicking natural

musculature using soft pneumatic

actuators.

•Regulates air delivery to produce

sinusoidal waves via four pairs of soft

actuators and pulse signals.

•Positive correlation between thrust

force and beat frequency; swimming

efficiency affected by thrust force and

body posture.

•Length: 0.53 units

•Width: 0.045 units

•Highest Velocity:19 cm s−1or

0.36 BL s−1.

•Velocity Range: 6.5 cm s−1

(0.122 BL s−1) to 10.5 cm s −1

(0.198 BL s−1)

•Lowest COT: 10.72

•Frequency for the highest velocity and

lowest COT: 1.25 Hz

•Suitable frequency range: 0.83–1.67 Hz

Feng et al [41]•Serial soft-actuator array for

anguilliform locomotion in

underwater robots

•Untethered eel-inspired soft robot

featuring bidirectional bending flexing

fluidic elastomer actuators,

miniaturized station with pumps,

batteries, and control components.

•Dragon skin-10 passive silicone rubber

tail attached to the FEA array.

•Mean forward speed: 0.12 l s−1, akin to

rigid anguilliform robots, slower than

real eels (0.6 l s−1).

•Larger head beating amplitude at 2.0 s

results in reduced speed, aligning with

computational fluid dynamics (CFD)

predictions.

Crespi et al [42]•Salamander-like robot, seen in [22],

having two degrees of freedom (DOFs)

for the spine and two for each leg.

•A six-segment walking robot with an

FPGA-based control system on board

was demonstrated in [23]. It is not

amphibious.

•Central pattern generator (CPG): Used

to drive locomotion as a network of

coupled nonlinear oscillators.

—

Crespi et al [43]•Amphibious snake-like robot

(AmphiBot I) for water and ground

movement.

•Sea snake and lamprey-like swimming;

snake-like lateral undulatory

movement on the ground.

•Figure 20 shows the robot in the

experimental setup and motion.

•Dimension:

0.07 m ×0.055 m ×0.033 m

•Speed: 0.035 m s−1

Stefanini et al [44]•Lamprey-inspired development

integrating biomechanical and

neuroscientific knowledge.

•Untethered, compliant body,

high-efficiency muscle-like actuators

replicating lamprey movements.

•Proprioceptive sensors for stretch

detection, stereoscopic vision for

enhanced environmental interaction.

•Dimension: 0.99 m

•Speed: 0.3 m s−1

(Continued.)

Bioinspir. Biomim. 19 (2024) 031002 A Prakash et al

Table 2. (Continued.)

Rollinson et al [45]•SEA snake line of elastic-actuated

snake robots with specifics on

architecture and design.

•Architecture allows exploration of

different robot topologies; joining

modules to a central chassis enables the

creation of legged robots with torque

and position control at each joint

•Dimension per module =0.064 m

Total module =1.174 m

Yu et al [46]•Designed for versatile locomotion on

land and in water using 9 Modular

Universal Units.

•Explores various gaits, including

traditional and novel ones like S-shape

rolling and helical rolling.

•Derived with each step under the

supervision of a reduced Serpenoid

Curve.

•Length: 1.17 m

•Speed: 0.07 m s−1

Liljeback et al [47]•Modular and reconfigurable for

versatile research applications,

featuring a waterproof design for

underwater locomotion.

•Length one section =0.089 m,

Total =1.1 m

•Joint speed: 429 deg s−1

Figure 19. Various caudal fin shapes utilized in the study and various views of its underwater snake-like robot. Reprinted from

[39], Copyright (2022), with permission from Elsevier.

Figure 20. The AmphiBot I in the (a) experimental setup and (b) in motion. Reprinted from [43], Copyright (2005), with

permission from Elsevier.

characterized by their larger size. A detailed descrip-

tion of a few robotic systems based on thunniform is

given in table 3. The kinematics and dynamics of their

undulatory motion, crucial for understanding their

propulsion mechanisms, often involve computational

fluid dynamics (CFD) to model swimming motion.

However, current dynamic models face limitations

due to certain assumptions, emphasizing the need for

improved models that consider both kinematics and

hydrodynamics without restrictive conditions [48].

Bioinspir. Biomim. 19 (2024) 031002 A Prakash et al

Table 3. Examples of systems based on thunniform.

Investigator Description System Parameters

Farideddin

Masoomi et al [48]

•Modeled after a tuna fish, featuring a

four-DOF design with inspiration

from thunniform swimming modes

assisted by a tail mechanism driven by

a single DC motor.

•Figure 21. Shows the prototype Fish

Robot UC-Ika 1 and its mechanical

drawing.

•Length: 0.213 m

•Cruising speed: 0.29 m s−1

•Angular: slight swing, 5.56◦amplitude

around the center of mass

•Mass: 4.12 Kg

Dong et al [52]•Innovative prototype modeled after the

whale shark, featuring controlled

pectoral fins, pitch adjustment,

buoyancy control, and actuation

systems.

•Investigates effects of pectoral fin

deflection and moveable mass

displacement on the gliding angle.

•Length: 0.47 m

•Width: 0.26 m

•Forward speed of gliding: 0.55 BL s−1

•Swimming speed: 0.77 BL s−1

•Turning radius: 0.72 BL

•Weight: 2.2 Kg

Tong et al [53]•Develop untethered high-performance

robotic tuna addressing speed and

maneuverability challenges in

bioinspired underwater robots.

•Establish a comprehensive

three-dimensional model and

accurately identify model parameters

through experimental data.

•Propose two strategies to enhance

steering performance, imitating tuna

and meeting operational demands.

•Length: 0.72 m

•High swimming speed: 2.26 m s−1

•Outstanding steering maneuverability:

0.48 body lengths in turning radius

Yan et al [50]•Implement autonomous navigation

with visual stabilization control and

develop a stereo vision-based

navigation network for guidance law.

•Novel sphyrnidae-inspired robotic

shark with superior motion

performance.

•Develop a navigation network to

generate strategies for dealing with

structural obstacles,

•Dimension(mm): 920 ×276 ×342

•Maximum swimming speed: 0.8 BL s−1

•Maximum Turning Radius: 20 cm

•Image Jitter Reduction: 26.02% improve-

ment compared to traditional methods

•Weight: 5.5 Kg

Mitin et al [54]•Thunniform swimming robot

mimicking tuna fish movement with a

novel elastic cord propulsion system.

Its mechanical design and motion are

shown in figure 22.

•Increased oscillation amplitude up to a

threshold enhances speed, but further

increases result in weak speed gains

with higher energy costs.

•Dimension(mm): 379.6 ×127.7 ×95.2

•Maximum Speed: 0.4 BL s−1

•Swimming efficiency: 30–70 J kg−1·m−1,

•Mass: 0.78 kg,

Wen et al [55]•Quantifying robotic fish thrust

efficiency using a novel experimental

approach.

•Identified peak efficiency within a

specific Strouhal number range,

emphasizing the significance of

optimizing robotic fish design.

•Length: 0.588 m

•Speed: 0.3 m s−1

Chen et al [56]•Biomimetic robotic fish propelled by

an IPMC actuator

•Passive plastic fin fastened to the IPMC

beam,

•Predict steady-state cruising speed

based on periodic actuation voltage,

supported by experimental results for

different tail dimensions.

•Length: 0.223 m

•Speed: 0.02 m s−1

(Continued.)

Bioinspir. Biomim. 19 (2024) 031002 A Prakash et al

Table 3. (Continued.)

Kopman and

Porfiri [57]

•robotic fish with a modular caudal fin

design.

•Investigates how different caudal fin

geometries impact thrust production.

•Conduct experiments to assess

performance, maneuverability, and

correlation between static thrust and

terminal speed.

•Length: 0.117 m (without caudal fin)

Shen et al [58]•Features 3-DOF flippers on each side

with versatile motion capabilities,

including lead-lag, feathering, and

up–down.

•Implements a multi-link oscillatory

propulsion system with four joints,

including a robust worm gear and a

powerful DC motor for high torque in

the first joint.

•Length: 1.2 m

Yu and Wei [59]•Mechatronic design and

implementation of a fast-swimming

dolphin-like robot.

•Modular slider-crank-based flapping

mechanism, leveraging continuous

high-speed rotation for enhanced

swimming speed.

•Length: 0.75 m

•Speed: 0.78 m s−1

Anderson [60]•Draper Lab VCUUV designed for

fish-swimming propulsion studies.

•Modeled after yellowfin tuna with a

rigid front hull and freely flooded

articulated robot tail.

•Utilizes vorticity control propulsion.

•Length: 2.4 m

•Speed: 1.2 m s−1

•Turning rates up to 75 degrees per second.

Wu et al [61]•Novel design for a gliding robotic

dolphin, combining features of robotic

dolphins and underwater gliders.

•Offers high maneuverability and long

endurance for diverse underwater

tasks.

•Controllable fins and a horizontal fluke

eliminate the need for internal

movable masses, saving space.

•Length: 1.125 m

•Gliding speed: 0.155 m s−1

Yu et al [62]•Achieve enhanced hydrodynamic

efficiency through a streamlined body

design. Ensure a high power-to-weight

ratio to optimize propulsion

•Equip the system with attitude and

depth sensors for enhanced control.

Increased Degrees of Freedom.

•72 cm long, 12 cm wide (at the shoulder),

13 cm high (at the shoulder),

•Minimum Exit Speed for Leap: 1.60 m s−1

(2.29 BL/s).

•Maximum swimming speed: 2.85 BL s−1

•Achieved speed: 2.05 ±0.03 m s−1at

k=0.25 Nm/◦,f=4.65 Hz, correspond-

ing to 0.61 BL Hz−1

•Weight: 4.7 kg.;

Lau et al [63]•Wire-driven robotic fish that is

bioinspired and can move both

forward and upward by swaying its

caudal fin.

•Derivation of a sway motion model for

the robot shark.

•Length: 0.6 m

•Maximum velocity: 129.6 mm s−1

(0.22 BL s−1)

•Bending amplitude: 35◦

•Flapping frequency: 2 Hz

•Maximum body pitch angle: 17.3◦

van den Berg et al

[64]

•Uses a DC motor’s continuous

rotation, which differs from a servo

motor’s typical back-and-forth action

and allows for higher frequencies and a

more sinusoidal waveform.

•Figure 23 shows OpenFish’s

mechanical design and movement

under experimental conditions.

•Velocity: 0.85 m s−1,

(Continued.)

Bioinspir. Biomim. 19 (2024) 031002 A Prakash et al

Table 3. (Continued.)

Investigator Description System Parameters

Mitin et al [65]•Propulsion system that combines an

elastic cord with a tail fin attached to it.

The tail fin is linked to a servomotor

with movable thrusts, simulating

muscle contraction.

—

CABRERA, Fausto

Ret al [66]

•Prototype of a robot replicating a white

shark’s locomotion using a caudal fin.

•The mathematical model accurately

represents the movement of the real

robot in short distances, making it

efficient.

—

Yu et al [67]•Improvements in the mechanical

structure of the robotic fish, which

endows it with large thrust and a wider

range of movements.

•Motion analysis is used to provide

information and parameters related to

the swimming state of the robotic fish

•Velocity: 1.14 m s−1

Figure 21. Thunniform robotic fish design [48]. Reproduced from [48]. CC BY 3.0.

Figure 22. (a) Fish robot UC-Ika 1 and (b) its digital design [54]. Reproduced from [54]. CC BY 4.0.

Tunable stiffness plays a pivotal role in the

performance of these fish-like robots, as real fish

optimize their swimming efficiency by adjusting

stiffness based on their speed. A study on stiff-

ness tuning reveals that it can lead to a doub-

ling of swimming efficiency, particularly at frequen-

cies resembling those of tuna, ranging from 0 to

6 hertz. The study’s model, supported by experi-

mental data, emphasizes the potential of stiffness

tuning in enhancing the performance of fish-like

robots, offering significant energy savings of 16%,

41%, and 55% when compared to fixed configura-

tions, an important metric in underwater robotics

development [49].

Bioinspir. Biomim. 19 (2024) 031002 A Prakash et al

Figure 23. Open fish design in experimental setup. Reproduced with permission from [64]. CC BY-NC-ND 4.0.

To address the problems with current stabilization

networks, especially in underwater conditions, a bio-

mimetic robotic hammerhead shark was developed

with an autonomous vision-based navigation and sta-

bility augmentation control system. By combining

vision-based navigation and cephalopodic stabiliza-

tion control, the suggested framework was success-

fully able to lower jitter by 26.02%, improving the

safety of autonomous navigation—a crucial measure

in underwater exploration. In addition to giving bio-

mimetic robotic fish greater intelligence and the abil-

ity to maneuver through a variety of obstacles and

situations, the research advances underwater percep-

tion and operation, a crucial area of study for the field

of underwater robotics. Future study objectives in this

particular field will be highlighted by the difficulties

in handling extreme vibrations and the necessity of

more research into computer efficiency [50].

A study was done to address the challenge of con-

trolling biomimetic underwater robots in nonlinear

fluid flow fields focusing on the attitude-holding task

of a robotic fish, using a learning-based approach

with a deep neural network mapping raw sensor data

to control signals. A data-driven simulation envir-

onment was created from an experimental dataset,

incorporating dynamic, sensor, and control mod-

ules. A DRL algorithm was trained on simulated

data to develop a control policy successfully trans-

ferred to a physical robotic fish and the experimental

results using a test robotic fish confirm the DRL’s

effectiveness [51].

3.3. System based on carangiform and

subcarangiform

Despite their great similarity, subcarangiform and

carangiform locomotion methods differ in the ini-

tiation point along the body; the former uses more

head movement and undulation near the head, while

the latter uses one-third of the posterior body for

undulation. These species, which exhibit a range of

sizes and weights, speed from 0.1 to 0.85 m s−1, and

endurance from 35 min to 4.5 h, are investigated for

their usefulness as power thrust fish [10]. A detailed

description of a few robotic systems based on sub-

carangiform & carangiform is given in table 4.

An illustrative example of biomimetic design for

underwater propulsion is presented in a micro-robot

fish with an embedded shape memory alloy (SMA)

wire-actuated flexible biomimetic fin as shown in

figure 24. Inspired by the musculature of squid and

cuttlefish fins, the design focuses on achieving flexible

bending akin to fish and cuttlefish propulsion modes.

The biomimetic fin, activated by SMA wires through

resistance heating and cooling, exhibits flexible bend-

ing to both sides, resembling the caudal fin com-

ponents of fish [34]. The integration of this biomi-

metic fin into a micro-robot fish design offers propul-

sion through flexible bending movements, showcas-

ing efficient biomimicry in underwater locomotion.

In the context of carangiform robotic fish, a study

explores dynamic simulation models and optimiza-

tion techniques for achieving precise values in a flex-

ible multi-joint propulsion mechanism resembling

an artificial spine system. The optimized parameters

imitate the body wave of a real fish’s motion, offering

flexible maneuverability, high propulsion efficiency,

and reduced noise. Implementations of route plan-

ning scenarios showcase the versatility of carangi-

form robotic fish in free swimming and destination-

reaching modes, where decision-making processes

determine turning directions and seek the shortest

route to the target, highlighting their potential in

autonomous underwater exploration [68].

3.4. System based on labriform

Labriform-based underwater systems have a unique

way of swimming characterized by the use of pec-

toral fin locomotion, designed for more agile swim-

ming at slower speeds. Labriform swimmers have

poor endurance when using their pectoral fins alone.

Bioinspir. Biomim. 19 (2024) 031002 A Prakash et al

Table 4. Examples of systems based on subcarangiform & carangiform.

Investigator Description System Parameters

Muralidharan

and Palani [69]

•Employs shape memory alloy (SMA)

subcarangiform propulsion

•Utilizes SMA spring actuators and a spring-based

propulsion mechanism, avoiding conventional

motors.

•Lightweight 3D printed components contribute to

the scalability and cost-effectiveness of the design.

•Length (without battery and electronic

control board): 25 cm

•Steady swimming velocity: 24.5 mm s−1

•Maximum force during fin oscillation:

0.39 N

•Weight: 416 gm

Wang et al [34]•Fish microrobots driven by a biomimetic fin

modeled after the muscular systems of squids and

cuttlefish.

•The fin is operated using shape memory alloy

(SMA) wires with an elastic substrate and

transverse SMA wires resembling muscles.

Mechanism for storing and exchanging elastic

energy to increase efficiency.

•Figure 25 shows a prototype of the Micro-robot

fish

•146 mm (length), 17 mm (width),

34 mm (height)

•Maximum swimming speed:

112 mm s−1

•Minimum turning radius: 136 mm

•Weight: 30 g

Gao and Techet

[70]

•Constructed using FDM rapid prototyping with a

spring-loaded wing mechanism.

•Analysis of lift forces demonstrates the

importance of keeping aerial-aquatic vehicles

small for maintaining altitude.

•Body length 0.25 m, volume 145 cm3,

•Using lighthill’s theory at 10 m s−1

requires a specific power output of

4150 W kg−1.

•Weight ∼145 g.

Hu et al [71]•The difficulties in replicating fish-like swim

patterns with a robotic system.

•Highlighting the significance of comprehending

fish biomechanics for effective use.

•Length: 0.52 m

•Speed: 0.6 m s−1

Yu et al [72] .

•Altering joint deflections to tune orientation and

varying joint oscillation frequency modulation to

change swimming speed.

•Utilizing a hybrid method and PID algorithm for

online speed control and a fuzzy logic controller

for orientation control.

•Utilizing an overhead vision system for real-time

visual feedback to enhance control effectiveness.

•Length: 0.4 m

•Speed: 0.32 m s−1

Yu et al [73]•Self-propelled robotic fish with multiple control

surfaces for enhanced aquatic maneuverability.

Figure 26 shows the assembled prototype and the

mechanical design of its various motions.

•Utilizing a central pattern generator (CPG)-based

method for motion control, with real-time

adjustments based on monocular underwater

vision feedback.

•Length: 0.68 m

•Speed: 0.71 m s−1

Ozmen Koca

et al [68]

•A flexible multi-joint propulsion mechanism that

resembles an artificial spine system is

incorporated using a dynamic simulation model

for the carangiform robotic fish.

•Achieving precise values for flapping frequency

and speed by optimizing parameters using a new

search method.

•Figure 27 shows the designed model and the

physical model during swimming.

•Length: 0.52 m

•Speed: 0.2 m s−1

Yu et al [74]•AmphiRobot-II is an amphibian-inspired design

that can move both on land and in water thanks to

a hybrid propelling system.

•Combines fin, wheel, and propeller movements to

enable flexible wheel-based crawling on land and

fish or dolphin-like swimming underwater.

•Proposes a body deformation steering approach to

achieve efficient turning on land with a minimal

turning radius.

•Length: 0.7 m

•Speed: 0.45 m s−1

(Continued.)

Bioinspir. Biomim. 19 (2024) 031002 A Prakash et al

Table 4. (Continued.)

Liu and

Huosheng [75]

•Identifies key challenges, especially in achieving

C-shape turning (CST), for replicating fish-like

swimming in robots.

•Introduces a novel 4-DOF tail kinematics model

to replicate CST behavior in a robotic fish.

•Length: 0.8 m

Ichikizaki and

Yamamoto [76]

•Utilizing rapid prototyping with stereolithography

and 3D laser measurements to create a realistic

outer shell for the robotic fish, minimizing fluid

resistance.

•Equipping the carp robot with servomotors, a

drive shaft, a joint in the waist area, and actuators

for ballast pumps.

•Length: 0.9 m

•Speed: 0.41 m s−1

Clapham and

Hu et al [77]

•Attaining total body synchronization in the

swimming motion of iSplash-I, including frontal,

mid-frontal, and rearward displacements.

•Introducing a novel mechanical drive system

operating in two swimming patterns to enhance

propulsion.

•Length: 0.251 m

•Speed: 0.85 m s−1

Katzschmann

et al [78]

•Uses hydraulic power for soft body deformation,

offering improved control and precision,

potentially surpassing pneumatic systems.

•Introduces a water-based transmission fluid for

reduced friction and energy consumption in the

aquatic environment.

•Smartly incorporates pectoral fins as dive planes,

suggesting dynamic maneuverability, including

diving, expanding the robotic fish’s range of

applications.

•Length: 0.45 m

•Speed: 0.1 m s−1

Zhong et al [79]•Transfers real fish actuation to biomimetic robots

using the lighthill model and fish muscle function.

•Suggests compliant parts contribute to robot fish

propulsion, reducing motor reliance.

•Offers a guide for building efficient robot fish with

fewer motors adaptable to various swimming

modes.

—

Marchese et al

[80]

•fish-mimicking soft-bodied robot with embedded

actuators for rapid, continuum-body motion.

•Utilizes a novel fluidic actuation system

integrating power, actuation, processing, and

control subsystems.

•A key feature is the ability to perform rapid escape

responses, resembling biological fish behavior

during escape maneuvers.

—

Zhong et al [81]•inspired by underwater gliders. Combines critical

modules of a glider, such as the buoyancy

regulation module (BRM) and attitude regulation

module (ARM), with the main structure of an

underactuated robotic fish.

•Uses a 1 R pseudo rigid-body model (PRBM) to

reflect the bending of the compliant tail during

swimming.

•Capable of two motion modes: swimming and

gliding. Utilizes an open and modular design

method

•Dimension:

1566 mm ×1080 mm ×530 mm

•Fastest swimming speed of 0.4 m s−1

(0.256 BL/s)

•Weighs 29.44 kg

•the fastest turning speed: 0.242 rad s−1

•minimum turning radius: 0.35 BL

However, they may periodically incorporate the use

of the caudal fin, especially during burst swimming

or when their pectoral muscles are fatigued. They

structurally have pectoral fins with stiff fin rays inside

the fin membrane, which allows for efficient swim-

ming. There are two different ways to swim in this

class: flapping, which uses the pectoral fin’s leading

edge for both upstroke and downstroke and is good

for close-quarters swimming, and rowing, which is

used for quicker speeds. Labriform AUVs typically

range in mass from 0.33 to 2.5 kg, achieving a speed

of 0.53 m s−1, with good turning and depth con-

trol capabilities [10]. A detailed description of a few

robotic systems based on labriform is given in table 5.

Bioinspir. Biomim. 19 (2024) 031002 A Prakash et al

Figure 24. The structure and bending of SMA

wire-actuated flexible biomimetic fin. Reprinted from [34],

Figure 25. Micro-robot fish prototype. Reprinted from

[34], Copyright (2008), with permission from Elsevier.

Several important metrics were examined in a

study that focused on improving the performance of

labriform swimming robots based on the morpho-

logical characteristics of pectoral fins. It was found

that 50◦was the ideal fin beginning angle for the

power stroke phase, suggesting a tactical placement

for effective forward propulsion. The amplitude of the

pectoral fins’ rotational displacement and the ensuing

forward thrust were shown to be linearly correlated,

offering important new information about the link

between fin movement and swimming efficiency. The

study also showed that the 50◦angle produced the

most thrust, highlighting the significance of this par-

ticular orientation for optimizing propulsive force.

Concurrently, the drag was found to be at its lowest

point at −50◦, leading to a complete angular displace-

ment of 100◦. This comprehensive exploration of dif-

ferent angles demonstrates a nuanced understand-

ing of how the orientation of the pectoral fins signi-

ficantly influences both thrust and drag forces dur-

ing swimming. Additionally, the relationship between

frequency and speed was studied, indicating a lin-

ear correlation at low frequencies. This insight into

the interplay between frequency and speed is cru-

cial for optimizing the swimming performance of

labriform AUVs, providing a foundation for fur-

ther advancements in biomimetic underwater robotic

design [82].

3.5. System based on ostraciiform and

tetraodontiform

Fish belonging to the ostraciiform classification, such

as boxfish and cowfish, use a special combination of

pectoral and dorsal/anal fin oscillations to control

movement, especially in tight quarters and reef hab-

itats. The body design of this class is a sturdy box that

allows for excellent maneuverability in small spaces

and well-tuned fin oscillations. Because of their box-

like body structure and rather poor fin actuation

for reaching high speeds, ostraciiform fishes, despite

their agility, have speed limits [10]. Even though these

fishes hardly ever employ their little caudal fin for

burst propulsion, their speed is still restricted when

they do. A detailed description of a few robotic sys-

tems based on ostraciiform is given in table 6.

A study was conducted looking into the pectoral

forms and locomotor tactics for creating a robotic

fish modeled after the ostraciiform morphology—

specifically, the boxfish (ostracion melagris)—in

the context of bio-inspired robotics. Three micro

servos—two on the pectoral fins and one on the

caudal fin—actuate a robotic fish prototype used in

the study. Different pectoral fin designs, including

rectangular, triangular, and quarter-ellipse shapes,

were examined while retaining the same aspect ratio

(AR) utilizing fused deposition modeling (FDM) or

3D printing. Important findings were obtained from

the comparison study of the robot fish’s pectoral

fin forms (triangle, rectangle, and quarter-ellipse)

to hydrodynamics, thrust, and speed. The pectoral

fin, shaped like a quarter-ellipse, produced the max-

imum thrust (0.06 N), outperforming the triangle

(0.04 N) and rectangle (0.038 N). Combining a delta-

shaped caudal fin with a quarter-elliptical pectoral fin

allowed greater maneuverability and the fastest speed

(0.106 m s−1). For the robot fish in this work, the

quarter-ellipse pectoral fin was the most efficient in

thrust, speed, and maneuverability, particularly when

paired with a delta-shaped caudal fin [91].

Bioinspir. Biomim. 19 (2024) 031002 A Prakash et al

Figure 26. Embedded vision guided robotic fish with multiple control surfaces-(a)The assembled prototype (b)its various

motions [73]. Reproduced under the terms of the Creative Common License Reproduced from [73]. CC BY 3.0.

Figure 27. (a)The designed model and (b) the physical model during swimming. Reprinted from [68], Copyright (2016), with

permission from Elsevier.

Another study built a morphable tail for an adapt-

able robotic fish modeled after the boxfish, butter-

fish, and big-eye. This allowed the fish to switch

between swimming modes and operate in quickly

changing underwater conditions. As illustrated in

figure 28, the tail comprises three articulated seg-

ments controlled by servomotors and a passive caudal

fin. Analytical functions represented fin deforma-

tions inspired by biological processes, and Simscape

was used to optimize the linkage lengths. Thrust

and forward speed were evaluated experimentally

throughout a range of lateral oscillation amplitudes

and undulation frequencies using a laser sensor and

load cell. The findings showed that various loco-

motion modes impact swimming performance and

that synchronizing locomotion parameters signific-

antly impacts overall thrust. The experiments sup-

ported lighthill’s notion of propulsion for elongated

entities. The artificial fishtail showed flexibility in

responding to shifting navigational needs by tuning

for various fish movement patterns [92].

Similar to the previous study, another study

also addresses the challenge faced by scientists and

engineers in modeling and controlling various fish

locomotion motions, focusing on cruising-straight,

cruising-turn, and fast turns. It also proposes a ver-

satile kinematic model synthesized using a nonlinear

oscillator and a traveling wave equation, defined by

four parameters that accommodate different swim-

ming motions. The model was validated through a

multi-joint robotic fish [93].

Large size is a defining characteristic of the tet-

raodontiform classification, which includes species

like the ocean sunfish, sharptail mola, and slender

sunfish. Despite their impressive size, species in this

class are considered to possess the least efficient

design among the fishes mentioned, utilizing two

large paddle fins—one dorsal and one anal—for

locomotion [11] and therefore it is notably underrep-

resented in the literature and lacks any robotic sys-

tems exclusively based on it.

3.6. System based on rajiforms

The field of aquatic robotics within the rajiform

category seeks to replicate the movement patterns

of marine animals, particularly those that resemble

Skates and Rays, which have extraordinarily flexible

bodies made of cartilage. Fish that move in a raji-

form fashion is characterized by their amazing undu-

lations in their enlarged pectoral fins. Their skeletons

are anatomically defined by fin ribs that grow from

the body into the pectoral fin. A detailed description

of a few robotic systems based on rajiform is given in

table 7.

The study by Yang et al [100] provides a fun-

damental contribution to our understanding of raji-

form locomotion in marine robotics. The goal of

the paper was to create the robotic fish known as

‘cownose ray-I,’ which has a body, two lateral fins,

and no tail. The study explores the complex interac-

tions between three important kinematic factors that

affect the robotic fish’s forward speed. The biggest

Bioinspir. Biomim. 19 (2024) 031002 A Prakash et al

Table 5. Examples of systems based on labriform.

Investigator Description System Parameters

Sitorus et al [83]•Highlighting screw-type propellers’

inefficiencies and maneuverability limitations

in underwater vehicles, attributed to issues like

vortices and sudden thrust forces.

•Length: 0.375 m

•Speed: 0.0351 m s−1

Naser and

Rashid [82]

•swimming robot with labriform swimming

mode, emphasizing effective steering.

•Utilizes a concave-shaped pectoral fin system

(1-DOF) with differential driving by adjusting

left and right fin velocities.

•Incorporates a mass sliding mechanism to alter

the center of gravity, allowing control of pitch

angle for underwater depth management. Two

independent actuators control the center of

gravity and pectoral fin actuation.

•Dimension:0.185 m ×0.06 m ×0.065 m

•Speed: 0.074 m s−1(0.40 BL/s)

•Body mass:0.650 Kg

Connor et al

[84]

•Modeled after the Scomber Scombrus fish,

emphasizing highly actuated pectoral fins and

labriform motion.

•Emulates the mackerel’s caudal tail for

propulsion, specifically focusing on labriform

motion over Carangiform locomotion.

•Rigid fins offer better lift but higher drag, while

bio-inspired fins, with lower lift, have

comparable drag. Challenges noted in

independent bio-inspired fin lift generation.

—

Behbahani and

Tan [85]

•Developed a flexible joint mechanism for

robotic fish pectoral fins, emphasizing aquatic

animal motion, rowing mode, and passive

feathering for efficient swimming.

•Utilizes blade element theory to create a

dynamic model, representing the joint as a

paired torsion spring and damper, enhancing

understanding of hydrodynamic forces.

•Speed: 0.53 m s−1

Zhang et al [86]•Create a robotic fish that mimics the

movements of insects and fish fins to improve

its underwater agility.

•Uses its two caudal fins at the tail for main

propulsion. It produces opposing lateral forces

by flapping its fins in opposite directions, which

makes for steady, high-performance swimming.

•Incorporates two pectoral fins mimicking

insect wings on each side to enhance

maneuverability in the vertical plane.

•Length: 0.44 m

•Speed: 0.53 m s−1

Costa et al [87]•a novel transmission system that transforms a

rotary actuator’s constant angular velocity into

the pitching-yawing rotation of fish pectoral

fins.

•The system employs biomimetic thrusters that

utilize a drag-based momentum transfer

mechanism from labriform swimmers to

generate the necessary steering torque.

•The inherent synchronization of the system

provided by the mechanism’s kinematics. The

proposed solution reduces the control system’s

effort, inertia, and encumbrance compared to a

system with multiple servomotors.

•For 3 Hz motor frequency and traveling

at a speed of 1Bl s−1, a turnabout radius

of 4.41 BL was achieved

Naser and

Rashid [88]

•Investigates the impact of pectoral fin geometry

on labriform-mode swimming mechanisms.

•Utilizes Computational Fluid Dynamics (CFD)

to simulate different fin shapes and angles,

starting with simulations before constructing a

robot prototype.

•Highest swimming velocity attained at

a 3:1 power-to-recovery ratio

(Continued.)

Bioinspir. Biomim. 19 (2024) 031002 A Prakash et al

Table 5. (Continued.)

Naser and

Rashid [89]

•Developed a swimming robot with efficient

steering in labriform mode.

•Utilizes a 1-DOF mechanism with two

concave-shaped pectoral fins, applying the

differential drive principle by varying fin

velocities for steering control and a mass

sliding mechanism

•Functions as a glider with independent

actuators for the center of gravity and pectoral

fin control, reported as one of the smallest

swimming-gliding robots in the literature.

•The robot achieved a minimum turning

radius of about 0.40 of its body length.

Li et al [90]•Inspired by the structure of a deep-sea snailfish.

•The robot is designed to be untethered,

eliminating the need for a rigid vessel.

•Careful design of the dielectric elastomer

material for the robot’s flapping fins allows

successful actuation.

•The work highlights the potential of designing

soft, lightweight devices for use in extreme

conditions.

•22 cm long (body length 11.5 cm, tail

length 10.5 cm) and 28 cm in wingspan.

•0.45 BL s−1

•Field Test in Mariana Trench, a depth of

10 900 m.

•Swimming in the South China Sea, a

depth of 3224 m.

amplitude, wave number, and phase discrepancies

between neighboring fin rays affect speed, whereas

the frequency of the fin beats has a significant effect

on the fish’s pectoral oscillatory propulsion. The

structural limitations of the fin rays and the com-

plex operation of the servo-motor were blamed for

the departure from accepted conclusions. The robotic

fish’s steering capability is notably achieved by intro-

ducing an asymmetric phase difference between the

left and right pectoral fins, emphasizing the signific-

ance of the phase difference in the propulsive wave for

effective steering. Acknowledging certain limitations

in the model design, evidenced by the slower for-

ward velocity of the robotic fish and existing defects,

the study propounds future research directions. It

advocates for an exploration of frequency analysis and

undulation studies within the context of a loaded flex-

ible beam model.

3.7. System based on amiiform, gymnotiform, and

balistiform

The unique use of an oscillating dorsal fin, as seen

in species such as the African aba aba and the bow-

fin, characterizes amiiform movement. When the sys-

tem is in this mode, it moves forward by undulat-

ing its long-base dorsal fin and keeping its body pos-

ture upright. Extensor and flexor muscles on both

sides contract to produce the intricate undulations of

the dorsal fin, which is made up of a flexible mem-

brane and tightly packed fin ribs or rays. This allows

for precise turning and multidirectional movement.

Amiiform locomotion could be faster, but it exhibits

good agility.

The opposite type of amiiform locomotion,

known as gymnotiform locomotion, is best repres-

ented by animals such as gymnotus carapo, whose

propulsion is provided by the undulations of a long

anal fin. This mode allows the body to bend and move

forward and backward, allowing gymnotiform spe-

cies to change their direction by varying the angle of

their fin and the waves of undulation. Pectoral fins

improve overall maneuverability by acting as control

surfaces for pitch and roll.

Balistiform locomotion, observed in fishes like

the picasso triggerfish, involves synchronized move-

ments of both dorsal and anal fins, collectively

referred to as MPF undulation locomotion. A detailed

description of a few robotic systems based on

amiiform, gymnotiform & balistiform is given in

table 8.

The thrust performance of a biomimetic under-

water undulating fin-driven propulsor inspired by

gymnotiform fish species is examined in the study

‘Experimental Study on Kinematic Parameter and

Undulating Pattern Influencing Thrust Performance

of Biomimetic Underwater Undulating Driven

Propulsor’, which was published in 2015. The robotic

fin uses a sinusoidal wave to propel itself and it is

made up of 16 connected fin rays with a rubber

membrane, simulating the swaying action of gym-

notiform fish. The investigation explores kinematic

parameters such as oscillation frequency, wavelength,

and amplitude envelope. Using varying amplitudes

(40 mm, 60 mm, 80 mm, 100 mm) and frequen-

cies (0.5 Hz, 0.75 Hz, 1 Hz, 1.25 Hz), four cyclic

motion patterns are investigated. According to exper-

imental results, oscillation frequency and amplitude

are almost directly proportional to static thrust force.

The impact of parameters on thrust force varies across

patterns, emphasizing the complexity of undulating

fin propulsion [115].

3.8. System based on jet propulsion

One common and effective form of propulsion for

AUVs and robotic systems is pulse jet propulsion.

This system takes the form of four different classes,

Bioinspir. Biomim. 19 (2024) 031002 A Prakash et al

Table 6. Examples of systems based on ostraciiform.

Investigator Description System Parameters

Lachat et al [94]•A versatile robot designed for swimming and

crawling, inspired by the boxfish, featuring

three fins and driven by DC motors.

•Utilizes a central pattern generator (CPG)

implemented as coupled nonlinear oscillators,

mimicking biological rhythmic activity for

controlling various locomotor behaviors.

•Length: 0.25 m

•Speed: 0.37 m s−1

Chen et al [95]•Developed a 6-DOF bionic boxfish-like robot

with 2-DOF pectoral fins and a caudal fin for

underwater swimming.

•CPG parameters on propulsion performance,

revealing the importance of amplitude and

frequency, with optimal propulsion observed at

1 Hz. Introduced time-asymmetric flapping

characteristic equation

•Figure 29 shows the mechatronics of the

boxfish robot and its various swimming

patterns.

•Length: 325 mm, width: 150 mm,

height: 140 mm

Qiu [96]•biomimetic robotic boxfish with 2-DOF

pectoral fins and a 1-DOF caudal fin for

lift-based and drag-based locomotion.

•Utilizes the Lattice–Boltzmann Method for

numerical calculations, exploring the effects of

motion parameters on hydrodynamics.

•Length 500 mm, width 300 mm, height

150 mm

•Straight swimming speed: 0.38 BL/s

•In-situ turning speed: 20◦/s.

•Weight (after counterweighting): 3.7 kg

Kodati [97]•Inspired by the boxfish, the Ostraciiform’s boxy

shape gives its trajectories a self-correcting

mechanism.

•Two 2-DOF pectoral fins and one 1-DOF tail

fin are used for locomotion.

•Two distinct halves are used in 3D prototyping

to create the robot’s final body design.

•Length: 0.15 m

•Speed: 0.0411 m s−1

Wang et al [98]•Utilizes a Central Pattern Generator (CPG)

model for motion control of an ostraciiform

fish robot.

•A boxfish robot with two motorized pectoral

fins and one motorized caudal fin is used to

demonstrate the technique.

•Integrates a vision sensor into the CPG-based

control, allowing real-time target tracking

behavior.

•Length: 0.35 m

•Speed: 0.38 m s−1

Wang and Xie

[99]

•Designing a Central Pattern Generator

(CPG)–based locomotion controller for a

boxfish-like robot to achieve flexible swimming

patterns and precise attitude control.

•Comprises two layers—a lower layer with

coupled linear oscillators and an upper layer

transforming locomotion stimuli into control

parameters for oscillators.

•Figure 30 shows the actual prototype of a

boxfish-like robot and the experimental pool

with a 3-axis force measurement system.

•Length: 0.33 m

•Speed: 0.347 m s−1

Mainong et al

[91]

•Design inspired by the boxfish in terms of

morphology.

•Three micro servos actuate the robotic fish

prototype, with two on the pectoral fins and

one on the caudal fin.

•Tested various pectoral fin shapes (rectangular,

triangular, quarter-ellipse) with the same

aspect ratio (AR).

•Length: 0.165 m

•Speed: 0.106 m s−1

Bioinspir. Biomim. 19 (2024) 031002 A Prakash et al

Figure 28. (a) CAD diagram of the robotic fish and (b) simulation of its undulatory deformation. Reprinted from [92], Copyright

(2022), with permission from Elsevier.

Figure 29. (a) Mechatronics of the boxfish robot and (b) its various swimming patterns [99]. Reproduced under the terms of the

Creative Common License. Reproduced from [99]. CC BY 3.0.

Figure 30. (a) Actual prototype of boxfish-like robot [95]. Reproduced from [95]. CC BY 4.0.

each of which represents a separate underwater mov-

ing strategy.

In the first class, bell constriction, creatures such

as jellyfish are representative. To propel itself forward,

the bell-shaped body contracts and releases water.

AUV designs have advanced as a result of this mech-

anism’s sophisticated simplicity.

Cephalopods, particularly octopuses, showcase

mantle constriction as their method for generating

thrust and achieving controlled movement. With the

versatile use of arms in both propulsion and manip-

ulation positions, cephalopods are potential assets in

diverse underwater tasks, including structural inspec-

tion, search and rescue operations, and ecological

monitoring [31].

Mollusks, typified by scallops, employ shell com-

pression as a means of achieving swift and agile jet

propulsion. The rapid closure of their shells facilitates

efficient movement. The adaptability of this mech-

anism offers insights for biomimetic applications in

robotics.

The fourth class involves a combination of jet

and fin undulation, observed in organisms like

squid and cuttlefish. This hybrid approach seam-

lessly integrates rapid jetting for bursts of speed

with fin undulation for precise and controlled move-

ments. The synergy between these two propul-

sion methods enables these organisms to navig-

ate their underwater environments with exceptional

agility.

Bioinspir. Biomim. 19 (2024) 031002 A Prakash et al

Table 7. Examples of systems based on rajiforms.

Investigator Description System Parameters

Zeng [101]•a robotic devil fish based on the MPF

movement pattern.

•Submerged underwater at a depth of 0.5 m for

72 h, the robotic devil fish continues to

function well, highlighting durability and

reliability.

•Recommends synchronizing the vibration

periods of the left and right fins for effective

turning without adjusting individual fin

vibration periods.

•Overall

Gliva et al [102]•The underwater robot is inspired by the

locomotion of rays and cuttlefish, featuring

lateral undulatory fins.

•Tube-shaped hull with laterally-mounted fin

propulsors. Each fin mechanism has three

individually actuated fin rays interconnected by

an elastic membrane.

•Overall length: 24 cm

•Overall width: 57 cm

•Speed: 20 cm s−1

Wang et al [103]•The micro biomimetic manta ray robot fish is

actuated by shape memory alloy (SMA) wire.

•Fish-like swimming is highlighted as superior

in terms of noise and maneuverability

compared to traditional motor actuators.

•Length:243 mm

•Width:220 mm

•Height: 66 mm

•Speed: 57 mm s−1

•Maximum amplitude: 40 mm

•Weight: 354 g

Krishnamurthy

et al [104]

•Taking inspiration from the Pacific electric ray,

the RayBot

•RayBot modeling and simulation are done

using a 6DOF impulse-based multi-body

technique. Nine bodies and nine hinge joints

make up the model, which can be divided into

separate pelvic fins, centrally and peripherally

actuated planes, and a tail with extra hinge

joints.

•Dimensions: 75 cm long x 50 cm wide,

maximal height 15 cm at the tip of the

caudal fin, disc height 9 cm.

•Top Speed: over 10 cm s−1

Gao et al [105]•Develop a fish robot mimicking Manta Rays’

swimming motion with a focus on flexible

oscillatory pectoral fins.

•pectoral fins, in oscillatory motion, generated

significant thrust, influenced by frequency and

amplitude.

•Length (L)=0.5 m, wing span =0.6 m

•Speed: 0.7 m s−1

Niu et al [106]•Robotic fish mimicking cownose ray with a

dorsoventrally flattened body for pitch stability.

•Utilizes a fuzzy logic method for automatic

depth control amid complex hydrodynamics

and environmental uncertainties.

•Equipped with an AHRS (Attitude and

Heading Reference System) for attitude angle

measurement.

•Length (L)=0.56 m, height

(H)=0.12 m, width (W)=0.643 m

•Speed: 0.32 m s−1

Chen et al [107]•Mimic manta ray swimming behavior using

artificial muscles.

•Robotic manta ray with two artificial pectoral

fins for undulatory flapping motions as shown

in figure 31

•Pectoral fins consist of IPMC muscle on the

leading edge and passive polydimethylsiloxane

membrane on the trailing edge.

•Length: 0.11 m

•Speed: 0.007 m s−1

•Fin attains a 40◦twist angle and a 100%

tip deflection.

y Alvarado et al

[108]

•Use flexible bodies with specific material

property distributions for desired locomotion

kinematics with minimal actuation.

•Length: 0.225 m

•Speed: 0.007 m s−1

Cai et al [109]•Flexible pectoral fins for fish-like propulsion.

Soft body with pneumatic muscles as driving

sources. Two ribs with distributed flexibility in

the propulsive mechanism.

•Length (L)=0.33, wing span =0.55

•Speed: 0.2 m s−1

(Continued.)

Bioinspir. Biomim. 19 (2024) 031002 A Prakash et al

Table 7. (Continued.)

Chew et al [110]•Pectoral fins designed to mimic manta ray’s fins

using passive flexibility. Experimentation

involves testing various fin materials and

adjusting flapping parameters.

•Length (L)=0.28 m, Wing

Span =0.58 m

•Speed: 0.45 m s−1

Zhou and Low

et al [111]

•Development of RoMan-II, a biologically

inspired underwater vehicle mimicking manta

ray locomotion using CPG.

•Length: 1 m

•Speed: 0.4 m s−1

Park et al [112]•Created to mimic batoid fish, featuring

tissue-engineered ray and optogenetics for

guidance and navigation.

•Rat cardiomyocytes are patterned on an

elastomeric body enclosing a microfabricated

gold skeleton. This approach replicates the

fish’s morphology at a smaller scale.

•Enables phototactic guidance, steering, and

turning, responding to light cues for

navigation.

•Diameter =0.01 m

•Speed: 0.001 m s−1

Yurugi et al

[113]

•The inspiration for the robot’s design comes

from the swimming principles, morphologies,

and softness of aquatic animals, specifically the

rajiform.

•The researchers test the hypothesis that

anisotropy in stiffness, achieved through

cartilage structures encapsulated in soft tissue,

contributes to efficient swimming in

underwater robots.

•They create a stingray robot using

silicone-based cartilage and soft tissue,

mirroring the anatomical features of the

rajiform fins.

•Flat cartilaginous: maximum speed

[mm s−1]:13.1

•Sharp cartilaginous:- maximum speed

[mm s−1]: 9.8

•Homogeneous soft (only soft tissue):-

maximum speed [mm s−1]: 7.7

•Homogeneous stiff (mixture of cartil-

age and soft tissue materials):- max-

imum speed [mm s−1]: 11.2

Huang et al

[114]

•A new design featuring two pectoral fins and

two vertical propellers for attitude adjustment,

aiming to enhance overall motion performance.

•The results suggest that the propellers

effectively adjust the robot’s posture and

enhance its stability.

•A 6 degrees of freedom (6DOF) motion

equation is established to describe the robot’s

movements and is validated through CFD

simulations.

•Average speeds after 16 s:

MATLAB Calculation: 0.223 m s−1

FLUENT Simulation: 0.226 m s−1

Figure 31. Robotic manta ray powered by ionic polymer–metal composite artificial muscles [107]. Reproduced from [107].

CC BY 4.0.

Bioinspir. Biomim. 19 (2024) 031002 A Prakash et al

Table 8. Examples of systems based on amiiform, gymnotiform & balistiform.

Investigator Description System Parameters

Low [116]•Examination of single fin segment geometry and

derivation of fin workspace, forming the design of

NKF-II robot.

•Comprises buoyancy tank, motor compartment,

and undulating fin modules mimicking black

ghost knifefish swimming gaits.

•Figure 32 shows the fully assembled Nanyang

knifefish II (NKF-II)

•Length 70 cm, Height 42 cm, Width

13.5 cm

•Total of seven segments,

•Mass: 11.62 kg

•Color: Black

•Material: mostly acrylic and delrin

Hu et al [117]•Developed motor-driven fin actuator as shown in

figure 33, inspired by gymnarchus niloticus,

mimicking the undulatory movement of the fish’s

dorsal fin.

•Introduction of a ruled surface-based model

describing undulation characteristics in

gymnarchus niloticus as a guide for biomimetic

mechanism design.

•Speed: 0.4 m s−1

Siahmansouri

et al [118]

•Biomimicry from knife fish, specifically focusing

on the anal fin for propulsion, aiming for

efficiency and agility in underwater robots.

•Robotic fish equipped with an undulating fin

producing a sinusoidal wave for mimicking

natural swimming motion.

•Speed: 0.24 m s−1

Curet et al [119]•Centered on South American electric knifefish

and their undulatory swimming as a neurobiology

model.

•Identified an optimal parameter space where the

robotic knifefish achieved the best performance.

•Optimal parameters resembled those observed in

the black ghost knifefish, affirming biomimicry.

•Length: 0.326 m

•Speed: 0.18 m s

Liu et al [120]•Commences with the development of a robotic

fish for practical experiments, validating

computational hydrodynamic models.

•Introduces a comprehensive computational model

to analyze the hydrodynamics of the undulating

fin, focusing on wave motion effects.

•Length: 1.2 m

•Speed: 0.25 m s−1

Figure 32. Fully assembled Nanyang knifefish II (NKF-II). Reprinted from [116], Copyright (2009), with permission from

Elsevier.

Figure 33. RoboGnilos, with nine-fin rays connected by a membrane surface. Reprinted from [117], Copyright (2009), with

permission from Elsevier.

Bioinspir. Biomim. 19 (2024) 031002 A Prakash et al

Several key features contribute to the enhanced

efficiency of pulsatile jet propulsion. These include

the presence of flexible propulsive structures, allow-

ing for dynamic adjustments during movement.

Variable jet orifice diameter adds a layer of adapt-

ability to the propulsion process. Vortex interac-

tions play a crucial role in optimizing thrust, and

the capture of energy during the refill phase and

enhance overall propulsion efficiency. The compre-

hensive understanding of these features not only illu-

minates the intricacies of natural locomotion but also

serves as a valuable resource for the development of

advanced AUVs and robotic systems [29]. A detailed

description of a few robotic systems based on Jet

Propulsion is given in table 9.

4. Marine bio-inspired sensory

technologies

Nature’s evolved sensing capabilities offer untapped

potential for advancing artificial devices. Various

marine species showcase impressive sensing sys-

tems, often surpassing human-engineered sensors,

especially under extreme environments of the ocean.

Translating these nature-derived solutions into artifi-

cial technologies promises groundbreaking advance-

ments in sensor design. Table 10 contains some

examples of marine bio-inspired sensory techno-

logies. These sensors in underwater robotics replicate

biological sensing mechanisms, enabling enhanced

perception and adaptation to aquatic environments

improving the robot’s ability to navigate, detect

obstacles, and respond to environmental changes. By

mimicking the sensory capabilities of marine organ-

isms, biomimetic sensors contribute to the devel-

opment of more effective and resilient underwater

robotic systems. One such advantage is explained in

a study [128], which investigates locomotion con-

trol in undulatory swimming, focusing on the inter-

play between central pattern generators (CPG) and

sensory feedback. The study employs a robot model

and simulations to reveal that both central and peri-

pheral mechanisms contribute redundantly to the

generation of undulatory swimming. The research

emphasizes the role of exteroceptive feedback, partic-

ularly in spinal transection recovery, and explores

how structural asymmetry and body properties

influence hydrodynamic forces, promoting forward

swimming.

5. Research trends and challenges

The trend of biomimicking robotics has experienced

a remarkable historical evolution, as demonstrated

by figure 34, which displays the number of biomim-

icking publications per year. This trend indicates a

transformative surge in interest that began to emerge

in 2003 and continues to the present day. A num-

ber of elements that have been crucial in influencing

the field’s growth are also driving the rising trend

of biomimicking robotics. Interest in biomimicking

robotics was greatly aided by the growing emphasis

on creating innovative products specifically for the

offshore and maritime industries as well as research

applications. Furthermore, one of the key reasons for

the continued expansion in this field has been the

use of marine robots to carry out complex indus-

trial and military tasks. The surge in robotic bio-

mimicry research is further justified by its pertin-

ence to space exploration endeavors envisioning the

deployment of various biomimicking probes tailored

for specific terrestrial, atmospheric, and aquatic

environments [147].

Energy provision and storage are key focal points

in recent research developments within biomimick-

ing robotics. Notably, the emergence of microbial fuel

cells (MFCs) and various other strides made towards

enhancing efficiency and provisioning of energy in

biomimicking robotic systems [147].

Bioinspiration within the marine sector has res-

ulted in many applications such as those shown in

figure 35, that capitalizes on the remarkable adapt-

ations observed in marine organisms. Biomimetic

approaches contribute to sustainable practices, from

adhesion inspired by barnacles and mussels for med-

ical and construction innovations to antifouling solu-

tions drawn from natural coatings. The protective

properties of fish scales and mollusk shells inspire

biomimetic designs, particularly in military armor.

Stealth adaptations in marine life, such as passive

camouflage and translucency, also inspire biomimetic

designs. The movement of marine life has inspired

various biomimetic designs spanning underwater

vehicles to robots. Sensory adaptations observed in

marine life have led to biomimetic designs, includ-

ing sonar and radar inspired by dolphins and whales,

sensors and artificial electroreceptors inspired by

sharks, and optical adaptations inspired by mantis

shrimp [3].

Exploring the frontier of biomimetic and bio-

inspired marine robotics presents an array of

opportunities, challenges, and promising research

directions:

•Biomaterials in marine robotics: Future advance-

ments in smart materials enable the creation of arti-

ficial muscles or soft actuators for more accurate

animal-like motion underwater [147]. The ongo-

ing development of biomaterials stands as a key

pillar to elevate the potential of soft-robotic solu-

tions. Soft robotics, characterized by utilizing flex-

ible and deformable materials mirroring the attrib-

utes of living organisms, promises a paradigm shift

in adaptability and functionality. Hybrid systems

of soft and hard materials are also important for

future research.

•Enhanced operation efficiency and energy util-

ization: To ensure their operational autonomy

Bioinspir. Biomim. 19 (2024) 031002 A Prakash et al

Table 9. Examples of systems based on jet propulsion.

Investigator Description System Parameters

Villanueva et al

[121]

•Developed a systematic method for fabricating

the robotic jellyfish ‘robojelly’ with SMA-based

BISMAC actuators.

•Resolved folding effect in silicone bell by

segmentation, with the best-performing

configuration including a segmented bell and a

passive flap structure.

•The addition of a passive flap to the segmented

bell increases average thrust by over 1300%.

•164 mm bell diameter.

•Average of 16.74 W over the 14th cycle.

•Thrust: averaged 2.8 ×10−3N for the

first 4 cycles

Najem et al [122]•Introducing the design, construction, and

testing of a biomimetic jellyfish robot that

drives itself with IPMCs acting as flexible

actuators.

•Flexible bell, central hub, eight spars, and

radially extending IPMC actuators defining the

jellyfish’s shape.

•Composed of a heat-shrinkable polymer film

for less weight and better shape retention.

•Diameter: 15 cm

•Height: 5.8 cm

•Eight actuators: maximum swimming

speed: 1.5 mm s

•Eight actuators: average power: 1.14 W.

Godaba et al

[123]

•Utilizes dielectric elastomer actuator for

muscle-like movement.

•Water ejection system mimics jellyfish

movement for efficient propulsion. When

voltage is applied, the membrane compresses

and expands, providing unique features like

high deformation (over 100%), high energy

density, quick response, and low energy

consumption

•Bell diameter: 118 mm

•Bell height: 140 mm

•Bell thickness: 2 mm

•Acceleration (a0): 0.09 cm s−2

•Maximum thrust:0.15 N

Dooley et al [124]•Bioinspired robotic jellyfish with tentacles and

steering mimicking natural jellyfish.

•Tentacles actuated by a pulley system, steering

via the center of mass adjustment with

weighted gears.

•Records surroundings for potential

applications like autonomous fish tracking.

•Length less than 20 cm.

•Runs on a 14.8 V LiPo battery for several

hours untethered

Arienti et al [125]•Mantel Constriction inspired by octopus

characteristics

•Silicone body integrating crawling and

swimming components, the three-bar

mechanism driving each leg, compliance for

navigating cramped environments, and narrow

apertures.

•PoseiDRONE, the first underwater soft robot,

exhibits multi-functional locomotion and basic

manipulation, including grasping objects and

holding onto submerged structures.

•Length: 0.20 m

•Mass: 0.755 Kg & 0.816 Kg

Wang et al [33]•Biomimetic cuttlefish robot replicating

streamlined shape and propulsion mechanisms

of cuttlefish.

•Biomimetic undulatory level pectoral fin and

mantle actuated by Shape Memory Alloy

(SMA) wires.

•Biomimetic cuttlefish robot combines

undulatory and jetting propulsion, showcasing

versatility in achieving different swimming

speeds and efficiencies.

•Biomimetic pectoral fin Speed:

0.18 m s−1

•Biomimetic mantle speed: 0.6 m s−1

•Biomimetic pectoral fin propulsive

force: 0.047 N

•Biomimetic mantle force: 0.79 N

Sfakiotakis et al

[31]

•Robotic swimming with multiple compliant

arms inspired by the octopus.

•Propulsive thrust and rapid acceleration

primarily achieved through arm movements.

•Complex pattern characterized by a sculling

profile involving a fast power stroke and a slow

recovery stroke.

•Up to 0.26 body lengths per second

•Weight: 2.68 kg

•Propulsive forces: up to 3.5 N

(Continued.)

Bioinspir. Biomim. 19 (2024) 031002 A Prakash et al

Table 9. (Continued.)

Cianchetti et al

[126]

•OCTOPUS is a fully soft robot inspired by the

octopus, featuring eight arms for manipulation

and locomotion and a central body housing

processing unit.

•Manipulation arms mimic the morphologi