Encyclopedia of Ocean Engineering

Living Edition
| Editors: Weicheng Cui, Shixiao Fu, Zhiqiang Hu

AUV/ROV/HOV Control Systems

  • Changhui SongEmail author
Living reference work entry
DOI: https://doi.org/10.1007/978-981-10-6963-5_282-1



Control system is a kind of active regulation system designed to improve or optimize specific performance or indictors for economic, social, or physical systems. AUV/ROV/HOV belongs to a multi-degree-of-freedom underwater platform system with high redundancy, coupling, and model nonlinearity. It works underactuated in complex underwater environment, and the body must maintain autonomous stability in the process of operation. Therefore, the design of appropriate control system to ensure the stability of the underwater platform during the operational process is of great significance to the successful and accurate completion of the underwater mission of AUV/ROV/HOV.

Scientific Fundamentals

Difficulties in the Design of Control System

Underwater vehicles have a large variety of types, and they are widely involved in undersea surveillance, inspection, and survey missions. Typically, gliders are common with a torpedo shape for long range missions, and human-occupied vehicles (HOVs) as well as remote operating vehicles (ROVs) are generally of a cubic shape used for hovering tasks. For some specific applications, undersea pipeline inspection, offshore infrastructure surveillance, and large vessel maintenance, AUV is preferred.

Regardless of modeling issues, the value of a model-based control approach depends on how robust and efficient the control scheme can adopt the hydrodynamic model. Potential trends of current methods focus on faster controllers to assist the pilot or the autopilot with better accuracy. Optimal controllers can reduce propelling actions to save the battery power as well as to increase the propeller lifespan. Moreover, numerous uncertainties should be considered, including parameter variations, nonlinear hydrodynamic damping effects, sensor transmit delays, and ocean current disturbances (Clement et al. 2016).

Each underwater vehicle is designed to perform specific functions. Even for the same type of underwater vehicle, their shapes may be very different. Therefore, the mathematical models of underwater vehicles will be much different. Each underwater vehicle pursues unique performance indicators. This makes it difficult to solve all control system design problems with a single theory or model system.

A decade of operational experience by numerous research groups has demonstrated ROVs to be ideal platforms for “rapid prototyping” and “rapid field deployment” of UUV subsystems. Recent examples include sonar imaging and survey, optical imaging and survey, navigation, control, oceanographic sampling, and subsea manipulation. Numerous research results first pioneered with ROVs, and towed vehicles are now commonly employed for autonomous underwater vehicles (AUV, HOV) and seafloor observatories (Smallwood et al. 1999).

Structure of Underwater Vehicle System

In the following chapters, the typical structure of the underwater vehicle system will be given. These components are the integral part for underwater vehicle to complete basal motion.

The submersible control system has two components, surface ship control system and underwater vehicle control system. The surface ship control system includes system GMT clock, ship dynamic positioning system, ship LBL navigation system, GPS satellite navigation system, real-time data logging system, science pay load interface, instruments, video distribution, and recording subsystem. The underwater vehicle control system includes vehicle control computer system, surface telemetry, pilot’s user interface (such as joystick, keyboard instruction), engineer user interface, manipulator user interface, work package interface, and pilot video display. Vehicle control computer system includes control process, positioning navigation process, LBL-Doppler navigation process, and power management and hotel process, shown in Fig. 1.
Fig. 1

Underwater vehicle system structure (Smallwood et al. 1999)

Surface Systems

The surface control system is the central “brain” of the ROV control system. It is comprised of a “core” system of safety-critical systems that are essential for safety and control of the ROV and an “extended” system providing non-safety-critical systems such as data logging and video recording. The safety-critical core system is comprised of the vehicle control computer and user interfaces for the vehicle pilot and engineer. The pilot station provides real-time video and navigation instruments, has joystick controls for closed-loop control of the vehicle reference trajectories and navigation waypoints, and has controls for the vehicle’s manipulator arms. The engineer station has a more comprehensive set of real-time vehicle status indicators and enables the engineer to control all vehicle subsystems. The modules are depicted in Fig. 1.

Concentrating the vehicle intelligence in the ship-board control computer dramatically simplifies and accelerates development – reprogramming an on-ship computer is significantly easier than reprogramming an embedded vehicle control computer.

Vehicle on-Board Core Control System

The on-board vehicle control system controls and monitors all vehicle sensors and actuators in response to real-time commands from a surface control system. We will adopt a “data concentrator” (DCON) type of architecture. Each data concentrator (DCON) module will independently control the power and data telemetry for an entire vehicle subsystem or scientific payload. The DCONs will receive commands from the surface control computer, monitor the status, and report data from on-board vehicle subsystems and instruments. Each data concentrator operates asynchronously and will communicate to the surface control computer via a high bandwidth fiber-optic telemetry link.

The data concentrator vehicle control system architecture employs relatively simple on-vehicle computer systems. We anticipate employing commercial off-the-shelf (COTS) embedded computers for the data concentrators. The simplicity of the data concentrator design will render them both highly reliable and easily reconfigurable (Smallwood et al. 1999).

Realization of Vehicle Control System

Generally, the underwater vehicle is equipped with scanning sonars and a color CCD camera, together with depth sensors and a fiber-optic gyroscope. This device is intended primarily as a research platform upon which to test novel sensing strategies and control methods. Autonomous navigation using the information provided by the vehicle’s on-board sensors represents one of the ultimate goals of the project (Williams et al. 2006).

Embedded Controller

At the heart of the robot control system is an embedded controller. Figure 2 shows a schematic diagram of the vehicle sensors and their connections. The vehicle uses a CompactPCI system running Windows CE that interfaces directly to the hardware and is used to control the motion of the robot and to acquire sensor data. While the Windows operating system doesn’t support hard real-time performance, it is suitable for soft real-time applications, and the wide range of development and debugging tools make it an ideal environment in which to test new navigation algorithms. Time-critical operations, such as sampling of the analog to digital converters, are performed on the hardware devices themselves and use manufacturer-supplied device drivers to transfer the data to the appropriate processes.
Fig. 2

Vehicle control system diagram system (Williams et al. 2006)

The sensor data is collated and sent to the surface using an Ethernet connection where a network of computers are used for further data processing and data logging and to provide the user with feedback about the state of the sub. Communications between the computers at the surface and the sub are via a tether. This tether also provides power to the robot, a coaxial cable for transmitting video data, and a leak detection circuit designed to shut off power to the vehicle in case of a leak. While some effort might have been spent on eliminating the tether, it was felt that the development of the navigational techniques was of more immediate interest.


Sonar is the primary sensor of interest on the vehicle. There are currently two sonars on the robot. A sonar unit has been mounted at the front of the vehicle. It is positioned such that its scanning head can be used as a forward and downward looking beam. This enables the altitude above the seafloor as well as the proximity of obstacles to be determined using the wide of the angle beam of the sonar.

The second sonar is an imaging sonar and has a dual frequency narrow beam sonar head that is mounted on top of the sub and is used to scan the environment in which the sub is operating. It can achieve 360° scan rates on the order of 0.25 Hz. The information returned from this sonar is used to build and maintain a feature map of the environment.

Internal Sensors

A laser Gyro has been included in the robot to allow the robot’s orientation to be determined. This sensor provides yaw rate information and is used to control the heading of the sub. We currently estimate the bias in the gyroscope reading prior to a mission. The bias compensated yaw rate is then integrated to provide an estimate of vehicle heading. Because the yaw rate signal will inevitably be noisy, the integration of this signal will cause the estimated heading to drift with time. At present, missions do not typically run for longer than 30 minutes, and yaw drift does not pose a significant problem on this time frame. In the future, a compass will be added to the system to allow us to periodically reset the heading for longer missions. This will also allow us to detect changes in the yaw rate bias that typically occur as a result of changes in the internal temperature of the unit.

A pressure sensor measures the external pressure experienced by the vehicle. This sensor provides a voltage signal proportional to the pressure and is sampled by an analogue to digital converter on the embedded controller. The pressure can then be converted to a depth below the surface of the ocean. Feedback from this sensor is used to control the depth of the sub.


A Panasonic camera in an underwater housing is mounted externally on the vehicle. It is used to provide video feedback of the underwater scenes in which the robot operates. This is a color camera and sends the video signal to the surface via the tether. A video acquisition card is then used to acquire the video signal for further image processing.


There are currently six thrusters on the vehicle. Four of these are oriented in the vertical direction, while the remaining two are directed horizontally. This gives the vehicle the ability to move itself up and down, control its yaw, pitch and roll and move forward and backward, and move lateral motion.

Vehicle Control Method

Control system of a mobile robot in six-dimensional space in an unstructured, dynamic environment that is found underwater can be a daunting and computationally intensive endeavor. Navigation and control both present difficult challenges in the subsea domain. We have developed a distributed, decoupled control architecture to help simplify the controller design.

The vehicle control architecture currently running on the sub is based on the Distributed Architecture for Mobile Navigation (DAMN) as proposed in Rosenblatt (1997). This behavior-based control architecture uses a centralized arbiter to combine votes from various behaviors running in the system in order to determine the optimal course of action to pursue. A behavior encapsulates the perception, planning, and task execution capabilities necessary to achieve one specific aspect of robot control and receives only that data specifically required for that task (Brooks 1986).

Decoupled Control

By decoupling the control problem, individual controller design is greatly simplified. In the case of the typical vehicle, vertical motion is controlled independently of its lateral motion using two separate PID controllers. These controllers are then tuned to provide the required performance in each case.

This division of control has been selected as it fits in with many of the anticipated missions to be undertaken by the vehicle. A typical mission might see the vehicle performing a survey of an area of the Great Barrier Reef while maintaining a fixed height above the seafloor (Williams et al. 1999). The task of performing the survey can then be made independent of maintaining the vehicle altitude. This also allows us to optimize the performance of the depth controller prior to deploying the navigation algorithms used for mapping of the environment to ensure that the vehicle has little risk of hitting the sea floor.

The behaviors that run on the vehicle are also divided into horizontal and vertical behaviors, and there are consequently two arbiters currently present. One is responsible for setting the desired depth of the vehicle, while the other sets the desired yaw and forward offset to achieve horizontal motion.

For a survey mission, the vertical behaviors would typically be responsible for keeping the sub from colliding with the seafloor. The vertical behaviors that run on the vehicle include maintain minimum depth, maintain minimum altitude, maintain depth, and maintain altitude. The combination of the outputs of these behaviors determines the depth at which the vehicle will operate. A large negative vote by the maintain minimum altitude behavior will keep the vehicle at a minimum distance from the seafloor. The horizontal behaviors that would run on a survey mission include follow line (sonar and/or vision), avoid obstacles, move to location, and perform survey. The combination of the outputs of these behaviors determines the orientation maintained by the vehicle as well as the forward offset applied to the two horizontal thrusters. This allows the vehicle to move forward while maintaining its heading.

Low- and High-Level Control

The control of the sub is further decoupled into low level- and high-level control processes. The low-level processes run on the embedded controller and are used to interface directly with the hardware (see Fig. 3). In addition, the PID controllers have been implemented at this level. This allows the PID controllers to respond quickly to changes in the state of the sub without being affected by the data processing and high-level control algorithms running at the surface.
Fig. 3

The low-level control processes that run on the vehicle controller. These processes include processes for sampling the internal sensor readings, computing the PID control outputs, and driving the thrusters (Williams et al. 2006)

The high-level processes run on a pair of Pentium machines at the surface control station. Information from the sub’s sensors is fed to a series of processes running on these machines. These processes use the data supplied by the sensors to determine the desired set points for the low-level control processes. The raw sensor data is preprocessed to produce virtual sensor information – information that is of interest to multiple behaviors in the system. The behaviors register their interest in the data being generated by the virtual sensors and send votes to the arbiters to specify their desired course of action. A task-level mission planner is used to enable and disable behaviors in the system depending on the current state of the mission and its desired objectives. The arbiter combines the votes from the behaviors and selects the optimal action to satisfy the goals of the system. A schematic representation of the data flow within the system is shown in Fig. 4. This control strategy relies on the ability of the sub to gather information about its environment and reason about its desired actions. By continuously monitoring the state of the environment, the sub is able to respond to changes as they occur.
Fig. 4

The high-level process and behaviors that run the vehicle. The sensor data is preprocessed to produce virtual sensor information available to the behaviors. The behaviors receive the virtual sensor information and send votes to the arbiters who send control signals to the low-level controllers (Williams et al. 2006)

Distributed Control

A number of processes have been developed to accomplish the tasks of gathering data from the robot’s sensors, processing this data, and reasoning about the course of action to be taken by the robot. These processes are distributed across a network of computers and communicate asynchronously via a TCP/IP socket-based interface using a message passing protocol developed at the center.

A central communications hub is responsible for routing messages between the distributed processes running on the vehicle and on the command station. Processes register their interest in messages being sent by other processes in the system, and the hub routes the messages when they arrive. While this communications structure has some drawbacks, such as potential communications bottlenecks and reliance on the performance of the central hub, it does provide some interesting possibilities for flexible configuration, especially during the development cycle of the system.

The above introduction mostly comes from Smallwood et al. 1999 and Williams (2006). More technical details or design method can be found in their related literature.

Key Applications

Over the past 20 years, with the widespread development of large-scale integrated circuit in the world, which plays an important role in the field of marine technology, many new submarine vehicles have been created to solve a wide range of problems. These devices have already demonstrated their effectiveness in performing emergency rescue, ocean survey, environment monitoring, inspections and operation, research, and other types of work.

Because the marine environment is unstructured and hazardous, the development of underwater vehicles will face many challenging scientific and engineering problems, and researchers have made great efforts to overcome these problems so far (Paull et al. 2014). The progress in new materials, sensor technology, computer technology, and advanced algorithms has greatly contributed to research and development activities in the underwater vehicle community.

With technological advances, underwater vehicles have the advanced control technology and intelligent control system that can efficiently perform scheduled tasks with its own decision-making and control capabilities.

Advanced Control Technology

The AUV control methodologies that have been proposed in the literature include linear control (see Yildiz et al. (2009)), sliding mode control (see Yang et al. (2013) and Kim et al. (2015)), and fuzzy and neural network control (see Khodayari and Balochian (2015) and Lakhekar et al. (2015)).

One simple and effective nonlinear control design approach that has been implemented in marine applications is dynamic inversion (DI), in which the control law is formulated to eliminate system nonlinearities by means of feedback. The DI control technique in turn allows to incorporate well-established linear control techniques. In spite of its flexibility and simplicity, standard DI lacks robustness to model uncertainties. Several modifications were added to the basic DI control structure to improve its robustness characteristics; see Steinicke and Michalka (2002) and Wang et al. (2012). Regardless of these attributes, DI has several shortcomings and limitations, including blind nonlinearity cancellation, large control effort, and numerical singularity configurations of square matrix inversion.

A new inversion-based control design methodology is Generalized Dynamic Inversion (GDI); see and Bajodah (2009). The methodology is of the left inversion type, and hence it does not involve deriving inverse equations of motion for the plant. In GDI, dynamic constraints are prescribed in the form of differential equation that encapsulates the control objective and is generalized inverted using the Moore-Penrose Generalized Inverse (MPGI)-based Greville formula. The GDI control technique has been an effective tool for several engineering problems; see Bajodah (2009), Bajodah (2010), Hameduddin and Bajodah (2012), and Gui et al. (2013). The Robust Generalized Dynamic Inversion (RGDI) control system is obtained by Ansari and Bajodah (2017).

Location and Navigation Technology

Navigation is one of the key AUV technologies because the localization, path tracking, and control of the vehicle are all based on precise navigation parameters. Some navigation methods commonly used for land and air are not suitable for underwater because of the attenuation effect of water on electromagnetic signals, and underwater navigation has become a challenging issue in AUV research (Paull et al. 2014). Among the many underwater navigation systems available, the inertial navigation system (INS) using inertial sensors typically acts as the central navigation system of AUVs because of its autonomy (Stutters et al. 2008).

Generally, the INS contains an inertial measurement unit (IMU), which consists of accelerometers measuring linear acceleration and gyroscopes measuring angular velocity; the accelerometers and the gyroscopes are usually made up of three mutually perpendicular accelerometers and three mutually perpendicular gyroscopes, respectively. For inertial navigation, the instantaneous speed and position of the vehicle are obtained by integrating the measured values of the accelerometers and gyroscopes. The errors of the IMU increase with increasing elapsed time due to the drift of accelerometers and gyroscopes. Theoretically, the velocity and heading errors accumulate linearly over time, and the position error accumulates exponentially over time (Ali and Mirza 2010). Therefore, the INS can provide relatively accurate navigation information within a short time, but it is physically impossible for a pure inertial navigation system to maintain the high-precision level throughout a mission. Aiding the INS with external information or measurements is an effective means of improving navigation accuracy and has been widely used. In AUV navigation, auxiliary sensors or other navigation systems, such as a Doppler Velocity Log (DVL), compass, pressure sensor, Global Positioning System (GPS), acoustic positioning system (APS), or geophysical navigation system, are usually combined with the INS to form an integrated navigation system (Kussat et al. 2005 and Vasilijevic et al. 2012) (Figs. 5 and 6).
Fig. 5

SINS/DVL integrated navigation structure (Bao et al. 2019)

Fig. 6

Principle diagram of geophysical navigation-assisted inertial navigation (Bao et al. 2019)

AUVs typically use the INS as their main navigation system; however, due to inherent error accumulation of inertial sensors over time, the pure INS has difficulty obtaining long-range precision navigation. Therefore, INS is often combined with other navigation systems or devices such as GPS, DVL, acoustic positioning systems, or geophysical navigation systems to improve the AUV navigation accuracy (Bao et al. 2019).



  1. Ali J, Mirza MRUB (2010) Performance comparison among some nonlinear filters for a low cost SINS/GPS integrated solution. Nonlinear Dyn 61(3):491–502CrossRefGoogle Scholar
  2. Ansari U, Bajodah AH (2017) Robust generalized dynamic inversion control of autonomous underwater vehicles. IFAC-PapersOnLine 50(1):10658–10665CrossRefGoogle Scholar
  3. Bajodah A (2009) Generalised dynamic inversion spacecraft control design methodologies. IET Control Theory Appl 3(2):239–251MathSciNetCrossRefGoogle Scholar
  4. Bajodah A (2010) Asymptotic generalised dynamic inversion attitude control. IET Control Theory Appl 4(5):827–840MathSciNetCrossRefGoogle Scholar
  5. Bao J, Li D, Qiao X, Rauschenbach T (2019) Integrated navigation for autonomous underwater vehicles in aquaculture: a review. Inf Proces Agric.  https://doi.org/10.1016/j.inpa.2019.04.003. Accepted Manuscript
  6. Clement B, Rui Y, Mansour A, Ming L (2016) A modeling and control approach for a cubic AUV. IFAC-PapersOnLine 49(23):279–284CrossRefGoogle Scholar
  7. Gui H, Jin L, Xu S (2013) Attitude maneuver control of a two-wheeled spacecraft with bounded wheel speeds. Acta Astronaut 88:98–107CrossRefGoogle Scholar
  8. Hameduddin I, Bajodah AH (2012) Nonlinear generalised dynamic inversion for aircraft manoeuvring control. Int J Control 85(4):437–450MathSciNetCrossRefGoogle Scholar
  9. Khodayari M, Balochian S (2015) Modeling and control of autonomous underwater vehicle (AUV) in heading and depth attitude via self-adaptive fuzzy PID controller. J Mar Sci Technol, 20(3):559–578CrossRefGoogle Scholar
  10. Kim M, Joe H, Kim J, Yu SC (2015) Integral sliding mode controller for precise manoeuvring of autonomous underwater vehicle in the presence of unknown environmental disturbances. Int J Control, 88(10):2055–2065MathSciNetCrossRefGoogle Scholar
  11. Kussat NH, Chadwell CD, Zimmerman R (2005) Absolute positioning of an autonomous underwater vehicle using GPS and acoustic measurements. IEEE J Ocean Eng 30(1):153–164CrossRefGoogle Scholar
  12. Lakhekar GV, Waghmare LM, Londhe PS (2015) Enhanced dynamic fuzzy sliding mode controller for autonomous underwater vehicles. In 2015 IEEE Underwater Technology (UT) (pp. 1–7). IEEEGoogle Scholar
  13. Liu X, Xu X, Liu Y, Wang L (2014) Kalman filter for cross-noise in the integration of SINS and DVL. Math Probl Eng 2014:1–8MathSciNetzbMATHGoogle Scholar
  14. Paull L, Saeedi S, Seto M, Li H (2014) AUV navigation and localization: a review. IEEE J Ocean Eng 39(1):131–149CrossRefGoogle Scholar
  15. Smallwood D, Bachmayer H, Whitcomb L (1999) A new remotely operated underwater vehicle for dynamics and control research. In: Proceedings of the 11th international symposium on unmanned untethered submersible technology, Durham, pp 370–377Google Scholar
  16. Steinicke A, Michalka G (2002) Improving transient performance of dynamic inversion missile autopilot by use of backstepping. In: AIAA guidance, navigation, and control conference and exhibit, MontereyGoogle Scholar
  17. Stutters L, Liu H, Tillman C, Brown DJ (2008) Navigation technologies for autonomous underwater vehicles. IEEE T Syst Man Cy C 38(4):581–589CrossRefGoogle Scholar
  18. Vasilijevic A, Borovic B, Vukic Z (2012) Underwater vehicle localization with complementary filter: performance analysis in the shallow water environment. J Intell Robot Syst 68(3–4):373–386CrossRefGoogle Scholar
  19. Wang Z, Liu L, Wang, Y, Wang Z. (2012) Dynamic integral sliding mode for launch vehicle attitude control system. In 2012 24th Chinese Control and Decision Conference (CCDC). Chengdu, China. IEEE. 1713–1718Google Scholar
  20. Williams S, Newman P, Majumder S, Rosenblatt J, Durrant-Whyte H (1999) Autonomous transect surveying of the great barrier reef. In Proc. Australian Conf. on Robotics and Automation, 16–20Google Scholar
  21. Williams SB, Newman P, Dissanayake G, Rosenblatt J, Durrant-Whyte H (2006) A decoupled, distributed AUV control architecture. In: International symposium on robotics, vol 31, pp 246–251Google Scholar
  22. Yang I, Byun S, Seo B, Lee D, Han DS (2013) Robust dynamic inversion based on sliding mode control for autonomous underwater vehicles. Intell Auton Veh 8:79–84Google Scholar

Copyright information

© Springer Nature Singapore Pte Ltd. 2019

Authors and Affiliations

  1. 1.School of EngineeringWestlake UniversityHangzhouChina

Section editors and affiliations

  • Changhui Song
    • 1
  1. 1.School of EngineeringWestlake UniversityHangzhouChina