Review on Unmanned Underwater Robotics, Structure Designs, Materials, Sensors, Actuators, and Navigation Control

Javier Neira; Cristhel Sequeiros; Richard Huamani; Elfer Machaca; Paola Fonseca; Wilder Nina

doi:10.1155/2021/5542920

Review on Unmanned Underwater Robotics, Structure Designs, Materials, Sensors, Actuators, and Navigation Control

Neira, Javier;Sequeiros, Cristhel;Huamani, Richard;Machaca, Elfer;Fonseca, Paola;Nina, Wilder 2021-07-06 00:00:00 Hindawi Journal of Robotics Volume 2021, Article ID 5542920, 26 pages https://doi.org/10.1155/2021/5542920 Review Article Review on Unmanned Underwater Robotics, Structure Designs, Materials, Sensors, Actuators, and Navigation Control Javier Neira, Cristhel Sequeiros, Richard Huamani, Elfer Machaca, Paola Fonseca, and Wilder Nina Veox Inc., Arequipa, Peru Correspondence should be addressed to Wilder Nina; wnina@veox.tech Received 21 February 2021; Accepted 7 June 2021; Published 6 July 2021 Academic Editor: L. Fortuna Copyright © 2021 Javier Neira et al. (is is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Since its beginning, around the 50s decade, until present days, the area of unmanned underwater vehicles (UUV) has considerably grown through time; those have been used for many tasks and applications, from bomb searching and recovery to sea exploration. Initially, these robots were used mainly for military and scientiﬁc purposes. However, nowadays, they are very much extended into civils, and it is not hard to ﬁnd them being used for recreation. In this context, the present research is an eﬀort to make a walkthrough of evolution in this area, showing a diversity of structure designs, used materials, sensor and instrumentation technologies, kinds and the number of actuators employed, navigation control techniques, and what is new in development trends. (e paper gives a clear starting point for those who are initializing into this research area; also, it brings some helpful knowledge for those who already have experience. to have an impact on the future survival of humanity. (at is 1. Introduction why researchers have strived to develop unmanned vehicles Robotics is a branch of engineering that involves the con- that are increasingly eﬃcient and can provide work and cept, design, manufacture, and operation of programmable maintenance services underwater, as well as explorations of machines, which can develop their autonomy in the exe- ever deeper marine environments. For example, there are an cution of a speciﬁc task. Submarine robotics can be sub- estimated 2 billion tons of manganese nodules in the Paciﬁc divided as shown in Figure 1, where one can identify those Ocean near the Hawaiian Islands [2]. related to underwater robotics (unmanned water vehicles). (e UUVs are identiﬁed in Figure 1. Mainly, they are Underwater robotics have been developed for decades, and subdivided into those that are remotely operated (ROV) and they were characterized by manned water vehicles devel- those that are autonomous (AUV). (e ROV is a vehicle opment. However, in the last decades, due to the cost joined to an umbilical cable, and, according to its applica- demanded by the construction and use of these vehicles, tion, it can be classiﬁed into observation, work, and special researchers led to the development of robots or unmanned purpose [3, 4]. ROVs play a vital role in military operations underwater vehicles (UUV), where each time the robots (e.g., torpedo and mine recovery), rescue operations (e.g., were optimized and used in a wide variety of applications for locating historic shipwrecks, such as the RMS Titanic), and the beneﬁt and sustainable development of the planet. Since critical oil and gas operations. In recent years, the range of there is an immense magnitude of the oceans and due to the vehicles and mimitizing the need for human operators have human diﬃculty for their exploration, the vast majority of increased due to research eﬀorts. However, we have au- them have not yet been explored (two-thirds of the planet) tonomous vehicles (AUVs); these are indispensable tools, and around 37% of the world’s population is less than mainly for ocean exploration tasks. (ey are built using 100 kilometres from the ocean [1]. (is means that various high-end technologies and equipped with several of knowledge of the marine environment has and will continue them [5, 6]. During the subsea exploration process, the AUV 2 Journal of Robotics Robots Unmanned Workers Zoomorphic Humanoids vehicles Industrial With wings Androids Space Of service With legs Aerial With ﬁns Terrestrial Aquatic Figure 1: Location of unmanned water vehicles within a basic robotics classiﬁcation. performs necessary maintenance and repairs, transports the movements, such as jellyﬁshes, stingrays, and sharks. Even necessary equipment, and monitors and records diﬀerent though underwater robots that use thrusters with symmetrical situations under exploration. In oceanographic research, the geometric shapes are generally more eﬃcient, many works in AUV may be used in various measurements; one of them is bioinspired robots are being developed and continue looking to for the observation, release, and recovery of instruments increase knowledge in this study area. related to the monitoring of submarine volcanoes, in ad- dition to carrying out surveys of the seabed and various 2.2. By Hull Type. Some robots work underwater; those are studies in areas such as biology and hydrology. of two hull types, open and close. (e open ones generally (e unmanned vehicles’ structural designs have been have a hermetic box that is over a frame, and the other based on their shape, actuators’ distribution, and materials components like engines, buoyancy elements, sensors, and used. (ey include monitoring and measurement tasks of lights are over it too, while the closest ones have a bigger case oceanographic variables and navigation of the UUV; to that covers all elements; they do not have a frame because all achieve these objectives, it will be necessary to have a di- elements are inside of the case. (e prototypes or designs versity of sensors. Everything presented is the subject of which have a close hull usually have a cylindrical or round analysis in this article. (is article is structured as follows: geometric shapes; for example, there are some which look Section 2analyzes the morphologies used which are char- like a torpedo or a sphere. acterized by their type of hull and geometric shape and those that are bioinspired; additionally, it considers the distri- bution of propellants, materials used, and hermeticity. 2.3. By Geometric Shape. One of the most famous shapes of Section 3presents the instrumentation and actuation ele- underwater robots is the cylindrical shape, and on the top ments used interchangeably in ROVs and AUVs. Section they have the appearance of a dome, which means they look 4analyzes the navigation, orientation, and movement’s like a torpedo. In 1973, one of the ﬁrst portable ROVs was control of both ROVs and AUVs. Finally, the conclusions of made, and it had a torpedo shape as shown in Figure 2. this work are presented in Section 5. (e robot was hydraulically controlled from the surface, and it began to be controlled with electricity. (at was the main reason of its popularity, ten years later, 500 prototypes were 2. Structural Designs and Used Materials sold. (ose kinds of shapes come from military technology 2.1. Morphologies. (e morphological review refers to un- developments like many other technological advances. In this speciﬁc situation, the form comes from torpedoes that were on derwater robots’ geometrical shape, which is dependent on many factors like internal electronic devices distribution, military ships; this is the case of MK-38. sensor, motors, structure stability, and drag coeﬃcient de- In the Applied Physics Laboratory of Washington ﬁned to move the robot, among others. Additionally, it must University, some modiﬁcations to a torpedo were made, and take the robot’s degrees of freedom into account, since this is it became an inspiration to create the REMUS robot in 1994 needed to locate motors in the structure correctly. [8]; a schematic of such robot is shown in Figure 3. (is kind of robot does not necessarily have a geometrical (e design was of a cylindrical shape, and it moved shape; there are many ROVs and AUVs with biomimetic thanks to only one propeller behind the robot and four ﬁns structure; (at is to say, they try to mimic aquatic animal’s behind the propellers; two of them were located on a vertical Journal of Robotics 3 (ere are new studies that have been done during the last few years about diﬀerent shapes of underwater robots. In comparison, those have advantages compared with tradi- tional shapes, since their forms are symmetrical. In an in- vestigation project, the goal was to control the position of the symmetrical robot in events of disturbances with a pro- portional change of the direction force from propellers; all their motors were inside of the closed hull. (e design was developed in 2012, and it had a diameter of 40 cm and weighted 6.3 kg. (is spherical robot is shown in Figure 8. (e assembly of the mechanical components which are inside of the robot was made manually, and it caused Figure 2: Snoopy ROV [7]. troubles in the robot’s operation underwater because the robot was less stable, and it could not move for long periods. axis, and they had their drive mechanism like the ones that Figure 9 shows some of the improvements made with a 3D printer to fabricate the mechanism for the water jets [15]. were located on a horizontal axis. (ey were there to give direction to the robot. (e robot had a length of 114 cm with (e robot has two servomotors and a water jet in each leg. a diameter of 18 cm and a weight of 40 kg. (e torpedo shape For its movements, one servomotor is ﬁrst turned vertically is still used because of hydrodynamic advantages. to lift the propeller of the water jet; then, the horizontal In 1997, the REMUS robot’s development continued with servomotor moves the propeller of the water jet to go for- one smaller model that a person could uphold and transport. ward; subsequently, the water jet drops vertically; ﬁnally the (e robot had a length of 134 cm with a diameter of 19 cm and robot swings to take a step. (ere is another way to get easy access to the robot’s a weight of 31 kg. (e robot had four ﬁns as shown in Figure 4. (e ﬁns were in the stern, and they had an aerodynamic shape components; those kinds of ROVs are used to have a square or rectangular frame, and some components lay over the called NACA 0012. (e ﬁns moves were done using a pinion, a toothed chain, and an engine. Behind the ﬁns was the propeller; frame, such as hermetic boxes, lights, propellers, and other all the system described gave the direction to the robot. devices. (e robot called CURV [7] was one of the ﬁrst to (is torpedo-like shape allows underwater robots to implement the distribution mentioned. It was designed by move at high speeds on the horizontal axis, and they descend the US Navy to recover lost bombs or torpedoes; in one of its diagonally with the help of their ﬁns. (ere are other kinds of missions, in 1966, the ROV was able to immerse up to 869 m; underwater robots which have a torpedo shape, but they it was more than the maximum depth the robot could have more propellers and do not have rear ﬁns; an example is achieve according to its design; in addition to this, it operates the Nessi robot in its sixth version [10]. It was built for the under positive buoyancy, which means that if its propellers were turned oﬀ, the robot was going to the surface auto- SAUC-2011 competition, with 174 cm length and a diameter of 28 cm. To be able to move underwater, it was equipped matically. (e robot is still working with some improve- ments, and it is sent to diﬀerent missions; Figure 10 shows with six propellers: two for lateral motion, two for vertical and pitch moves, and two on the robot sides to control the the robot CURV II that is similar to CURV. In Figure 10, it is forward and yaw moving. A picture of Nessi’s sixth version possible to appreciate the ROV size compared with a per- can be seen in Figure 5. son’s size. (erefore, it was necessary to use a pulley on the Another development of a torpedo-shaped robot was boat to place it on the water. made in 2017, which had many propellers, as shown in With the passing of the years, underwater vehicles which Figure 6, but this was bigger than others shown in this paper. have similar shapes to CURV have been developed for subsea exploration and they can be carried by one person It had 534 cm length, 62 cm diameter, and a weight of 380 kg [11]. It had ﬁve propellers: one located on the top, two on the due to their smaller form. An ROV with a similar appearance to that of CURV was developed in 2010; the robot was called extremes to immerse, and the last two across it to control yaw movements. If there were any troubles and the robot Nessie IV [16], and it had a rectangular shape, aluminum frame, two hermetic chambers over the frame with a di- needed to come back to the surface, it had a ballast tank. (ere were other investigations of more complex ameter of 22 cm, sensors, and some safety components. models, which had a torpedo shape; an example of the One of the hermetic cameras had the motors and research was the development of a robot in 2018. It had four drivers’ electric connections and batteries; the other had ﬁns located on the sides including four propellers [12] and sensors interfaces and batteries electric connections. (e two propellers to immerse. (e robot had a length of 140 cm, robot worked using ﬁve propellers: two of them were a diameter of 20 cm, and a weight of 32 kg. Figure 7 shows vertically located to control depth and pitch, two others were located on the sides to control forward and yaw the position of those elements. (e robot was constructed to improve the propulsion system, which did not generate moves, and the last one was over the vertical axis to guide translation over this axis. (e Nessie ROV shown in countless bubbles because it was a trouble for the front camera to see; for this reason, the design had two more Figure 11 was an evolution of past versions and it was made to participate in a competition known as “SAUC-E” where propellers and it had four ﬁns to go forward; the movement was based on the turtles’ movement [13]. it was the winner robot. 4 Journal of Robotics 1.14” (.3’) Depth and heading Flux gate Sensor interface and control ﬁns compass data logger Homing array Command and DC brushless motor and control computer motor controller Transmitter 0.8” (7.17’) Temperature, conductivity, Hollow drive sha Lcoustic system dissolved oxygen, and pressure sensor heads Lead acid batteries Highly eﬃcient propeller Figure 3: REMUS (1994) [8]. depth and roll and the other four were located to allow forward movements, yaw, and translation over the vertical axis. Figure 13 shows the robot DaryaBird, showing the propellers distribution where they are approximated to over travel axis forward. (ere was a design proposed in 2019 which had a rectangular shape and positive buoyancy; this investigation [20] was more about the mechanic design, the design fea- Figure 4: REMUS (1997) [9]. tures to achieve stability, and a CFD analysis of its ﬁnal shape. (e robot had two pipes on the top from 2 to 4 inches, respectively, and a frame made of aluminum which sup- ported the ROV’s elements. (e structure guaranteed sta- bility due to the location of the mass center in the design, which was under the buoyancy center. (e top pipes were located there to obtain better stability while the robot rolled and to get the buoyancy center of the robot nearer to the top, as is shown in Figure 14. (e proposed design aligned mass center and buoyancy center over the horizontal axis to the movements on pitch and yaw could be easier. Figure 5: Nessi (2011) [10]. Due to the fact that the mass center and buoyancy center were not aligned over the vertical axis, the immersion Some ROVs have cubic-like frames such as CISCREA that propellers were distributed to get a pitch control as is shown is shown in Figure 12. It had six propellers: four of them were in Figure 15. outside of the vehicle's longitudinal axis, which gave the robot a (ere was a robot proposed in 2019 to work in shallow better turning moment, more lengthwise stability, and better waters [21]; it had a cubic shape, and its principal structure movements over the horizontal axis, and the other two were to was anticollision. (e robot’s design had an acrylic plate on control depth. (e robot had a length of 52.2 cm, a width of the front of it to locate a camera. Two propellers controlled 40.6 cm, a height of 39.5 cm, and a weight of 22 kg [18]. the forward and yaw movements. (e force given by the Some ROVs have been developed, having a frame to propellers should be designed greater than 1 kg and the support their components as the ones mentioned; one of immersion propellers should be between 2 kg and 8 kg. (e them was a robot called DaryaBird that was developed in robot was designed to have a hermetic chamber just for the 2016; it was made to participate in a competition for un- battery and another for electronic parts; Figure 16 shows the derwater robots, and one of its hermetic sections had a ﬁnal design with all its parts. torpedo shape. Over the frames were some elements such as its gripper, that is, a mechanical hand to carry objects, a battery, a pneumatic tank, and its hermetic chambers. 2.4. Bioinspired. In the last decades, there has been a lot of It had 83 cm length, 50.6 cm width, 41.3 cm height, and interest in improving the performance of underwater robots 32 kg weight [19]. It had six propellers; two were to control by taking a bioinspired optimization approach in aquatic Journal of Robotics 5 Figure 6: Large torpedo [11]. rusters Flapping fins (e union of soft and low rigidity materials with in- telligent actuators is called smart soft composites (SSC); its manufacture is carried out through the use of multinozzled Forward camera Tail 3D printer, for this UV curable polydimethylsiloxane (PDMS) is used, which can be 3D printed by direct ink writing (DIW) or stereolithography (SLA), where cables, connections, communication modules, batteries, and in- Body telligent actuators or anisotropic materials are placed layer Downward camera by layer in the ﬁnal prints [23, 24]. See Figure 19. Head In the use of these technologies, a new low-cost printing Figure 7: Hybrid (2018) [12]. method is presented in [25], where printing tests were carried out, and a material switcher PDMS was developed from laser diﬀraction thanks to the use of a T union by two currents: air and a second ﬂuid (water or oil); it should be noted that they do not use mirrors or lenses that are nor- mally in stereolithography (SLA). Although some robots are made based on these intel- ligent actuators, their movement speed is slow and limited compared to the living beings from which the inspiration was obtained or robots that use conventional actuators, since these have the advantage of being more compact, in addition to achieving better mimicry and acquiring important bio- logical characteristics, which are what they were looking for. A robot with a dolphin shape was made in 2016 [26]; it was an autonomous underwater vehicle (AUV) developed to Figure 8: Sphere (2012) [14]. measure water quality with a biomimetic system. (e prototype is shown in Figure 20, and the design was inspired animals, which after natural selection have evolved due to by an orca. (ey were looking to mimic the high hydro- dynamic and stability performance of an orca, because the their habitat to the point of having physical characteristics and excellent morphology. evolution of those animals has allowed them to develop those features correctly. (e dolphin could mimic some In 2012, a review of diﬀerent biologically inspired under- water robots presented and addressed a key problem in them, peculiarities on an orca and it could monitor and move in mentioned in [22], which is to achieve adequate movement in diﬃcult environments. real environmental conditions and classiﬁcation of robots by (e robot’s design had a rigid head, a hollow shell to the their way of swimming, which can be seen in Figure 17. (e components, a cabinet for the waist joint, and a chamber for proposal presented for the problem was the use of intelligent the caudal joint. It had a hydrodynamic shape to reduce the actuators, which are a viable alternative to be able to achieve the drag force, its waist was joint to get the necessary thrust to move, and a dorsal ﬁn and pectoral ﬁns to improve stability ﬂexible and complex movements that a camouﬂaged robot requires without the need for additional parts. also allowed performance turning manoeuvres. In the same year, a mechanical design of an autonomous vehicle (AUV) (e smart actuators are classiﬁed on shape memory alloy (SMA), ionic polymer-metal compound (IPMC), and a mix was presented inspired by a shark that had a vision system to between SMA and IPMC. Usually, the materials are com- take the information of the underwater environment, for bined with other materials; an example of this is presented in future applications as monitoring, searching objects, and Figure 18. (is embedding gives better mechanical features trajectory tracking. for robots. (e shark robot’s structure is shown in Figure 21 and it Normally, for the creation of parts of a bioinspired robot, had two main principal parts. On the front was a rigid body with an aerodynamic it is necessary to use polymers together with the alloys presented above, which serve to simulate elasticity and contour of a shark to get better performance swimming and two pectoral ﬁns to make movements as diving, emerging, deformability to achieve complex shapes of the living being to be mimicked. and swinging. Further, it had space where there were parts 6 Journal of Robotics Xilinx Zynq-7000 SoC Battery and Servomotor Hip joint circuit control module Knee joint Water-jet propeller Figure 9: Sphere (2014) [15]. Figure 10: CURV II [7]. Figure 12: CISCREA [18]. Mission module Aluminum pressure hull USB camera Aluminum ruster T-sloted frame Doppler velocity log Figure 11: Nessie IV [17]. Grabber Battery hull like cameras, circuits, sensors, and batteries. On the back was Figure 13: DaryaBird [19]. a multiarticulated body with three servomotors connected with an aluminum structure that allowed better movements when turning and achieved more manoeuvrability and thrust. because they had a gyratory axis with a bearing and a Glyd (e shell was made of ABS, and it had 482 mm length, Ring; they needed a dynamic seal in each ﬁn. (ere was an 208 mm width, 125 mm height, and 1.3 kg weight [27]. (e underwater manipulator which was built inspired by a robot used two impermeable systems: the ﬁrst one for the snake. (e robot’s purpose was to make exploration activ- top where there was a removable chamber that was a seal ities and intervene in underwater infrastructures, thanks to with a ﬂuoride rubber gasket and the second for pectoral ﬁns its shape, which gave ﬂexibility and the ability to get into Journal of Robotics 7 (e shell optimization was done diﬀerently on typical applications identifying the appropriate parameters to modify and improve with a biological model that meets the needs. For the selected models, simulations were done in a scalar level (1 : 3) inside of an air chamber. A better structure under the drag was implemented on a functional level (1 :1) to do a test on water and the designs that have passed by CFD simulations to get better data. (e other focus was the implementation of an alternative thrust system based on biological moves of ﬁns in the ﬁsh. (e better models were selected and made with a 3D printer; they were put to the test and, ﬁnally, the one that had a better performance was determined. (e implementation of the system is shown in Figure 23. Figure 14: Mechanical design of a rectangular ROV [20]. (e experimental results showed a drag force reduction between 50% and 85% in case of comparison to the basic OpenROV and the thruster system; although they were not T1 T2 that eﬃcient as the thrusters, they were better on the ﬂex- ibility, frequency, and amplitude. In 2019, another design of a robot called “CasiTuna” was inspired by tuna, a ﬁsh capable T5 T6 of high-velocity movements and manoeuvrability; addi- tionally, it can swim huge distances. Front (e robot tried to mimic the features of tuna, focusing on three main aspects; the ﬁrst one was a rigid body in the front where the electronic systems and a new thruster system of two engines exist, something that was not common on those T4 kinds of robots due to the quantity and the position. (ey are T3 accompanied by dorsal ﬁns and an internal system to adjust Figure 15: Mechanical design of a rectangular ROV [20]. the buoyancy. (e second main point was a lightweight body on the back; the section had two articulates, and, to achieve the movements, it had a mechanical system of 4 bars and a group of bevel gears that avoid involuntary ripples thanks to the new position of the engines. (e third main aspect was a rigid tail ﬁn that allowed it to generate greater thrust. (e robot “CasiTuna” is shown in Figure 24. (e robot had 520 mm length, 100 mm width, 130 mm height, and 2.6 kg weight; principally, its materials were ABS and PP. (e robot went through simulations in ADAMS and tests underwater validating design’s eﬀectiveness focus on displacement speed and stability, due to its aerodynamic Figure 16: Underwater robot for shallow depth [21]. shape. Table 1 shows some features of the mentioned robots. diﬃcult spaces. (e robot was designed by the University of Science and Technology of Norway. 2.5. (ruster Distribution. (e thruster’s location and the quantity depend on with how many degrees of freedom the (e robot is in the category of AUV and it had a serial robot is going to be designed; there is a case where, to achieve link mechanism with diﬀerent modules, which had tunnel all degrees of freedom, the thrusters have actuators to move thrusters and a stern thruster to move forward. Further, if them and change the direction of the force. needed, the robot could mimic the swimming of an eel due An ROV had ﬁve thrusters; some actuators on them were to good ﬂexibility. (e prototype is shown in Figure 22. (e designed in 2015; two thrusters were located on the extremes design was submitted on a very complex simulation envi- ronment with Matlab, which allowed to coordinate diﬀerent to move over the horizontal axis and yaw movements, and the other thrusters located on the sides had a servomotor; actuators on the articulations, getting good results of the proposed design, evidencing that the simulation environ- each one has the objective of changing the direction force and controlling roll movements as pitch and yaw, but they ment developed had high potential for this kind of tests [28]. (e robot OpenROV was an open code ROV developed in were employed to immerse [31]. (e last thruster was designed to operate all time and get stability on pitch 2019 with a bioinspired focus; it was made for the City movements. (e thrusters are shown in Figure 25. University of Applied Sciences in Bremen; its work was (e underwater robots must have an emerge and im- focused more on getting better stability, reducing the drag merse system for that reason; they use thrusters to do those force, and evaluating alternative thrust systems [29]. 8 Journal of Robotics Oscillatory (BCA-O) BCA Swimming mode body/caudal actuation biomimetic underwater robot Undulatory (BCA-U) MPA Oscillatory (MPA-O) median/paired actuation Undulatory (MPA-U) JET Jet propulsion jet propulsion Figure 17: Classiﬁcation of bioinspired robots by their swimming mode [22]. is possible to see locations of their four thrusters, which gave Pectoral fin (latex) them 5 degrees of freedom [33]. Body (ere was another research where the robot’s shape had more DOFs (degrees of freedom) than other distributions, Tail besides having just four engines as is shown in Figure 29. (e thrusters are distributed vertically and horizontally; the diﬀerence in forces between each pair allows the robot to SMA wires Skin have more degrees of freedom. For example, if the motor Holder above has more force than the one below, the robot will be able to change direction achieving a 90 turn [34]. Middle board Elastic substrate (e robot shown in Figure 30 was employed in research Biomimetic fin (SMA) for the best control system for a robot of this type according Figure 18: Manta ray robot [22]. to the mathematical model obtained by a CFD program [35]. (e robot had six thrusters: two of them were applied to dive and the other four located out of the longitudinal tasks or implement a ballast tank. (ere was a robot de- axis were approximately 45 concerning the forward axis. veloped for investigation purposes by Norway in 2017; it had three thrusters for the robot’s movements underwater. (e (e roll and pitch did not need to be controlled because the distribution element gave enough stability to those degrees prototype is shown in Figure 26. (e objective of the ROV mentioned in Figure 26 was the of freedoms. (ere are robots with more than ﬁve thrusters. (e robot development of a low-cost robot to monitor ﬁshes for shown in Figure 31 had eight thrusters, which allowed the aquaculture [32]. (e thrusters were around a cylinder robot to move on 6 degrees of freedom, like with six placed 120 from one another; this allowed them to move in thrusters, but giving the advantage of more thrust to dive. two dimensions. (e robot had negative buoyancy. To Figure 31 is an upgrade of Bluerobotics’ ROV [36], with control its depth, it had a rope attached to a platform on the eight thrusters: four of them were locatedvertically outside surface. (e robot’s stability had better control due to the the frame with an external protection to prevent the pro- prototypes’ symmetry; therefore, the pitch and roll caused by pellers from being damaged during their operation or the the pendulum eﬀect were not aﬀected. In 2019, a design was tether cable can be tangled or damaging one of themwhile proposed for aquaculture applications; the shape and they are working or tethered could tangle and damage on thrusters were located to avoid lurching eﬀect, which means that a buoyancy object loses its stability due to a strong water them. (e thrusters were placed in the same position as the original Bluerobotics’ ROV. stream. Figure 27 shows assembled robots and in Figure 28 it Journal of Robotics 9 Anisotropic Circuit material Polymer printing matrix Supporter (a) Supporter deposition (b) Part deposition (c) Circuit printing (d) Deposition of anisotropic materials Comm. Controller module Battery Actuator Repeat processes (e) Component (f ) Part deposition (g) Support removal (h) So morphing embedment of robot/structure Figure 19: SSC manufacturing process [23]. Hollow shell Fins Head Fluke GPS antenna Stern propeller module Tunnel thruster module Waterproof Caudal joint camera Waist joint cabinet Joint module cabinet Flipper Figure 20: Dolphin (2016) [26]. Figure 22: Underwater manipulator design (2016) [28]. Multilink propulsive units Dorsal fin Right pectoral fin Peduncle caudal fin Switch Left pectoral fin Camera Infrared sensors Figure 23: Bioinspired alternative shell and drive design (2016) [29]. Figure 21: Shark robot prototype (2016) [27]. 10 Journal of Robotics DC motors Battery Motor controllers Anterior shell Caudal fin Peduncle Pectoral ﬁn e buoyancy Transmission Posterior shell adjusting mechanism mechanism Figure 24: Almost tuna (2019) robot design [30]. Table 1: Characteristics of biomimetic robots. Year Name Form Propulsion Dimension (cm) Material 2011 RoboJelly Jellyﬁsh Membrane Not registered Silicone, SMA, steel 2012 TurtleLike Turtle Pectoral ﬁns Not registered ABS, PDMS, SMA 2016 Dolphin Killer whale Tail ﬁn 77.1 × 13.2 × 35.2 POM, ﬁberglass 2016 Nameless Shark Tail ﬁn 48.3 × 20.8 ×12.5 ABS, rubber, aluminum 2016 Nameless Snake Stern thruster/tunnel thruster/articulated swimming Not registered Not registered 2019 OpenRov Fish Pair of propellers and ﬁn Not registered PLA, varnish 2019 CasiTuna Tuna Tail ﬁn 52.0 ×10.0 ×13.0 ABS, PP D frame for carrying and launching 3. Materials and Hermeticity (e material selected for an ROV manufacturer must withstand the pressure to which it is going to be subjected; the deeper the robot reaches, the greater the external Servo pressure is during the robot’s operation. (e robots that have S2 motor housing a torpedo shape as [9] are made of aluminum, with chambers S1 ﬁlled with pressure air, and their frontal sections are made of plastic, such as the REMUS robot. (ere are more robots Front with torpedo shape; an example is Nessie [10], which was camera mounting built externally with aluminum, and the internal structure was made of PVC to support some parts, such as sensors. Nessie’s distribution is shown in Figure 32. Figure 25: 5 thrusters [31]. In 2016, a robot that had a hermetic chamber made of acrylic and PVC was proposed [34]; the acrylic part was placed in front of the camera to record outside and its thrusters supports were 3D printed. Figure 33 shows the parts. Another robot that had some of its parts 3D printed was introduced in [15]. (e parts are in Figure 9. (e 3D printer was an enormous advantage to the robot fabrication because the thrusters can save space due to better distri- bution; also, the mechanical fasteners are designed in a better way by 3D printing, and the mechanical resistance is better compared to the other propulsion system of the ro- bot’s water jets. One way to reduce the fabrication cost is using 3D print; the prototype [32] was an investigation project of an ROV in which the thrusters shell was 3D printed, and most of its structural parts were made of PE (polyethene) and PMMA (polymethylmethacrylate) to allow the camera recording. One of the most important aspects taken into account in the manufactured material of the ROVs is the density; since materials with low density make the buoyancy increase in Figure 26: 3 thrusters [32]. proportion to its weight, this means that more force is Journal of Robotics 11 (a) (b) Figure 27: 4 assembled thrusters [33]. [27] which was made of ABS (acrylonitrile butadiene sty- rene), with an impermeable system that used o-rings of ﬂuoride rubbers and Glyd Ring for dynamic sealing; further the posterior thrust part was connected to the mobile joints by a skeleton made of aluminum. CasiTuna robot [30] was similar to the shark with an an- terior and posterior body made of ABS, but this prototype had a ﬂow ﬁn made of PP (polypropylene). Another shape of this kind of robot is a turtle [22]; its body was of ABS and the head of (a) PDMS (polydimethylsiloxane) and the structure ﬁn was a t″ t combination between ABS and SMA (shape memory alloy). OpenROV robot bioinspiration came from a shell [29] and an alternative thruster made of PLA with a 3D printer, a further ﬁll- in process, varnish, and sanding to get a smooth surface. t′ (ere are a wide variety of shapes for this class of robots. t′ As an example, there was a robot called RoboJelly [22] and it had a bell matrix structure made of silicon and 8 BISMAC actuators formed by a steel spring, silicone, and SMA wires. t″ 1 Table 2 shows the main features of the aquatic robots dis- cussed in this section. (b) Figure 28: 4 thrusters [33]. 4. Instrumentation and Actuators If we refer to the number of external devices, we should required to submerge the robot. (e underwater robot in- know that these increase or decrease with the arrival of new vestigation [30] focused on its structural design where a applications for unmanned aquatic robots. Mainly these are 6036T6 aluminum alloy was selected for the structure divided into two types: autonomous robots and AUVs supported motors and other 1 cm diameter elements; hence, (autonomous underwater vehicles) and ROVs (remotely the hermetic chamber was made of AL5053. Within the operated vehicles), which also are divided by the type of aforementioned robot’s design, deformation analysis was application that they develop, such as intervention and made when the hermetic chamber was exposed to the inspection. (e diﬀerence between these two classes lies in pressure under the sea and the deformation of its supports. the diﬀerent use of resources, size, and weight that the robot Another prototype designed for aquaculture was in- has in each division. troduced in [33]; the design had a hermetic chamber made of In this section, we will explain and focus on the use of acrylic with a width of 5 cm. (e hermetical chamber’s instrumentation and actuators, collecting information from material beside the features mentioned above should have both types of unmanned robots. good corrosion resistance as mentioned in the design of [21]. (e design proposed a hermetic chamber made of aluminum alloy 4032-T6 and in the forepart an acrylic dome was in- 4.1. Sensors stalled for the camera. (ere is a kind of underwater robot that uses something called biomimetics to move underwater 4.1.1. Measurement of Oceanographic Variables and it is made of diﬀerent materials, for example, a dolphin structure [26] made of ﬁbreglass and the rest of the body (1) Temperature. To measure and process oceanographic made of paraformaldehyde (POM); this material has variables in advanced systems, it is well known that the ﬁrst lightweight features. Another structure’s robot is a shark relevant variable to be accounted for is the environmental 12 Journal of Robotics (a) (b) Figure 29: 4 thrusters located horizontally and vertically [34]. Attimeter Thrusters Enclosure Dockin hoop DVL (a) (b) Figure 30: 6 thrusters [35]. III III III III II III III III III III (a) (b) Figure 31: BlueROV2 and Bluerobotics’ ROV with 8 thrusters [36]. temperature; this happens due to the changes in the tem- helps ﬁshermen to set places and times propitious to ﬁshing perature in the ocean, which not only inﬂuence the dynamics and also to know the distribution of fauna. of the sea and the atmosphere but also intervene in the Likewise, sensors of depth, altitude, and temperature are distribution of marine organisms and their metabolism; for always used in the design of aquatic robots. When we couple this reason, some boats use a temperature sensor, which these 3 sensors, we will be able to obtain the data of Journal of Robotics 13 (a) (b) Figure 32: PVC frame [10]. has sensors attached to its structure for reading chlorophyll and nitrate, where water samples can be used for the cali- bration of these parameters, as shown in Figure 34. (4) Conductivity. Conductivity is the measurement of electrical resistivity, a property that quantiﬁes how many dissolved substances, chemicals, and minerals are present in water. (is means that a large amount of these impurities determines a higher conductivity. (e use of CTD sensors allows water measurement of temperature, conductivity, and pressure. El Dorado AUV has a CTD sensor [38], which, like the SOTAB-I robot, is attached to its frame, with a sampling frequency of up to 16 Hz enabling a high spatial resolution, with a consumption of 3.4 W [39]. Figure 33: Printed thrusters mount [34]. (5) Total ATP (adenosine triphosphate). (ere are some other oceanographic variables and the related variations between kinds of sensors that ROVs have been carrying on recently; them. However, the measurement of the temperature sensor they are called microﬂuidics, which deal with the manip- can be used not only to know better surroundings but also to ulation of concerning particles or droplets temporal dy- compensate for the operation of the gyroscope and accel- namics, velocity, and spatial ﬂow patterns in microchannels erometer [37]. (at is the case of the UUV or SUR-II. [40]; although it is a new multidisciplinary ﬁeld, it has the potential to inﬂuence areas as biological analysis. (ey are (2) Pressure. It is well known that the control of ballasts used also called LOC (lab on a chip) to be used as screen in- in the stability of the structure of UUVs is directly related to struments in cell biology, chemical synthesis, and bio- the pressure sensors; this is due to the assembly of com- analysis. (ey have some advantages because they are pressed air tanks, which are regulated by the constant portable, and they can be done with low-cost fabrication variation of atmospheric pressure in the environment materials. An example of an ROV is given below. (symmetrical robots allow changing their center of gravity). (e total ATP is a useful biochemical parameter for (is kind of behavior is observed with DaryaBird, a UUV detecting biomass or biochemical activity anomalies in the that employs a YOKOGAWA’s pressure sensor that mea- natural environment; since dissolved ATP is an important sures depth and a gyroscope that measures the azimuth angle carbon and phosphorus for marine microbes is also related and altitude angle [19]. (ese two sensors help in the remote to microbial activity, the total ATP is a useful parameter function, turning the UUV into an AUV (autonomous indicative of the presence of biogeochemical events, such underwater vehicle) or in some cases it could help in turning submarine volcanism, hydrocarbon seepages, and occasional it into a ROV that can be remotely operated; this type of supply of organic resources [41]. (at is why, to obtain these UUV can recollect data and send it through an umbilical variables in real time, a new version of an ATP analyzer was cable that reaches the surface to be visualized in a computer. developed and evaluated in situ using an ROV, achieving a depth of 200m in the tests carried out. (3) Nitrate. (e standard methods of obtaining oceano- Figure 35 shows the analyzer; it has a microﬂuidic device graphic samples were through the samples collection with a and analyzer module, which is the core component, and a single bottle, with the main tests being temperature, pres- photometry module for the bioluminescence intensity sure, and salinity. In an advance to automation, “El Dorado” measurements based on the L-L reaction. was presented in 2016, which is an AUV that recollects up to (e measurements taken with the in situ ATP analyzer 10 shots of water in a bottle called “Gulper” [38]. El Dorado were consistent with those measured manually, which 14 Journal of Robotics Table 2: Evaluated underwater robots features. Year Name Shape Hull (rs DoF Dims (cm) Material 1973 Snoopy Torpedoes Closed NA NA Not registered Not registered 1994 REMUS Torpedoes Closed 1 2 (roll, yaw) 114.0 ×18.0 Not registered 1997 REMUS Torpedoes Closed 1 2 (roll, yaw) 134.0 ×19.0 Aluminum 2011 Nessie Torpedoes Closed 6 5 (roll, yaw, Y, Z, X) 174.0 × 28.0 Aluminum and PVC 2017 Nameless Torpedoes Closed 5 5 (roll, yaw, Y, Z, X) 534.0 × 62.0 Not registered 2017 Hybrid Torpedoes Closed 4 4 (roll, yaw, Y, Z) 140.0 × 20.0 Not registered 2012 Nameless Sphere Closed 3 3 (yaw, Y, Z) 40.0 Acrylic 1966 CURV Rectangular Open NA NA Not registered Not registered 2010 Nessie IV Rectangular Open 5 6 (pitch, roll, yaw, Y, Z, X) Not registered Aluminum 2014 CISCREA Cubic Open 6 5 (roll, yaw, Y, Z, X) 52.2 × 40.6 × 39.5 Not registered 2016 DayaBird Rectangular Open 6 5 (roll, yaw, Y, Z, X) 80.0 × 50.6 × 41.3 Aluminum 2019 Nameless Rectangular Open 6 5 (pitch, yaw, Y, Z, X) Not registered Aluminum and PVC 2019 Nameless Cubic Open 4 4 (roll, yaw, Y, Z) Not registered Aluminum 4032-T6 and acrylic 2015 Nameless Rectangular Open 5 6 (pitch, roll, yaw, Y, Z, X) 70.0 × 40.0 Aluminum AL 5053 and 6036T6 2017 Nameless Cylinder Closed 3 3 (yaw, Y, X) 30.0 × 20.0 ×15.0 PPE and PMMA 2019 Nameless Cylinder Closed 4 5 (roll, yaw, Y, Z, X) 44.0 × 26.0 × 24.8 Acrylic 2019 X4-ROV Cylinder Closed 4 3 (yaw, Y, X) Not registered PPVC and acrylic 2018 BlueROV2 Heavy Rectangular Open 8 6 (pitch, roll, yaw, Y, Z, X) 25.4 × 57.5 × 45.7 Aluminum and acrylic (a) (b) Figure 34: Images of El Dorado AUV, Gulper system [38]. position, direction, and speed. Also, the attitude and heading Microfluidic device reference system (AHRS), the inertial navigation system (INS), or the hydroacoustic position reference (HPR) system was implemented to maintain better control of positioning PC and stability. (e sensors that make up these systems and the applications in the diﬀerent UVVs are described below. Analysis Photometry (1) Inertial Sensors. (e inertial measurement unit, better module module known as IMU, is a sensor that detects linear acceleration using one or more accelerometers, as well as the rotational speed using one or more gyroscopes. Some of these devices include a magnetometer that is used as the main reference. Reagent and waste bags Sample inlet In 2008, the ROV Nessie III was ﬁrst introduced as an AUV that uses the 3-gyro reference system for navigation Figure 35: Total ATP analyzer with a microﬂuidic device [42]. targets [43]. In 2010 AMOUR was introduced, which was a medium ROV destined for the investigation of maritime areas; this ROV used the coupling of an inertial sensor that demonstrated that a portable, simple, and reliable ﬂow analysis system such as its microﬂuidic device can be used in estimates the position and depth that uses a record of 10 data raws unprocessed, corresponding to the sensors (a pressure extreme environments for real-time biochemical analyses. sensor, 3 magnetic ﬁeld sensors, 3 accelerometers, and 3 gyroscopes) [37]. On the other hand, MINERVA appeared 4.1.2. Navigation Instruments. When the UUVs were built, in 2014, which was an intervention ROV that mixed two diﬀerent data acquisition methods were used such as positioning systems and used an inertial sensor as the main Journal of Robotics 15 (5) Doppler Sonars. Doppler velocity sensor (DVS) uses the sensor and a depth sensor [44]. (is indicates that not only is a single-precision algorithm needed for navigation, but also Doppler eﬀect to measure the octagonal velocity; its limi- tations are based on the calculation of the integration of other sensors are needed to compensate for the errors of a single system. velocity and the time of calculating the position; this type of positioning control system will be explained in the next (2) Compass (Magnetometer). (e magnetometer works by section; however, its operation can be up to 300 m. measuring the magnetic ﬁeld variation in three referential As mentioned before, MINERVA mixes two positioning axes that are subtracted from the Earth’s magnetic ﬁeld; systems, in which, apart from using an inertial sensor, it also despite its wide applications on the UUV development ﬁeld, uses the hydroacoustic positioning system through the use of the operation of this sensor is sensitive to the noise caused by a Doppler velocity record (DVL) to measure its velocity [44]. other sources like the operation of other sensors, motors, Figure 36 is an example of the calculated position through the use of a hydrophone array. (is use is carried out in and others. (is means that you cannot only rely on the values of a diﬀerent UUVs, such as DaryaBird, which employs an al- titude sensor TRAX [19], which was installed to control the magnetometer, as there are also several parts to be con- sidered in underwater navigation such as a pressure sensor movement in DVL sensors. to measure depth, a gyroscope, and an accelerometer to control altitude and locomotion [45]. (is also gives the 4.1.3. Optic Sensors solution of integrating dedicated digital sensors to increase the accuracy, modules that can evaluate the heading di- (1) Video Cameras. Most UUVs have standardized a rection with a minimum diﬀerence of degrees, which are still complement of one video camera for transmission from compensated and calibrated to support magnetic distortions depth to the surface; the images are important in envi- with the combination of other sensors, as is the case of a ronmental analysis. In 2010, in the construction of the TSL MEMS accelerometer 3-axis sensor and a 3-axis magneto- (Tunnel Sea Lion) robot, 2 video cameras were incorpo- resistive sensor [46]. rated: one with a direct view on the bow and the other in a vertically downward orientation [48]. Likewise, another (3) GPS. (e global positioning system is managed by direct application to the mounted camera is to use it as a new communication with satellites; the use of these devices addressing method by an AUV; this was observed in 2018 underwater does not allow their correct operation, so the when tests were carried out on an interactive technique best application to UUVs is through collection or recovery of shown in Figure 37 which makes the construction of a these robots when they reach the surface. coordinated system with a route drawn from the image (e use is described by Choyekh Mahdi, indicating that taking through video [49]. the tracking of the SOTAB-I robot on the sea surface is Another application by the transmission of images was ensured by a global positioning system (GPS) receiver that established in the detection of marine animals with varied serves to determine the absolute position of the robot. In the visibility from a new dataset; video capture consists of three case where the robot is submerged, the Ultrashort Baseline cameras and three lights. (e colour cameras have a reso- (USBL) system ensures tracking [39]. lution of up to 1080 ×1920 pixels, and the frame rate is up to 30 fps [50]; the camera direction is diagonally downward (4) Sonars. (e sonar’s performance is through sound, where towards the riverbed. the propagation of waves underwater allows navigation, communication, and detection of submerged objects. Since their use is standardized in underwater vehicles for opera- 4.2. Actuators. (e underwater robot principle of move- tion in low-visibility conditions, there are a wide range of ment is based on the use of propellers; the type, power, and UUVs using this device; an example is Nessie III, which sent weight of the motor used with these propellers depend on a speciﬁc signal from the vehicle to a transponder that re- the robot’s work, as explained at the beginning of the section. sponds; the delay in the vehicle that receives this response gives the bidirectional ﬂight time for the signal; the range 4.2.1. Engines and (rusters used was between 60 kHz and 90 kHz [43], and the result was to obtain the raw data of speed, distance, and distance time. In 2017, the navigation compensation of an ROV was (1) Water Jets. Among the variety of motors that we use to presented through the comparison of data extracted from a move the robot, we can ﬁnd the propulsion based on a high- sonar with the use of the dead reckoning methodology and its pressure water jet. For a long time, this type of propulsion compensated error [47], which details the use of the sonar was normally used because of its comfortable design and its when the ROV does not have any movement due to the great propulsion with higher weights. In 2000, this type of propulsion was used in the development of a torpedo- interference that occurs with the operation of the engines, for its previous compensation. (ere are passive sonar systems, in shaped robot called TSL, where the bow and stem pro- pulsion systems had a tunnel for the performance of water which hydrophone-based communication participates in points not so far away between the robot and a boat or surface, jets [48]. giving sound pulses to ﬁnd the distance between both objects In spherical-shaped AUVs, this type of drive is used in a and calculate the angle of the sound source. [19]. vertical direction; two actuators can be controlled by one 16 Journal of Robotics Y (0, y , 0) (5) Brushless. Brushless motors became popular in small and O (0, 0, 0) Hydrophone y medium ROVs; their use dates back to the 80s and 90s. An array example is ABE, an AUV destined for benthic species ex- ploration that uses brushless motors (brushless) with oil [52] X (x , 0, 0) x for pressure compensation. A clear example is ARMOUR, where each drive is made up of a motor controller and a brushless DC motor [53]. Another related example is the operation of Jeﬀ, a small AUV designed for inspection and Z (0, 0, z ) swarm joint work; the propulsion system has mainly two DC T (x , y , z ) t t t Pinger motors with a custom magnetic coupling design to avoid corrosion and short circuits. [45] (6) Hydraulic Systems. In the design and development of Figure 36: Acoustic detection plane [19]. UUVs, a new structure was chosen for the steering man- agement in the 6-DOF; this particular conﬁguration in parallel can be seen in Figure 38; it has two main thrusters in Y the front and another in the rear that handles the steering of CS camera Trajectory the robot [54], and the union between the two parts is through the hydraulic system. 4.2.2. Luminaries. (e functions performed by the UUVs include the inspection, manipulation, and data collection; all the robots have a video transmission system implemented, so it is necessary to develop a lighting system to acquire Y images since the underwater environment does not have P visibility conditions due to the lack of a light source. (e ROV system presented by Jinwoo uses two panoramic halogen lights and two LED lights to acquire high-quality images [55], as shown in Figure 39. Figure 37: Construction of external orthogonal coordinate system In another application in image acquisition, there are [49]. the recognition and detection of objects, for rescue or supervision robots, and a clear example of detection is found in the research of Pedersen et al., where the illu- thruster and one servomotor; the jet-based thruster decides minated area is needed for the detection of pelagic species, the value of the driving force, and the servomotor controls where 3 LED lamps of 1900 lumens and an approximate the height of the thruster [37, 51]. resistance of 10 bar are used [50], to properly visualize the case study. (2) Stepper Motor. In the UUV’s direction management, a form of steering management was introduced, which in- volved the torpedo structure (Xianbo et al., 2017); this 4.2.3. Manipulators. (e term “manipulators” is used to structure has two propellers: one for horizontal movement describe a mechanical device with mobile joints intended for and the other for vertical motion; these two are attached to the manipulation of tools, parts, or special devices to per- the main propeller which is driven by a DC motor and four form various tasks. (is meaning applied to robotics results rudders driven by stepper motors [11]. in an automatic handling machine, reprogrammable in ei- ther a moving or ﬁxed position. In 2011, a hybrid underwater (3) Brushes. Nessie III and DaryaBird used the same brush robot was made, with a crab and lobster structure, where its motor-based propeller oﬀered by the SeaBotix brand, with a legs acted as manipulators and its main function was to consumption of 110 W and an ability to withstand depth up inspect underwater structures and shipwrecks in shallow to 150 m; these were used in directional movement in the -xy waters, where activities such as cable cutting, grinding, and plane; in the case of the second robot mentioned, it used a drilling are required [56]. RoboPlus Hibikino thruster with a power of 90 W for the (e most common way to implement manipulators is in robot descent control. intervention class ROVs, for example, MINERVA, where its manipulator allows the samples collection. Table 3 sum- (4) Servomotors. ARMOUR has a closed torpedo-shaped hull marizes all the aquatic robots seen in this section. structure, which utilizes 10 propellers for handling 6 DOF (degrees of freedom); however, small spherical ROVs, such 5. Navigation and Control as SUR-II, use vector water jet thrusters composed of wa- terproof housing, two servomotors, and a support frame Navigation, in simple terms, conforms to particular [37]. methods that allow someone to know where they are and Journal of Robotics 17 150mm F 100mm 100mm 50mm 50mm 0mm CG 0mm F y CG (a) (b) Figure 38: Direction change operation diagram [54]. (a) (b) Figure 39: (a) (e use of two LED lights. (b) (e use of two halogen lights with their respective results [55]. how to get to a new point. Depending on the type of en- conducting maritime exploration which covers large areas of vironment and available reference points, these methods several hundred square kilometres. Navigation plays a vital could be simple; however, they could become complex role here that if it is not properly executed, it could not only results in a hostile, changing, and unpredictable environ- aﬀect the fulﬁlment of the mission but also aﬀect the safety of ment, also, to reference points that are not visible [6]. (e the vehicle [11] and, in the worst scenario, could lead to the tasks that the UUVs must perform require navigation to loss of the robot, causing economic loses and contamination displace to diﬀerent location points to complete their duty. of the explored environment. Due to its autonomy, navi- gation must be accomplished under the control of a com- Navigation can be done by the vehicle itself (in the case of AUVs) or by operators (in ROVs cases). Usually, the tasks puter embedded in the vehicle. performed by ROVs demand heavy work and can be conﬁned to smaller spaces where navigation is probably not very complex and could be performed directly by the 5.1. Navigation Methods. Navigation in AUVs represents a great challenge for most researchers due to the impossibility of operators through a joystick and the use of one or more video cameras installed on the ROV. using a global positioning system (GPS) underwater. (e electromagnetic radiation waves emitted by satellites are Moreover, autonomous vehicles have more tasks and must carry out missions that take several hours, days, or even absorbed when they come into contact with water, so a GPS signal receiver cannot capture the waves underwater [58]. months [57]. Most of these missions are focused on 18 Journal of Robotics Table 3: Summary of the evaluated UUVs. UUVs Type Application Equipped sensors Actuators Compass, low-frequency sonar, angular 3 aft, 2 vertical and 2 ABE (1992) AUV Seabed supervision velocity sensor horizontal thrusters 2 horizontal thrusters, 1 Aqua Explorer Inspection of telecommunication Gyroscope, altimeter, depth meter, AUV vertical with brushless 1000 (1992) signals and cables accelerometer, acoustic transponder motor Depth gauge, CTDO sensor, INS with A main thruster, 2 vertical R1 (1995) AUV Monitoring near the seabed Doppler sonar, acoustic transponder water jet thrusters MINERVA ROV No register a DVL, HPR and IMU 5 thrusters (2014) ARMOUR UUV Reef survey and other applications IMU, GPS, DVL 5 thrusters (2010) Environmental study and Depth sensor, leak sensor, camera, FOSN, 4 thrusters and an 840 W (2017) URV surveillance task in the mid-range of and mini-AHRS main propeller shallow waters Monitoring of nuclear storage ponds SUR 3 servomotors and 3 water AUV and wastewater treatment facilities to No register (2013–2015) jet propellers prevent leaks Application of parallel robots in the REMUS I IMU, pressure sensor, immersion sensor, LED lights, 1 thruster at UPR underwater environment requires (2011) temperature sensor, camera the back studies GPS satellite navigation system, USBL Consists of six thrusters, positioning system and autonomous on- TSL (2000) AUV Tunnel inspection providing arbitrary board navigation system, TV system, and movements in 3 axes IFSSI scanning sonar Nessie III Designed to participate in SAUC-E Altimeter, IMU, camera, transponder, 5 80 W propellers, AUV (2008) competition battery temperature sensor brushless motors with oil Check and evaluate new navigation (2018) AUV Stereo camera, does not register other sensors Does not register methods 4 100 W thrusters and 2 DaryaBird Pressure sensor, DVL, USB camera, altitude AUV Does not register main thrusters brushless (2016) sensor, and hydrophone motors with oil. (erefore, underwater devices have been used to establish a acceleration in the vehicle’s direction. If there is no need to local positioning system. (anks to advanced technology, execute the proposed algorithm, they use traditional dead many of these devices have been improved and optimized reckoning to estimate the vehicle’s heading direction and its considerably in terms of dimensions and performance. (is position. Another interesting work about accelerometers and how motivated the investigation and improvement of methods that allow a better estimation of the location and thus more exact they are used corresponds to the authors Yan et al. [59]. (ey navigation. worked on a dead reckoning navigation system based on neural networks using only accelerometers, due to the cost of using other sensors such as a DVL or the dependence on an 5.1.1. Proprioceptive Navigation. If the travel speed of the acoustic system. (e errors that have been generated by AUV is known, new positions can be estimated by con- using inertial units are reﬂected in rapid changes in the secutive integrations of speed. To perform velocity measures, measured angles by gyroscopes; this considerably increases Doppler velocity log (DVL) is normally used in conjunction the error of the dead reckoning system. (e use of neural with inertial systems and a compass; the estimated position networks to estimate the pitch angles through an exploration solely depends on the movement of the vehicle and hence the between the measured orientations and the measured ac- proprioceptive name. (is type of methodology is known as celerations varying in time allows estimating the vehicle’s dead reckoning [59]. (is type of navigation system cor- positions and reducing the errors caused by the gyroscopes. responds to the research written by Itzik Klein and Roee Normally, the proprietary navigation systems work in Diamant; they developed a system that estimates the tra- conjunction with external systems to correct themselves and jectory travelled by a water vehicle that moves freely in the reduce the accumulated error. Some of the work done on direction of the marine currents [60]. Because these robots dead reckoning which works in conjunction with other work very closely to the sea surface, they are easily sus- reference systems corresponds [61] to Kepper et al. who ceptible to orientation change, which creates problems in the developed a death reckoning model based on an inertial path. (rough acceleration measures, the investigator can measurement unit (IMU), which works in conjunction with constrain the execution of a proposed algorithm based on an acoustic measurement system to reduce the error ac- the principal component analysis to calculate the cumulated by the IMU. Due to the noise generated by the Journal of Robotics 19 (is method works quite well for a single vehicle. In the case of the navigation with several vehicles, variants of this method have been proposed to eliminate the consultation Query signals, converting the communication in one direction and thus removing the dependence of the time intervals for Query vehicle location updates. However, both the beacons and the underwater vehicles must be synchronized [63]. Reply (e standard conﬁguration of the LBL system and its variants (Figure 41) allows establishing an absolute posi- Reply tioning for either one or several underwater vehicles. However, the task of implementing and calibrating polyg- onal beacon arrangements is expensive and diﬃcult. Figure 40: LBL standard mode [62]. (erefore, it was decided to improve these systems even more and only one beacon has been achieved to determine IMU, the raw data captured was ﬁltered using an extended the position of a vehicle; this system has been called Ul- trashort Baseline (USBL). (is conﬁguration is illustrated in Kalman ﬁlter. For the model eﬀectiveness, implementation was evaluated in data collected in 3 diﬀerent environments Figure 42. (is type of conﬁguration works similarly to the for ﬁeld experiments and in an open ocean environment. standard LBL conﬁguration; however, the vehicles have Correct navigation requires a good position estimate, so multiple acoustic receivers, because they must determine not the instruments and the algorithms used must obtain the only the distance at which they are from the beacon but also most exact ubication. the angle with which the replica of the signal arrives. (e query signal was issued. In this way, the need for using several beacons for the trilateration calculation is avoided. 5.1.2. Acoustic Navigation. (e main problem of proprio- (e vehicle’s work area is conﬁned to the entire radius ceptive navigation is that the error increases limitlessly as the generated by the beacon. distance travelled by the vehicle increases. If an external We can cite the work done by Hidaka et al. [19]. (ey reference system is not considered, the navigation becomes implemented this acoustic navigation system which inten- critical for the vehicle and its mission. As a solution to this ded to use an array of hydrophones that were very close. (e problem, acoustic navigation is employed. Acoustics waves angle was calculated to the arriving sound from the oﬀset are appropriate for underwater propagation due to minimal that occurred between the hydrophones. Besides, the sonar attenuation. Hence, they are employed for underwater system uses an electronic circuit for signal ampliﬁcation, communication and positioning the underwater vehicles. phase comparison, and digital to analog conversion (D/A). For example, underwater vehicles use data of the placed beacons for estimation of their positions in the work zone. 5.1.3. Optical. Optical navigation uses optical devices such (e most used acoustics methods for underwater location as video cameras or optical diodes from which morpho- are Long Baseline (LBL) and Ultrashort Baseline (USBL) logical data of the seabed are recorded. M. Carreras et al. [62]. presented a localization approach for an underwater robot (e standard LBL method is characterized by beacons or based on vision and in an environment structured like a transborder, which is ﬁxed as shown in Figure 40. (e image water tank [64]. In the work, the location algorithm details shows the conﬁguration system for the vehicle and through some graphic results and the precision of the transponders. system. (e algorithm allows obtaining a 3D position, First, transponders listen to the pings emitted for the orientation, and speed of the vehicle by detecting reference vehicles, and distance estimation is obtained from TAT points from the bottom of the tank. (e location estimates (turnaround time) at a speciﬁc frequency. (us, the vehicle are highly accurate without drift, allowing them to be used as can estimate its position by algorithms based on recursive feedback measurements for low-level speed-based control- least squares (RLS) or using extended Kalman ﬁlters. (e lers. Its computing system is 12.5 Hz in real time. vehicle must save transponders positions. Some works related to the implementation of an LBL system correspond to Christopher von Alt et al. (ose who 5.2. Orientation and Motion Control. It is necessary to take developed REMUS [8], a torpedo-shaped underwater robot controlling the orientation and movement of underwater created for exploration of marine resources, presented two vehicles into account, and it may demand an exploration modes of operation: autonomous and nonautonomous. In mission or some work that requires manipulation or ex- the autonomous mode, the robot had to follow a path traction on the seabed. However, due to the presence of formed by acoustic transponders implanted on the seabed, external disturbances and uncertainties in the marine en- and REMUS acted as a target hunter. (e transponders vironment, linear control methods are not very eﬃcient, so it distribution deﬁned the navigation path of REMUS; on the is necessary to apply advanced robust control methods. (e other hand, in the nonautonomous mode, the navigation objective of an orientation control is to retain the required was carried out with the help of a boat, and REMUS followed orientation regardless of swell and unpredictable distur- it through an acoustic communication. bances in the environment. (at is why a hydrodynamic 20 Journal of Robotics Ping Ping Ping 1 Ping (a) (b) Figure 41: Variants of the standard LBL conﬁguration. (a) An LBL conﬁguration without query pings. (b) (e conﬁguration shown allows the beacons to obtain their locations using GPS [62]. Ping Ping Figure 42: Ultrashort Baseline system [62]. model and mathematical parameters of the structure must hidden in the structure. However, most researchers make use of be obtained ﬁrst. Computational Fluid Dynamics (CFD) simulations on the Likewise, to establish a control system, the following behavior of their framework to reduce the error due to changing environmental conditions which are diﬃcult to predict. points must be taken into account: the performance of the system is limited, adding that the behavior of the control system must be robust in terms of both stability and per- 5.2.1. Sliding Mode Control. (e sliding mode control formance, since it takes into account the energy manage- (SMC) is a robust control for modelling uncertainty and ment and optimization of the entire system [65]. (is parameter variations and has good disturbance rejection approach takes into account the inevitable imperfection in characteristics. (ere have been a wide variety of applica- physical systems and variables; one of the investigations on tions of the same [67–71]. However, it inherits a discon- the performance of a new control strategy for imperfect tinuous control action; therefore, the chattering systems is observed in [66], starting from an electrome- phenomenon that occurs when the system operates close to chanical system based on a light structure that acts as a the sliding surface will occur. Sometimes this discontinuous support and supply for the simple coils found in the control action can even make system performance unstable. structure; the purpose of this research is the simulation of control systems for imperfect systems that, thanks to the peculiar properties in the structure, the eﬀects of vibration 5.2.2. Adaptive Control. Side Zhao and Junku Yuh proposed signals on the hidden dynamic system of the imperfect an adaptive control based on a disturbance observer [69, 72]; system can be observed. the control scheme of this system has an adaptive controller Given the premise on nonlinear control systems in im- based on a nonregressor, and it is the outer loop of the control perfect systems with more than two variables, it is considered scheme, while the inner loop controller is the disturbance that most of the research carried out within the ﬁeld of hy- observer. (ese two elements mentioned above are the drodynamics and the behavior of an ROV is established in only components of the adaptive control system proposed by the the movement controls, guaranteeing the movement of the authors, which is robust against external disturbances and robot in the established route without considering the dynamics unpredictable behaviors due to the self-adjustment of its Journal of Robotics 21 + d y++ Nonregressor- ∗ + u u y based adaptive P –+ controller +– d –1 f P ξ n y Disturbance observer Figure 43: Diagram of the adaptive control system based on a disturbance observer [72]. ∆E – ∆e ∆e z –1 Depth rate of + uZ Limiter FLC change f t(E, ∆E) Motion Limiter z Limiter Figure 44: Fuzzy control system diagram for depth [74]. ∆ E – ∆e ∆e Ψ –1 Yaw angle rate + uN Limiter FLC f (E, ∆E) Rotation Limiter Ψ Limiter Figure 45: Fuzzy control system diagram for guidance [74]. Table 4: Classiﬁcation of orientation and movement control methods most used in UUV. Control methods Contribution of the method (i) Path control in the horizontal and vertical plane in an AUV, using 6 degrees of freedom [77–79]. (ii) Integrated PID control with a backstepping control for trajectory control of an underactuated AUV [80]. PID (iii) Implementation of a self-adaptive fuzzy PID controller [81]. (iv) Proposal of a self-tuneable PID control, using neural networks [82]. (i) Avoid collision in marine vessels through an intelligent decision-making system [83]. (ii) Linear approximation control for tuning parameters of a fuzzy controller [84]. Fuzzy (iii) Features a torque controller, calculated with a trajectory compensation technique [85]. (iv) Adaptive fuzzy control for a multiple-input multiple-output (MIMO) system [86]. (i) Stabilize the motion control of an AUV disturbed by unknown hydrodynamic coeﬃcients [87]. Adaptive (ii) Introducing an enhanced composite model reference adaptive control method to control AUV motion [88]. (iii) Adaptive control based on sliding mode control and fuzzy logic [89]. Sliding modes (i) Improved response, insensitive to parameter variation and disturbance [90–93]. (i) A bioinspired neurodynamic model is presented, used for a kinematic controller [94]. Neural networks (ii) Adaptive neural network controller combining hidden single-layer neural network and sliding mode control [95]. 22 Journal of Robotics control parameters. (ey have implemented three controllers, factors in the design such as hydrodynamic drag, propulsion PID, PID plus Dob, and the ADOB (adaptive controller based force, and energy consumption, giving room to achieving better results with further study. It is important to mention on a disturbance observer), to compare and evaluate the eﬃciency and performance, as shown in Figure 43. that the biomimetic form of a robot not only implies im- provements for itself but also reduces the degree of risk to possible alterations to a natural biological environment at 5.2.3. Neural Networks. Recently, neural networks have the time of the interaction. (e constant improvement of gained considerable attention in robotic systems control due biomimetic technology has broken the trend of only to their versatile properties, such as nonlinear mapping, implementing robots based on propulsion by caudal or learning ability, and parallel processing [67, 69, 73]. (e pectoral ﬁn; studies have opened a new window for the use of most useful feature of neural networks in control is their intelligent actuators, materials capable of providing better ability to approximate arbitrary linear or nonlinear mapping mechanical characteristics, such as greater ﬂexibility under through learning. Due to this property, neural networks have speciﬁc conditions, getting closer to the eﬃciency of real been proven to be a suitable tool to control complex non- biological models with diverse morphological characteris- linear dynamic systems. However, due to their arithmetic tics. (e use of a pressure sensor has become much more complexity, their implementation in engineering is not easy. standardized in the manufacture of any UUV, simply to obtain the depth data. However, some of these robots still have a dedicated depth sensor, thus achieving a greater 5.2.4. Fuzzy Control. Control based on fuzzy logic or fuzzy comparison range between points. In most current ROVs, control (FC) in English is a control that has supplanted we can observe the constant use of an IMU sensor with the conventional technologies in many applications combination of sonar to ﬁnd the underwater positioning, [35, 68, 74–76]. An important property of fuzzy logic is its also applying a ﬁlter for the correct interpretation of data. ability to express ambiguity in human thought. (erefore, GPS modules are used more in AUVs than in ROVs, because when the mathematical model of the process does not exist ROVs present a physical connection between the robot and or does exist with uncertainties, the FC becomes an alter- controller, while the AUVs are programmed with a path or native way for dealing with the unknown process. However, route to follow; that is why they emerge to the surface to the large number of fuzzy rules for high-order systems obtain their position before making a submersion. (e use of makes the analysis complex. A fuzzy-based depth control brushless motors has become very popular with the inte- scheme is illustrated in Figure 44 and a fuzzy-based yaw gration of propellers. It is found in diﬀerent types of UUVs angle control scheme is illustrated in Figure 45. long before the 20th century. (e advantage of this type of Table 4 summarizes the main contributions of some motor is adequate cost, better quality, and less maintenance additional navigation and orientation control methods that than other motors. (e application of new sensors for the correspond to those most used by UUVs. In the ﬁrst row of acquisition of oceanographic data in robots has become Table 4, some linear methods of proportional-integral-de- increasingly common, as a result, mainly due to the growing rivative (PID) type are included, which work in conjunction interest in the study of marine ecosystems and the con- with the other previously reviewed methods. servation of species. Many of the works reviewed, related to the control of direction or displacement with diﬀerent en- 6. Conclusions gines, do not show much detail in the electronic components (e ROVs ﬁrst shape was rectangular with an open hull and used, making it diﬃcult to trace an evolutionary timeline of emerging technologies of electronic components used in positive buoyancy; it was so big that it could not be transported by a single person, and it was necessary to place UUVs. (e methods and algorithms for navigating UUVs a pulley in the water. (e torpedo-shaped underwater ro- are mainly implemented in autonomous vehicles (AUVs). bot’s design allowed robots to be faster during their un- You can see the trend towards map-based navigation derwater operation. Inside the investigated robots, it was methods as opposed to those that use ﬁxed beacons around their exploration environment. (e orientation and move- found that only four thrusters provide 5 degrees of freedom compared to others that need six thrusters to reach 5 de- ment control is applicable for both ROVs and AUVs, highlighting the routes control and trajectory tracking to- grees. (e studied robots determine that, to achieve all the degrees of freedom, the robots must have eight thrusters wards autonomous vehicles. (e trend of new control methods is to apply combinations of more than one method installed or ﬁve thrusters with two actuators to change the force direction. One of the most used materials in the to improve their characteristics and achieve ﬁner control. manufacture of aquatic robots is aluminum, because it does not deform at high pressures; it is a dense material and is not Conflicts of Interest corrosive. Most of the researches evaluated are designed to operate under positive buoyancy; to be able to submerge, (e authors declare that they have no conﬂicts of interest. they must activate the immersion thrusters; and to return to the surface it is enough that the thrusters are deactivated. Acknowledgments (e improvement of aerodynamic and hydrodynamic characteristics of aquatic robots with a biomimetic approach (is research work is part of the project identiﬁed with the has accumulated great results, improving very important study BM-PNIPA-PES-SIADE-PP (No. 000027; Contract Journal of Robotics 23 Mechatronics and Automation, pp. 1382–1387, IEEE, Tianjin, No. 256-2018), which is supported by PNIPA and World China, August 2014. Bank. (e authors thank companies MASTER PROVIDER [17] F. Maurelli, J. Cartwright, N. Johnson, and Y. Petillot, “Nessie and ACEMAR for help. iv autonomous underwater vehicle wins the sauc-e compe- tition,” in Proceedings of the 10th International Conference on References Mobile Robots and Competitions (ROBOTICA 2010), Leiria, Portugal, March 2010. [1] J. E. Cohen, C. Small, A. Mellinger, J. Gallup, and J. Sachs, [18] R. Yang, B. Clement, A. Mansour, H. Li, M. Li, and N. Wu, “Estimates of coastal populations,” Science, vol. 278, no. 5341, “Modeling of a complex-shaped underwater vehicle,” in pp. 1209–1213, 1997. Proceedings of the 2014 IEEE International Conference on [2] J. Yuh, “Design and develop of an autonomous underwater Autonomous Robot Systems and Competitions (ICARSC), vehicle: a survey,” Autonomous Robots, vol. 8, no. 1, pp. 7–24, pp. 36–41, IEEE, Espinho, Portugal, May 2014. [19] S. Hidaka, S. Kawashima, S. Nam et al., “System design and [3] E. V. Lewis, “Principles of naval architecture,” Seakeeping and hardware development of autonomous underwater robot Controllability, vol. 3, Society of Naval Architects & Marine “daryabird,” in Proceedings of the AUVSI&ONR’s 18th Annual Engineers, Jersey City, NJ, USA, 1988. RoboSub Competition Journal Paper, San Diego, CA, USA, [4] S. Chutia, N. M. Kakoty, and D. Deka, “A review of underwater July 2015. robotics, navigation, sensing techniques and applications,” in [20] E. H. Binugroho, B. Setyawan, R. S. Dewanto, Proceedings of the Advances in Robotics, pp. 1–6, New Delhi, D. Pramadihanto et al., “Static and dynamic analysis of erov India, July 2017. mechanical design using CFD,” in Proceedings of the 2019 [5] B. K. Sahu and B. Subudhi, “(e state of art of autonomous International Electronics Symposium (IES), pp. 628–632, IEEE, underwater vehicles in current and future decades,” in Pro- Surabaya, Indonesia, September 2019. ceedings of the 2014 First International Conference on Auto- [21] Z. Yunhui, S. Lina, Q. Lu, and F. Qiang, “(e frame structure mation, Control, Energy and Systems (ACES), pp. 1–6, IEEE, of underwater Robot for shallow water,” in Proceedings of the Adisaptagram, India, February 2014. 2019 International Conference on Precision Machining, Non- [6] R. B. Wynn, V. A. I. Huvenne, T. P. Le Bas et al., “Auton- traditional Machining and Intelligent Manufacturing (PNTIM omous underwater vehicles (AUVS): their past, present and 2019), Atlantis Press, Xi’an, China, September 2019. future contributions to the advancement of marine geo- [22] W.-S. Chu, K.-T. Lee, S.-H. Song et al., “Review of biomimetic science,” Marine Geology, vol. 352, pp. 451–468, 2014. underwater robots using smart actuators,” International [7] R. D. Christ, “(e ROV manual a user guide for remotely Journal of Precision Engineering and Manufacturing, vol. 13, operated vehicles,” in (e ROV Manual, pp. 57-58, IEEE, no. 7, pp. 1281–1292, 2012. Piscataway, NJ, USA, 2014. [23] S.-H. Ahn, K.-T. Lee, H.-J. Kim, R. Wu, J.-S. Kim, and [8] C. von Alt, B. Allen, T. Austin, and R. Stokey, “Remote en- S.-H. Song, “Smart soft composite: an integrated 3D soft vironmental measuring units,” in Proceedings of the IEEE morphing structure using bend-twist coupling of anisotropic Symposium on Autonomous Underwater Vehicle Technology materials,” International Journal of Precision Engineering and (AUV’94), pp. 13–19, IEEE, Cambridge, MA, USA, July 1995. Manufacturing, vol. 13, no. 4, pp. 631–634, 2012. [9] B. Allen, R. Stokey, T. Austin et al., “Remus: a small, low cost [24] W.-S. C. Hyung-Soo Kim, J.-Y. Lee, and S.-H. Ahn, “Design AUV; system description, ﬁeld trials and performance re- and fabrication of soft morphing ray propulsor: undulator and sults,” in Proceedings of the Oceans’97. MTS/IEEE Conference, oscillator,” Soft Robotics, vol. 4, no. 1, pp. 49–60, 2016. pp. 994–1000, IEEE, Halifax, Canada, October 1997. [25] F. Cairone, S. Gagliano, D. C. Carbone, G. Recca, and [10] M. Radojevic, M. M. Nawaf, F. Maurelli et al., Nessie VI M. Bucolo, “Micro-optoﬂuidic switch realized by 3D printing Autonomous Underwater Vehicle, 2011, https://braincadet. technology,” Microﬂuid and Nanoﬂuidics, vol. 20, 2016. com/papers/nessie11.pdf. [26] J. Liu, Z. Wu, and J. Yu, “Design and implementation of a [11] X. Xiang, C. Yu, and Q. Zhang, “On intelligent risk analysis robotic dolphin for water quality monitoring,” in Proceedings and critical decision of underwater robotic vehicle,” Ocean of the 2016 IEEE International Conference on Robotics and Engineering, vol. 140, pp. 453–465, 2017. Biomimetics (ROBIO), pp. 835–840, IEEE, Qingdao, China, [12] X. Ai, S. Kang, and W. Chou, “System design and experiment December 2016. of the hybrid underwater vehicle,” in Proceedings of the 2018 [27] X. Yang, Z. Wu, and J. Yu, “Design and implementation of a International Conference on Control and Robots (ICCR), Hong robotic shark with a novel embedded vision system,” in Kong, China, September 2018. Proceedings of the 2016 IEEE International Conference on [13] G. Yao, J. Liang, T. Wang et al., “Development of a turtle-like Robotics and Biomimetics (ROBIO), pp. 841–846, IEEE, underwater vehicle using central pattern generator,” in Pro- Qingdao, China, December 2016. ceedings of the 2013 IEEE International Conference on Robotics [28] J. Sverdrup-(ygeson, E. Kelasidi, K. Y. Pettersen, and and Biomimetics (ROBIO), Shenzhen, China, December 2013. J. T. Gravdahl, “(e underwater swimming manipulator-a [14] X. Lin and S. Guo, “Development of a spherical underwater bio-inspired AUV,” in Proceedings of the 2016 IEEE/OES robot equipped with multiple vectored water-jet-based thrusters,” Journal of Intelligent & Robotic Systems, vol. 67, Autonomous Underwater Vehicles (AUV), pp. 387–395, IEEE, Tokyo, Japan, November 2016. no. 3-4, pp. 307–321, 2012. [15] S. Guo, Y. He, L. Shi et al., “Modeling and experimental [29] F. Hoﬀmann and A. B. Kesel, “Biologically inspired optimi- zation of underwater vehicles hull geometries and ﬁn pro- evaluation of an improved amphibious robot with compact structure,” Robotics and Computer-Integrated Manufacturing, pulsion,” in Proceedings of the OCEANS 2019, pp. 1–4, IEEE, Marseille, France, June 2019. vol. 51, pp. 37–52, 2018. [16] Y. He, S. Guo, L. Shi, S. Pan, and Z. Wang, “3D printing [30] S. Du, Z. Wu, J. Wang, S. Qi, and J. Yu, “Design and control of a two-motor-actuated tuna-inspired robot system,” IEEE technology-based an amphibious spherical robot,” in Pro- ceedings of the 2014 IEEE International Conference on Transactions on Systems, Man, and Cybernetics: Systems, 2019. 24 Journal of Robotics [31] H. Saha and Parhi, “Conceptual design of an underwater Journal of Nuclear Science and Technology, vol. 55, no. 3, robot,” in 2015 International Conference on Man and Machine pp. 1–10, 2017. Interfacing (MAMI), pp. 1–6, IEEE, Bhubaneswar, India, [48] M. D. Ageev, A. A. Boreyko, V. E. Gornak et al., “Modernized December 2015. TSL-underwater robot for tunnel and shallow-water inspec- [32] O. L. Osen, R.-I. Sandvik, V. Rogne, and H. Zhang, “A novel tion,” Institute for Marine Techonology Problems FEB RAS, low cost rov for aquaculture application,” in Proceedings of the Vladivostok, Russia, 2000. [49] V. Bobkov, M. Sporyshev, and A. Scherbatyuk, “Evaluation of OCEANS 2017, pp. 1–7, IEEE, Anchorage, Scotland, June 2017. the eﬀectivenes of the visual navigation technique for an autonomous underwater vehicle,” in Proceedings of the 2018 [33] M. Yang, Z. Sheng, Y. Che, J. Hu, K. Hu, and Y. Du, “Design of small monitoring rov for aquaculture,” in Proceedings of the OCEANS—MTS/IEEE Kobe Techno-Ocean (OTO), IEEE, OCEANS 2019, pp. 1–9, IEEE, Marseille, France, June 2019. Kobe, Japan, May 2018. [34] Z. M. Zain, M. M. Noh, K. A. Ab Rahim, and N. Harun, [50] M. Pedersen, J. Bruslund, R. Gade, and T. Moeslund, “De- “Design and development of an X4-ROV,” in Proceedings of tection of marine animals in a new underwater dataset with the 2016 IEEE International Conference on Underwater System varying visibility,” in Section of Biology and Enviromental Technology: (eory and Applications (USYS), pp. 207–211, ScienceAalborg University, Aalborg, Denmark, 2019. [51] C. Yeng, S. Guo, L. Shi, and J. Du, “Characteristics evaluation IEEE, Penang, Malaysia, December 2016. [35] C. S. Chin and W. P. Lin, “Robust genetic algorithm and fuzzy of the vertical motion of a spherical underwater,” in Pro- ceeding of the 2012 IEEE International Conference on Robotics inference mechanism embedded in a sliding-mode controller for an uncertain underwater robot,” IEEE/ASME Transactions and Biomimetics, Guangzhou, China, December 2012. [52] B. W. D. R. Yoeger and A. M. Bradley, (e Autonomous on Mechatronics, vol. 23, no. 2, pp. 655–666, 2018. [36] R. Jehangir, Bluerov2 Heavy Conﬁguration, Seascape Subsea Benthic explorer (ABE): an AUV Optimized for Deep Seaﬂoor B.V., Sleeuwijk, Netherlands, 2018. Studies, Woods Hole Oceanographic Institution, Dept. of [37] C. Yue, S. Guo, M. Li, Y. Li, H. Hirata, and H. Ishihara, Applied Ocean Physics and Engineering, Falmouth, MA, Mechatronic System and Experiments of a Spherical Underwater USA, 1991. Robot: SUR-II, Springer Science+Business Media, Dordrecht, [53] M. Doniec, I. Vasilescu, C. Detweiler, and D. Rus, “Complete Netherlands, 2015. SE3 underwater robot control with arbitrary thruster con- [38] F. C. J. T. Pennington and M. Blum, “Seawater sampling by an ﬁgurations,” in Proceedings of the 2010 IEEE International autonomous underwater vehicle: “gulper” sample validation Conference on Robotics and Automation Anchorage Conven- for nitrate, chlorophyll, phytoplankton, and primary pro- tion District, p. 6, IEEE, Anchorage, AL, USA, May 2010. duction,” Limnology and Oceanography: Methods, vol. 14, [54] R. Saltaren, C. Garcia, C. Alvarez, and R. Aracil, “Experiences no. 1, pp. 14–23, 2016. and results from designing and developing a 6 DOF under- [39] M. Choyekh, “Development and operation of underwater water parallel robot,” Robotics and Autonomous Systems, robot for autonomous tracking and monitoring of subsea vol. 59, no. 2, pp. 101–112, 2010. plumes after oil spill and gas leak from seabed, and analyses of [55] J. Choi, T. Lee, T. Kim, J. Jung, and H. Choi, “Development of measured data,” Doctoral thesis, Osaka University Knowledge a ROV for visual inspection of harbor structures,” in Pro- ceedings of the 2017 IEEE Underwater Technology (UT), Busan, Archive, School of Engineering Osaka University, Osaka, Japan, 2016. South Korea, February 2017. [56] P. L. B. Jun and H. Shim, “Approximated generalized torques [40] F. Sapuppo, F. Schembri, L. Fortuna, and M. Bucolo, “Microﬂuidic circuits and systems,” IEEE Circuits and Systems by the hydrodynamic forces acting on legs of an underwater Magazine, vol. 9, no. 3, pp. 6–19, 2009. walking robot,” in International Journal of Ocean System [41] T. Fukuba, T. Noguchi, K. Okamura, and T. Fujii, “Adenosine Engineering, vol. 1, no. 4, pp. 222–229, IEEE, 2011. triphosphate measurement in deep sea using a microﬂuidic [57] K. P. Valavanis, D. Gracanin, M. Matijasevic, R. Kolluru, and G. A. Demetriou, “Control architectures for autonomous device,” Micromachines, vol. 9, no. 8, 2018. [42] F. S. Florinda Schembri and M. Bucolo, “Experimental underwater vehicles,” IEEE Control Systems Magazine, vol. 17, no. 6, pp. 48–64, 1997. classiﬁcation of nonlinear dynamics in microﬂuidic bubbles’ ﬂow,” Nonlinear Dinamics, vol. 4, pp. 2807–2819, 2012. [58] Y. Zhang and L. Cheng, “A distributed protocol for multi-hop underwater robot positioning,” in Proceedings of the 2004 [43] A. Limanes, J. Cartwright, B. Davis et al., Nessie III Auton- omous Underwater VehicleHeriot-Watt University, Scotland, IEEE International Conference on Robotics and Biomimetics, UK, 2008. pp. 480–484, IEEE, Dali, China, December 2004. [44] M. B. Bo Zhao, S. Roger, and F. Dukan, “Particle ﬁlter for fault [59] Z. Yan, S. Peng, J. Zhou, J. Xu, and H. Jia, “Research on an diagnosis and robust navigation of underwater robot,” Improved dead reckoning for AUV navigation,” in Pro- Transactions on Control Systems Technology, vol. 22, no. 6, ceedings of the 2010 Chinese Control and Decision Conference, 2014. pp. 1793–1797, IEEE, Xuzhou, China, May 2010. [45] S. Mintchev, E. Donati, S. Marrazza, and C. Stefanini, [60] I. Klein and R. Diamant, “Dead reckoning for trajectory es- Mechatronic Design of a Miniature Underwater Robot for timation of underwater drifters under water currents,” Journal of Marine Science and Engineering, vol. 8, no. 3, p. 205, 2020. Swarm Operations, Hong Kong Convention and Exhibition Center, Hong Kong, China, 2014. [61] J. H. Kepper IV, B. C. Claus, and J. C. Kinsey, “A navigation solution using a MEMS IMU, model-based dead-reckoning, [46] A. Jebelli, M. Yagoub, and B. Dhillon, “Design and control of underwater robots with rotating thrusters,” IAES Interna- and one-way-travel-time acoustic range measurements for tional Journal of Robotics and Automation, vol. 5, no. 4, autonomous underwater vehicles,” IEEE Journal of Oceanic pp. 284–294, 2016. Engineering, vol. 44, no. 3, pp. 664–682, 2018. [47] G. Y. M. Dong and W. Chou, “A new navigation strategy for [62] J. J. Leonard and A. Bahr, “Autonomous underwater vehicle underwater robot in reactor pool combined propeller speed navigation,” in Springer Handbook of Ocean Engineering, detection and dynamics analysis with sonar data correction,” pp. 341–358, Springer, Cham, Switzerland, 2016. Journal of Robotics 25 [63] A. Alcocer, P. Oliveira, and A. Pascoal, “Study and imple- [79] W. Zhang, H. Wang, X. Bian, Z. Yan, and G. Xia, “(e ap- mentation of an ekf gib-based underwater positioning sys- plication of self-tuning fuzzy PID control method to recov- tem,” Control Engineering Practice, vol. 15, no. 6, pp. 689–701, ering AUV,” in Proceedings of the 2012 OCEANS, pp. 1–5, 2007. IEEE, Virginia Beach, VA, USA, October 2012. [64] M. Carreras, P. Ridao, R. Garc´ıa, and T. Nicosevici, “Vision- [80] N. H. Z. M. Zain, “A backstepping based PID controller for based localization of an underwater robot in a structured stabilizing an underactuated X4-AUV,” ARPN Journal of Engineering and Applied Sciences, vol. 10, no. 21, 2015. environment,” vol. 1, pp. 971–976, in Proceedings of the 2003 IEEE International Conference on Robotics and Automation [81] M. H. Khodayari and S. Balochian, “Modeling and control of autonomous underwater vehicle (AUV) in heading and depth (Cat. No. 03CH37422), vol. 1, pp. 971–976, IEEE, Taipei, China, September 2003. attitude via self-adaptive fuzzy PID controller,” Journal of [65] L. Fortuna, A. Buscarino, M. Frasca, and C. Famoso, “Chapter Marine Science and Technology, vol. 20, no. 3, pp. 559–578, 1: imperfect uncertain systems,” in Control of Imperfect 2015. Nonlinear Electromechanical Large Scale Systems, World [82] R. Hernandez-Alvarado, ´ L. G. Garc´ıa-Valdovinos, T. Salgado- Jimenez, ´ A. Gomez-Espinosa, ´ and F. Fonseca-Navarro, Scientiﬁc Series on Nonlinear Science, Springer, Boston, MA, USA, 2017. “Neural network-based self-tuning PID control for under- water vehicles,” Sensors, vol. 16, no. 9, p. 1429, 2016. [66] M. Bucolo, A. Buscarino, C. Famoso, L. Fortuna, and M. Frasca, “Control of Imperfect dynamical systems,” Non- [83] L. Perera, J. Carvalho, and C. G. Soares, “Fuzzy logic based decision making system for collision avoidance of ocean linear Dynamics volume, vol. 98, pp. 2989–2999, 2019. [67] T. Chatchanayuenyong and M. Parnichkun, “Neural network navigation under critical collision conditions,” Journal of based-time optimal sliding mode control for an autonomous Marine Science and Technology, vol. 16, no. 1, pp. 84–99, 2011. underwater robot,” Mechatronics, vol. 16, no. 8, pp. 471–478, [84] M. Aras, M. Shahrieel, H. I. Jaafar, and M. K. Anuar, “Tuning 2006. process of single input fuzzy logic controller based on linear [68] S. Guo, J. Du, X. Lin, and C. Yue, “Adaptive fuzzy sliding control surface approximation method for depth control of mode control for spherical underwater robots,” in Proceedings underwater remotely operated vehicle,” Journal of Engineering and Applied Sciences, vol. 8, no. 6, pp. 208–214, 2013. of the 2012 IEEE International Conference on Mechatronics and Automation, pp. 1681–1685, IEEE, Chengdu, China, [85] Y. Chen, R. Zhang, X. Zhao, and J. Gao, “Tracking control of underwater vehicle subject to uncertainties using fuzzy in- August 2012. [69] M.-j. Zhang and Z.-z. Chu, “Adaptive sliding mode control verse desired trajectory compensation technique,” Journal of based on local recurrent neural networks for underwater Marine Science and Technology, vol. 21, no. 4, pp. 624–650, robot,” Ocean Engineering, vol. 45, pp. 56–62, 2012. 2016. [70] S. Jin, J. Bak, J. Kim, T. Seo, and H. S. Kim, “Switching pd- [86] H. F. Ghavidel and A. A. Kalat, “Robust control for mimo based sliding mode control for hovering of a tilting-thruster hybrid dynamical system of underwater vehicles by composite underwater robot,” PLoS One, vol. 13, no. 3, 2018. adaptive fuzzy estimation of uncertainties,” Nonlinear Dy- [71] A. J. Healey and D. Lienard, “Multivariable sliding mode namics, vol. 89, no. 4, pp. 2347–2365, 2017. control for autonomous diving and steering of unmanned [87] H. R. Kooﬁgar, “Robust adaptive motion control with envi- ronmental disturbance rejection for perturbed underwater underwater vehicles,” IEEE Journal of Oceanic Engineering, vol. 18, no. 3, pp. 327–339, 1993. vehicles,” Journal of Marine Science and Technology, vol. 22, [72] S. Zhao and J. Yuh, “Experimental study on advanced un- no. 4, pp. 455–462, 2014. derwater robot control,” IEEE Transactions on Robotics, [88] C. D. Makavita, H. D. Nguyen, D. Ranmuthugala, and vol. 21, no. 4, pp. 695–703, 2005. S. G. Jayasinghe, “Composite model reference adaptive [73] M. Chen, B. Jiang, J. Zou, and X. Feng, “Robust adaptive control for an unmanned underwater vehicle,” Underwater tracking control of the underwater robot with input non- Technology, vol. 33, no. 2, pp. 81–93, 2015. linearity using neural networks,” International Journal of [89] S. Pezeshki, A. Ghiasi, M. Badamchizadeh, and K. Sabahi, Computational Intelligence Systems, vol. 3, no. 5, pp. 646–655, “Adaptive robust control of autonomous underwater vehicle,” Journal of Control, Automation and Electrical Systems, vol. 27, [74] I. Akkizidis, G. Roberts, P. Ridao, and J. Batlle, “Designing a no. 3, pp. 250–262, 2016. fuzzy-like pd controller for an underwater robot,” Control [90] A. Bagheri and J. J. Moghaddam, “Simulation and tracking Engineering Practice, vol. 11, no. 4, pp. 471–480, 2003. control based on neural-network strategy and sliding-mode [75] X. Xiang, C. Yu, and Q. Zhang, “Robust fuzzy 3d path fol- control for underwater remotely operated vehicle,” Neuro- lowing for autonomous underwater vehicle subject to un- computing, vol. 72, no. 7–9, pp. 1934–1950, 2009. certainties,” Computers & Operations Research, vol. 84, [91] D. Zhu and B. Sun, “(e bio-inspired model based hybrid pp. 165–177, 2017. sliding-mode tracking control for unmanned underwater [76] C. Yu, X. Xiang, L. Lapierre, and Q. Zhang, “Nonlinear vehicles,” Engineering Applications of Artiﬁcial Intelligence, guidance and fuzzy control for three-dimensional path fol- vol. 26, no. 10, pp. 2260–2269, 2013. lowing of an underactuated autonomous underwater vehicle,” [92] J. Xu, M. Wang, and L. Qiao, “Dynamical sliding mode Ocean Engineering, vol. 146, pp. 457–467, 2017. control for the trajectory tracking of underactuated un- [77] I. Kaminer, A. Pascoal, E. Hallberg, and C. Silvestre, “Tra- manned underwater vehicles,” Ocean Engineering, vol. 105, jectory tracking for autonomous vehicles: an integrated ap- pp. 54–63, 2015. proach to guidance and control,” Journal of Guidance, [93] Z. Chu and D. Zhu, “Adaptive sliding mode heading control Control, and Dynamics, vol. 21, no. 1, pp. 29–38, 1998. for autonomous underwater vehicle including actuator dy- [78] G. Antonelli, F. Caccavale, S. Chiaverini, and G. Fusco, “A namics,” in Proceedings of the OCEANS 2016, pp. 1–5, IEEE, novel adaptive control law for underwater vehicles,” IEEE Shanghai, China, September 2016.. Transactions on Control Systems Technology, vol. 11, no. 2, [94] B. Sun, D. Zhu, F. Ding, and S. X. Yang, “A novel tracking pp. 221–232, 2003. control approach for unmanned underwater vehicles based on 26 Journal of Robotics bio-inspired neurodynamics,” Journal of Marine Science and Technology, vol. 18, no. 1, pp. 63–74, 2013. [95] J. Gao, P. Wu, B. Yang, and F. Xia, “Adaptive neural network control for visual servoing of underwater vehicles with pose estimation,” Journal of Marine Science and Technology, vol. 22, no. 3, pp. 470–478, 2017. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Robotics Hindawi Publishing Corporation http://www.deepdyve.com/lp/hindawi-publishing-corporation/review-on-unmanned-underwater-robotics-structure-designs-materials-Tg5xsOAFYP

Loading next page...

References (107)

Shuxiang Guo, Juan Du, Xichuan Lin, Chunfeng Yue (2012)
Adaptive fuzzy sliding mode control for spherical underwater robots
2012 IEEE International Conference on Mechatronics and Automation
Mahdi Choyekh, N. Kato, Y. Yamaguchi, Ryan Dewantara, H. Chiba, H. Senga, M. Yoshie, Toshinari Tanaka, E. Kobayashi, Timothy Short (2017)
Development and operation of underwater robot for autonomous tracking and monitoring of subsea plumes after oil spill and gas leak from seabed, and analyses of measured data
B. Sahu, B. Subudhi (2014)
The state of art of Autonomous Underwater Vehicles in current and future decades
2014 First International Conference on Automation, Control, Energy and Systems (ACES)
A. Alcocer, P. Oliveira, A. Pascoal (2004)
Study and implementation of an EKF GIB-based underwater positioning system
Control Engineering Practice, 15
I. Akkizidis, G. Roberts, P. Ridao, J. Batlle (2003)
Designing a Fuzzy-like PD controller for an underwater robot
Control Engineering Practice, 11
J. Sverdrup-Thygeson, E. Kelasidi, K. Pettersen, J. Gravdahl (2016)
The underwater swimming manipulator - a bio-inspired AUV
2016 IEEE/OES Autonomous Underwater Vehicles (AUV)
Hesam Ghavidel, A. Kalat (2017)
Robust control for MIMO hybrid dynamical system of underwater vehicles by composite adaptive fuzzy estimation of uncertainties
Nonlinear Dynamics, 89
Bo Zhao, R. Skjetne, M. Blanke, Fredrik Dukan (2014)
Particle Filter for Fault Diagnosis and Robust Navigation of Underwater Robot
IEEE Transactions on Control Systems Technology, 22
Zhang Yunhui, S. Lina, Qiao Lu, Feng Qiang (2019)
The Frame Structure of Underwater Robot for Shallow Water
Proceedings of the 2019 International Conference on Precision Machining, Non-Traditional Machining and Intelligent Manufacturing (PNTIM 2019)
Mingjun Zhang, Zhenzhong Chu (2012)
Adaptive sliding mode control based on local recurrent neural networks for underwater robot
Ocean Engineering, 45
F. Sapuppo, F. Schembri, Luigi Fortuna, M. Bucolo (2009)
Microfluidic circuits and systems
IEEE Circuits and Systems Magazine, 9
S. Mintchev, Elisa Donati, S. Marrazza, C. Stefanini (2014)
Mechatronic design of a miniature underwater robot for swarm operations
2014 IEEE International Conference on Robotics and Automation (ICRA)
(2014)
e ROV manual a user guide for remotely operated vehicles
A. Jebelli, M. Yagoub, B. Dhillon (2016)
Design and Control of Underwater Robots with Rotating Thrusters
, 5
F. Schembri, F. Sapuppo, M. Bucolo (2012)
Experimental classification of nonlinear dynamics in microfluidic bubbles’ flow
Nonlinear Dynamics, 67
L. Perera, Joao Carvalho, C. Soares (2011)
Fuzzy logic based decision making system for collision avoidance of ocean navigation under critical collision conditions
Journal of Marine Science and Technology, 16
F. Hoffmann, A. Kesel (2019)
Biologically inspired optimization of underwater vehicles hull geometries and fin propulsion
OCEANS 2019 - Marseille
Z. Sheng M. Yang
Design of small monitoring rov for aquaculture,
Proceedings of the OCEANS 2019
M. Aras, S. Abdullah, H. Jaafar, M. Aziz, A. Kassim (2013)
Tuning process of single input fuzzy logic controller based on linear control surface approximation method for depth control of underwater remotely operated vehicle
Journal of Engineering and Applied Sciences, 8
Zheping Yan, Shuping Peng, Jiajia Zhou, Jian Xu, Heming Jia (2010)
Research on an improved dead reckoning for AUV navigation
2010 Chinese Control and Decision Conference
A. Buscarino L. Fortuna
Chapter 1: imperfect uncertain systems,
Control of Imperfect Nonlinear Electromechanical Large Scale Systems, World Scientific Series on Nonlinear Science
K. Valavanis, D. Gračanin, M. Matijasevic, R. Kolluru, Georgios Demetriou (1997)
Control architectures for autonomous underwater vehicles
IEEE Control Systems Magazine, 17
(2018)
“Evaluation of the eﬀectivenes of the visual navigation technique for an autonomous underwater vehicle,”
Ryota Nakanishi, Jonghyun Ahn, Katsuaki Suzuki, Yuki Soejima, K. Masuda, Yuya Nishida (2015)
System Design and Hardware Development of Autonomous Underwater Robot “ DaryaBird ”
R. Hernández-Alvarado, L. García-Valdovinos, T. Salgado-Jiménez, A. Gómez-Espinosa, Fernando Fonseca-Navarro (2016)
Neural Network-Based Self-Tuning PID Control for Underwater Vehicles
Sensors (Basel, Switzerland), 16
C. Alt, B. Allen, T. Austin, R. Stokey (1994)
Remote environmental measuring units
Proceedings of IEEE Symposium on Autonomous Underwater Vehicle Technology (AUV'94)
Xiang Yang, Zhengxing Wu, Junzhi Yu (2016)
Design and implementation of a robotic shark with a novel embedded vision system
2016 IEEE International Conference on Robotics and Biomimetics (ROBIO)
J. Yuh (2000)
Design and Control of Autonomous Underwater Robots: A Survey
Autonomous Robots, 8
R. Wynn, V. Huvenne, T. Bas, B. Murton, D. Connelly, B. Bett, H. Ruhl, K. Morris, J. Peakall, D. Parsons, E. Sumner, S. Darby, R. Dorrell, J. Hunt (2014)
Autonomous Underwater Vehicles (AUVs): Their past, present and future contributions to the advancement of marine geoscience
Marine Geology, 352
M. Carreras, P. Ridao, Rafael García, T. Nicosevici (2003)
Vision-based localization of an underwater robot in a structured environment
2003 IEEE International Conference on Robotics and Automation (Cat. No.03CH37422), 1
(2018)
Bluerov2 Heavy Configuration
Swagat Chutia, N. Kakoty, D. Deka (2017)
A Review of Underwater Robotics, Navigation, Sensing Techniques and Applications
B. Jiang M. Chen (2010)
Robust adaptive tracking control of the underwater robot with input nonlinearity using neural networks,
International Journal of Computational Intelligence Systems, 3
F. C. J. T. Pennington and M. Blum (2016)
Seawater sampling by an autonomous underwater vehicle: “gulper
Limnology and Oceanography: Methods, 14
E. Binugroho, B. Setyawan, Wafiqqurochman, R. Dewanto, D. Pramadihanto (2019)
Static and Dynamic Analysis of eROV Mechanical Design using CFD
2019 International Electronics Symposium (IES)
R. Saltarén, Cecilia Cena, Cesar Arocha, R. Aracil (2011)
Experiences and results from designing and developing a 6 DoF underwater parallel robot
Robotics Auton. Syst., 59
Jincun Liu, Zhengxing Wu, Junzhi Yu (2016)
Design and implementation of a robotic dolphin for water quality monitoring
2016 IEEE International Conference on Robotics and Biomimetics (ROBIO)
Shuxiang Guo, Yanlin He, Liwei Shi, Shaowu Pan, Rui Xiao, Kun Tang, Ping Guo (2018)
Modeling and experimental evaluation of an improved amphibious robot with compact structure
Robotics and Computer-Integrated Manufacturing
Sangrok Jin, J. Bak, Jongwon Kim, Taewon Seo, H. Kim (2018)
Switching PD-based sliding mode control for hovering of a tilting-thruster underwater robot
PLoS ONE, 13
Homagni Saha, Parhi (2015)
Conceptual design of an underwater robot
2015 International Conference on Man and Machine Interfacing (MAMI)
C. Makavita, H. Nguyen, D. Ranmuthugala, S. Jayasinghe (2015)
Composite model reference adaptive control for an unmanned underwater vehicle
Underwater Technology, 33
M. Bucolo, A. Buscarino, Carlo Famoso, L. Fortuna, M. Frasca (2019)
Control of imperfect dynamical systems
Nonlinear Dynamics, 98
(1997)
By the sea ...
Jinwoo Choi, Yeongjun Lee, Tae-Jin Kim, Jongdae Jung, Hyun-Taek Choi (2017)
Development of a ROV for visual inspection of harbor structures
2017 IEEE Underwater Technology (UT)
Chunfeng Yue, Shuxiang Guo, Liwei Shi, Juan Du (2012)
Characteristics evaluation of the vertical motion of a spherical underwater robot
2012 IEEE International Conference on Robotics and Biomimetics (ROBIO)
Sung-hoon Ahn, Kyung-Tae Lee, Hyungjung Kim, Renzhe Wu, Ji-Soo Kim, Sung-Hyuk Song (2012)
Smart soft composite: An integrated 3D soft morphing structure using bend-twist coupling of anisotropic materials
International Journal of Precision Engineering and Manufacturing, 13
M. Ageev, A.A. Boreyko, V.E. Gornak, Y. Matvienko, A. Scherbatyuk, Yu.V. Vanlin, V. Zolotarev (2000)
Modernized TSL-underwater robot for tunnel and shallow-water inspection
Proceedings of the 2000 International Symposium on Underwater Technology (Cat. No.00EX418)
J. Cohen, C. Small, A. Mellinger, J. Gallup, J. Sachs (1997)
Estimates of Coastal Populations
Science, 278
Sheng Du, Zhengxing Wu, Jian Wang, Suwen Qi, Junzhi Yu (2021)
Design and Control of a Two-Motor-Actuated Tuna-Inspired Robot System
IEEE Transactions on Systems, Man, and Cybernetics: Systems, 51
M. Yagoub A. Jebelli (2016)
Design and control of underwater robots with rotating thrusters,
IAES International Journal of Robotics and Automation, 5
Xianbo Xiang, Caoyang Yu, Qin Zhang (2017)
On intelligent risk analysis and critical decision of underwater robotic vehicle
Ocean Engineering, 140
H. Wang W. Zhang
The application of self-tuning fuzzy PID control method to recovering AUV,
Proceedings of the 2012 OCEANS
W. Chu, Kyung-Tae Lee, Sung-Hyuk Song, Min-Woo Han, Jang-Yeob Lee, Hyungsoo Kim, Min-Soo Kim, Yong-Jai Park, Kyu-Jin Cho, Sung-hoon Ahn (2012)
Review of biomimetic underwater robots using smart actuators
International Journal of Precision Engineering and Manufacturing, 13
(1988)
Society of Naval Architects & Marine Engineers
Side Zhao, J. Yuh (2005)
Experimental study on advanced underwater robot control
IEEE Transactions on Robotics, 21
M. Radojević, M. Nawaf, F. Maurelli, Tummas Tomassson, Chengjiang Wang, Jonathan Clay, Y. Pétillot (2011)
Nessie VI Autonomous Underwater Vehicle
A. Bayer (2016)
Principles Of Naval Architecture
Marek Doniec, I. Vasilescu, Carrick Detweiler, D. Rus (2010)
Complete SE3 underwater robot control with arbitrary thruster configurations
2010 IEEE International Conference on Robotics and Automation
I. Klein, R. Diamant (2020)
Dead Reckoning for Trajectory Estimation of Underwater Drifters under Water Currents †
Journal of Marine Science and Engineering
J. Leonard, A. Bahr (2016)
Autonomous Underwater Vehicle Navigation
Mohammad Khodayari, S. Balochian (2015)
Modeling and control of autonomous underwater vehicle (AUV) in heading and depth attitude via self-adaptive fuzzy PID controller
Journal of Marine Science and Technology, 20
I. Vasilescu M. Doniec
Complete SE3 underwater robot control with arbitrary thruster configurations,
Proceedings of the 2010 IEEE International Conference on Robotics and Automation Anchorage Convention District
Ottar Osen, R. Sandvik, J⊘rgen Trygstad, Vegard Rogne, Houxiang Zhang (2017)
A novel low cost ROV for aquaculture application
OCEANS 2017 – Anchorage
Miao Yang, Zhibin Sheng, Yanhe Che, Jintong Hu, Ke Hu, Yixiang Du (2019)
Design of Small Monitoring ROV for Aquaculture
OCEANS 2019 - Marseille
Mou Chen, B. Jiang, J. Zou, Xing Feng (2010)
Robust Adaptive Tracking Control of the Underwater Robot with Input Nonlinearity Using Neural Networks
Int. J. Comput. Intell. Syst., 3
Yuecheng Zhang, Liang Cheng (2004)
A Distributed Protocol for Multi-hop Underwater Robot Positioning
2004 IEEE International Conference on Robotics and Biomimetics
Rui Yang, B. Clement, A. Mansour, Huajun Li, Ming Li, Nailong Wu (2014)
Modeling of a complex-shaped underwater vehicle
2014 IEEE International Conference on Autonomous Robot Systems and Competitions (ICARSC)
G. Antonelli, F. Caccavale, S. Chiaverini, G. Fusco (2003)
A novel adaptive control law for underwater vehicles
IEEE Trans. Control. Syst. Technol., 11
R. Stokey B. Allen
Remus: a small, low cost AUV; system description, field trials and performance results,
Proceedings of the Oceans’97. MTS/IEEE Conference
Jian Gao, Puguo Wu, Bo Yang, Fei Xia (2017)
Adaptive neural network control for visual servoing of underwater vehicles with pose estimation
Journal of Marine Science and Technology, 22
Guocai Yao, Jianhong Liang, Tianmiao Wang, Xingbang Yang, Qi Shen, Yucheng Zhang, Hailiang Wu, Weicheng Tian (2013)
Development of a turtle-like underwater vehicle using central pattern generator
2013 IEEE International Conference on Robotics and Biomimetics (ROBIO)
D. Yoerger, A. Bradley, B. Walden (2000)
The Autonomous Benthic Explorer ( ABE ) : An AUV Optimized for Deep Seafloor Studies
Hyungsoo Kim, Jang-Yeob Lee, W. Chu, Sung-hoon Ahn (2017)
Design and Fabrication of Soft Morphing Ray Propulsor: Undulator and Oscillator.
Soft robotics, 4 1
Jian Xu, Man Wang, Lei Qiao (2015)
Dynamical sliding mode control for the trajectory tracking of underactuated unmanned underwater vehicles
Ocean Engineering, 105
J. Cartwright F. Maurelli
Nessie iv autonomous underwater vehicle wins the sauc-e competition,
Proceedings of the 10th International Conference on Mobile Robots and Competitions (ROBOTICA 2010)
Mingjie Dong, Wusheng Chou, Guodong Yao (2018)
A new navigation strategy for underwater robot in reactor pool combined propeller speed detection and dynamics analysis with sonar data correction
Journal of Nuclear Science and Technology, 55
Harun Fadzillah, M. Zainah (2015)
A Backstepping Based Pid Controller for Stabilizing an Underactuated X4-AUV
F. Cairone, S. Gagliano, D. Carbone, G. Recca, M. Bucolo (2016)
Micro-optofluidic switch realized by 3D printing technology
Microfluidics and Nanofluidics, 20
Caoyang Yu, Xianbo Xiang, L. Lapierre, Qin Zhang (2017)
Nonlinear guidance and fuzzy control for three-dimensional path following of an underactuated autonomous underwater vehicle
Ocean Engineering, 146
J. Pennington, M. Blum, F. Chavez (2016)
Seawater sampling by an autonomous underwater vehicle: “Gulper” sample validation for nitrate, chlorophyll, phytoplankton, and primary production
Limnology and Oceanography: Methods, 14
C. Chin, W. Lin (2018)
Robust Genetic Algorithm and Fuzzy Inference Mechanism Embedded in a Sliding-Mode Controller for an Uncertain Underwater Robot
IEEE/ASME Transactions on Mechatronics, 23
Z. Chu and D. Zhu
Adaptive sliding mode heading control for autonomous underwater vehicle including actuator dynamics,
Proceedings of the OCEANS 2016
Yanlin He, Shuxiang Guo, Liwei Shi, Shaowu Pan, Zhe Wang (2014)
3D printing technology-based an amphibious spherical robot
2014 IEEE International Conference on Mechatronics and Automation
Bing Sun, Daqi Zhu, F. Ding, Simon Yang (2013)
A novel tracking control approach for unmanned underwater vehicles based on bio-inspired neurodynamics
Journal of Marine Science and Technology, 18
F. Hoffmann and A. B. Kesel
Biologically inspired optimization of underwater vehicles hull geometries and fin propulsion,
Proceedings of the OCEANS 2019
Xianbo Xiang, Caoyang Yu, Qin Zhang (2017)
Robust fuzzy 3D path following for autonomous underwater vehicle subject to uncertainties
Comput. Oper. Res., 84
B. Jun, H. Shim, Pan-Mook Lee (2011)
Approximated Generalized Torques by the Hydrodynamic Forces Acting on Legs of an Underwater Walking Robot
, 1
Wei Zhang, Hongtao Wang, X. Bian, Zheping Yan, G. Xia (2012)
The application of self-tuning fuzzy PID control method to recovering AUV
2012 Oceans
I. Kaminer, A. Pascoal, E. Hallberg, Carlos Silvestre (1998)
Trajectory Tracking for Autonomous Vehicles: An Integrated Approach to Guidance and Control
Journal of Guidance Control and Dynamics, 21
ZAINAH ZAIN, M. Noh, Khairul Rahim, N. Harun (2016)
Design and development of an X4-ROV
2016 IEEE International Conference on Underwater System Technology: Theory and Applications (USYS)
T. Fukuba, T. Noguchi, K. Okamura, T. Fujii (2018)
Adenosine Triphosphate Measurement in Deep Sea Using a Microfluidic Device
Micromachines, 9
J. Iv, B. Claus, J. Kinsey (2019)
A Navigation Solution Using a MEMS IMU, Model-Based Dead-Reckoning, and One-Way-Travel-Time Acoustic Range Measurements for Autonomous Underwater Vehicles
IEEE Journal of Oceanic Engineering, 44
Zhenzhong Chu, Daqi Zhu (2016)
Adaptive sliding mode heading control for autonomous underwater vehicle including actuator dynamics
OCEANS 2016 - Shanghai
Xichuan Lin, Shuxiang Guo (2012)
Development of a Spherical Underwater Robot Equipped with Multiple Vectored Water-Jet-Based Thrusters
Journal of Intelligent & Robotic Systems, 67
B. Allen, R. Stokey, T. Austin, N. Forrester, R. Goldsborough, M. Purcell, C. Alt (1997)
REMUS: a small, low cost AUV; system description, field trials and performance results
Oceans '97. MTS/IEEE Conference Proceedings, 2
Daqi Zhu, Bing Sun (2013)
The bio-inspired model based hybrid sliding-mode tracking control for unmanned underwater vehicles
Eng. Appl. Artif. Intell., 26
T. Chatchanayuenyong, M. Parnichkun (2006)
Neural network based-time optimal sliding mode control for an autonomous underwater robot
Mechatronics, 16
H. Koofigar (2013)
Robust Adaptive Motion Control with Environmental Disturbance Rejection for Perturbed Underwater Vehicles
Journal of Marine Science and Technology, 22
Chunfeng Yue, Shuxiang Guo, Maoxun Li, Yaxin Li, H. Hirata, H. Ishihara (2015)
Mechatronic System and Experiments of a Spherical Underwater Robot: SUR-II
Journal of Intelligent & Robotic Systems, 80
Yuan Chen, Rongmin Zhang, Xin-Gang Zhao, Jun Gao (2016)
Tracking control of underwater vehicle subject to uncertainties using fuzzy inverse desired trajectory compensation technique
Journal of Marine Science and Technology, 21
S. Pezeshki, A. Ghiasi, M. Badamchizadeh, K. Sabahi (2016)
Adaptive Robust Control of Autonomous Underwater Vehicle
Journal of Control, Automation and Electrical Systems, 27
A. Bagheri, J. Moghaddam (2009)
Simulation and tracking control based on neural-network strategy and sliding-mode control for underwater remotely operated vehicle
Neurocomputing, 72
A. Healey, D. Lienard (1993)
Multivariable sliding mode control for autonomous diving and steering of unmanned underwater vehicles
IEEE Journal of Oceanic Engineering, 18
Malte Pedersen, Joakim Haurum, Rikke Gade, T. Moeslund (2019)
Detection of Marine Animals in a New Underwater Dataset with Varying Visibility
Xiaojie Ai, Song Kang, Wusheng Chou (2018)
System Design and Experiment of the Hybrid Underwater Vehicle
2018 International Conference on Control and Robots (ICCR)
R.-I. Sandvik O. L. Osen
A novel low cost rov for aquaculture application,
Proceedings of the OCEANS 2017
J. Yuh (2000)
Design and develop of an autonomous underwater vehicle: a survey,
Autonomous Robots, 8

Publisher: Hindawi Publishing Corporation
Copyright: Copyright © 2021 Javier Neira et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
ISSN: 1687-9600
eISSN: 1687-9619
DOI: 10.1155/2021/5542920
Publisher site: See Article on Publisher Site

Abstract

Hindawi Journal of Robotics Volume 2021, Article ID 5542920, 26 pages https://doi.org/10.1155/2021/5542920 Review Article Review on Unmanned Underwater Robotics, Structure Designs, Materials, Sensors, Actuators, and Navigation Control Javier Neira, Cristhel Sequeiros, Richard Huamani, Elfer Machaca, Paola Fonseca, and Wilder Nina Veox Inc., Arequipa, Peru Correspondence should be addressed to Wilder Nina; wnina@veox.tech Received 21 February 2021; Accepted 7 June 2021; Published 6 July 2021 Academic Editor: L. Fortuna Copyright © 2021 Javier Neira et al. (is is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Since its beginning, around the 50s decade, until present days, the area of unmanned underwater vehicles (UUV) has considerably grown through time; those have been used for many tasks and applications, from bomb searching and recovery to sea exploration. Initially, these robots were used mainly for military and scientiﬁc purposes. However, nowadays, they are very much extended into civils, and it is not hard to ﬁnd them being used for recreation. In this context, the present research is an eﬀort to make a walkthrough of evolution in this area, showing a diversity of structure designs, used materials, sensor and instrumentation technologies, kinds and the number of actuators employed, navigation control techniques, and what is new in development trends. (e paper gives a clear starting point for those who are initializing into this research area; also, it brings some helpful knowledge for those who already have experience. to have an impact on the future survival of humanity. (at is 1. Introduction why researchers have strived to develop unmanned vehicles Robotics is a branch of engineering that involves the con- that are increasingly eﬃcient and can provide work and cept, design, manufacture, and operation of programmable maintenance services underwater, as well as explorations of machines, which can develop their autonomy in the exe- ever deeper marine environments. For example, there are an cution of a speciﬁc task. Submarine robotics can be sub- estimated 2 billion tons of manganese nodules in the Paciﬁc divided as shown in Figure 1, where one can identify those Ocean near the Hawaiian Islands [2]. related to underwater robotics (unmanned water vehicles). (e UUVs are identiﬁed in Figure 1. Mainly, they are Underwater robotics have been developed for decades, and subdivided into those that are remotely operated (ROV) and they were characterized by manned water vehicles devel- those that are autonomous (AUV). (e ROV is a vehicle opment. However, in the last decades, due to the cost joined to an umbilical cable, and, according to its applica- demanded by the construction and use of these vehicles, tion, it can be classiﬁed into observation, work, and special researchers led to the development of robots or unmanned purpose [3, 4]. ROVs play a vital role in military operations underwater vehicles (UUV), where each time the robots (e.g., torpedo and mine recovery), rescue operations (e.g., were optimized and used in a wide variety of applications for locating historic shipwrecks, such as the RMS Titanic), and the beneﬁt and sustainable development of the planet. Since critical oil and gas operations. In recent years, the range of there is an immense magnitude of the oceans and due to the vehicles and mimitizing the need for human operators have human diﬃculty for their exploration, the vast majority of increased due to research eﬀorts. However, we have au- them have not yet been explored (two-thirds of the planet) tonomous vehicles (AUVs); these are indispensable tools, and around 37% of the world’s population is less than mainly for ocean exploration tasks. (ey are built using 100 kilometres from the ocean [1]. (is means that various high-end technologies and equipped with several of knowledge of the marine environment has and will continue them [5, 6]. During the subsea exploration process, the AUV 2 Journal of Robotics Robots Unmanned Workers Zoomorphic Humanoids vehicles Industrial With wings Androids Space Of service With legs Aerial With ﬁns Terrestrial Aquatic Figure 1: Location of unmanned water vehicles within a basic robotics classiﬁcation. performs necessary maintenance and repairs, transports the movements, such as jellyﬁshes, stingrays, and sharks. Even necessary equipment, and monitors and records diﬀerent though underwater robots that use thrusters with symmetrical situations under exploration. In oceanographic research, the geometric shapes are generally more eﬃcient, many works in AUV may be used in various measurements; one of them is bioinspired robots are being developed and continue looking to for the observation, release, and recovery of instruments increase knowledge in this study area. related to the monitoring of submarine volcanoes, in ad- dition to carrying out surveys of the seabed and various 2.2. By Hull Type. Some robots work underwater; those are studies in areas such as biology and hydrology. of two hull types, open and close. (e open ones generally (e unmanned vehicles’ structural designs have been have a hermetic box that is over a frame, and the other based on their shape, actuators’ distribution, and materials components like engines, buoyancy elements, sensors, and used. (ey include monitoring and measurement tasks of lights are over it too, while the closest ones have a bigger case oceanographic variables and navigation of the UUV; to that covers all elements; they do not have a frame because all achieve these objectives, it will be necessary to have a di- elements are inside of the case. (e prototypes or designs versity of sensors. Everything presented is the subject of which have a close hull usually have a cylindrical or round analysis in this article. (is article is structured as follows: geometric shapes; for example, there are some which look Section 2analyzes the morphologies used which are char- like a torpedo or a sphere. acterized by their type of hull and geometric shape and those that are bioinspired; additionally, it considers the distri- bution of propellants, materials used, and hermeticity. 2.3. By Geometric Shape. One of the most famous shapes of Section 3presents the instrumentation and actuation ele- underwater robots is the cylindrical shape, and on the top ments used interchangeably in ROVs and AUVs. Section they have the appearance of a dome, which means they look 4analyzes the navigation, orientation, and movement’s like a torpedo. In 1973, one of the ﬁrst portable ROVs was control of both ROVs and AUVs. Finally, the conclusions of made, and it had a torpedo shape as shown in Figure 2. this work are presented in Section 5. (e robot was hydraulically controlled from the surface, and it began to be controlled with electricity. (at was the main reason of its popularity, ten years later, 500 prototypes were 2. Structural Designs and Used Materials sold. (ose kinds of shapes come from military technology 2.1. Morphologies. (e morphological review refers to un- developments like many other technological advances. In this speciﬁc situation, the form comes from torpedoes that were on derwater robots’ geometrical shape, which is dependent on many factors like internal electronic devices distribution, military ships; this is the case of MK-38. sensor, motors, structure stability, and drag coeﬃcient de- In the Applied Physics Laboratory of Washington ﬁned to move the robot, among others. Additionally, it must University, some modiﬁcations to a torpedo were made, and take the robot’s degrees of freedom into account, since this is it became an inspiration to create the REMUS robot in 1994 needed to locate motors in the structure correctly. [8]; a schematic of such robot is shown in Figure 3. (is kind of robot does not necessarily have a geometrical (e design was of a cylindrical shape, and it moved shape; there are many ROVs and AUVs with biomimetic thanks to only one propeller behind the robot and four ﬁns structure; (at is to say, they try to mimic aquatic animal’s behind the propellers; two of them were located on a vertical Journal of Robotics 3 (ere are new studies that have been done during the last few years about diﬀerent shapes of underwater robots. In comparison, those have advantages compared with tradi- tional shapes, since their forms are symmetrical. In an in- vestigation project, the goal was to control the position of the symmetrical robot in events of disturbances with a pro- portional change of the direction force from propellers; all their motors were inside of the closed hull. (e design was developed in 2012, and it had a diameter of 40 cm and weighted 6.3 kg. (is spherical robot is shown in Figure 8. (e assembly of the mechanical components which are inside of the robot was made manually, and it caused Figure 2: Snoopy ROV [7]. troubles in the robot’s operation underwater because the robot was less stable, and it could not move for long periods. axis, and they had their drive mechanism like the ones that Figure 9 shows some of the improvements made with a 3D printer to fabricate the mechanism for the water jets [15]. were located on a horizontal axis. (ey were there to give direction to the robot. (e robot had a length of 114 cm with (e robot has two servomotors and a water jet in each leg. a diameter of 18 cm and a weight of 40 kg. (e torpedo shape For its movements, one servomotor is ﬁrst turned vertically is still used because of hydrodynamic advantages. to lift the propeller of the water jet; then, the horizontal In 1997, the REMUS robot’s development continued with servomotor moves the propeller of the water jet to go for- one smaller model that a person could uphold and transport. ward; subsequently, the water jet drops vertically; ﬁnally the (e robot had a length of 134 cm with a diameter of 19 cm and robot swings to take a step. (ere is another way to get easy access to the robot’s a weight of 31 kg. (e robot had four ﬁns as shown in Figure 4. (e ﬁns were in the stern, and they had an aerodynamic shape components; those kinds of ROVs are used to have a square or rectangular frame, and some components lay over the called NACA 0012. (e ﬁns moves were done using a pinion, a toothed chain, and an engine. Behind the ﬁns was the propeller; frame, such as hermetic boxes, lights, propellers, and other all the system described gave the direction to the robot. devices. (e robot called CURV [7] was one of the ﬁrst to (is torpedo-like shape allows underwater robots to implement the distribution mentioned. It was designed by move at high speeds on the horizontal axis, and they descend the US Navy to recover lost bombs or torpedoes; in one of its diagonally with the help of their ﬁns. (ere are other kinds of missions, in 1966, the ROV was able to immerse up to 869 m; underwater robots which have a torpedo shape, but they it was more than the maximum depth the robot could have more propellers and do not have rear ﬁns; an example is achieve according to its design; in addition to this, it operates the Nessi robot in its sixth version [10]. It was built for the under positive buoyancy, which means that if its propellers were turned oﬀ, the robot was going to the surface auto- SAUC-2011 competition, with 174 cm length and a diameter of 28 cm. To be able to move underwater, it was equipped matically. (e robot is still working with some improve- ments, and it is sent to diﬀerent missions; Figure 10 shows with six propellers: two for lateral motion, two for vertical and pitch moves, and two on the robot sides to control the the robot CURV II that is similar to CURV. In Figure 10, it is forward and yaw moving. A picture of Nessi’s sixth version possible to appreciate the ROV size compared with a per- can be seen in Figure 5. son’s size. (erefore, it was necessary to use a pulley on the Another development of a torpedo-shaped robot was boat to place it on the water. made in 2017, which had many propellers, as shown in With the passing of the years, underwater vehicles which Figure 6, but this was bigger than others shown in this paper. have similar shapes to CURV have been developed for subsea exploration and they can be carried by one person It had 534 cm length, 62 cm diameter, and a weight of 380 kg [11]. It had ﬁve propellers: one located on the top, two on the due to their smaller form. An ROV with a similar appearance to that of CURV was developed in 2010; the robot was called extremes to immerse, and the last two across it to control yaw movements. If there were any troubles and the robot Nessie IV [16], and it had a rectangular shape, aluminum frame, two hermetic chambers over the frame with a di- needed to come back to the surface, it had a ballast tank. (ere were other investigations of more complex ameter of 22 cm, sensors, and some safety components. models, which had a torpedo shape; an example of the One of the hermetic cameras had the motors and research was the development of a robot in 2018. It had four drivers’ electric connections and batteries; the other had ﬁns located on the sides including four propellers [12] and sensors interfaces and batteries electric connections. (e two propellers to immerse. (e robot had a length of 140 cm, robot worked using ﬁve propellers: two of them were a diameter of 20 cm, and a weight of 32 kg. Figure 7 shows vertically located to control depth and pitch, two others were located on the sides to control forward and yaw the position of those elements. (e robot was constructed to improve the propulsion system, which did not generate moves, and the last one was over the vertical axis to guide translation over this axis. (e Nessie ROV shown in countless bubbles because it was a trouble for the front camera to see; for this reason, the design had two more Figure 11 was an evolution of past versions and it was made to participate in a competition known as “SAUC-E” where propellers and it had four ﬁns to go forward; the movement was based on the turtles’ movement [13]. it was the winner robot. 4 Journal of Robotics 1.14” (.3’) Depth and heading Flux gate Sensor interface and control ﬁns compass data logger Homing array Command and DC brushless motor and control computer motor controller Transmitter 0.8” (7.17’) Temperature, conductivity, Hollow drive sha Lcoustic system dissolved oxygen, and pressure sensor heads Lead acid batteries Highly eﬃcient propeller Figure 3: REMUS (1994) [8]. depth and roll and the other four were located to allow forward movements, yaw, and translation over the vertical axis. Figure 13 shows the robot DaryaBird, showing the propellers distribution where they are approximated to over travel axis forward. (ere was a design proposed in 2019 which had a rectangular shape and positive buoyancy; this investigation [20] was more about the mechanic design, the design fea- Figure 4: REMUS (1997) [9]. tures to achieve stability, and a CFD analysis of its ﬁnal shape. (e robot had two pipes on the top from 2 to 4 inches, respectively, and a frame made of aluminum which sup- ported the ROV’s elements. (e structure guaranteed sta- bility due to the location of the mass center in the design, which was under the buoyancy center. (e top pipes were located there to obtain better stability while the robot rolled and to get the buoyancy center of the robot nearer to the top, as is shown in Figure 14. (e proposed design aligned mass center and buoyancy center over the horizontal axis to the movements on pitch and yaw could be easier. Figure 5: Nessi (2011) [10]. Due to the fact that the mass center and buoyancy center were not aligned over the vertical axis, the immersion Some ROVs have cubic-like frames such as CISCREA that propellers were distributed to get a pitch control as is shown is shown in Figure 12. It had six propellers: four of them were in Figure 15. outside of the vehicle's longitudinal axis, which gave the robot a (ere was a robot proposed in 2019 to work in shallow better turning moment, more lengthwise stability, and better waters [21]; it had a cubic shape, and its principal structure movements over the horizontal axis, and the other two were to was anticollision. (e robot’s design had an acrylic plate on control depth. (e robot had a length of 52.2 cm, a width of the front of it to locate a camera. Two propellers controlled 40.6 cm, a height of 39.5 cm, and a weight of 22 kg [18]. the forward and yaw movements. (e force given by the Some ROVs have been developed, having a frame to propellers should be designed greater than 1 kg and the support their components as the ones mentioned; one of immersion propellers should be between 2 kg and 8 kg. (e them was a robot called DaryaBird that was developed in robot was designed to have a hermetic chamber just for the 2016; it was made to participate in a competition for un- battery and another for electronic parts; Figure 16 shows the derwater robots, and one of its hermetic sections had a ﬁnal design with all its parts. torpedo shape. Over the frames were some elements such as its gripper, that is, a mechanical hand to carry objects, a battery, a pneumatic tank, and its hermetic chambers. 2.4. Bioinspired. In the last decades, there has been a lot of It had 83 cm length, 50.6 cm width, 41.3 cm height, and interest in improving the performance of underwater robots 32 kg weight [19]. It had six propellers; two were to control by taking a bioinspired optimization approach in aquatic Journal of Robotics 5 Figure 6: Large torpedo [11]. rusters Flapping fins (e union of soft and low rigidity materials with in- telligent actuators is called smart soft composites (SSC); its manufacture is carried out through the use of multinozzled Forward camera Tail 3D printer, for this UV curable polydimethylsiloxane (PDMS) is used, which can be 3D printed by direct ink writing (DIW) or stereolithography (SLA), where cables, connections, communication modules, batteries, and in- Body telligent actuators or anisotropic materials are placed layer Downward camera by layer in the ﬁnal prints [23, 24]. See Figure 19. Head In the use of these technologies, a new low-cost printing Figure 7: Hybrid (2018) [12]. method is presented in [25], where printing tests were carried out, and a material switcher PDMS was developed from laser diﬀraction thanks to the use of a T union by two currents: air and a second ﬂuid (water or oil); it should be noted that they do not use mirrors or lenses that are nor- mally in stereolithography (SLA). Although some robots are made based on these intel- ligent actuators, their movement speed is slow and limited compared to the living beings from which the inspiration was obtained or robots that use conventional actuators, since these have the advantage of being more compact, in addition to achieving better mimicry and acquiring important bio- logical characteristics, which are what they were looking for. A robot with a dolphin shape was made in 2016 [26]; it was an autonomous underwater vehicle (AUV) developed to Figure 8: Sphere (2012) [14]. measure water quality with a biomimetic system. (e prototype is shown in Figure 20, and the design was inspired animals, which after natural selection have evolved due to by an orca. (ey were looking to mimic the high hydro- dynamic and stability performance of an orca, because the their habitat to the point of having physical characteristics and excellent morphology. evolution of those animals has allowed them to develop those features correctly. (e dolphin could mimic some In 2012, a review of diﬀerent biologically inspired under- water robots presented and addressed a key problem in them, peculiarities on an orca and it could monitor and move in mentioned in [22], which is to achieve adequate movement in diﬃcult environments. real environmental conditions and classiﬁcation of robots by (e robot’s design had a rigid head, a hollow shell to the their way of swimming, which can be seen in Figure 17. (e components, a cabinet for the waist joint, and a chamber for proposal presented for the problem was the use of intelligent the caudal joint. It had a hydrodynamic shape to reduce the actuators, which are a viable alternative to be able to achieve the drag force, its waist was joint to get the necessary thrust to move, and a dorsal ﬁn and pectoral ﬁns to improve stability ﬂexible and complex movements that a camouﬂaged robot requires without the need for additional parts. also allowed performance turning manoeuvres. In the same year, a mechanical design of an autonomous vehicle (AUV) (e smart actuators are classiﬁed on shape memory alloy (SMA), ionic polymer-metal compound (IPMC), and a mix was presented inspired by a shark that had a vision system to between SMA and IPMC. Usually, the materials are com- take the information of the underwater environment, for bined with other materials; an example of this is presented in future applications as monitoring, searching objects, and Figure 18. (is embedding gives better mechanical features trajectory tracking. for robots. (e shark robot’s structure is shown in Figure 21 and it Normally, for the creation of parts of a bioinspired robot, had two main principal parts. On the front was a rigid body with an aerodynamic it is necessary to use polymers together with the alloys presented above, which serve to simulate elasticity and contour of a shark to get better performance swimming and two pectoral ﬁns to make movements as diving, emerging, deformability to achieve complex shapes of the living being to be mimicked. and swinging. Further, it had space where there were parts 6 Journal of Robotics Xilinx Zynq-7000 SoC Battery and Servomotor Hip joint circuit control module Knee joint Water-jet propeller Figure 9: Sphere (2014) [15]. Figure 10: CURV II [7]. Figure 12: CISCREA [18]. Mission module Aluminum pressure hull USB camera Aluminum ruster T-sloted frame Doppler velocity log Figure 11: Nessie IV [17]. Grabber Battery hull like cameras, circuits, sensors, and batteries. On the back was Figure 13: DaryaBird [19]. a multiarticulated body with three servomotors connected with an aluminum structure that allowed better movements when turning and achieved more manoeuvrability and thrust. because they had a gyratory axis with a bearing and a Glyd (e shell was made of ABS, and it had 482 mm length, Ring; they needed a dynamic seal in each ﬁn. (ere was an 208 mm width, 125 mm height, and 1.3 kg weight [27]. (e underwater manipulator which was built inspired by a robot used two impermeable systems: the ﬁrst one for the snake. (e robot’s purpose was to make exploration activ- top where there was a removable chamber that was a seal ities and intervene in underwater infrastructures, thanks to with a ﬂuoride rubber gasket and the second for pectoral ﬁns its shape, which gave ﬂexibility and the ability to get into Journal of Robotics 7 (e shell optimization was done diﬀerently on typical applications identifying the appropriate parameters to modify and improve with a biological model that meets the needs. For the selected models, simulations were done in a scalar level (1 : 3) inside of an air chamber. A better structure under the drag was implemented on a functional level (1 :1) to do a test on water and the designs that have passed by CFD simulations to get better data. (e other focus was the implementation of an alternative thrust system based on biological moves of ﬁns in the ﬁsh. (e better models were selected and made with a 3D printer; they were put to the test and, ﬁnally, the one that had a better performance was determined. (e implementation of the system is shown in Figure 23. Figure 14: Mechanical design of a rectangular ROV [20]. (e experimental results showed a drag force reduction between 50% and 85% in case of comparison to the basic OpenROV and the thruster system; although they were not T1 T2 that eﬃcient as the thrusters, they were better on the ﬂex- ibility, frequency, and amplitude. In 2019, another design of a robot called “CasiTuna” was inspired by tuna, a ﬁsh capable T5 T6 of high-velocity movements and manoeuvrability; addi- tionally, it can swim huge distances. Front (e robot tried to mimic the features of tuna, focusing on three main aspects; the ﬁrst one was a rigid body in the front where the electronic systems and a new thruster system of two engines exist, something that was not common on those T4 kinds of robots due to the quantity and the position. (ey are T3 accompanied by dorsal ﬁns and an internal system to adjust Figure 15: Mechanical design of a rectangular ROV [20]. the buoyancy. (e second main point was a lightweight body on the back; the section had two articulates, and, to achieve the movements, it had a mechanical system of 4 bars and a group of bevel gears that avoid involuntary ripples thanks to the new position of the engines. (e third main aspect was a rigid tail ﬁn that allowed it to generate greater thrust. (e robot “CasiTuna” is shown in Figure 24. (e robot had 520 mm length, 100 mm width, 130 mm height, and 2.6 kg weight; principally, its materials were ABS and PP. (e robot went through simulations in ADAMS and tests underwater validating design’s eﬀectiveness focus on displacement speed and stability, due to its aerodynamic Figure 16: Underwater robot for shallow depth [21]. shape. Table 1 shows some features of the mentioned robots. diﬃcult spaces. (e robot was designed by the University of Science and Technology of Norway. 2.5. (ruster Distribution. (e thruster’s location and the quantity depend on with how many degrees of freedom the (e robot is in the category of AUV and it had a serial robot is going to be designed; there is a case where, to achieve link mechanism with diﬀerent modules, which had tunnel all degrees of freedom, the thrusters have actuators to move thrusters and a stern thruster to move forward. Further, if them and change the direction of the force. needed, the robot could mimic the swimming of an eel due An ROV had ﬁve thrusters; some actuators on them were to good ﬂexibility. (e prototype is shown in Figure 22. (e designed in 2015; two thrusters were located on the extremes design was submitted on a very complex simulation envi- ronment with Matlab, which allowed to coordinate diﬀerent to move over the horizontal axis and yaw movements, and the other thrusters located on the sides had a servomotor; actuators on the articulations, getting good results of the proposed design, evidencing that the simulation environ- each one has the objective of changing the direction force and controlling roll movements as pitch and yaw, but they ment developed had high potential for this kind of tests [28]. (e robot OpenROV was an open code ROV developed in were employed to immerse [31]. (e last thruster was designed to operate all time and get stability on pitch 2019 with a bioinspired focus; it was made for the City movements. (e thrusters are shown in Figure 25. University of Applied Sciences in Bremen; its work was (e underwater robots must have an emerge and im- focused more on getting better stability, reducing the drag merse system for that reason; they use thrusters to do those force, and evaluating alternative thrust systems [29]. 8 Journal of Robotics Oscillatory (BCA-O) BCA Swimming mode body/caudal actuation biomimetic underwater robot Undulatory (BCA-U) MPA Oscillatory (MPA-O) median/paired actuation Undulatory (MPA-U) JET Jet propulsion jet propulsion Figure 17: Classiﬁcation of bioinspired robots by their swimming mode [22]. is possible to see locations of their four thrusters, which gave Pectoral fin (latex) them 5 degrees of freedom [33]. Body (ere was another research where the robot’s shape had more DOFs (degrees of freedom) than other distributions, Tail besides having just four engines as is shown in Figure 29. (e thrusters are distributed vertically and horizontally; the diﬀerence in forces between each pair allows the robot to SMA wires Skin have more degrees of freedom. For example, if the motor Holder above has more force than the one below, the robot will be able to change direction achieving a 90 turn [34]. Middle board Elastic substrate (e robot shown in Figure 30 was employed in research Biomimetic fin (SMA) for the best control system for a robot of this type according Figure 18: Manta ray robot [22]. to the mathematical model obtained by a CFD program [35]. (e robot had six thrusters: two of them were applied to dive and the other four located out of the longitudinal tasks or implement a ballast tank. (ere was a robot de- axis were approximately 45 concerning the forward axis. veloped for investigation purposes by Norway in 2017; it had three thrusters for the robot’s movements underwater. (e (e roll and pitch did not need to be controlled because the distribution element gave enough stability to those degrees prototype is shown in Figure 26. (e objective of the ROV mentioned in Figure 26 was the of freedoms. (ere are robots with more than ﬁve thrusters. (e robot development of a low-cost robot to monitor ﬁshes for shown in Figure 31 had eight thrusters, which allowed the aquaculture [32]. (e thrusters were around a cylinder robot to move on 6 degrees of freedom, like with six placed 120 from one another; this allowed them to move in thrusters, but giving the advantage of more thrust to dive. two dimensions. (e robot had negative buoyancy. To Figure 31 is an upgrade of Bluerobotics’ ROV [36], with control its depth, it had a rope attached to a platform on the eight thrusters: four of them were locatedvertically outside surface. (e robot’s stability had better control due to the the frame with an external protection to prevent the pro- prototypes’ symmetry; therefore, the pitch and roll caused by pellers from being damaged during their operation or the the pendulum eﬀect were not aﬀected. In 2019, a design was tether cable can be tangled or damaging one of themwhile proposed for aquaculture applications; the shape and they are working or tethered could tangle and damage on thrusters were located to avoid lurching eﬀect, which means that a buoyancy object loses its stability due to a strong water them. (e thrusters were placed in the same position as the original Bluerobotics’ ROV. stream. Figure 27 shows assembled robots and in Figure 28 it Journal of Robotics 9 Anisotropic Circuit material Polymer printing matrix Supporter (a) Supporter deposition (b) Part deposition (c) Circuit printing (d) Deposition of anisotropic materials Comm. Controller module Battery Actuator Repeat processes (e) Component (f ) Part deposition (g) Support removal (h) So morphing embedment of robot/structure Figure 19: SSC manufacturing process [23]. Hollow shell Fins Head Fluke GPS antenna Stern propeller module Tunnel thruster module Waterproof Caudal joint camera Waist joint cabinet Joint module cabinet Flipper Figure 20: Dolphin (2016) [26]. Figure 22: Underwater manipulator design (2016) [28]. Multilink propulsive units Dorsal fin Right pectoral fin Peduncle caudal fin Switch Left pectoral fin Camera Infrared sensors Figure 23: Bioinspired alternative shell and drive design (2016) [29]. Figure 21: Shark robot prototype (2016) [27]. 10 Journal of Robotics DC motors Battery Motor controllers Anterior shell Caudal fin Peduncle Pectoral ﬁn e buoyancy Transmission Posterior shell adjusting mechanism mechanism Figure 24: Almost tuna (2019) robot design [30]. Table 1: Characteristics of biomimetic robots. Year Name Form Propulsion Dimension (cm) Material 2011 RoboJelly Jellyﬁsh Membrane Not registered Silicone, SMA, steel 2012 TurtleLike Turtle Pectoral ﬁns Not registered ABS, PDMS, SMA 2016 Dolphin Killer whale Tail ﬁn 77.1 × 13.2 × 35.2 POM, ﬁberglass 2016 Nameless Shark Tail ﬁn 48.3 × 20.8 ×12.5 ABS, rubber, aluminum 2016 Nameless Snake Stern thruster/tunnel thruster/articulated swimming Not registered Not registered 2019 OpenRov Fish Pair of propellers and ﬁn Not registered PLA, varnish 2019 CasiTuna Tuna Tail ﬁn 52.0 ×10.0 ×13.0 ABS, PP D frame for carrying and launching 3. Materials and Hermeticity (e material selected for an ROV manufacturer must withstand the pressure to which it is going to be subjected; the deeper the robot reaches, the greater the external Servo pressure is during the robot’s operation. (e robots that have S2 motor housing a torpedo shape as [9] are made of aluminum, with chambers S1 ﬁlled with pressure air, and their frontal sections are made of plastic, such as the REMUS robot. (ere are more robots Front with torpedo shape; an example is Nessie [10], which was camera mounting built externally with aluminum, and the internal structure was made of PVC to support some parts, such as sensors. Nessie’s distribution is shown in Figure 32. Figure 25: 5 thrusters [31]. In 2016, a robot that had a hermetic chamber made of acrylic and PVC was proposed [34]; the acrylic part was placed in front of the camera to record outside and its thrusters supports were 3D printed. Figure 33 shows the parts. Another robot that had some of its parts 3D printed was introduced in [15]. (e parts are in Figure 9. (e 3D printer was an enormous advantage to the robot fabrication because the thrusters can save space due to better distri- bution; also, the mechanical fasteners are designed in a better way by 3D printing, and the mechanical resistance is better compared to the other propulsion system of the ro- bot’s water jets. One way to reduce the fabrication cost is using 3D print; the prototype [32] was an investigation project of an ROV in which the thrusters shell was 3D printed, and most of its structural parts were made of PE (polyethene) and PMMA (polymethylmethacrylate) to allow the camera recording. One of the most important aspects taken into account in the manufactured material of the ROVs is the density; since materials with low density make the buoyancy increase in Figure 26: 3 thrusters [32]. proportion to its weight, this means that more force is Journal of Robotics 11 (a) (b) Figure 27: 4 assembled thrusters [33]. [27] which was made of ABS (acrylonitrile butadiene sty- rene), with an impermeable system that used o-rings of ﬂuoride rubbers and Glyd Ring for dynamic sealing; further the posterior thrust part was connected to the mobile joints by a skeleton made of aluminum. CasiTuna robot [30] was similar to the shark with an an- terior and posterior body made of ABS, but this prototype had a ﬂow ﬁn made of PP (polypropylene). Another shape of this kind of robot is a turtle [22]; its body was of ABS and the head of (a) PDMS (polydimethylsiloxane) and the structure ﬁn was a t″ t combination between ABS and SMA (shape memory alloy). OpenROV robot bioinspiration came from a shell [29] and an alternative thruster made of PLA with a 3D printer, a further ﬁll- in process, varnish, and sanding to get a smooth surface. t′ (ere are a wide variety of shapes for this class of robots. t′ As an example, there was a robot called RoboJelly [22] and it had a bell matrix structure made of silicon and 8 BISMAC actuators formed by a steel spring, silicone, and SMA wires. t″ 1 Table 2 shows the main features of the aquatic robots dis- cussed in this section. (b) Figure 28: 4 thrusters [33]. 4. Instrumentation and Actuators If we refer to the number of external devices, we should required to submerge the robot. (e underwater robot in- know that these increase or decrease with the arrival of new vestigation [30] focused on its structural design where a applications for unmanned aquatic robots. Mainly these are 6036T6 aluminum alloy was selected for the structure divided into two types: autonomous robots and AUVs supported motors and other 1 cm diameter elements; hence, (autonomous underwater vehicles) and ROVs (remotely the hermetic chamber was made of AL5053. Within the operated vehicles), which also are divided by the type of aforementioned robot’s design, deformation analysis was application that they develop, such as intervention and made when the hermetic chamber was exposed to the inspection. (e diﬀerence between these two classes lies in pressure under the sea and the deformation of its supports. the diﬀerent use of resources, size, and weight that the robot Another prototype designed for aquaculture was in- has in each division. troduced in [33]; the design had a hermetic chamber made of In this section, we will explain and focus on the use of acrylic with a width of 5 cm. (e hermetical chamber’s instrumentation and actuators, collecting information from material beside the features mentioned above should have both types of unmanned robots. good corrosion resistance as mentioned in the design of [21]. (e design proposed a hermetic chamber made of aluminum alloy 4032-T6 and in the forepart an acrylic dome was in- 4.1. Sensors stalled for the camera. (ere is a kind of underwater robot that uses something called biomimetics to move underwater 4.1.1. Measurement of Oceanographic Variables and it is made of diﬀerent materials, for example, a dolphin structure [26] made of ﬁbreglass and the rest of the body (1) Temperature. To measure and process oceanographic made of paraformaldehyde (POM); this material has variables in advanced systems, it is well known that the ﬁrst lightweight features. Another structure’s robot is a shark relevant variable to be accounted for is the environmental 12 Journal of Robotics (a) (b) Figure 29: 4 thrusters located horizontally and vertically [34]. Attimeter Thrusters Enclosure Dockin hoop DVL (a) (b) Figure 30: 6 thrusters [35]. III III III III II III III III III III (a) (b) Figure 31: BlueROV2 and Bluerobotics’ ROV with 8 thrusters [36]. temperature; this happens due to the changes in the tem- helps ﬁshermen to set places and times propitious to ﬁshing perature in the ocean, which not only inﬂuence the dynamics and also to know the distribution of fauna. of the sea and the atmosphere but also intervene in the Likewise, sensors of depth, altitude, and temperature are distribution of marine organisms and their metabolism; for always used in the design of aquatic robots. When we couple this reason, some boats use a temperature sensor, which these 3 sensors, we will be able to obtain the data of Journal of Robotics 13 (a) (b) Figure 32: PVC frame [10]. has sensors attached to its structure for reading chlorophyll and nitrate, where water samples can be used for the cali- bration of these parameters, as shown in Figure 34. (4) Conductivity. Conductivity is the measurement of electrical resistivity, a property that quantiﬁes how many dissolved substances, chemicals, and minerals are present in water. (is means that a large amount of these impurities determines a higher conductivity. (e use of CTD sensors allows water measurement of temperature, conductivity, and pressure. El Dorado AUV has a CTD sensor [38], which, like the SOTAB-I robot, is attached to its frame, with a sampling frequency of up to 16 Hz enabling a high spatial resolution, with a consumption of 3.4 W [39]. Figure 33: Printed thrusters mount [34]. (5) Total ATP (adenosine triphosphate). (ere are some other oceanographic variables and the related variations between kinds of sensors that ROVs have been carrying on recently; them. However, the measurement of the temperature sensor they are called microﬂuidics, which deal with the manip- can be used not only to know better surroundings but also to ulation of concerning particles or droplets temporal dy- compensate for the operation of the gyroscope and accel- namics, velocity, and spatial ﬂow patterns in microchannels erometer [37]. (at is the case of the UUV or SUR-II. [40]; although it is a new multidisciplinary ﬁeld, it has the potential to inﬂuence areas as biological analysis. (ey are (2) Pressure. It is well known that the control of ballasts used also called LOC (lab on a chip) to be used as screen in- in the stability of the structure of UUVs is directly related to struments in cell biology, chemical synthesis, and bio- the pressure sensors; this is due to the assembly of com- analysis. (ey have some advantages because they are pressed air tanks, which are regulated by the constant portable, and they can be done with low-cost fabrication variation of atmospheric pressure in the environment materials. An example of an ROV is given below. (symmetrical robots allow changing their center of gravity). (e total ATP is a useful biochemical parameter for (is kind of behavior is observed with DaryaBird, a UUV detecting biomass or biochemical activity anomalies in the that employs a YOKOGAWA’s pressure sensor that mea- natural environment; since dissolved ATP is an important sures depth and a gyroscope that measures the azimuth angle carbon and phosphorus for marine microbes is also related and altitude angle [19]. (ese two sensors help in the remote to microbial activity, the total ATP is a useful parameter function, turning the UUV into an AUV (autonomous indicative of the presence of biogeochemical events, such underwater vehicle) or in some cases it could help in turning submarine volcanism, hydrocarbon seepages, and occasional it into a ROV that can be remotely operated; this type of supply of organic resources [41]. (at is why, to obtain these UUV can recollect data and send it through an umbilical variables in real time, a new version of an ATP analyzer was cable that reaches the surface to be visualized in a computer. developed and evaluated in situ using an ROV, achieving a depth of 200m in the tests carried out. (3) Nitrate. (e standard methods of obtaining oceano- Figure 35 shows the analyzer; it has a microﬂuidic device graphic samples were through the samples collection with a and analyzer module, which is the core component, and a single bottle, with the main tests being temperature, pres- photometry module for the bioluminescence intensity sure, and salinity. In an advance to automation, “El Dorado” measurements based on the L-L reaction. was presented in 2016, which is an AUV that recollects up to (e measurements taken with the in situ ATP analyzer 10 shots of water in a bottle called “Gulper” [38]. El Dorado were consistent with those measured manually, which 14 Journal of Robotics Table 2: Evaluated underwater robots features. Year Name Shape Hull (rs DoF Dims (cm) Material 1973 Snoopy Torpedoes Closed NA NA Not registered Not registered 1994 REMUS Torpedoes Closed 1 2 (roll, yaw) 114.0 ×18.0 Not registered 1997 REMUS Torpedoes Closed 1 2 (roll, yaw) 134.0 ×19.0 Aluminum 2011 Nessie Torpedoes Closed 6 5 (roll, yaw, Y, Z, X) 174.0 × 28.0 Aluminum and PVC 2017 Nameless Torpedoes Closed 5 5 (roll, yaw, Y, Z, X) 534.0 × 62.0 Not registered 2017 Hybrid Torpedoes Closed 4 4 (roll, yaw, Y, Z) 140.0 × 20.0 Not registered 2012 Nameless Sphere Closed 3 3 (yaw, Y, Z) 40.0 Acrylic 1966 CURV Rectangular Open NA NA Not registered Not registered 2010 Nessie IV Rectangular Open 5 6 (pitch, roll, yaw, Y, Z, X) Not registered Aluminum 2014 CISCREA Cubic Open 6 5 (roll, yaw, Y, Z, X) 52.2 × 40.6 × 39.5 Not registered 2016 DayaBird Rectangular Open 6 5 (roll, yaw, Y, Z, X) 80.0 × 50.6 × 41.3 Aluminum 2019 Nameless Rectangular Open 6 5 (pitch, yaw, Y, Z, X) Not registered Aluminum and PVC 2019 Nameless Cubic Open 4 4 (roll, yaw, Y, Z) Not registered Aluminum 4032-T6 and acrylic 2015 Nameless Rectangular Open 5 6 (pitch, roll, yaw, Y, Z, X) 70.0 × 40.0 Aluminum AL 5053 and 6036T6 2017 Nameless Cylinder Closed 3 3 (yaw, Y, X) 30.0 × 20.0 ×15.0 PPE and PMMA 2019 Nameless Cylinder Closed 4 5 (roll, yaw, Y, Z, X) 44.0 × 26.0 × 24.8 Acrylic 2019 X4-ROV Cylinder Closed 4 3 (yaw, Y, X) Not registered PPVC and acrylic 2018 BlueROV2 Heavy Rectangular Open 8 6 (pitch, roll, yaw, Y, Z, X) 25.4 × 57.5 × 45.7 Aluminum and acrylic (a) (b) Figure 34: Images of El Dorado AUV, Gulper system [38]. position, direction, and speed. Also, the attitude and heading Microfluidic device reference system (AHRS), the inertial navigation system (INS), or the hydroacoustic position reference (HPR) system was implemented to maintain better control of positioning PC and stability. (e sensors that make up these systems and the applications in the diﬀerent UVVs are described below. Analysis Photometry (1) Inertial Sensors. (e inertial measurement unit, better module module known as IMU, is a sensor that detects linear acceleration using one or more accelerometers, as well as the rotational speed using one or more gyroscopes. Some of these devices include a magnetometer that is used as the main reference. Reagent and waste bags Sample inlet In 2008, the ROV Nessie III was ﬁrst introduced as an AUV that uses the 3-gyro reference system for navigation Figure 35: Total ATP analyzer with a microﬂuidic device [42]. targets [43]. In 2010 AMOUR was introduced, which was a medium ROV destined for the investigation of maritime areas; this ROV used the coupling of an inertial sensor that demonstrated that a portable, simple, and reliable ﬂow analysis system such as its microﬂuidic device can be used in estimates the position and depth that uses a record of 10 data raws unprocessed, corresponding to the sensors (a pressure extreme environments for real-time biochemical analyses. sensor, 3 magnetic ﬁeld sensors, 3 accelerometers, and 3 gyroscopes) [37]. On the other hand, MINERVA appeared 4.1.2. Navigation Instruments. When the UUVs were built, in 2014, which was an intervention ROV that mixed two diﬀerent data acquisition methods were used such as positioning systems and used an inertial sensor as the main Journal of Robotics 15 (5) Doppler Sonars. Doppler velocity sensor (DVS) uses the sensor and a depth sensor [44]. (is indicates that not only is a single-precision algorithm needed for navigation, but also Doppler eﬀect to measure the octagonal velocity; its limi- tations are based on the calculation of the integration of other sensors are needed to compensate for the errors of a single system. velocity and the time of calculating the position; this type of positioning control system will be explained in the next (2) Compass (Magnetometer). (e magnetometer works by section; however, its operation can be up to 300 m. measuring the magnetic ﬁeld variation in three referential As mentioned before, MINERVA mixes two positioning axes that are subtracted from the Earth’s magnetic ﬁeld; systems, in which, apart from using an inertial sensor, it also despite its wide applications on the UUV development ﬁeld, uses the hydroacoustic positioning system through the use of the operation of this sensor is sensitive to the noise caused by a Doppler velocity record (DVL) to measure its velocity [44]. other sources like the operation of other sensors, motors, Figure 36 is an example of the calculated position through the use of a hydrophone array. (is use is carried out in and others. (is means that you cannot only rely on the values of a diﬀerent UUVs, such as DaryaBird, which employs an al- titude sensor TRAX [19], which was installed to control the magnetometer, as there are also several parts to be con- sidered in underwater navigation such as a pressure sensor movement in DVL sensors. to measure depth, a gyroscope, and an accelerometer to control altitude and locomotion [45]. (is also gives the 4.1.3. Optic Sensors solution of integrating dedicated digital sensors to increase the accuracy, modules that can evaluate the heading di- (1) Video Cameras. Most UUVs have standardized a rection with a minimum diﬀerence of degrees, which are still complement of one video camera for transmission from compensated and calibrated to support magnetic distortions depth to the surface; the images are important in envi- with the combination of other sensors, as is the case of a ronmental analysis. In 2010, in the construction of the TSL MEMS accelerometer 3-axis sensor and a 3-axis magneto- (Tunnel Sea Lion) robot, 2 video cameras were incorpo- resistive sensor [46]. rated: one with a direct view on the bow and the other in a vertically downward orientation [48]. Likewise, another (3) GPS. (e global positioning system is managed by direct application to the mounted camera is to use it as a new communication with satellites; the use of these devices addressing method by an AUV; this was observed in 2018 underwater does not allow their correct operation, so the when tests were carried out on an interactive technique best application to UUVs is through collection or recovery of shown in Figure 37 which makes the construction of a these robots when they reach the surface. coordinated system with a route drawn from the image (e use is described by Choyekh Mahdi, indicating that taking through video [49]. the tracking of the SOTAB-I robot on the sea surface is Another application by the transmission of images was ensured by a global positioning system (GPS) receiver that established in the detection of marine animals with varied serves to determine the absolute position of the robot. In the visibility from a new dataset; video capture consists of three case where the robot is submerged, the Ultrashort Baseline cameras and three lights. (e colour cameras have a reso- (USBL) system ensures tracking [39]. lution of up to 1080 ×1920 pixels, and the frame rate is up to 30 fps [50]; the camera direction is diagonally downward (4) Sonars. (e sonar’s performance is through sound, where towards the riverbed. the propagation of waves underwater allows navigation, communication, and detection of submerged objects. Since their use is standardized in underwater vehicles for opera- 4.2. Actuators. (e underwater robot principle of move- tion in low-visibility conditions, there are a wide range of ment is based on the use of propellers; the type, power, and UUVs using this device; an example is Nessie III, which sent weight of the motor used with these propellers depend on a speciﬁc signal from the vehicle to a transponder that re- the robot’s work, as explained at the beginning of the section. sponds; the delay in the vehicle that receives this response gives the bidirectional ﬂight time for the signal; the range 4.2.1. Engines and (rusters used was between 60 kHz and 90 kHz [43], and the result was to obtain the raw data of speed, distance, and distance time. In 2017, the navigation compensation of an ROV was (1) Water Jets. Among the variety of motors that we use to presented through the comparison of data extracted from a move the robot, we can ﬁnd the propulsion based on a high- sonar with the use of the dead reckoning methodology and its pressure water jet. For a long time, this type of propulsion compensated error [47], which details the use of the sonar was normally used because of its comfortable design and its when the ROV does not have any movement due to the great propulsion with higher weights. In 2000, this type of propulsion was used in the development of a torpedo- interference that occurs with the operation of the engines, for its previous compensation. (ere are passive sonar systems, in shaped robot called TSL, where the bow and stem pro- pulsion systems had a tunnel for the performance of water which hydrophone-based communication participates in points not so far away between the robot and a boat or surface, jets [48]. giving sound pulses to ﬁnd the distance between both objects In spherical-shaped AUVs, this type of drive is used in a and calculate the angle of the sound source. [19]. vertical direction; two actuators can be controlled by one 16 Journal of Robotics Y (0, y , 0) (5) Brushless. Brushless motors became popular in small and O (0, 0, 0) Hydrophone y medium ROVs; their use dates back to the 80s and 90s. An array example is ABE, an AUV destined for benthic species ex- ploration that uses brushless motors (brushless) with oil [52] X (x , 0, 0) x for pressure compensation. A clear example is ARMOUR, where each drive is made up of a motor controller and a brushless DC motor [53]. Another related example is the operation of Jeﬀ, a small AUV designed for inspection and Z (0, 0, z ) swarm joint work; the propulsion system has mainly two DC T (x , y , z ) t t t Pinger motors with a custom magnetic coupling design to avoid corrosion and short circuits. [45] (6) Hydraulic Systems. In the design and development of Figure 36: Acoustic detection plane [19]. UUVs, a new structure was chosen for the steering man- agement in the 6-DOF; this particular conﬁguration in parallel can be seen in Figure 38; it has two main thrusters in Y the front and another in the rear that handles the steering of CS camera Trajectory the robot [54], and the union between the two parts is through the hydraulic system. 4.2.2. Luminaries. (e functions performed by the UUVs include the inspection, manipulation, and data collection; all the robots have a video transmission system implemented, so it is necessary to develop a lighting system to acquire Y images since the underwater environment does not have P visibility conditions due to the lack of a light source. (e ROV system presented by Jinwoo uses two panoramic halogen lights and two LED lights to acquire high-quality images [55], as shown in Figure 39. Figure 37: Construction of external orthogonal coordinate system In another application in image acquisition, there are [49]. the recognition and detection of objects, for rescue or supervision robots, and a clear example of detection is found in the research of Pedersen et al., where the illu- thruster and one servomotor; the jet-based thruster decides minated area is needed for the detection of pelagic species, the value of the driving force, and the servomotor controls where 3 LED lamps of 1900 lumens and an approximate the height of the thruster [37, 51]. resistance of 10 bar are used [50], to properly visualize the case study. (2) Stepper Motor. In the UUV’s direction management, a form of steering management was introduced, which in- volved the torpedo structure (Xianbo et al., 2017); this 4.2.3. Manipulators. (e term “manipulators” is used to structure has two propellers: one for horizontal movement describe a mechanical device with mobile joints intended for and the other for vertical motion; these two are attached to the manipulation of tools, parts, or special devices to per- the main propeller which is driven by a DC motor and four form various tasks. (is meaning applied to robotics results rudders driven by stepper motors [11]. in an automatic handling machine, reprogrammable in ei- ther a moving or ﬁxed position. In 2011, a hybrid underwater (3) Brushes. Nessie III and DaryaBird used the same brush robot was made, with a crab and lobster structure, where its motor-based propeller oﬀered by the SeaBotix brand, with a legs acted as manipulators and its main function was to consumption of 110 W and an ability to withstand depth up inspect underwater structures and shipwrecks in shallow to 150 m; these were used in directional movement in the -xy waters, where activities such as cable cutting, grinding, and plane; in the case of the second robot mentioned, it used a drilling are required [56]. RoboPlus Hibikino thruster with a power of 90 W for the (e most common way to implement manipulators is in robot descent control. intervention class ROVs, for example, MINERVA, where its manipulator allows the samples collection. Table 3 sum- (4) Servomotors. ARMOUR has a closed torpedo-shaped hull marizes all the aquatic robots seen in this section. structure, which utilizes 10 propellers for handling 6 DOF (degrees of freedom); however, small spherical ROVs, such 5. Navigation and Control as SUR-II, use vector water jet thrusters composed of wa- terproof housing, two servomotors, and a support frame Navigation, in simple terms, conforms to particular [37]. methods that allow someone to know where they are and Journal of Robotics 17 150mm F 100mm 100mm 50mm 50mm 0mm CG 0mm F y CG (a) (b) Figure 38: Direction change operation diagram [54]. (a) (b) Figure 39: (a) (e use of two LED lights. (b) (e use of two halogen lights with their respective results [55]. how to get to a new point. Depending on the type of en- conducting maritime exploration which covers large areas of vironment and available reference points, these methods several hundred square kilometres. Navigation plays a vital could be simple; however, they could become complex role here that if it is not properly executed, it could not only results in a hostile, changing, and unpredictable environ- aﬀect the fulﬁlment of the mission but also aﬀect the safety of ment, also, to reference points that are not visible [6]. (e the vehicle [11] and, in the worst scenario, could lead to the tasks that the UUVs must perform require navigation to loss of the robot, causing economic loses and contamination displace to diﬀerent location points to complete their duty. of the explored environment. Due to its autonomy, navi- gation must be accomplished under the control of a com- Navigation can be done by the vehicle itself (in the case of AUVs) or by operators (in ROVs cases). Usually, the tasks puter embedded in the vehicle. performed by ROVs demand heavy work and can be conﬁned to smaller spaces where navigation is probably not very complex and could be performed directly by the 5.1. Navigation Methods. Navigation in AUVs represents a great challenge for most researchers due to the impossibility of operators through a joystick and the use of one or more video cameras installed on the ROV. using a global positioning system (GPS) underwater. (e electromagnetic radiation waves emitted by satellites are Moreover, autonomous vehicles have more tasks and must carry out missions that take several hours, days, or even absorbed when they come into contact with water, so a GPS signal receiver cannot capture the waves underwater [58]. months [57]. Most of these missions are focused on 18 Journal of Robotics Table 3: Summary of the evaluated UUVs. UUVs Type Application Equipped sensors Actuators Compass, low-frequency sonar, angular 3 aft, 2 vertical and 2 ABE (1992) AUV Seabed supervision velocity sensor horizontal thrusters 2 horizontal thrusters, 1 Aqua Explorer Inspection of telecommunication Gyroscope, altimeter, depth meter, AUV vertical with brushless 1000 (1992) signals and cables accelerometer, acoustic transponder motor Depth gauge, CTDO sensor, INS with A main thruster, 2 vertical R1 (1995) AUV Monitoring near the seabed Doppler sonar, acoustic transponder water jet thrusters MINERVA ROV No register a DVL, HPR and IMU 5 thrusters (2014) ARMOUR UUV Reef survey and other applications IMU, GPS, DVL 5 thrusters (2010) Environmental study and Depth sensor, leak sensor, camera, FOSN, 4 thrusters and an 840 W (2017) URV surveillance task in the mid-range of and mini-AHRS main propeller shallow waters Monitoring of nuclear storage ponds SUR 3 servomotors and 3 water AUV and wastewater treatment facilities to No register (2013–2015) jet propellers prevent leaks Application of parallel robots in the REMUS I IMU, pressure sensor, immersion sensor, LED lights, 1 thruster at UPR underwater environment requires (2011) temperature sensor, camera the back studies GPS satellite navigation system, USBL Consists of six thrusters, positioning system and autonomous on- TSL (2000) AUV Tunnel inspection providing arbitrary board navigation system, TV system, and movements in 3 axes IFSSI scanning sonar Nessie III Designed to participate in SAUC-E Altimeter, IMU, camera, transponder, 5 80 W propellers, AUV (2008) competition battery temperature sensor brushless motors with oil Check and evaluate new navigation (2018) AUV Stereo camera, does not register other sensors Does not register methods 4 100 W thrusters and 2 DaryaBird Pressure sensor, DVL, USB camera, altitude AUV Does not register main thrusters brushless (2016) sensor, and hydrophone motors with oil. (erefore, underwater devices have been used to establish a acceleration in the vehicle’s direction. If there is no need to local positioning system. (anks to advanced technology, execute the proposed algorithm, they use traditional dead many of these devices have been improved and optimized reckoning to estimate the vehicle’s heading direction and its considerably in terms of dimensions and performance. (is position. Another interesting work about accelerometers and how motivated the investigation and improvement of methods that allow a better estimation of the location and thus more exact they are used corresponds to the authors Yan et al. [59]. (ey navigation. worked on a dead reckoning navigation system based on neural networks using only accelerometers, due to the cost of using other sensors such as a DVL or the dependence on an 5.1.1. Proprioceptive Navigation. If the travel speed of the acoustic system. (e errors that have been generated by AUV is known, new positions can be estimated by con- using inertial units are reﬂected in rapid changes in the secutive integrations of speed. To perform velocity measures, measured angles by gyroscopes; this considerably increases Doppler velocity log (DVL) is normally used in conjunction the error of the dead reckoning system. (e use of neural with inertial systems and a compass; the estimated position networks to estimate the pitch angles through an exploration solely depends on the movement of the vehicle and hence the between the measured orientations and the measured ac- proprioceptive name. (is type of methodology is known as celerations varying in time allows estimating the vehicle’s dead reckoning [59]. (is type of navigation system cor- positions and reducing the errors caused by the gyroscopes. responds to the research written by Itzik Klein and Roee Normally, the proprietary navigation systems work in Diamant; they developed a system that estimates the tra- conjunction with external systems to correct themselves and jectory travelled by a water vehicle that moves freely in the reduce the accumulated error. Some of the work done on direction of the marine currents [60]. Because these robots dead reckoning which works in conjunction with other work very closely to the sea surface, they are easily sus- reference systems corresponds [61] to Kepper et al. who ceptible to orientation change, which creates problems in the developed a death reckoning model based on an inertial path. (rough acceleration measures, the investigator can measurement unit (IMU), which works in conjunction with constrain the execution of a proposed algorithm based on an acoustic measurement system to reduce the error ac- the principal component analysis to calculate the cumulated by the IMU. Due to the noise generated by the Journal of Robotics 19 (is method works quite well for a single vehicle. In the case of the navigation with several vehicles, variants of this method have been proposed to eliminate the consultation Query signals, converting the communication in one direction and thus removing the dependence of the time intervals for Query vehicle location updates. However, both the beacons and the underwater vehicles must be synchronized [63]. Reply (e standard conﬁguration of the LBL system and its variants (Figure 41) allows establishing an absolute posi- Reply tioning for either one or several underwater vehicles. However, the task of implementing and calibrating polyg- onal beacon arrangements is expensive and diﬃcult. Figure 40: LBL standard mode [62]. (erefore, it was decided to improve these systems even more and only one beacon has been achieved to determine IMU, the raw data captured was ﬁltered using an extended the position of a vehicle; this system has been called Ul- trashort Baseline (USBL). (is conﬁguration is illustrated in Kalman ﬁlter. For the model eﬀectiveness, implementation was evaluated in data collected in 3 diﬀerent environments Figure 42. (is type of conﬁguration works similarly to the for ﬁeld experiments and in an open ocean environment. standard LBL conﬁguration; however, the vehicles have Correct navigation requires a good position estimate, so multiple acoustic receivers, because they must determine not the instruments and the algorithms used must obtain the only the distance at which they are from the beacon but also most exact ubication. the angle with which the replica of the signal arrives. (e query signal was issued. In this way, the need for using several beacons for the trilateration calculation is avoided. 5.1.2. Acoustic Navigation. (e main problem of proprio- (e vehicle’s work area is conﬁned to the entire radius ceptive navigation is that the error increases limitlessly as the generated by the beacon. distance travelled by the vehicle increases. If an external We can cite the work done by Hidaka et al. [19]. (ey reference system is not considered, the navigation becomes implemented this acoustic navigation system which inten- critical for the vehicle and its mission. As a solution to this ded to use an array of hydrophones that were very close. (e problem, acoustic navigation is employed. Acoustics waves angle was calculated to the arriving sound from the oﬀset are appropriate for underwater propagation due to minimal that occurred between the hydrophones. Besides, the sonar attenuation. Hence, they are employed for underwater system uses an electronic circuit for signal ampliﬁcation, communication and positioning the underwater vehicles. phase comparison, and digital to analog conversion (D/A). For example, underwater vehicles use data of the placed beacons for estimation of their positions in the work zone. 5.1.3. Optical. Optical navigation uses optical devices such (e most used acoustics methods for underwater location as video cameras or optical diodes from which morpho- are Long Baseline (LBL) and Ultrashort Baseline (USBL) logical data of the seabed are recorded. M. Carreras et al. [62]. presented a localization approach for an underwater robot (e standard LBL method is characterized by beacons or based on vision and in an environment structured like a transborder, which is ﬁxed as shown in Figure 40. (e image water tank [64]. In the work, the location algorithm details shows the conﬁguration system for the vehicle and through some graphic results and the precision of the transponders. system. (e algorithm allows obtaining a 3D position, First, transponders listen to the pings emitted for the orientation, and speed of the vehicle by detecting reference vehicles, and distance estimation is obtained from TAT points from the bottom of the tank. (e location estimates (turnaround time) at a speciﬁc frequency. (us, the vehicle are highly accurate without drift, allowing them to be used as can estimate its position by algorithms based on recursive feedback measurements for low-level speed-based control- least squares (RLS) or using extended Kalman ﬁlters. (e lers. Its computing system is 12.5 Hz in real time. vehicle must save transponders positions. Some works related to the implementation of an LBL system correspond to Christopher von Alt et al. (ose who 5.2. Orientation and Motion Control. It is necessary to take developed REMUS [8], a torpedo-shaped underwater robot controlling the orientation and movement of underwater created for exploration of marine resources, presented two vehicles into account, and it may demand an exploration modes of operation: autonomous and nonautonomous. In mission or some work that requires manipulation or ex- the autonomous mode, the robot had to follow a path traction on the seabed. However, due to the presence of formed by acoustic transponders implanted on the seabed, external disturbances and uncertainties in the marine en- and REMUS acted as a target hunter. (e transponders vironment, linear control methods are not very eﬃcient, so it distribution deﬁned the navigation path of REMUS; on the is necessary to apply advanced robust control methods. (e other hand, in the nonautonomous mode, the navigation objective of an orientation control is to retain the required was carried out with the help of a boat, and REMUS followed orientation regardless of swell and unpredictable distur- it through an acoustic communication. bances in the environment. (at is why a hydrodynamic 20 Journal of Robotics Ping Ping Ping 1 Ping (a) (b) Figure 41: Variants of the standard LBL conﬁguration. (a) An LBL conﬁguration without query pings. (b) (e conﬁguration shown allows the beacons to obtain their locations using GPS [62]. Ping Ping Figure 42: Ultrashort Baseline system [62]. model and mathematical parameters of the structure must hidden in the structure. However, most researchers make use of be obtained ﬁrst. Computational Fluid Dynamics (CFD) simulations on the Likewise, to establish a control system, the following behavior of their framework to reduce the error due to changing environmental conditions which are diﬃcult to predict. points must be taken into account: the performance of the system is limited, adding that the behavior of the control system must be robust in terms of both stability and per- 5.2.1. Sliding Mode Control. (e sliding mode control formance, since it takes into account the energy manage- (SMC) is a robust control for modelling uncertainty and ment and optimization of the entire system [65]. (is parameter variations and has good disturbance rejection approach takes into account the inevitable imperfection in characteristics. (ere have been a wide variety of applica- physical systems and variables; one of the investigations on tions of the same [67–71]. However, it inherits a discon- the performance of a new control strategy for imperfect tinuous control action; therefore, the chattering systems is observed in [66], starting from an electrome- phenomenon that occurs when the system operates close to chanical system based on a light structure that acts as a the sliding surface will occur. Sometimes this discontinuous support and supply for the simple coils found in the control action can even make system performance unstable. structure; the purpose of this research is the simulation of control systems for imperfect systems that, thanks to the peculiar properties in the structure, the eﬀects of vibration 5.2.2. Adaptive Control. Side Zhao and Junku Yuh proposed signals on the hidden dynamic system of the imperfect an adaptive control based on a disturbance observer [69, 72]; system can be observed. the control scheme of this system has an adaptive controller Given the premise on nonlinear control systems in im- based on a nonregressor, and it is the outer loop of the control perfect systems with more than two variables, it is considered scheme, while the inner loop controller is the disturbance that most of the research carried out within the ﬁeld of hy- observer. (ese two elements mentioned above are the drodynamics and the behavior of an ROV is established in only components of the adaptive control system proposed by the the movement controls, guaranteeing the movement of the authors, which is robust against external disturbances and robot in the established route without considering the dynamics unpredictable behaviors due to the self-adjustment of its Journal of Robotics 21 + d y++ Nonregressor- ∗ + u u y based adaptive P –+ controller +– d –1 f P ξ n y Disturbance observer Figure 43: Diagram of the adaptive control system based on a disturbance observer [72]. ∆E – ∆e ∆e z –1 Depth rate of + uZ Limiter FLC change f t(E, ∆E) Motion Limiter z Limiter Figure 44: Fuzzy control system diagram for depth [74]. ∆ E – ∆e ∆e Ψ –1 Yaw angle rate + uN Limiter FLC f (E, ∆E) Rotation Limiter Ψ Limiter Figure 45: Fuzzy control system diagram for guidance [74]. Table 4: Classiﬁcation of orientation and movement control methods most used in UUV. Control methods Contribution of the method (i) Path control in the horizontal and vertical plane in an AUV, using 6 degrees of freedom [77–79]. (ii) Integrated PID control with a backstepping control for trajectory control of an underactuated AUV [80]. PID (iii) Implementation of a self-adaptive fuzzy PID controller [81]. (iv) Proposal of a self-tuneable PID control, using neural networks [82]. (i) Avoid collision in marine vessels through an intelligent decision-making system [83]. (ii) Linear approximation control for tuning parameters of a fuzzy controller [84]. Fuzzy (iii) Features a torque controller, calculated with a trajectory compensation technique [85]. (iv) Adaptive fuzzy control for a multiple-input multiple-output (MIMO) system [86]. (i) Stabilize the motion control of an AUV disturbed by unknown hydrodynamic coeﬃcients [87]. Adaptive (ii) Introducing an enhanced composite model reference adaptive control method to control AUV motion [88]. (iii) Adaptive control based on sliding mode control and fuzzy logic [89]. Sliding modes (i) Improved response, insensitive to parameter variation and disturbance [90–93]. (i) A bioinspired neurodynamic model is presented, used for a kinematic controller [94]. Neural networks (ii) Adaptive neural network controller combining hidden single-layer neural network and sliding mode control [95]. 22 Journal of Robotics control parameters. (ey have implemented three controllers, factors in the design such as hydrodynamic drag, propulsion PID, PID plus Dob, and the ADOB (adaptive controller based force, and energy consumption, giving room to achieving better results with further study. It is important to mention on a disturbance observer), to compare and evaluate the eﬃciency and performance, as shown in Figure 43. that the biomimetic form of a robot not only implies im- provements for itself but also reduces the degree of risk to possible alterations to a natural biological environment at 5.2.3. Neural Networks. Recently, neural networks have the time of the interaction. (e constant improvement of gained considerable attention in robotic systems control due biomimetic technology has broken the trend of only to their versatile properties, such as nonlinear mapping, implementing robots based on propulsion by caudal or learning ability, and parallel processing [67, 69, 73]. (e pectoral ﬁn; studies have opened a new window for the use of most useful feature of neural networks in control is their intelligent actuators, materials capable of providing better ability to approximate arbitrary linear or nonlinear mapping mechanical characteristics, such as greater ﬂexibility under through learning. Due to this property, neural networks have speciﬁc conditions, getting closer to the eﬃciency of real been proven to be a suitable tool to control complex non- biological models with diverse morphological characteris- linear dynamic systems. However, due to their arithmetic tics. (e use of a pressure sensor has become much more complexity, their implementation in engineering is not easy. standardized in the manufacture of any UUV, simply to obtain the depth data. However, some of these robots still have a dedicated depth sensor, thus achieving a greater 5.2.4. Fuzzy Control. Control based on fuzzy logic or fuzzy comparison range between points. In most current ROVs, control (FC) in English is a control that has supplanted we can observe the constant use of an IMU sensor with the conventional technologies in many applications combination of sonar to ﬁnd the underwater positioning, [35, 68, 74–76]. An important property of fuzzy logic is its also applying a ﬁlter for the correct interpretation of data. ability to express ambiguity in human thought. (erefore, GPS modules are used more in AUVs than in ROVs, because when the mathematical model of the process does not exist ROVs present a physical connection between the robot and or does exist with uncertainties, the FC becomes an alter- controller, while the AUVs are programmed with a path or native way for dealing with the unknown process. However, route to follow; that is why they emerge to the surface to the large number of fuzzy rules for high-order systems obtain their position before making a submersion. (e use of makes the analysis complex. A fuzzy-based depth control brushless motors has become very popular with the inte- scheme is illustrated in Figure 44 and a fuzzy-based yaw gration of propellers. It is found in diﬀerent types of UUVs angle control scheme is illustrated in Figure 45. long before the 20th century. (e advantage of this type of Table 4 summarizes the main contributions of some motor is adequate cost, better quality, and less maintenance additional navigation and orientation control methods that than other motors. (e application of new sensors for the correspond to those most used by UUVs. In the ﬁrst row of acquisition of oceanographic data in robots has become Table 4, some linear methods of proportional-integral-de- increasingly common, as a result, mainly due to the growing rivative (PID) type are included, which work in conjunction interest in the study of marine ecosystems and the con- with the other previously reviewed methods. servation of species. Many of the works reviewed, related to the control of direction or displacement with diﬀerent en- 6. Conclusions gines, do not show much detail in the electronic components (e ROVs ﬁrst shape was rectangular with an open hull and used, making it diﬃcult to trace an evolutionary timeline of emerging technologies of electronic components used in positive buoyancy; it was so big that it could not be transported by a single person, and it was necessary to place UUVs. (e methods and algorithms for navigating UUVs a pulley in the water. (e torpedo-shaped underwater ro- are mainly implemented in autonomous vehicles (AUVs). bot’s design allowed robots to be faster during their un- You can see the trend towards map-based navigation derwater operation. Inside the investigated robots, it was methods as opposed to those that use ﬁxed beacons around their exploration environment. (e orientation and move- found that only four thrusters provide 5 degrees of freedom compared to others that need six thrusters to reach 5 de- ment control is applicable for both ROVs and AUVs, highlighting the routes control and trajectory tracking to- grees. (e studied robots determine that, to achieve all the degrees of freedom, the robots must have eight thrusters wards autonomous vehicles. (e trend of new control methods is to apply combinations of more than one method installed or ﬁve thrusters with two actuators to change the force direction. One of the most used materials in the to improve their characteristics and achieve ﬁner control. manufacture of aquatic robots is aluminum, because it does not deform at high pressures; it is a dense material and is not Conflicts of Interest corrosive. Most of the researches evaluated are designed to operate under positive buoyancy; to be able to submerge, (e authors declare that they have no conﬂicts of interest. they must activate the immersion thrusters; and to return to the surface it is enough that the thrusters are deactivated. Acknowledgments (e improvement of aerodynamic and hydrodynamic characteristics of aquatic robots with a biomimetic approach (is research work is part of the project identiﬁed with the has accumulated great results, improving very important study BM-PNIPA-PES-SIADE-PP (No. 000027; Contract Journal of Robotics 23 Mechatronics and Automation, pp. 1382–1387, IEEE, Tianjin, No. 256-2018), which is supported by PNIPA and World China, August 2014. Bank. (e authors thank companies MASTER PROVIDER [17] F. Maurelli, J. Cartwright, N. Johnson, and Y. Petillot, “Nessie and ACEMAR for help. iv autonomous underwater vehicle wins the sauc-e compe- tition,” in Proceedings of the 10th International Conference on References Mobile Robots and Competitions (ROBOTICA 2010), Leiria, Portugal, March 2010. [1] J. E. Cohen, C. Small, A. Mellinger, J. Gallup, and J. Sachs, [18] R. Yang, B. Clement, A. Mansour, H. Li, M. Li, and N. Wu, “Estimates of coastal populations,” Science, vol. 278, no. 5341, “Modeling of a complex-shaped underwater vehicle,” in pp. 1209–1213, 1997. Proceedings of the 2014 IEEE International Conference on [2] J. Yuh, “Design and develop of an autonomous underwater Autonomous Robot Systems and Competitions (ICARSC), vehicle: a survey,” Autonomous Robots, vol. 8, no. 1, pp. 7–24, pp. 36–41, IEEE, Espinho, Portugal, May 2014. [19] S. Hidaka, S. Kawashima, S. Nam et al., “System design and [3] E. V. Lewis, “Principles of naval architecture,” Seakeeping and hardware development of autonomous underwater robot Controllability, vol. 3, Society of Naval Architects & Marine “daryabird,” in Proceedings of the AUVSI&ONR’s 18th Annual Engineers, Jersey City, NJ, USA, 1988. RoboSub Competition Journal Paper, San Diego, CA, USA, [4] S. Chutia, N. M. Kakoty, and D. Deka, “A review of underwater July 2015. robotics, navigation, sensing techniques and applications,” in [20] E. H. Binugroho, B. Setyawan, R. S. Dewanto, Proceedings of the Advances in Robotics, pp. 1–6, New Delhi, D. Pramadihanto et al., “Static and dynamic analysis of erov India, July 2017. mechanical design using CFD,” in Proceedings of the 2019 [5] B. K. Sahu and B. Subudhi, “(e state of art of autonomous International Electronics Symposium (IES), pp. 628–632, IEEE, underwater vehicles in current and future decades,” in Pro- Surabaya, Indonesia, September 2019. ceedings of the 2014 First International Conference on Auto- [21] Z. Yunhui, S. Lina, Q. Lu, and F. Qiang, “(e frame structure mation, Control, Energy and Systems (ACES), pp. 1–6, IEEE, of underwater Robot for shallow water,” in Proceedings of the Adisaptagram, India, February 2014. 2019 International Conference on Precision Machining, Non- [6] R. B. Wynn, V. A. I. Huvenne, T. P. Le Bas et al., “Auton- traditional Machining and Intelligent Manufacturing (PNTIM omous underwater vehicles (AUVS): their past, present and 2019), Atlantis Press, Xi’an, China, September 2019. future contributions to the advancement of marine geo- [22] W.-S. Chu, K.-T. Lee, S.-H. Song et al., “Review of biomimetic science,” Marine Geology, vol. 352, pp. 451–468, 2014. underwater robots using smart actuators,” International [7] R. D. Christ, “(e ROV manual a user guide for remotely Journal of Precision Engineering and Manufacturing, vol. 13, operated vehicles,” in (e ROV Manual, pp. 57-58, IEEE, no. 7, pp. 1281–1292, 2012. Piscataway, NJ, USA, 2014. [23] S.-H. Ahn, K.-T. Lee, H.-J. Kim, R. Wu, J.-S. Kim, and [8] C. von Alt, B. Allen, T. Austin, and R. Stokey, “Remote en- S.-H. Song, “Smart soft composite: an integrated 3D soft vironmental measuring units,” in Proceedings of the IEEE morphing structure using bend-twist coupling of anisotropic Symposium on Autonomous Underwater Vehicle Technology materials,” International Journal of Precision Engineering and (AUV’94), pp. 13–19, IEEE, Cambridge, MA, USA, July 1995. Manufacturing, vol. 13, no. 4, pp. 631–634, 2012. [9] B. Allen, R. Stokey, T. Austin et al., “Remus: a small, low cost [24] W.-S. C. Hyung-Soo Kim, J.-Y. Lee, and S.-H. Ahn, “Design AUV; system description, ﬁeld trials and performance re- and fabrication of soft morphing ray propulsor: undulator and sults,” in Proceedings of the Oceans’97. MTS/IEEE Conference, oscillator,” Soft Robotics, vol. 4, no. 1, pp. 49–60, 2016. pp. 994–1000, IEEE, Halifax, Canada, October 1997. [25] F. Cairone, S. Gagliano, D. C. Carbone, G. Recca, and [10] M. Radojevic, M. M. Nawaf, F. Maurelli et al., Nessie VI M. Bucolo, “Micro-optoﬂuidic switch realized by 3D printing Autonomous Underwater Vehicle, 2011, https://braincadet. technology,” Microﬂuid and Nanoﬂuidics, vol. 20, 2016. com/papers/nessie11.pdf. [26] J. Liu, Z. Wu, and J. Yu, “Design and implementation of a [11] X. Xiang, C. Yu, and Q. Zhang, “On intelligent risk analysis robotic dolphin for water quality monitoring,” in Proceedings and critical decision of underwater robotic vehicle,” Ocean of the 2016 IEEE International Conference on Robotics and Engineering, vol. 140, pp. 453–465, 2017. Biomimetics (ROBIO), pp. 835–840, IEEE, Qingdao, China, [12] X. Ai, S. Kang, and W. Chou, “System design and experiment December 2016. of the hybrid underwater vehicle,” in Proceedings of the 2018 [27] X. Yang, Z. Wu, and J. Yu, “Design and implementation of a International Conference on Control and Robots (ICCR), Hong robotic shark with a novel embedded vision system,” in Kong, China, September 2018. Proceedings of the 2016 IEEE International Conference on [13] G. Yao, J. Liang, T. Wang et al., “Development of a turtle-like Robotics and Biomimetics (ROBIO), pp. 841–846, IEEE, underwater vehicle using central pattern generator,” in Pro- Qingdao, China, December 2016. ceedings of the 2013 IEEE International Conference on Robotics [28] J. Sverdrup-(ygeson, E. Kelasidi, K. Y. Pettersen, and and Biomimetics (ROBIO), Shenzhen, China, December 2013. J. T. Gravdahl, “(e underwater swimming manipulator-a [14] X. Lin and S. Guo, “Development of a spherical underwater bio-inspired AUV,” in Proceedings of the 2016 IEEE/OES robot equipped with multiple vectored water-jet-based thrusters,” Journal of Intelligent & Robotic Systems, vol. 67, Autonomous Underwater Vehicles (AUV), pp. 387–395, IEEE, Tokyo, Japan, November 2016. no. 3-4, pp. 307–321, 2012. [15] S. Guo, Y. He, L. Shi et al., “Modeling and experimental [29] F. Hoﬀmann and A. B. Kesel, “Biologically inspired optimi- zation of underwater vehicles hull geometries and ﬁn pro- evaluation of an improved amphibious robot with compact structure,” Robotics and Computer-Integrated Manufacturing, pulsion,” in Proceedings of the OCEANS 2019, pp. 1–4, IEEE, Marseille, France, June 2019. vol. 51, pp. 37–52, 2018. [16] Y. He, S. Guo, L. Shi, S. Pan, and Z. Wang, “3D printing [30] S. Du, Z. Wu, J. Wang, S. Qi, and J. Yu, “Design and control of a two-motor-actuated tuna-inspired robot system,” IEEE technology-based an amphibious spherical robot,” in Pro- ceedings of the 2014 IEEE International Conference on Transactions on Systems, Man, and Cybernetics: Systems, 2019. 24 Journal of Robotics [31] H. Saha and Parhi, “Conceptual design of an underwater Journal of Nuclear Science and Technology, vol. 55, no. 3, robot,” in 2015 International Conference on Man and Machine pp. 1–10, 2017. Interfacing (MAMI), pp. 1–6, IEEE, Bhubaneswar, India, [48] M. D. Ageev, A. A. Boreyko, V. E. Gornak et al., “Modernized December 2015. TSL-underwater robot for tunnel and shallow-water inspec- [32] O. L. Osen, R.-I. Sandvik, V. Rogne, and H. Zhang, “A novel tion,” Institute for Marine Techonology Problems FEB RAS, low cost rov for aquaculture application,” in Proceedings of the Vladivostok, Russia, 2000. [49] V. Bobkov, M. Sporyshev, and A. Scherbatyuk, “Evaluation of OCEANS 2017, pp. 1–7, IEEE, Anchorage, Scotland, June 2017. the eﬀectivenes of the visual navigation technique for an autonomous underwater vehicle,” in Proceedings of the 2018 [33] M. Yang, Z. Sheng, Y. Che, J. Hu, K. Hu, and Y. Du, “Design of small monitoring rov for aquaculture,” in Proceedings of the OCEANS—MTS/IEEE Kobe Techno-Ocean (OTO), IEEE, OCEANS 2019, pp. 1–9, IEEE, Marseille, France, June 2019. Kobe, Japan, May 2018. [34] Z. M. Zain, M. M. Noh, K. A. Ab Rahim, and N. Harun, [50] M. Pedersen, J. Bruslund, R. Gade, and T. Moeslund, “De- “Design and development of an X4-ROV,” in Proceedings of tection of marine animals in a new underwater dataset with the 2016 IEEE International Conference on Underwater System varying visibility,” in Section of Biology and Enviromental Technology: (eory and Applications (USYS), pp. 207–211, ScienceAalborg University, Aalborg, Denmark, 2019. [51] C. Yeng, S. Guo, L. Shi, and J. Du, “Characteristics evaluation IEEE, Penang, Malaysia, December 2016. [35] C. S. Chin and W. P. Lin, “Robust genetic algorithm and fuzzy of the vertical motion of a spherical underwater,” in Pro- ceeding of the 2012 IEEE International Conference on Robotics inference mechanism embedded in a sliding-mode controller for an uncertain underwater robot,” IEEE/ASME Transactions and Biomimetics, Guangzhou, China, December 2012. [52] B. W. D. R. Yoeger and A. M. Bradley, (e Autonomous on Mechatronics, vol. 23, no. 2, pp. 655–666, 2018. [36] R. Jehangir, Bluerov2 Heavy Conﬁguration, Seascape Subsea Benthic explorer (ABE): an AUV Optimized for Deep Seaﬂoor B.V., Sleeuwijk, Netherlands, 2018. Studies, Woods Hole Oceanographic Institution, Dept. of [37] C. Yue, S. Guo, M. Li, Y. Li, H. Hirata, and H. Ishihara, Applied Ocean Physics and Engineering, Falmouth, MA, Mechatronic System and Experiments of a Spherical Underwater USA, 1991. Robot: SUR-II, Springer Science+Business Media, Dordrecht, [53] M. Doniec, I. Vasilescu, C. Detweiler, and D. Rus, “Complete Netherlands, 2015. SE3 underwater robot control with arbitrary thruster con- [38] F. C. J. T. Pennington and M. Blum, “Seawater sampling by an ﬁgurations,” in Proceedings of the 2010 IEEE International autonomous underwater vehicle: “gulper” sample validation Conference on Robotics and Automation Anchorage Conven- for nitrate, chlorophyll, phytoplankton, and primary pro- tion District, p. 6, IEEE, Anchorage, AL, USA, May 2010. duction,” Limnology and Oceanography: Methods, vol. 14, [54] R. Saltaren, C. Garcia, C. Alvarez, and R. Aracil, “Experiences no. 1, pp. 14–23, 2016. and results from designing and developing a 6 DOF under- [39] M. Choyekh, “Development and operation of underwater water parallel robot,” Robotics and Autonomous Systems, robot for autonomous tracking and monitoring of subsea vol. 59, no. 2, pp. 101–112, 2010. plumes after oil spill and gas leak from seabed, and analyses of [55] J. Choi, T. Lee, T. Kim, J. Jung, and H. Choi, “Development of measured data,” Doctoral thesis, Osaka University Knowledge a ROV for visual inspection of harbor structures,” in Pro- ceedings of the 2017 IEEE Underwater Technology (UT), Busan, Archive, School of Engineering Osaka University, Osaka, Japan, 2016. South Korea, February 2017. [56] P. L. B. Jun and H. Shim, “Approximated generalized torques [40] F. Sapuppo, F. Schembri, L. Fortuna, and M. Bucolo, “Microﬂuidic circuits and systems,” IEEE Circuits and Systems by the hydrodynamic forces acting on legs of an underwater Magazine, vol. 9, no. 3, pp. 6–19, 2009. walking robot,” in International Journal of Ocean System [41] T. Fukuba, T. Noguchi, K. Okamura, and T. Fujii, “Adenosine Engineering, vol. 1, no. 4, pp. 222–229, IEEE, 2011. triphosphate measurement in deep sea using a microﬂuidic [57] K. P. Valavanis, D. Gracanin, M. Matijasevic, R. Kolluru, and G. A. Demetriou, “Control architectures for autonomous device,” Micromachines, vol. 9, no. 8, 2018. [42] F. S. Florinda Schembri and M. Bucolo, “Experimental underwater vehicles,” IEEE Control Systems Magazine, vol. 17, no. 6, pp. 48–64, 1997. classiﬁcation of nonlinear dynamics in microﬂuidic bubbles’ ﬂow,” Nonlinear Dinamics, vol. 4, pp. 2807–2819, 2012. [58] Y. Zhang and L. Cheng, “A distributed protocol for multi-hop underwater robot positioning,” in Proceedings of the 2004 [43] A. Limanes, J. Cartwright, B. Davis et al., Nessie III Auton- omous Underwater VehicleHeriot-Watt University, Scotland, IEEE International Conference on Robotics and Biomimetics, UK, 2008. pp. 480–484, IEEE, Dali, China, December 2004. [44] M. B. Bo Zhao, S. Roger, and F. Dukan, “Particle ﬁlter for fault [59] Z. Yan, S. Peng, J. Zhou, J. Xu, and H. Jia, “Research on an diagnosis and robust navigation of underwater robot,” Improved dead reckoning for AUV navigation,” in Pro- Transactions on Control Systems Technology, vol. 22, no. 6, ceedings of the 2010 Chinese Control and Decision Conference, 2014. pp. 1793–1797, IEEE, Xuzhou, China, May 2010. [45] S. Mintchev, E. Donati, S. Marrazza, and C. Stefanini, [60] I. Klein and R. Diamant, “Dead reckoning for trajectory es- Mechatronic Design of a Miniature Underwater Robot for timation of underwater drifters under water currents,” Journal of Marine Science and Engineering, vol. 8, no. 3, p. 205, 2020. Swarm Operations, Hong Kong Convention and Exhibition Center, Hong Kong, China, 2014. [61] J. H. Kepper IV, B. C. Claus, and J. C. Kinsey, “A navigation solution using a MEMS IMU, model-based dead-reckoning, [46] A. Jebelli, M. Yagoub, and B. Dhillon, “Design and control of underwater robots with rotating thrusters,” IAES Interna- and one-way-travel-time acoustic range measurements for tional Journal of Robotics and Automation, vol. 5, no. 4, autonomous underwater vehicles,” IEEE Journal of Oceanic pp. 284–294, 2016. Engineering, vol. 44, no. 3, pp. 664–682, 2018. [47] G. Y. M. Dong and W. Chou, “A new navigation strategy for [62] J. J. Leonard and A. Bahr, “Autonomous underwater vehicle underwater robot in reactor pool combined propeller speed navigation,” in Springer Handbook of Ocean Engineering, detection and dynamics analysis with sonar data correction,” pp. 341–358, Springer, Cham, Switzerland, 2016. Journal of Robotics 25 [63] A. Alcocer, P. Oliveira, and A. Pascoal, “Study and imple- [79] W. Zhang, H. Wang, X. Bian, Z. Yan, and G. Xia, “(e ap- mentation of an ekf gib-based underwater positioning sys- plication of self-tuning fuzzy PID control method to recov- tem,” Control Engineering Practice, vol. 15, no. 6, pp. 689–701, ering AUV,” in Proceedings of the 2012 OCEANS, pp. 1–5, 2007. IEEE, Virginia Beach, VA, USA, October 2012. [64] M. Carreras, P. Ridao, R. Garc´ıa, and T. Nicosevici, “Vision- [80] N. H. Z. M. Zain, “A backstepping based PID controller for based localization of an underwater robot in a structured stabilizing an underactuated X4-AUV,” ARPN Journal of Engineering and Applied Sciences, vol. 10, no. 21, 2015. environment,” vol. 1, pp. 971–976, in Proceedings of the 2003 IEEE International Conference on Robotics and Automation [81] M. H. Khodayari and S. Balochian, “Modeling and control of autonomous underwater vehicle (AUV) in heading and depth (Cat. No. 03CH37422), vol. 1, pp. 971–976, IEEE, Taipei, China, September 2003. attitude via self-adaptive fuzzy PID controller,” Journal of [65] L. Fortuna, A. Buscarino, M. Frasca, and C. Famoso, “Chapter Marine Science and Technology, vol. 20, no. 3, pp. 559–578, 1: imperfect uncertain systems,” in Control of Imperfect 2015. Nonlinear Electromechanical Large Scale Systems, World [82] R. Hernandez-Alvarado, ´ L. G. Garc´ıa-Valdovinos, T. Salgado- Jimenez, ´ A. Gomez-Espinosa, ´ and F. Fonseca-Navarro, Scientiﬁc Series on Nonlinear Science, Springer, Boston, MA, USA, 2017. “Neural network-based self-tuning PID control for under- water vehicles,” Sensors, vol. 16, no. 9, p. 1429, 2016. [66] M. Bucolo, A. Buscarino, C. Famoso, L. Fortuna, and M. Frasca, “Control of Imperfect dynamical systems,” Non- [83] L. Perera, J. Carvalho, and C. G. Soares, “Fuzzy logic based decision making system for collision avoidance of ocean linear Dynamics volume, vol. 98, pp. 2989–2999, 2019. [67] T. Chatchanayuenyong and M. Parnichkun, “Neural network navigation under critical collision conditions,” Journal of based-time optimal sliding mode control for an autonomous Marine Science and Technology, vol. 16, no. 1, pp. 84–99, 2011. underwater robot,” Mechatronics, vol. 16, no. 8, pp. 471–478, [84] M. Aras, M. Shahrieel, H. I. Jaafar, and M. K. Anuar, “Tuning 2006. process of single input fuzzy logic controller based on linear [68] S. Guo, J. Du, X. Lin, and C. Yue, “Adaptive fuzzy sliding control surface approximation method for depth control of mode control for spherical underwater robots,” in Proceedings underwater remotely operated vehicle,” Journal of Engineering and Applied Sciences, vol. 8, no. 6, pp. 208–214, 2013. of the 2012 IEEE International Conference on Mechatronics and Automation, pp. 1681–1685, IEEE, Chengdu, China, [85] Y. Chen, R. Zhang, X. Zhao, and J. Gao, “Tracking control of underwater vehicle subject to uncertainties using fuzzy in- August 2012. [69] M.-j. Zhang and Z.-z. Chu, “Adaptive sliding mode control verse desired trajectory compensation technique,” Journal of based on local recurrent neural networks for underwater Marine Science and Technology, vol. 21, no. 4, pp. 624–650, robot,” Ocean Engineering, vol. 45, pp. 56–62, 2012. 2016. [70] S. Jin, J. Bak, J. Kim, T. Seo, and H. S. Kim, “Switching pd- [86] H. F. Ghavidel and A. A. Kalat, “Robust control for mimo based sliding mode control for hovering of a tilting-thruster hybrid dynamical system of underwater vehicles by composite underwater robot,” PLoS One, vol. 13, no. 3, 2018. adaptive fuzzy estimation of uncertainties,” Nonlinear Dy- [71] A. J. Healey and D. Lienard, “Multivariable sliding mode namics, vol. 89, no. 4, pp. 2347–2365, 2017. control for autonomous diving and steering of unmanned [87] H. R. Kooﬁgar, “Robust adaptive motion control with envi- ronmental disturbance rejection for perturbed underwater underwater vehicles,” IEEE Journal of Oceanic Engineering, vol. 18, no. 3, pp. 327–339, 1993. vehicles,” Journal of Marine Science and Technology, vol. 22, [72] S. Zhao and J. Yuh, “Experimental study on advanced un- no. 4, pp. 455–462, 2014. derwater robot control,” IEEE Transactions on Robotics, [88] C. D. Makavita, H. D. Nguyen, D. Ranmuthugala, and vol. 21, no. 4, pp. 695–703, 2005. S. G. Jayasinghe, “Composite model reference adaptive [73] M. Chen, B. Jiang, J. Zou, and X. Feng, “Robust adaptive control for an unmanned underwater vehicle,” Underwater tracking control of the underwater robot with input non- Technology, vol. 33, no. 2, pp. 81–93, 2015. linearity using neural networks,” International Journal of [89] S. Pezeshki, A. Ghiasi, M. Badamchizadeh, and K. Sabahi, Computational Intelligence Systems, vol. 3, no. 5, pp. 646–655, “Adaptive robust control of autonomous underwater vehicle,” Journal of Control, Automation and Electrical Systems, vol. 27, [74] I. Akkizidis, G. Roberts, P. Ridao, and J. Batlle, “Designing a no. 3, pp. 250–262, 2016. fuzzy-like pd controller for an underwater robot,” Control [90] A. Bagheri and J. J. Moghaddam, “Simulation and tracking Engineering Practice, vol. 11, no. 4, pp. 471–480, 2003. control based on neural-network strategy and sliding-mode [75] X. Xiang, C. Yu, and Q. Zhang, “Robust fuzzy 3d path fol- control for underwater remotely operated vehicle,” Neuro- lowing for autonomous underwater vehicle subject to un- computing, vol. 72, no. 7–9, pp. 1934–1950, 2009. certainties,” Computers & Operations Research, vol. 84, [91] D. Zhu and B. Sun, “(e bio-inspired model based hybrid pp. 165–177, 2017. sliding-mode tracking control for unmanned underwater [76] C. Yu, X. Xiang, L. Lapierre, and Q. Zhang, “Nonlinear vehicles,” Engineering Applications of Artiﬁcial Intelligence, guidance and fuzzy control for three-dimensional path fol- vol. 26, no. 10, pp. 2260–2269, 2013. lowing of an underactuated autonomous underwater vehicle,” [92] J. Xu, M. Wang, and L. Qiao, “Dynamical sliding mode Ocean Engineering, vol. 146, pp. 457–467, 2017. control for the trajectory tracking of underactuated un- [77] I. Kaminer, A. Pascoal, E. Hallberg, and C. Silvestre, “Tra- manned underwater vehicles,” Ocean Engineering, vol. 105, jectory tracking for autonomous vehicles: an integrated ap- pp. 54–63, 2015. proach to guidance and control,” Journal of Guidance, [93] Z. Chu and D. Zhu, “Adaptive sliding mode heading control Control, and Dynamics, vol. 21, no. 1, pp. 29–38, 1998. for autonomous underwater vehicle including actuator dy- [78] G. Antonelli, F. Caccavale, S. Chiaverini, and G. Fusco, “A namics,” in Proceedings of the OCEANS 2016, pp. 1–5, IEEE, novel adaptive control law for underwater vehicles,” IEEE Shanghai, China, September 2016.. Transactions on Control Systems Technology, vol. 11, no. 2, [94] B. Sun, D. Zhu, F. Ding, and S. X. Yang, “A novel tracking pp. 221–232, 2003. control approach for unmanned underwater vehicles based on 26 Journal of Robotics bio-inspired neurodynamics,” Journal of Marine Science and Technology, vol. 18, no. 1, pp. 63–74, 2013. [95] J. Gao, P. Wu, B. Yang, and F. Xia, “Adaptive neural network control for visual servoing of underwater vehicles with pose estimation,” Journal of Marine Science and Technology, vol. 22, no. 3, pp. 470–478, 2017.

Journal

Journal of Robotics – Hindawi Publishing Corporation

Published: Jul 6, 2021

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Review on Unmanned Underwater Robotics, Structure Designs, Materials, Sensors, Actuators, and Navigation Control

Review on Unmanned Underwater Robotics, Structure Designs, Materials, Sensors, Actuators, and Navigation Control

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Review on Unmanned Underwater Robotics, Structure Designs, Materials, Sensors, Actuators, and Navigation Control

Review on Unmanned Underwater Robotics, Structure Designs, Materials, Sensors, Actuators, and Navigation Control

References (107)

Abstract

Journal

Recommended Articles

There are no references for this article.

Our policy towards the use of cookies