Autonomous underwater vehicle for subsea equipment inspection

Abstract

A bio-inspired autonomous underwater vehicle (BAUV) for inspecting subsea equipment, the BAUV including a body extending between a front end and a rear end, a caudal fin coupled to the rear end of the body for providing thrust to the BAUV, an electric motor positioned within the body and including a mechanical motor output coupled to the caudal fin for adjusting a magnitude of the thrust provided by the caudal fin, an electromechanical servo positioned within the body and including a mechanical servo output coupled to caudal fin for adjusting a direction of the thrust provided by the caudal fin, a camera supported by the body for providing images of the external environment, and an onboard controller in signal communication with the motor, the servo, and the camera for controlling the operation of the motor and the servo based on images captured by the camera.

Description

AUTONOMOUS UNDERWATER VEHICLE FOR SUBSEA EQUIPMENT INSPECTIONCROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims benefit of U.S. provisional patent application Serial No. 63 / 626,321 filed January 29, 2024, and entitled "Bio-Inspired Autonomous Underwater Vehicle for Subsea Equipment Inspection," which is hereby incorporated herein by reference in its entirety for all purposes.STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT

[0001] This invention was made with government support under Contract No. 582-15- 57593 awarded by Texas Commission on Environmental Quality (TCEQ) through the Subsea Systems Institute (SSI). The Government has certain rights in the invention.BACKGROUND

[0002] The disclosure relates generally to the inspection of subsea equipment such as subsea pipelines and other subsea fluid conduits. More particularly, the disclosure relates to autonomous underwater vehicles including autonomous underwater vehicles (AUVs) and bio-inspired autonomous underwater vehicles (BAUVs) that may be utilized for inspecting subsea equipment for efficient and cost-effective detection of fluid leakage.

[0003] Subsea equipment (e g., subsea pipelines, platforms, manifolds, wells, and other equipment) such as subsea pipelines and other fluid conduits are typically laid on or buried below the seabed and are used for conveying materials (e.g., liquids, gas, and / or loose solids) in manifold applications. As an example, hydrocarbons such as oil and / or gas are routinely transported through subsea pipelines. Depending on the application, subsea equipment such as subsea pipelines may be used to connect subsea wellheads, manifolds, platforms, and other subsea equipment within a particular subsea hydrocarbon development field. As another example, such subsea equipment may be used to collect and convey hydrocarbons ashore for further processing by land-based equipment.SUMMARY

[0004] An embodiment of a BALIV for inspecting subsea equipment comprises a body extending between a front end and a rear end, a caudal fin coupled to the rear end of the body for providing thrust to the BAUV, an electric motor positioned within the body and comprising a mechanical motor output coupled to the caudal fin for adjusting a magnitude of the thrust provided by the caudal fin, an electromechanical servo positioned within the body and comprising a mechanical servo output coupled to caudal fin for adjusting a direction of the thrust provided by the caudal fin, a camera supported by the body for providing images of the external environment, and an onboard controller in signal communication with the motor, the servo, and the camera for controlling the operation of the motor and the servo based on images captured by the camera. In some embodiments, the BAUV comprises a control arm extending from the caudal fin and connected to both the motor output at a motor connector of the control arm and the servo output at a servo connector of the control arm that is spaced from the motor connector. In some embodiments, the servo connector comprises a pivot joint about which the control arm and the caudal fin are both pivotable, and wherein a position or an orientation of the pivot joint are adjustable by the servo. In certain embodiments, the BAUV comprises a wireless transceiver positioned within the body and configured to wirelessly transmit data between the controller and one or more external communication devices. In certain embodiments, the BAUV comprises a buoyancy control system supported by the body and in signal communication with the onboard controller for adjusting by the onboard controller a buoyancy of the BAUV. In some embodiments, the buoyancy control system comprises a ballast pump controllable by the onboard controller and a ballast tank fillable with water from the external environment. In some embodiments, the BAUV comprises a localization system supported on the body and in signal communication with the onboard controller, the localization system comprising at least one of a global positioning system (GPS) unit and an inertial measurement unit (IMU) for estimating an orientation or a position of the BAUV. In certain embodiments, the onboard controller comprises a field-programmable gate array (FPGA) for implementing a neural network onboard the AUV to facilitate detecting target objects in the external environment from the images provided by the camera.

[0005] An embodiment of an AUV for inspecting subsea equipment comprises a body extending between a front end and a rear end, a propulsion system coupled to thebody for providing thrust to the AUV, a camera supported by the body for providing images of the external environment, and an onboard controller in signal communication with the camera for controlling the operation of the propulsion system based on images captured by the camera, the onboard controller comprising a movement control engine for directing the AUV along a pathway defined by a path module of the movement control engine based on the images provided by the camera, and an equipment identification engine configured to identify subsea equipment in the external environment by applying a trained neural network to the images provided by the camera. In certain embodiments, the pathway extends along a seabed. In some embodiments, the movement control engine comprises an image projection module configured to determine a target angle formed between a lens axis of the camera and a target line extending from the camera and which intercepts a target object. In some embodiments, the onboard controller comprises a field-programmable gate array for implementing the neural network onboard the AUV. In certain embodiments, the neural network comprises a deep neural network. In certain embodiments, the equipment identification engine comprises an image restoration module for providing corrected images from the images provided by the camera in which haze is reduced in the corrected images. In some embodiments, the equipment identification engine comprises an image deblurring module for providing deblurred images from the corrected images received from the image restoration module in which motion blur is reduced in the corrected images. In some embodiments, the AUV comprises a motor supported on the body and having a mechanical motor output for driving the propulsion system. In certain embodiments, the onboard controller comprises a camera orientation module to adjust a position of the camera relative to the body of the AUV based on the operation of the motor. In certain embodiments, the AUV comprises a BAUV and the propulsion system comprises a caudal fin.

[0006] An embodiment of a method for inspecting subsea equipment using one or more autonomous AUVs comprises (a) deploying an AUV into a subsea environment proximal an offshore system comprising subsea equipment, (b) autonomously identifying by an onboard controller of the AUV, subsea equipment of the offshore system, (c) autonomously directing by the onboard controller the AUV through the subsea environment towards the identified subsea equipment, and (d) autonomously determining by the onboard controller the presence of a fluid leak in the identified subsea equipment. In some embodiments, the method comprises (e) autonomouslytransmitting by the onboard controller a message to an external communication device.

[0007] Embodiments described herein comprise a combination of features and characteristics intended to address various shortcomings associated with certain prior devices, systems, and methods. The foregoing has outlined rather broadly the features and technical characteristics of the disclosed embodiments in order that the detailed description that follows may be better understood. The various characteristics and features described above, as well as others, will be readily apparent to those skilled in the art upon reading the following detailed description, and by referring to the accompanying drawings. It should be appreciated that the conception and the specific embodiments disclosed may be readily utilized as a basis for modifying or designing other structures for carrying out the same purposes as the disclosed embodiments. It should also be realized that such equivalent constructions do not depart from the spirit and scope of the principles disclosed herein.BRIEF DESCRIPTION OF THE DRAWINGS

[0008] For a detailed description of various exemplary embodiments, reference will now be made to the accompanying drawings in which:

[0009] FIG. 1 is a schematic representation of an embodiment of a subsea equipment inspection system in accordance with principles disclosed herein;

[0010] FIG. 2 is a block diagram of an embodiment of an AUV in accordance with principles disclosed herein;

[0011] FIG. 3 is a block diagram of an AUV controller in accordance with principles disclosed herein;

[0012] FIGS. 4-6 are schematic representations of another embodiment of an AUV in accordance with principles disclosed herein;

[0013] FIGS. 7 and 8 are schematic representations of another embodiment of an AUV in accordance with principles disclosed herein;

[0014] FIG. 9 is a schematic representation of another embodiment of an AUV in accordance with principles disclosed herein;

[0015] FIGS. 10-12 are camera images from an embodiment of an AUV camera in accordance with principles disclosed herein;

[0016] FIG. 13 is a block diagram of an embodiment of a field-programmable gate array in accordance with principles disclosed herein;

[0017] FIG. 14 is a side view of another embodiment of an AUV in accordance with principles disclosed herein;

[0018] FIG. 15 is a partial cross-sectional view of the AUV of FIG. 14;

[0019] FIG. 16 is a flowchart of an embodiment of a method for inspecting subsea equipment using one or more AUVs in accordance with principles disclosed herein; and

[0020] FIG. 17 is a block diagram of an embodiment of a computer system in accordance with principles disclosed herein.DESCRIPTION OF THE DISCLOSED EMBODIMENTS

[0021] The following discussion is directed to various exemplary embodiments. However, one skilled in the art will understand that the examples disclosed herein have broad application, and that the discussion of any embodiment is meant only to be exemplary of that embodiment, and not intended to suggest that the scope of the disclosure, including the claims, is limited to that embodiment.

[0022] Certain terms are used throughout the following description and claims to refer to particular features or components. As one skilled in the art will appreciate, different persons may refer to the same feature or component by different names. This document does not intend to distinguish between components or features that differ in name but not function. The drawing figures are not necessarily to scale. Certain features and components herein may be shown exaggerated in scale or in somewhat schematic form and some details of conventional elements may not be shown in interest of clarity and conciseness.

[0023] In the following discussion and in the claims, the terms "including" and "comprising" are used in an open-ended fashion, and thus should be interpreted to mean "including, but not limited to...’’ Also, the term "couple" or "couples" is intended to mean either an indirect or direct connection. Thus, if a first device couples to a second device, that connection may be through a direct connection, or through an indirect connection via other devices, components, and connections. In addition, as used herein, the terms "axial" and "axially" generally mean along or parallel to a central axis (e.g., central axis of a body or a port), while the terms "radial" and "radially" generally mean perpendicular to the central axis. For instance, an axial distance refers to a distance measured along or parallel to the central axis, and a radial distance means a distance measured perpendicular to the central axis. Further, as used herein,the terms “approximately,” “about,” “substantially,” and the like mean within 10% (i.e. , plus or minus 10%) of the recited value. Thus, for example, a recited angle of “about 80 degrees” refers to an angle ranging from 72 degrees to 88 degrees.

[0024] As described above, subsea equipment such as subsea pipelines and the like are used to transport various materials between different subsea locations and / or ashore. For instance, in the Gulf of Mexico and around the world, subsea pipelines for conveying hydrocarbons have been deployed for decades. Particularly, thousands of active subsea wells and tens of thousands of miles of subsea pipelines have been deployed in the Gulf of Mexico. Moreover, as the search for hydrocarbons move into deeper waters, projects involving the deployment of subsea pipelines and other subsea equipment are expected to increase significantly in the future.

[0025] In addition, unlike at least some onshore equipment, subsea equipment is subjected to harsh environmental conditions including temperature variations, deep sea pressures, and the like that increase the risk of damage to the subsea equipment. With time, these environmental conditions may cause the underwater equipment to weaken, weld defects, corrosion pits, cracks, buckling, leakage, etc., and undermine safe operation. A leakage from cracked or ruptured pipes carrying harmful fluids can negatively impact the environment and potentially trigger an ecosystem disaster. For example, the Pipeline and Hazardous Materials Safety Administration (PHMSA) reported 334 fatalities and more than $10 billion in losses due directly to 12,782 pipeline accidents over the past 20 years. Timely inspection and thus, early detection of subsea pipeline failure is the key to preventing environmental disasters caused by spills or leakages. Current inspection techniques often involve trained human personnel operating a remotely controlled underwater vehicle (ROV) to take images and videos of the subsea pipeline for offline rupture detection. However, such inspection may be time consuming and expensive.

[0026] Accordingly, embodiments of subsea equipment inspection systems and associated methods employing ALIVs and BAUVs are described herein to provide energy-efficient, time-efficient and cost-effective autonomous underwater vehicles for real-time inspection and monitoring of subsea equipment. For example, embodiments of AUVs disclosed herein may autonomously acquire images of subsea equipment via an onboard camera thereof, autonomously process the images onboard the AUV, and detect autonomously and onboard the AUV subsea pipeline anomalies or leakages at early stages and in real time (e.g., within one second), allowing operators to maketimely decisions on maintenance and repairs, and thus, avoid or minimize any environmental impact. The ALIVs may comprise an onboard controller comprising an edge computing device configured to control the orientation and movement of the AUV as well as identify selected target objects (e.g., subsea equipment) onboard and autonomously such that the AUV is not reliant on instructions from external communication devices in order to identify the target object, travel towards the identified target object, and detect whether the target object is suffering from a fluid leak or other issue. For instance, the onboard controller may employ one or more neural networks to facilitate the identification and tracking of target objects as will be discussed further herein.

[0027] In some embodiments, the AUV comprises a BAUV having, for instance, a robotic fish design having a body extending between a front end and an opposed rear end, and a caudal fin or flapping tail coupled to the rear end of the body and driven by an electric motor and a servo system. The BAUV may allow the BAUV to maintain depth control and efficiently travel underwater, due to the buoyancy forces acting on the body of the BAUV.

[0028] In some embodiments, the BAUV may include a BAUV body and a caudal fin to provide relatively faster and more nimble maneuvering capabilities, more controllability under less steady conditions, and less impact on marine live compared to torpedo-shaped AUVs. Additionally, embodiments of BAUVs may also generate significantly less acoustic noise compared to propeller driven commercial AUVs while having significant advantages in operating in enclosed spaces such as confined and grassy environments. Furthermore, embodiments of BAUVs disclosed herein may achieve better stealth and higher propulsion efficiency than AUVs that use rotating propellers with distinct noise signatures.

[0029] In some embodiments, the BAUV subsea equipment inspection system comprises a body with a robotic fish design, a caudal fin coupled to a rear end of the body and driven by an electric (e.g., direct current) motor and a servo system. In addition, the BAUV may include an onboard controller which includes a microcontroller, a vision detection system (VDS) and a localization system. Furthermore, the BAUV may include an antenna for exchanging data with a remote center which may coordinate the monitoring tasks executed by multiple BAUVs. In some embodiments, the BAUV subsea equipment inspection system comprises an onboard vision-based data acquisition system which includes a computer visionsystem comprising an imaging device, an image-based visual servoing (IBVS) control, and an image correction and recognition module. As used herein, the term “servoing” refers to the mechanism of connecting one object movement to another. In some embodiments, the image recognition module comprises a neural network such as a DNN. In other embodiments, the BALIV subsea equipment inspection system includes mobile edge computing devices comprising memory devices storing DNN-based neural network infrastructure for processing vision data onsite and in real-time.

[0030] Referring to FIG. 1 , an embodiment of an exemplary subsea equipment inspection system 20 in accordance with principles disclosed herein is shown. Particularly, inspection system 20 may be used to inspect subsea equipment of an offshore system 10. In this exemplary embodiment, offshore system 10 generally includes a plurality of subsea production wells 11 which penetrate into a subsurface region 2 extending beneath a seabed 3, a plurality of subsea fluid conduits or pipelines 12, a plurality of subsea production manifolds 14, a plurality of marine risers 16, and a surface storage vessel 18 located at the waterline 4 of a subsea environment extending between waterline 4 and the seabed 3. Offshore system 10 may be used for extracting hydrocarbons from the subsurface region 2. Offshore system 10 serves merely as an example of an offshore system the subsea equipment of which may be inspected by inspection system 20, as will be discussed further herein. Thus, offshore systems and / or subsea equipment inspectable by inspection system 20 may vary significantly from that of offshore system 10 shown in FIG. 1. For instance, inspection system 20 may be used to inspect subsea equipment utilized in applications other than the capture and / or transportation of hydrocarbons.

[0031] In this exemplary embodiment, storage vessel 18 processes and stores the hydrocarbons extracted from production wells 11 and may be moored to or otherwise supported by the seabed 3. For instance, storage vessel 18 may comprise a floating production storage and offloading system (FPSO). The subsea pipelines 12 of offshore system 10 fluidically connect the plurality of productions wells 11 to the storage vessel 18 using the plurality of production manifolds 14 located along the seabed 3 and the corresponding plurality of marine risers 16 extending vertically between the production manifolds 14 and the storage vessel 18.

[0032] The inspection system 20 may be used to inspect subsea equipment of the offshore system 10 including for example, subsea pipelines 12 in order to detect the presence of fluid leaks therealong which may result in the contamination of thesurrounding aquatic environment. In this manner, inspection system 20 may detect the presence of leaks early in their formation whereby such detected leaks may be quickly remediated to minimize the amount of hydrocarbons and / or other materials of offshore system 10 leaked to the surrounding aquatic environment. As will be discussed further herein, the operation of inspection system 20 is at least partially if not fully automated to maximize the speed of inspection system 20 in detecting the presence of subsea leaks in offshore system 10 shortly following their formation. In this exemplary embodiment, inspection system 20 generally includes a surface transceiver 22 located at the waterline 4, a remote computer system 24 in signal communication with the surface transceiver 22 via a communication network 6, and one or more ALIVs 50 in signal communication with the surface transceiver 22. Generally, ALIVs 50 autonomously travel along the subsea equipment of offshore system 10 (e.g., subsea pipelines 12), detect leaks in the subsea equipment, intelligently interpret results through unsupervised machine learning and edge computing, and communicate with surface transceiver 22. Using this detection system 20, subsea equipment anomalies due to seismic activity, offshore drilling, turbulence, and ship anchoring may be detected at early stages allowing operators to make informed decisions on maintenance and repairs of the subsea equipment.

[0033] Referring to FIG. 2, an embodiment of an exemplary AUV 50 in accordance with principles disclosed herein is shown. In this exemplary embodiment, AUV 50 extends between a front end 51 and an opposing rear end 53 and generally includes an AUV body 52, an onboard power supply 54, a motor 56, a propulsion system 58, a buoyancy control system 60, an onboard AUV controller 64, an onboard localization system 66, an onboard wireless transceiver 72, and an onboard AUV camera 80 (e.g., a video camera). AUV body 52 of AUV 50 physically supports the different onboard components of AUV 50 including, for example, power supply 54, motor 56, propulsion system 58, buoyancy control system 60, AUV controller 64, localization system 66, wireless transceiver 72, and AUV camera 80. These components may be at least partially received within an interior of the AUV body 52 and / or mounted external the AUV body 52. Additionally, the AUV body 52 may take various shapes and may comprise various materials depending upon the given application.

[0034] The power supply 54 of AUV 50 provides power (e.g., electrical power) to the different components of AUV 50 including, for example, motor 56, propulsion system 58, buoyancy control system 60, AUV controller 64, localization system 66, wirelesstransceiver 72, and AUV camera 80. The physical configuration, power storage capacity, and other attributes of power supply 54 may vary depending on the given application. For instance, in some embodiments, power supply 54 comprises one or more electrical batteries such as lithium batteries and the like. However, power supply 54 may comprise power sources other than electrical batteries in other embodiments.

[0035] The motor 56 of AUV 50 receives power from power supply 54 and converts that power into mechanical motion that is transmitted from the motor 56 to the propulsion system 58 for controlling the orientation or pose and trajectory of the AUV 50. For example, motor 56 may comprise an electrical motor such as a direct current (DC) motor that converts electrical power provided by power supply 54 into mechanical torque and rotation that is supplied to the propulsion system 58. However, motor 56 may convert the power received thereby into various types of mechanical motion and / orforces for driving the operation of propulsion system 58. In turn, the propulsion system 58 receives mechanical energy from the motor 56 and converts that into physical motion of the AUV 50 within the surrounding subsea environment. For instance, propulsion system 58 may comprise one or more propellers or jets configured to apply a thrust (indicated by arrow 55 in FIG. 2) to the AUV 50 the direction of which may be adjusted (e.g., three-dimensionally) by actuation of the motor 56 and / or propulsion system 58. As will be discussed further herein, in other embodiments, propulsion system 58 may comprise a fin such as a caudal fin for applying the directable thrust 55 to the AUV 50. Thus, in some embodiments, AUV 50 may comprise a BAUV. In this manner, the pose as well as the velocity of the AUV 50 may be continuously controlled and adjusted via the operation of motor 56 and / or propulsion system 58.

[0036] The buoyancy control system 60 of AUV 50 actively controls the buoyancy of AUV 50 so as to facilitate continuous control over the magnitude and direction (e.g., vertically rising or sinking) of a buoyancy force (indicated by arrow 57 in FIG. 2) applied to the AUV 50. In this exemplary embodiment, buoyancy control system includes a ballast pump 61 powered by power supply 54 and a ballast tank 63 that is fillable with ballast in the form of water from the surrounding subsea environment. In this manner, the magnitude and / or direction of buoyancy force 57 may be adjusted by operating ballast pump 61 to add or remove ballast from ballast tank 63. However, the configuration and / or operation of buoyancy control system 60 may vary in other embodiments. In still other embodiments, AUV 50 may not include buoyancy controlsystem 60 and instead may, for instance, only include propulsion system 58 for controlling the pose and / or movement of AUV 50.

[0037] While in some embodiments the vertical depth of AUV 50 within the subsea environment may also be controlled or adjusted by propulsion system 58, buoyancy control system 60 provides another mechanism by which AUV 50 may control its vertical depth. For instance, by adjusting the magnitude and / or direction of the buoyancy force 57, buoyancy control system 60 may permit AUV 50 to vertically rise and fall without needing to move horizontally and / or adjust an orientation of the AUV 50. In this manner, buoyancy control system 60 may be used to control the vertical depth of the AUV 50 subsea passively without reliance on the propulsion system 58. Such operation may be more energy efficient than operating propulsion system 58 to similarly control vertical depth of AUV 50, thereby saving onboard power and increasing the energy efficiency of AUV 50. For example, an operational range and / or operational timespan of the AUV 50 may be maximized by utilizing the buoyancy control system 60 for controlling the vertical depth of the AUV 50 in lieu of the propulsion system 58.

[0038] The AUV controller 64 of AUV 50 is in signal communication with and controls the operation of various controllable components of the AUV 50 including, for example, motor 56, buoyancy control system 60, localization system 66, wireless transceiver 72, and AUV camera 80. AUV controller 64 comprises a computer system that may, in at least some embodiments, control the operation of the aforementioned components of AUV 50 autonomously or at least semi-autonomously whereby AUV 50 need not rely on constant direction from an outside source (e.g., remote computer system 24 shown in FIG. 1 ) in order to accomplish its various tasks of identifying specific pieces of subsea equipment and whether such identified subsea equipment is actively suffering from a leak and / or other undesirable issue. Thus, in some embodiments, AUV controller 64 comprises an edge computing system configured to minimize the amount of data that must be transmitted via wireless transceiver 72 to the external environment in order to accomplish its mission (e.g., detecting fluid leaks in selected subsea equipment), instead performing such computational tasks (e.g., identifying specific subsea equipment, detecting the presence of leaks) onboard the AUV 50.

[0039] The localization system 66 of AUV 50 provides positional and / or orientation information to the AUV controller 64 to assist the AUV controller 64 in facilitating theaccomplishment of the AUV 50's mission. In this exemplary embodiment, localization system 66 includes a global positioning system (GPS) unit 68, and an inertial measurement unit (IMU) 70. GPS unit 68 may provide current absolute positional information of the AUV 50 to the AUV controller 64 while IMU 70 may provide information regarding the current orientation and motion (e.g., velocity, rotation, and acceleration) of the AUV 50. For instance, IMU 70 may comprise one or more accelerometers for measuring linear acceleration, one or more gyroscopes for measuring angular rotation rates, one or more magnetometers for measuring the strength and direction of Earth’s magnetic field, and one or more pressure sensors for measuring the fluid pressure directly external AUV 50 in the subsea environment.

[0040] The wireless transceiver 72 facilitates two-way wireless communication between the AUV 50 (e.g., the AUV controller 64) and external communication devices such as, for example, the surface transceiver 22 shown in FIG. 1. In some embodiments, wireless transceiver 72 comprises an antenna or antenna array configured to communicate electromagnetically with external communication devise. However, the configuration of wireless transceiver 72 may vary in other embodiments. For instance, in other embodiments, wireless transceiver 72 may facilitate wireless communication via sound waves communicated through the subsea environment.

[0041] In some embodiments, wireless transceiver 72 may employ (e.g., in conjunction with AUV controller 64) various techniques for limiting the amount of data that must be communicated thereby external the AUV 50 which may otherwise slow or hinder the operation of AUV 50 in performing its various tasks. For example, wireless transceiver 72 and / or AUV controller 64 may employ a pruning technique to reduce latency and / or apply quantization-aware training to a neural network of the AUV controller 64.

[0042] The AUV camera 80 of AUV 50 is located at the front end 51 of AUV 50 in this exemplary embodiment and is configured to capture a series of images of the surrounding ambient environment to guide the operation of AUV controller 64 in controlling the orientation and / or movement of AUV 50 to accomplish the AUV 50’s mission. AUV camera 80 has a lens 81 oriented along a lens axis 83 projecting externally from the AUV 50. In this exemplary embodiment, lens 81 comprises a zoom lens permitting the AUV camera 80 to focus at different distances from the AUV 50. Additionally, AUV camera 80 is coupled to a camera mount 59 of the AUV body 52 via a positioner or actuator 82 of the AUV camera 80 for controlling the orientation of lens 81 . For instance, positioner 82 permits the lens axis 83 to pan horizontally (indicatedby arrow 85 in FIG. 2) about a vertically extending axis and / or tilt vertically (indicated by arrow 87 in FIG. 2) about a horizontally extending axis. In this manner, the orientation of lens axis 83 may be controlled independently (e.g., by AUV controller 64) of the orientation of the AUV body 52 of AUV 50.

[0043] Referring to FIGS. 2 and 33, an embodiment of an onboard AUV controller 100 is shown. In some embodiments, AUV controller 64 includes features in common or may be configured similarly as AUV controller 100. In other embodiments, the configuration of AUV controller 64 may vary from AUV controller 100. In this exemplary embodiment, AUV controller 100 includes a camera orientation module 102, a movement control engine 110, and an equipment identification engine 120. In this exemplary embodiment, AUV controller 100 also includes one or more trained neural networks 128 that may interface with the equipment identification engine 120 or comprise a feature or component thereof. In at least some embodiments, AUV controller 100 comprises an edge computing device whereby the functionalities of camera orientation module 102, movement control engine 110, equipment identification engine 120, and trained neural network 128 are each implemented in real-time onboard the AUV comprising the AUV controller 100. Further, AUV controller 100 may include features in addition to those shown in FIG. 3.

[0044] The camera orientation module 102 is configured to control the position or orientation of a AUV camera of the AUV controlled by AUV controller 100. For example, image acquisition by the AUV camera may be disturbed as the AUV camera oscillates or otherwise moves relative to the surrounding ambient environment due to motion of the AUV itself, such as motion created by a propulsion system of the AUV. Particularly, in embodiments where the propulsion system comprises an oscillating member such as a fin, oscillating motion may be induced in the AUV body as the AUV travels through the subsea environment. Generally, camera orientation module 102 is configured to cancel out such body motion of the AUV to permit the AUV camera to focus on objects of interest within the surrounding ambient environment. As an example with respect to AUV 50, a camera orientation module 102 of AUV controller 64 may receive as an input propulsion signal one or more parameters of the motor 56 and / or propulsion system 58 (e.g., driveshaft speed of motor 56, oscillating frequency of propulsion system 58, and the like) and may generate an output or control signal that is applied to the camera positioner 82 for controlling the position of camera 80 based on the input propulsion signal (e.g., the speed and / or magnitude of themechanical motion induced by the propulsion system 58) received by the AUV controller 64. In this manner, AUV controller 64 may cancel out the motion induced in AUV body 52 by the operation of motor 56 and / or propulsion system 58 with respect to the camera 80 and the orientation of its lens axis 83.

[0045] Referring briefly to FIGS. 4-6, additional details of the operating principle of camera orientation module 102 are shown. Particularly, FIGS. 4-6 each illustrate an AUV 130 including an AUV body 132 and an AUV camera 134 coupled to but moveable (e.g., via the operation of a camera positioner thereof) relative to the AUV body 132. FIG. 4 illustrates a focus or field of view (FOV) 135 of the AUV camera 134 without the operation of camera orientation module 102 while FIG. 5 illustrates the FOV 135 of AUV camera 134 with camera orientation module 102 activated to cancel out or attenuate motion from the AUV body 132. Particularly, FIGS 4 and 5 each illustrate the FOV 135 of AUV camera 134 at a first point in time (FOV 135-1 ) and at a second point in time (FOV 135-2). The first FOV 135-1 of camera 134 is substantially different from the second FOV 135-2 of camera 134 in FIG. 4 due to oscillating AUV body motion (indicated by arrow 137 in FIGS. 4-6) whereas the degree of difference between FOVs 135-1 and 135-2 is substantially eliminated in FIG 5 via the controlled camera motion (indicated by arrow 139-1 in Fig. 5) of the AUV camera 134 relative to the AUV body 132 induced by camera orientation module 102. Particularly, camera motion 139 cancels out or dampens the body motion 137 of AUV 130 to stabilize the FOV 135 of camera 134.

[0046] In addition to stabilizing the FOV 135 of camera 134 relative to AUV body 132, the controlled camera motion 139 induced in camera 134 relative to AUV body 132 by camera orientation module 102 may be used to orient the FOV 135 at a desired bias angle 131 relative to, for example, a longitudinal axis 133 of AUV 130 as shown particularly in FIG. 6. In other words, the output or control signal applied by camera orientation module 102 to the AUV camera 134 may both dampen the AUB body motion 137 while also controlling or adjusting the bias angle 131 of camera 134 to direct the FOV 135 of camera 134 as desired around the surrounding ambient environment without needing to rotate or otherwise reorient the AUV body 132 of AUV 130 itself. Thus, by adjusting bias angle 131 , the energy efficiency of AUV 130 (or any AUV implementing camera orientation module 102) may be maximized by reducing the need to reposition or reorient AUV 130 during operation thereof.

[0047] Referring again to FIGS. 2 and 3, the movement control engine 110 of AUV controller 100 is configured to autonomously control the orientation and / or movement of the AUV comprising AUV controller 100 (e.g., AUV 50 in some embodiments). Movement control engine 110 of AUV controller 100 may guide the operation of a propulsion system and / or buoyancy control system of the AUV based on an input signal in the form of images provided by the AUV camera of the AUV. For example, movement control engine 110 may guide the AUV towards or along selected subsea equipment identified by the equipment identification engine 120 of AUV controller 100 whereby the AUV camera may image the identified subsea equipment in order to detect the presence of fluid leaks therealong.

[0048] In this exemplary embodiment, movement control engine 1 10 of AUV controller 100 includes both a path module 112, an image projection module 114, and a disturbance rejection module 116. The path module 112 and the image projection module 1 14 may work in tandem in order to define an AUV pathway (e.g., via path module 112) in the surrounding ambient environment and controlling the movement of the AUV (e.g., via image projection module 114) to maintain the position of the AUV along the AUV pathway. For instance, modules 112 and 114 may work in conjunction to determine the difference between a current trajectory and a desired trajectory that follows the AUV pathway and intercepts a desired or target object external the AUV such as subsea equipment identified by the equipment identification engine 120.

[0049] Referring briefly to FIGS. 7 and 8, additional details of the operating principle of path module 112 and image projection module 114 are shown. Particularly, FIG. 7 includes a diagram 140 illustrating an AUV 141 comprising an AUV camera 142. AUV 141 is shown in diagram 140 from a sideview travelling along a trajectory 143 relative to a seabed 3. Additionally, diagram illustrates an AUV pathway 144 defined by path module 112 along which the AUV 141 is directed to travel, the AUV pathway 144 extending along the seabed 3.

[0050] Further, diagram 140 illustrates a desired or target trajectory 147 which projects from the AUV 141 and intercepts a target object 145 (subsea equipment located along the seabed 3) external the AUV 141 and which is located along the AUV pathway 144. In some embodiments, the AUV pathway 144 is linear in shape and may follow the seabed 3 providing a direction along the seabed 3 which AUV 141 is directed to travel (e.g., as controlled by an AUV controller) in order to arrive at or vertically above the target object 145. In other embodiments, AUV pathway 144 may not be positionedalong the seabed 3 and / or may not be linear in shape. In certain embodiments, a movement control engine (e.g. movement control engine 110) of AUV 141 may control a propulsion system thereof to minimize an error between a current position and / or trajectory 143 of the AUV 141 and the AUV pathway 144.

[0051] Diagram 140 also illustrates a horizontal axis 146 extending parallel the seabed 3 and / or AUV pathway 144. In some embodiments, horizontal axis 146 corresponds to a lens axis of the AUV camera 142 and thus may also be referred to as lens axis 146. In other embodiments, the lens axis 146 may not extend parallel the seabed 3 and / or AUV pathway 144. The current trajectory 143 of AUV 141 is disposed at a first or current angle 148 from the lens axis 146 while the target trajectory 147 is disposed at a second or target angle 149 from the lens axis 146. Given that the target trajectory 147 is spaced or angled from the current trajectory 143, the target angle 149 is different from (e.g., shown as greater than in FIG. 7) the current angle 148.

[0052] FIG. 8 includes a diagram 155 that includes both an exemplary image 160 captured by the camera 142 of AUV 141. Additionally, diagram 155 illustrates a two- dimensional (2D) side projection of the camera image 160. Among other features, camera image 160 illustrates a vanishing line 161 corresponding to the interface (from the perspective of camera 142) between the subsea environment and the seabed 3, and a direction line 162 corresponding to the direction of the trajectory 143 of AUV 141 from the perspective of camera 142. Vanishing line 161 and direction line 162 may be added or superimposed onto camera image 160 by an image projection module (e.g., image projection module 114). In this example, the vanishing line 161 and direction line 162 intercept at an image or camera center 163 corresponding to, for example, a lens axis of AUV camera 142.

[0053] The target object 145 is also captured in the camera image 160 (labeled as Pmiin camera image 160) and is spaced from both the vanishing line 161 and the direction line 162. Particularly, target object 145 is spaced by a distance 164 (labeled as di in camera image 160) from the direction line 162 whereby an axis extending parallel vanishing line 162 and which intercepts the target object 145 also intercepts the direction line 162 at an intercept point 165 (labeled as Lm1in camera image 160) that is spaced (along with target object 145 itself) by a length 166 from the vanishing line 161 (labeled as / 1 in camera image 160).

[0054] As, described above, diagram 170 represents a 2D side projection of the camera image 160 which is shown from its side in 2D projection diagram 170. Thus,2D projection diagram 170 illustrates camera center 163 and intercept point 165 as lying along the side projection of camera image 160. 2D projection diagram 170 illustrates lens axis 146 of AUV camera 142 as intercepting the camera center 173. Additionally, camera image 160 is spaced along the lens axis 146 from the AUV camera 142 by a focal length 174 (labeled as fin camera image 160). Further, camera 142 is spaced by a vertical height 177 (labeled as h in camera image 160) from the seabed 3 with camera 142 positioned along the seabed 3 at a seabed location 178.

[0055] 2D projection diagram 170 also illustrates a 2D projected target line 171 that extends from the camera 142, intercepts intercept point 165, and terminates at the target object 145 (labeled as Lwiin 2D projection diagram 170). Given that the focal length 174 of AUV camera 142 is known, the current angle 143 and target angle 147 shown in diagram 140 may be calculated based on the positions of the target object 145 and intercept point 165 as captured in the camera image 160. Additionally, the image projection module of AUV 141 may use the calculated target angle 149 as a control input for controlling the operation of the propulsion system of AUV 141.

[0056] Returning to FIGS. 2 and 3, the AUV comprising AUV controller 100 may be subject to disturbances from the external environment such as water currents, wave action, water turbulence, and the like. Disturbance rejection module 116 may monitor said external disturbances and dampen such external disturbances when controlling the orientation and motion of the AUV in attempting to maintain the target angle 149 so as to arrive at the target object 145. For example, the AUV may comprise one or more sensors for estimating said external disturbances to facilitate their rejection. In other embodiments, AUV may be in signal communication with one or more remote sensors located distal AUV but in contact with the subsea environment for monitoring conditions of the subsea environment. The remote sensors may communicate their measurements to the AUV via the wireless transceiver of the AUV.

[0057] While movement control engine 1 10 is configured to guide the orientation and movement of the AUV in relation to a target object (e.g., target object 145 shown in FIGS. 7 and 8), the equipment identification engine 120 of AUV controller 100 is configured to initially detect or identify the target object (e.g., a selected piece of subsea equipment such as a target subsea pipeline and the like) which may then be utilized by movement control engine 110 for controlling the motion of the AUV. Particularly, equipment identification engine 120 is configured to autonomously identify different target objects (e.g., different pieces of subsea equipment) in a subseaenvironment based on images provided by an AUV camera of the AUV. In some embodiments, equipment identification engine 120 may classify the type of target object identified thereby such as whether the target object corresponds to a subsea pipeline, a subsea production well, a production manifold, a marine riser, and the like. In certain embodiment, equipment identification engine 120 may identify specific pieces of subsea equipment such as a specific subsea pipeline of an offshore system comprising a plurality of separate subsea pipelines and / or other subsea equipment. In certain embodiments, equipment identification engine 120 determines a confident interval of the identification made thereby.

[0058] In this exemplary embodiment, equipment identification engine 120 includes an image restoration module 122, an image deblurring module 124, and an identification module 126. Generally, modules 122 and 124 are configured for enhancing the quality of images provided by the AUV camera. Particularly, image restoration module 122 is generally configured to filter intervening light and other noise or artifacts (e.g., light scattering, haze) that is included in the camera image while image deblurring module 124 is configured to reduce or eliminate camera blur in the camera images resulting from motion of the AUV and the relatively long exposure times for the AUV camera to form the camera image given the low ambient light available in the surrounding ambient environment. In this manner, image restoration module 122 may receive a raw camera image from the AUV camera as an input and produce a corrected image as an output in which haze, and / or other noise or artifacts have been corrected. Additionally, the image deblurring module 124 may receive the corrected image as an input and produce a deblurred image as an output which is then provided to the identification module 126 for identifying subsea equipment and / or other target objects in the deblurred image.

[0059] Referring to FIG. 9, additional details of the operating principle of image restoration module 122 is shown. Particularly, FIG. 9 includes a diagram 180 illustrating an AUV 182 comprising an AUV camera 184 and positioned in the subsea environment positioned above subsea equipment in the form of a subsea pipeline 186. Initially, light reflects off of or radiates from the subsea pipeline 186 as scene radiance 188. The scene radiance 188 includes information of interest for identifying the subsea pipeline 186 by the equipment identification engine 120. The scene radiance 188 travels through the transmission medium 190 (e.g., the subsea environment) from the subsea pipeline 186 until it is captured by the AUV camera 184. Additionally, as thescene radiance 188 travels along the transmission medium 190, it is subject to interference from both the subsea environment itself as well as ambient light 192 present in the subsea environment. Such interference may add substantial haze to the camera image making it substantially more difficult to identify the presence of subsea pipeline 186 therein. Thus, the scene radiance 188 is received by the camera at an observed intensity 194 that is different from the initial scene radiance 188 prior to its travel through the transmission medium 190.

[0060] In some embodiments, the image restoration module 122 estimates the magnitude of ambient light 192 and the transmission time required for the scene radiance 188 to travel from the subsea pipeline 186 (or other subsea equipment) to the AUV camera 182. In certain embodiments, image restoration module 122 estimates the ambient light 192 by producing a Laplacian pyramid from one or more camera images and then producing a multi-scale gradient map from the Laplacian pyramid. Generally, the Laplacian pyramid is a multi-scale image representation that decomposes the camera image into a set of band-pass filtered versions, capturing details at different spatial frequencies. For instance, the Laplacian pyramid may be produced from a camera image by creating a series of smaller images therefrom by iteratively applying a Gaussian filter and downsampling (producing a Gaussian pyramid, and then computing the difference between successive levels of the Gaussian pyramid, upsampling the lower-resolution image and subtracting it from the higher-resolution image whereby the resulting Laplacian pyramid comprises the resulting series of images produced from the initial camera image. Further, the multiscale gradient map may be produced from the Laplacian pyramid by calculating a gradient for each level of the Laplacian pyramid to thereby capture directional intensity changes for each layer. In some embodiments, the gradients from all levels of the Laplacian pyramid may be aggregated to produce the multi-scale gradient map encoding information across various spatial frequencies.

[0061] In certain embodiments, in addition to producing a multi-scale gradient map, a depth map may also be produced by the image restoration module 122 by modeling the linear relationship between the observed intensity 194 and depth. Further, the image restoration module 122 may estimate ambient light by averaging the intensity value of the top percent farthest pixels in the produced depth map. A corrected image may be produced by the image restoration module 122 by accounting for the estimated ambient light and / or and the light absorption in the subsea environment whenestimating the transmission time. In certain embodiments, the corrected image may also correct for color shift in the camera image resulting from the subsea environment.

[0062] Returning to FIGS. 2 and 3, an image deblurring module 124 receives the corrected image from the image restoration module 122 to produce a deblurred image therefrom. In some embodiments, image deblurring module 124 employs the trained neural network 128 in producing the deblurred image from the corrected image. The trained neural network 128 may comprise a convolutional neural network (CNN) having an encoder-decoder structure, where the encoder extracts features from the corrected image through successive convolutional and pooling layers, capturing both local and global context. The decoder of the trained neural network 128 may then reconstruct the corrected image back to its original resolution as a deblurred image using upsampling layers, with the help of, for example, skip connections that pass detailed spatial information from the encoder to the decoder of the trained neural network 128.

[0063] The deblurring process may be initiated by feeding a corrected image to the trained neural network 128 which, through its various layers, identifies and learns patterns of blurring, such as motion blur or defocus, and applies its learned filters to suppress the blur while reconstructing sharp edges and textures when generating the deblurred image. In some embodiments, trained neural network 128 is initially trained on paired datasets of blurred and sharp images, where a loss function (e.g., Mean Squared Error or perceptual loss) measures the difference between the neural network’s output and the ground truth sharp image. This allows the neural network to learn a mapping from blurred inputs to their sharp counterparts. However, the configuration of trained neural network may vary from the description provided above in other embodiments. In still other embodiments, image deblurring module 124 may not employ or comprise a trained neural network such as trained neural network 128.

[0064] As described above, the identification module 126 receives the deblurred image from the image deblurring module 124 and identifies the presence of subsea equipment (e.g., subsea pipelines 12 shown in FIG. 1 ) and / or other target objects present in the deblurred image. For example, and referring to FIGS. 10-12, different camera images 200, 205, and 210 are shown obtained by an AUV camera and processed by equipment identification image 120. Particularly, FIG. 10 illustrates a raw camera image 200 in which a subsea pipeline 201 experiencing a rupture or leak 202 is present but obscured by substantial haze and / or other artifacts. FIG. 11illustrates a corrected image 205 produced by image restoration module 122 from the raw camera image 200 in which the haze and / or other artifacts present in raw camera image 200 have largely been eliminated, making the subsea pipeline 201 and its leak 202 more visible in the corrected image 205. Finally, FIG. 12 illustrates a deblurred image 210 that has been processed by the identification module 126 to identify the presence of the subsea pipeline 201 as indicated by identification box 211. Additionally, deblurred image 210 provides a classification or label 212 of the subsea equipment identified therein, and a confidence score 214 of the identification made by the identification module 126.

[0065] In some embodiments, equipment identification engine 120 facilitates real-time processing of the raw images 200 captured by the AUV camera such that the subsea pipeline 201 , for example, may be identified within one second of capturing the raw image 200. In certain embodiments, the identification module 126 also identifies the presence of the fluid leak 202 and may associate fluid leak 202 with the subsea pipeline 201. This information may be communicated to one or more external communication devices in wireless signal communication with the AUV such as in the form of an alarm and the like.

[0066] Referring now to FIG. 13, a block diagram of an exemplary field-programmable gate array (FPGA) is shown for implementing a deep neural network (DNN) such as the trained neural network 128 is shown. As disclosed herein, a deep learning-based infrastructure implemented using FPGA 230 may be pre-trained to recognize and detect an object. The deep learning-based infrastructure may include a portable network infrastructure, a pruning technique to reduce latency, and quantization-aware training for vision data processing in a mobile computing device. For example, FPGA 230 may be used to facilitate the forward inference of a CNN. The FPGA 230 may be configured as a dedicated DNN accelerator focusing only on DNN inference, thus, making the system more energy and cost efficient compared to other processors.

[0067] In this exemplary embodiment, FPGA 230 generally includes a weight buffer 232, a processing element (PE) array 234, a resistivity and boundary module 236, a normalization module 238, a pooling layer 240, an output buffer 242, and an input buffer 244. The configuration of FPGA 230 may vary from that shown in FIG. 13 in other embodiments. Generally, input buffer 244 of FPGA 230 serves as the entry point, temporarily storing input data and feeding it into the computational pipeline of the FPGA 230 at a suitable rate to match the processing speed of the FPGA 230.Weight buffer 232 stores the weights of the neural network (e.g., DNN) implemented by FPGA 230, often fetched from external memory (e.g., memory device 220 shown in FIG. 13) and loaded into the FPGA 230 for computation to reduce the latency and bandwidth bottleneck by ensuring that weights are readily available when required by the PE array 234.

[0068] The PE array 234 of FPGA 230 comprises the core computational engine of FPGA 230, responsible for executing the matrix multiplications and other operations that dominate DNN workloads. Each processing element within PE array 234 performs arithmetic operations, such as multiply-accumulate (MAC), on chunks of data in parallel, achieving high throughput for neural network layers. Resistivity and boundary module 236 ensures stability in computations performed by PE array 234 by handling boundary conditions and / or implementing constraints specific to certain neural network operations, such as preventing values from exceeding specified limits. Normalization module 238 normalizes outputs from the resistivity and boundary module 236, ensuring that the intermediate activations are scaled appropriately to prevent issues like vanishing or exploding gradients. Pooling layer 240 performs downsampling operations like max pooling or average pooling to reduce the spatial dimensions of feature maps, preserving important features while reducing computational load. Finally, the output buffer 242 of FPGA 230 collects and temporarily stores the results of computations, ensuring that processed data is sent to the next stage of the pipeline or offloaded efficiently to external memory (e.g., memory device 220). Together, these components allow the FPGA 230 to accelerate DNN computations by leveraging parallelism, minimizing data movement, and optimizing resource utilization. This modular design ensures high efficiency, scalability, and adaptability for various neural network architectures.

[0069] As described above, the AUV 50 shown in FIG. 2 may comprise a BAUV in some embodiments. As an example, and referring to FIGS. 14 and 15, an embodiment of a BAUV 300 in the form of a robotic fish (e.g., shark) design is shown leveraging the mobility and superior maneuverability of a fish in confined environments. BAUV 300 may include features in common with AUV 50 shown in FIG. 2, and shared features are labeled similarly. Particularly, in this exemplary embodiment, BAUV 300 generally includes a BAUV body 302, power supply 54, an electric motor 310, an electromechanical servo 320, a propulsion system 340, AUV controller 64, wireless transceiver 72, and AUV camera 80. BAUV 300 extends between a front end 301 atwhich the AUV camera 80 is located and an opposing rear end 303 at which the propulsion system 340 is located. Additionally, in this exemplary embodiment, BALIV body 302 of BAUV 300 includes a pair of pectoral fins 304 proximal the front end 301 of the BAUV 300. The BAUV body 302 of may comprise a housing which may serve as storage for other components of the BAUV 300.

[0070] In this exemplary embodiment, BAUV 300 includes electric motor 310 that receives electrical power from power supply 54 in the form of a pair of electrical batteries. Electrical motor 310 includes a mechanical output 312 that connects to the propulsion system 340 while servo 320 similarly includes a mechanical output 322 that connects to the propulsion system 340, each of which is controllable by the AUV controller 64. Motor 310 may control the magnitude of propulsion (e.g., thrust) generated by propulsion system 340 while servo 320 may control the direction of the thrust generated by propulsion system 340.

[0071] In this exemplary embodiment, propulsion system 340 of BAUV 300 comprises a caudal fin 342 projecting from the rear end 303 of the BAUV 300, and a control arm 344 extending into the interior of the BAUV body 302. The mechanical output 312 of motor 310 connects to a motor connector 346 of the control arm 344 positioned at a terminal or free end thereof, while the mechanical output 322 of servo 320 connects to a servo connector 348 positioned along control arm 344 between the motor connector 346 and the caudal fin 342. In some embodiments, servo connector 348 comprises a pivot joint about which the control arm 344 may be oscillated in response to the activation of motor 310. For instance, motor 310 may control the oscillation rate and magnitude of control arm 344 to in turn control the magnitude of thrust produced by caudal fin 342. In certain embodiments, servo 320 may be activated to control a position or orientation of the pivot joint itself (e.g., rotate the pivot joint about one or more different axes) in order to adjust the direction of the thrust produced by the caudal fin 342 to permit the BAUV 300 to travel vertically and / or horizontally through the subsea environment as controlled by the AUV controller 64. In some embodiments, the AUV controller 64 comprises a linear state-feedback controller and / or an errorbased proportional integral derivative (PID) controller for controlling the steering of the BAUV 300.

[0072] Referring to FIG. 16, an exemplary method 380 for inspecting subsea equipment using one or more AUVs. Beginning at block 382, method 380 includes deploying an AUV (e.g., AUV 50 shown in FIG. 2 and / or BAUV 300 shown in FIGS.14 and 15) into a subsea environment (e.g., subsea environment 5 shown in FIG. 1 ) proximal an offshore system (e.g., offshore system 10 shown in FIG. 1 ) comprising subsea equipment (e.g., subsea equipment 11 , 12, 14, and 16 shown in FIG. 1 ).

[0073] At block 384, method 380 comprises autonomously identifying by an onboard controller (e.g., onboard controller 64 shown in FIG. 2 and / or onboard AUV controller 100 shown in FIG. 3) of the AUV subsea equipment of the offshore system. At block 386, method 380 comprises autonomously directing by the onboard controller the AUV through the subsea environment towards the identified subsea equipment. At block 388, method 380 comprises autonomously determining by the onboard controller the presence of a fluid leak (e.g., fluid leak 202 shown in FIG. 12) in the identified subsea equipment (e.g., subsea pipeline 201 identified in FIG. 12).

[0074] Referring to FIG. 17, a block diagram of an embodiment of an exemplary computer system 400 in accordance with principles disclosed herein is shown. For example, the AUV controllers 64 and 100 shown in FIGS. 2 and 3 may comprise the computer system 400 or may incorporate at least some of the features of computer system 400. The computer system 400 of FIG. 17 includes a processor 402 (which may be referred to as a central processor unit or CPU) that is in communication with memory devices including secondary storage 404, read only memory (ROM) 406, random access memory (RAM) 408, input / output (I / O) devices 410, and network connectivity devices 412. The processor 402 may be implemented as one or more CPU chips. It is understood that by programming and / or loading executable instructions onto the computer system 400, at least one of the CPU 402, the RAM 408, and the ROM 406 are changed, transforming the computer system 400 in part into a particular machine or apparatus having the novel functionality taught by the present disclosure.

[0075] Additionally, after the system 400 is turned on or booted, the CPU 402 may execute a computer program or application. For example, the CPU 402 may execute software or firmware stored in the ROM 406 or stored in the RAM 408. In some cases, on boot and / or when the application is initiated, the CPU 402 may copy the application or portions of the application from the secondary storage 404 to the RAM 408 or to memory space within the CPU 402 itself, and the CPU 402 may then execute instructions that the application is comprised of. In some cases, the CPU 402 may copy the application or portions of the application from memory accessed via the network connectivity devices 412 or via the I / O devices 410 to the RAM 408 or to memory space within the CPU 402, and the CPU 402 may then execute instructions that the applicationis comprised of. During execution, an application may load instructions into the CPU 402, for example load some of the instructions of the application into a cache of the CPU 402. In some contexts, an application that is executed may be said to configure the CPU 402 to do something, e.g., to configure the CPU 402 to perform the function or functions promoted by the subject application. When the CPU 402 is configured in this way by the application, the CPU 402 becomes a specific purpose computer or a specific purpose machine.

[0076] Secondary storage 404 may be used to store programs which are loaded into RAM 408 when such programs are selected for execution. The ROM 406 is used to store instructions and perhaps data which are read during program execution. ROM 406 is a non-volatile memory device which typically has a small memory capacity relative to the larger memory capacity of secondary storage 404. The secondary storage 404, the RAM 408, and / or the ROM 406 may be referred to in some contexts as computer readable storage media and / or non-transitory computer readable media. I / O devices 410 may include printers, video monitors, liquid crystal displays (LCDs), touch screen displays, keyboards, keypads, switches, dials, mice, track balls, voice recognizers, card readers, paper tape readers, or other well-known input devices.

[0077] The network connectivity devices 412 may take the form of modems, modem banks, Ethernet cards, universal serial bus (USB) interface cards, wireless local area network (WLAN) cards, radio transceiver cards, and / or other well-known network devices. The network connectivity devices 412 may provide wired communication links and / or wireless communication links. These network connectivity devices 412 may enable the processor 402 to communicate with the Internet or one or more intranets. With such a network connection, it is contemplated that the processor 402 might receive information from the network, or might output information to the network. Such information, which may include data or instructions to be executed using processor 402 for example, may be received from and outputted to the network, for example, in the form of a computer data baseband signal or signal embodied in a carrier wave.

[0078] The processor 402 executes instructions, codes, computer programs, scripts which it accesses from hard disk, floppy disk, optical disk, flash drive, ROM 406, RAM 408, or the network connectivity devices 412. While only one processor 402 is shown, multiple processors may be present. Thus, while instructions may be discussed as executed by a processor, the instructions may be executed simultaneously, serially, or otherwise executed by one or multiple processors. Instructions, codes, computerprograms, scripts, and / or data that may be accessed from the secondary storage 404, for example, hard drives, floppy disks, optical disks, and / or other device, the ROM 406, and / or the RAM 408 may be referred to in some contexts as non-transitory instructions and / or non-transitory information.

[0079] In an embodiment, the computer system 400 may comprise two or more computers in communication with each other that collaborate to perform a task. For example, but not by way of limitation, an application may be partitioned in such a way as to permit concurrent and / or parallel processing of the instructions of the application. Alternatively, the data processed by the application may be partitioned in such a way as to permit concurrent and / or parallel processing of different portions of a data set by the two or more computers. In an embodiment, the functionality disclosed above may be provided by executing the application and / or applications in a cloud computing environment. Cloud computing may comprise providing computing services via a network connection using dynamically scalable computing resources.

[0080] While embodiments of the disclosure have been shown and described, modifications thereof can be made by one skilled in the art without departing from the scope or teachings herein. The embodiments described herein are exemplary only and are not limiting. Many variations and modifications of the systems, apparatus, and processes described herein are possible and are within the scope of the disclosure. For example, the relative dimensions of various parts, the materials from which the various parts are made, and other parameters can be varied. Accordingly, the scope of protection is not limited to the embodiments described herein, but is only limited by the claims that follow, the scope of which shall include all equivalents of the subject matter of the claims. Unless expressly stated otherwise, the steps in a method claim may be performed in any order. The recitation of identifiers such as (a), (b), (c) or (1 ), (2), (3) before steps in a method claim are not intended to and do not specify a particular order to the steps, but rather are used to simplify subsequent reference to such steps.

Claims

CLAIMSWhat is claimed is:

1. A bio-inspired autonomous underwater vehicle (BAUV) for inspecting subsea equipment, the BAUV comprising: a body extending between a front end and a rear end; a caudal fin coupled to the rear end of the body for providing thrust to the BAUV; an electric motor positioned within the body and comprising a mechanical motor output coupled to the caudal fin for adjusting a magnitude of the thrust provided by the caudal fin; an electromechanical servo positioned within the body and comprising a mechanical servo output coupled to caudal fin for adjusting a direction of the thrust provided by the caudal fin; a camera supported by the body for providing images of the external environment; and an onboard controller in signal communication with the motor, the servo, and the camera for controlling the operation of the motor and the servo based on images captured by the camera.

2. The BAUV of claim 1 , further comprising a control arm extending from the caudal fin and connected to both the motor output at a motor connector of the control arm and the servo output at a servo connector of the control arm that is spaced from the motor connector.

3. The BAUV of claim 2, wherein the servo connector comprises a pivot joint about which the control arm and the caudal fin are both pivotable, and wherein a position or an orientation of the pivot joint are adjustable by the servo.

4. The BAUV of claim 1 , further comprising a wireless transceiver positioned within the body and configured to wirelessly transmit data between the controller and one or more external communication devices.

5. The BAUV of claim 1 , further comprising a buoyancy control system supported by the body and in signal communication with the onboard controller for adjusting bythe onboard controller a buoyancy of the BAUV.

6. The BALIV of claim 5, wherein the buoyancy control system comprises a ballast pump controllable by the onboard controller and a ballast tank tillable with water from the external environment.

7. The BALIV of claim 1 , further comprising a localization system supported on the body and in signal communication with the onboard controller, the localization system comprising at least one of a global positioning system (GPS) unit and an inertial measurement unit (IMU) for estimating an orientation or a position of the BALIV.

8. The BALIV of claim 1 , wherein the onboard controller comprises a field- programmable gate array (FPGA) for implementing a neural network onboard the AUV to facilitate detecting target objects in the external environment from the images provided by the camera.

9. An autonomous underwater vehicle (AUV) for inspecting subsea equipment, the AUV comprising: a body extending between a front end and a rear end; a propulsion system coupled to the body for providing thrust to the AUV; a camera supported by the body for providing images of the external environment; and an onboard controller in signal communication with the camera for controlling the operation of the propulsion system based on images captured by the camera, the onboard controller comprising: a movement control engine for directing the AUV along a pathway defined by a path module of the movement control engine based on the images provided by the camera; and an equipment identification engine configured to identify subsea equipment in the external environment by applying a trained neural network to the images provided by the camera.

10. The AUV of claim 9, wherein the pathway extends along a seabed.11 . The AUV of claim 9, wherein the movement control engine comprises an image projection module configured to determine a target angle formed between a lens axis of the camera and a target line extending from the camera and which intercepts a target object.

12. The AUV of claim 9, wherein the onboard controller comprises a field- programmable gate array (FPGA) for implementing the neural network onboard the AUV.

13. The AUV of claim 9, wherein the neural network comprises a deep neural network (DNN).

14. The AUV of claim 9, wherein the equipment identification engine comprises an image restoration module for providing corrected images from the images provided by the camera in which haze is reduced in the corrected images.

15. The AUV of claim 14, wherein the equipment identification engine comprises an image deblurring module for providing deblurred images from the corrected images received from the image restoration module in which motion blur is reduced in the corrected images.

16. The AUV of claim 9, further comprising a motor supported on the body and having a mechanical motor output for driving the propulsion system.

17. The AUV of claim 16, wherein the onboard controller comprises a camera orientation module to adjust a position of the camera relative to the body of the AUV based on the operation of the motor.

18. The AUV of claim 9, wherein the AUV comprises a bio-inspired autonomous underwater vehicle (BAUV) and the propulsion system comprises a caudal fin.

19. A method for inspecting subsea equipment using one or more autonomous underwater vehicle (AUVs), the method comprising:(a) deploying an AUV into a subsea environment proximal an offshoresystem comprising subsea equipment;(b) autonomously identifying by an onboard controller of the AUV, subsea equipment of the offshore system;(c) autonomously directing by the onboard controller the AUV through the subsea environment towards the identified subsea equipment; and(d) autonomously determining by the onboard controller the presence of a fluid leak in the identified subsea equipment.

20. The method of claim 19, further comprising:(e) autonomously transmitting by the onboard controller a message to an external communication device indicating the presence of the fluid leak in the subsea equipment.

Images