Technical Design and Data Analysis of Autonomous Underwater Vehicle-Based Side-Scan Sonar Operations

Abstract

In response to the absence of standardized work practices, work safety measures, efficient work procedures, and suitable line planning methods for exploring seabed topography using autonomous underwater vehicles (AUVs) equipped with side-scan sonar systems, this paper proposes a design framework for side-scan sonar measurement technology based on AUVs. Detailed research has been conducted on line planning methods that differ significantly from those used in shipborne operations. The new line planning method was developed based on the performance characteristics of AUV equipment, the principles of motion modes, and shipborne side-scan sonar operations. The validity of the line planning method was verified through actual sea experiments using different types of AUVs in various terrains. The experimental results demonstrate that the new survey line planning method is distinct from existing AUV path planning methods and is highly applicable for conducting seabed topography exploration operations. Furthermore, the technical design scheme studied in this article is applicable in different sea areas and ensures the safety, operational efficiency, and data quality of AUVs. It holds significant importance for marine science engineering and practical applications. The article also features an analysis of successful data collected from AUVs equipped with side-scan sonar for conducting seabed topography exploration operations.

1. Introduction

The assessment of seabed topography plays a pivotal role in supporting safe navigation and marine resource development [1]. In this regard, side-scan sonar (SSS) is recognized as a crucial tool for examining underwater terrains and, hence, has found extensive application in tasks such as cartography, surveying, and search operations. Notably, its underwater exploration capabilities have a significant role in detecting diverse seafloor objects.

As a notable advancement in underwater vehicle technology, autonomous underwater vehicles (AUVs) have evolved into versatile underwater platforms with extensive applications in scientific research and military operations [2]. AUVs possess the capability to navigate significant depths underwater, providing flexible yet precise attitude control. Equipped with side-scan sonar systems, AUVs effectively minimize the distance between these systems and the seabed, consequently enhancing the resolution and clarity of geomorphological surveys. Furthermore, AUVs are distinguished by their extended operational ranges and high levels of autonomy, making them exceptionally well-suited for missions in remote and vulnerable marine environments. This underscores the substantial potential for AUVs equipped with side-scan sonar technology, transcending various applications, including underwater topographic mapping, seafloor object retrieval, deep-sea resource exploration, underwater patrolling and surveillance, and beyond [3,4,5].

The use of an AUV equipped with side-scan sonar allows for maximum utilization of the high-resolution capabilities of side-scan sonar. The AUV’s large underwater depth enables the side-scan sonar to effectively reach a height of approximately 10 m from the seabed, ensuring high data resolution. Additionally, the AUV’s flexible posture enhances operational efficiency and reduces time, while maintaining stable positioning and ensuring consistent resolution and clarity. The AUV’s long range capabilities enable remote deployment and covert measurements, making it suitable for detecting landforms in sensitive sea areas. Furthermore, the AUV’s strong autonomy minimizes the need for human resources. The use of an AUV equipped with side-scan sonar to approach the seabed also reduces the impact of reverberation, noise interference, and uneven underwater conditions, resulting in improved quality of the original acoustic image. Moreover, the side-scan sonar is directly embedded into the AUV device, eliminating the need for positioning cables and simplifying the calculation of positioning information.

Various scholars from different countries have conducted experiments and research on the application of AUVs. For example, in 2012, the Japan Marine Technology and Engineering Center investigated ocean carbon dioxide using the “Jinbei” AUV equipped with side-scan sonar and a multi-beam system, obtaining corresponding experimental data [6]. Scholars such as Pan Shiying and Zhang Hanbin have studied AUV path and trajectory planning under different conditions [7,8,9,10,11]. Chinese scholars have researched the multi-beam DVL time-sharing strategy model and proposed a wide-narrow alternating operation method [12]. In 2006, the National Institute of Subsea Science and Technology of the United States conducted mission system testing on the “Eagle ray” AUV carrying multiple beams and performed multi-beam terrain measurement experiments [13]. The Helmholtz Research Center in Germany employed the AUV Abyss equipped with multiple beams to conduct lawn power pattern scanning in the sea area and processed the multi-beam data using MB system software and Python programs [14]. The focus of the aforementioned research experiments primarily revolved around debugging the carrier task control system [15], improving AUV positioning [16], planning the path of pure AUV equipment itself [17,18,19,20,21], and integrating AUVs with various loads and conducting experiments in different sea areas [13,14]. Meanwhile, research by domestic scholars has been relatively limited, with more emphasis on the current situation abroad and on the control, navigation, and path planning of AUV devices themselves.

In relation to the relevant research on the application of autonomous underwater vehicles (AUVs) equipped with side-scan sonar, the book “Technology and Application of Autonomous Underwater Vehicles” discusses the utilization of AUVs with side-scan sonar for seabed resource detection. It also addresses the feasibility and practical validation of this objective [2]. Additionally, Benoit Zen’s research examines the use of AUVs with side-scan sonar to obtain images for precise positioning correction [22], while L. J. Hamilton and J. Cleary focus on autonomous data processing by unmanned AUVs equipped with side-scan sonar [23]. Our institute, led by Asada Ming et al., conducts research on enhancing and classifying AUV side-scan sonar image processing [24]. John Fawcett explores the effectiveness and applicability of combining AUVs with side-scan sonar for target detection in complex environments [25]. Tang Yulin et al. also conduct research on AUVs equipped with side-scan sonar for target detection [26], with a primary emphasis on real-time processing of side-scan sonar data and target detection. While research has explored AUV applications and the use of various sensors, including side-scan sonar, there remains a lack of systematic focus on the operational methods and specifications for AUVs equipped with side-scan sonar. Consequently, in-depth exploration is necessary to propose a comprehensive production process for AUVs equipped with side-scan sonar. The most critical aspects of this exploration are the technical design and operational implementation processes.

The above content underscores that incorporating side-scan sonar into AUVs is a supplementary strategy. In practical applications, utilizing side-scan sonar-equipped AUVs for seabed topography exploration requires technical specifications, operational guidance, and requirements and limitations under various operating conditions. Presently, there is a dearth of research in this area, necessitating a thorough investigation and analysis of the technical design scheme for AUVs conducting side-scan sonar operations.

The principal contributions and innovations of this article encompass the following:

• Design of an operational process for AUVs equipped with side-scan sonar for seabed topography detection, involving technical design, operational preparation, and execution. This work has enabled the preliminary development of the capability to utilize unmanned autonomous underwater platforms for seabed topography detection. It conducts a systematic analysis of AUV performance characteristics and their advantages for side-scan sonar operations. By integrating AUV equipment performance characteristics with side-scan sonar operational principles, a technical design scheme for AUVs equipped with side-scan sonar measurement operations is proposed. This design differs from schemes designed for shipborne operations, offering efficient scenarios for applying AUVs in combination with side-scan sonar for seabed topography detection and target scanning. The standardized technical guidance facilitates the enhancement of AUV efficiency while ensuring their safety and load safety, leveraging the advantages of AUVs and payloads, and ensuring data quality.
• To validate the suitability of the survey line planning scheme, a series of targeted validation tests were designed, and the actual sea trial process of AUVs equipped with side-scan sonar for measurement operations is described in detail. The feasibility, advantages, and disadvantages of the proposed technical design scheme are discussed and analyzed, and issues encountered during actual side-scan sonar operations using AUVs are summarized, along with corresponding solutions. Additionally, a detailed analysis is conducted regarding the characteristics of AUV-acquired side-scan sonar image data and its impact on data processing.

In a novel approach, this paper introduces a practical technical framework that offers effective guidance for operating AUV models equipped with side-scan technology. It provides a detailed plan from pre-operation preparations to execution, as well as post-mission inspection and maintenance. Moreover, this paper offers a comprehensive framework meticulously outlining the operational parameters, prerequisites, and established standards, culminating in a detailed analysis of the output imagery. The article structure is shown in Figure 1.

2. Materials and Methods

In the technical design scheme, the crucial aspect pertaining to autonomous underwater vehicle (AUVs) is the line planning method design. The proposition of this method relies on examining the distinctions between AUV and shipborne operations in side-scan sonar endeavors. It is imperative to analyze these disparities by considering the performance of AUV and adhering to the principles of side-scan sonar operations.

2.1. Features of AUV Performance and AUV-Based Side-Scan Sonar Operations

The operation of an autonomous underwater vehicle (AUV) equipped with side-scan sonar is influenced by several factors, with AUV performance playing a crucial role in determining the operational mode. Specifically, the performance of the AUV primarily manifests in its movement mode.

2.1.1. AUV Performance Features

Conventional AUVs typically encompass a range of various key functions, such as depth measurement, altitude measurement, obstacle detection, self-propulsion, steering, buoyancy control, communication, and endurance. Collectively, these functionalities define their capacity to conduct side-scan sonar operations. Excluding communication and endurance, the autonomous mobility toolkit, comprised of these competencies, distinguishes AUV-based side-scan sonar operations from their shipborne counterparts. This unique suite of capabilities empowers AUVs to explore deeper depths at precise altitudes, necessitating customized operational planning for deploying side-scan sonar payloads, a process far more challenging than conventional vessel-based operations.

In the hardware systems of an AUV, navigation, control, and safety systems are pivotal for underwater motion.

To address the challenge of precise positioning of AUVs both above and below water, a sophisticated design combination has been implemented for the navigation system. The navigation system holds significant importance in formulating technical design solutions, and its fundamental working principle is depicted in Figure 2.

When on the water surface, the primary hardware used for positioning and navigation in the navigation system is a satellite positioning system. However, when the AUV is moving underwater, an inertial navigation system, composed of accelerometers, gyroscopes, and heading sensors, is primarily employed in conjunction with a Doppler Velocity Log (DVL) for precise positioning and navigation.

The AUV’s operational mechanism comprises the main thruster, horizontal rudder, and vertical rudder. The main thruster rotates the propeller via an electric motor, generating the necessary thrust for the AUV to move forward or backward. During AUV navigation, the horizontal and vertical rudders deflect at specific angles, allowing the rudder plate to generate lift and thereby facilitating changes in the AUV’s pitch and bow states. The AUV can be controlled with four degrees of freedom: forward motion, heading adjustments, pitch changes, and vertical movement (heave).

The control system is a critical hardware system that governs the AUV’s vertical motion mode, and it primarily receives action responses from the depth and altitude meter. Once the depth and altitude meter measurements are obtained, response parameters are relayed to the control system, which in turn alters the AUV’s current motion state.

Among the most important safety systems are collision avoidance units and load shedding devices. The collision avoidance unit is employed to autonomously control the AUV, ensuring it avoids any unknown obstacles during navigation, thus safeguarding the AUV’s operation in complex marine environments. It encompasses three key technologies: forward-looking sonar information processing, environmental information modeling, and real-time collision avoidance strategies. In terms of collision avoidance strategies, the AUV independently evaluates its maneuverability to determine if changes to its original operational state are necessary to optimize obstacle avoidance behavior. This project utilizes the artificial potential field method as the collision avoidance strategy, known for its fast calculation speed and excellent real-time performance. Additionally, to overcome the issue of local optimal solutions that this method may encounter, temporary false targets are employed to compensate.

The load shedding device ensures that, in case of a severe malfunction or unspecified interference event, the long-range AUV triggers an emergency response by discarding a portion of its heavy load. This action enables the AUV to surface, serving as the final safety barrier for its operations. To maintain low power consumption while ensuring system safety and reliability, a mechanical load shedding structure is employed in this design.

Impact of AUV Dimensions on Performance

a.

Performance Features of Small-Size AUVs

When creating the structural design of smaller AUVs, which require efficient use of space to balance endurance and dynamic loading needs, buoyancy control is typically achieved through the coordinated action of rudders and propellers. As the AUV descends, it employs its propellers to generate forward momentum and adjusts its rudders to lift the stern. The descent primarily relies on the thrust from the propellers to submerge beneath the sea surface. Conversely, acceleration is not required during the ascent, but the AUV deploys its rudders to lower the stern before utilizing its buoyancy and propeller thrust to rise to the surface. This submersion method preserves the AUV’s natural buoyancy, necessitating continuous adjustment of its angular attitude to maintain thrust equilibrium with buoyant forces. Consequently, in this configuration, all payload beams are not perpendicular to the horizontal plane, potentially resulting in inaccuracies in bottom tracking range measurements. In the context of constant-altitude navigation, the AUV will continuously adjust its pitch to regulate its altitude above the seafloor. Nevertheless, these adjustments may introduce variations in the overall attitude, potentially affecting the vehicle’s ability to consistently uphold a specific depth above changing seafloor topographies in the constant-altitude mode.

Compact AUVs often prioritize the use of DVLs among their payloads, as the data generated by DVLs significantly influence the AUV’s intricate maneuvers. Due to the limited space available on compact AUVs, smaller DVL models are favored, covering an operational range of approximately 100 m.

b.

Performance Features of Large-Size AUVs

In the case of larger AUVs, sufficient interior space is available to easily accommodate the installation of additional payloads. These more spacious AUVs are equipped with rudders and propellers and feature a water-permeable head and oil bladders, enabling precise control over the buoyancy of the front section by adjusting their volume. The middle section typically comprises a waterproof compartment housing a sliding rail mechanism. Skillful manipulation of the slider’s position within this mechanism allows for regulating the vehicle’s center of gravity, consequently controlling the overall vehicle’s attitude. Simultaneously, the rudders and propellers function like AUVs that lack buoyancy control. This approach facilitates a remarkable steadiness in the AUV’s overall attitude, eliminating pitch oscillations. The constant-altitude mode is suitable for sustaining the submersible at a specific depth above the seafloor. In cases of emergency, a safeguard mechanism can be activated to bring the submersible to the surface through a combination of jettisoning and bladder inflation.

Large-size AUVs can accommodate an advanced iteration of the DVL, commonly referred to as an acoustic Doppler current profiler (ADCP). The ADCP exhibits similar functionality and performance to the DVL but distinguishes itself with a significantly extended operational range, typically capable of reaching depths of approximately 300 m. These instruments’ operational range depends on the AUV’s physical dimensions. For instance, shipborne ADCPs benefit from more installation space, enabling them to achieve significantly broader operational ranges, with some reaching depths of up to approximately 1000 m.

The size of an AUV directly affects the quantity and variety of payloads it can accommodate, consequently having an indirect impact on the AUV’s performance, particularly its underwater motion capabilities in the vertical plane.

AUV Underwater Motion Control Modes in the Horizontal Plane

a.

Fixed-Point Mode

It requires the integration of bottom tracking and satellite positioning systems. Before submersion, the AUV synchronizes with satellite signals to determine its initial coordinates. Once submerged, it switches to the Doppler Velocity Log (DVL) for calculating velocity and direction relative to the seafloor. This dual approach ensures that as long as the AUV remains within the DVK’s operational range, it can maintain precise positioning, even when satellite signals are unavailable [27]. Nevertheless, one limitation of this mode is its reliance on the AUV to remain within the operational constraints of the DVL. The fixed-point mode includes a time-bound safeguard mechanism, necessitating the AUV to reach its predetermined destination within a specified timeframe. Failure to do so will promptly activate a surfacing alarm.

When conducting surveys along a predetermined trajectory, the desired outcome is to align the AUV’s path with the planned survey line. However, the underwater environment may impact the AUV’s navigation, leading to deviations from the planned line, as depicted in Figure 3.

In this diagram, A and B denote the starting and ending points of a specific measuring line in fixed point navigation mode. C and C₁ represent positions corresponding to two different yaw situations. Distances from the point to the planned measuring line are indicated by CD and C₁D. The heading angle is denoted as t, and its calculation formula is as follows:

$$t=arc\tan \frac{y_B-y_C}{x_B-x_C}$$

(1)

Here, ($x_B$,$y_B$) and ($x_C$,$y_C$) refer to the coordinates of point B and point C, respectively, under the Mercator projection.

The calculation formula for CD and C₁D is as follows:

$$d=-\frac{(y_A-y_B)x_C+(x_B-x_A)y_C+(x_A{y}_B-x_B{y}_A)}{\sqrt{{(y_A-y_B)}^2+{(x_B-x_A)}^2}}$$

(2)

Here, ($x_A$,$y_A$) denotes the coordinate of point A under the Mercator projection. d represents the distance value, with CD being a positive value and C₁D being a negative value.

The AUV line correction algorithm is a dynamic correction algorithm, and its formula is as follows:

$$t\prime =t+d$$

(3)

Here, $t\prime$ represents the heading of the AUV during dynamic correction. It is worth noting that a limit is set for d, preventing it from exceeding $\pm 5$°. This limitation is in place to avoid large-angle turns by the AUV.

b.

Fixed-Direction Mode

In this specific mode, the AUV’s movement is dictated by preset time and direction parameters. This approach depends on the initial satellite-based positioning, usually determined before submersion. Once the AUV is submerged, there is no requirement for the DVL to continuously track the seafloor. This allows AUV’s flexibility to operate at depths exceeding the measuring range of the DVL. However, its pronounced dependence on inertial navigation data has been found to result in inaccurate positioning [28]. When planning continuous survey lines, the timing for changing direction must be meticulously calculated. This mode is particularly vulnerable to external factors, such as water currents, which can introduce deviations in the submersible’s position. Figure 4 illustrates the fixed-point and fixed-direction motion control modes for visual reference.

AUV Underwater Motion Control Modes in the Vertical Plane

• Constant-Altitude Mode

It is designed to maintain the AUV at a predetermined depth above the seafloor, but the submersible’s movements are influenced by seafloor topography. As underwater depths vary, the separation between the vehicle and the seafloor is determined either by the DVL or single-beam systems. This mode incorporates an altitude maintenance alarm mechanism, whereby the AUV will resurface if it deviates from a specified altitude for an extended duration. The principle revolves around the height measurement equipment which measures the distance to the seabed at a fixed working frequency, typically 2 Hz. If the measured height continuously exceeds the predetermined set height for a certain period of time, usually between 60 and 180 s (1 to 3 min), an alarm mechanism is triggered. This duration signifies continuous height measurements. If the equipment’s performance prevents it from reaching the set height for an extended period, regardless of whether the desired height has been previously reached or maintained, it indicates that the intended height operation cannot be completed as planned, and an alarm will be generated.

The principal formula for height control is as follows:

$$pitc{h}_{out}=-pitch\times p+pitc{h}_{angle\_speed}\times d$$

(4)

$$altitud{e}_{out}=\frac{(altitud{e}_{t\arg et}-altitude)\times 180}{p\times i}$$

(5)

$$h_{out}=pitc{h}_{out}+altitud{e}_{out}$$

(6)

Among them, $h_{out}$ is the correction angle of the horizontal rudder, $pitch$ is the current equipment pitch angle, $altitude$ is the current equipment altitude, $pitc{h}_{angle\_speed}$ is the maximum pitch angular velocity of the AUV, $altitud{e}_{t\arg et}$ is the height set for constant-altitude, $p$, $i$, and $d$ are the three control parameters of the equipment. Their values are determined by the response parameters of the AUV when there is a target depth difference or depth change in depth measurement, $pitc{h}_{out}$ and $altitud{e}_{out}$ respectively represent the influence of $pitch$ and $altitude$ on $h_{out}$ when $altitude$ is matched to $altitud{e}_{t\arg et}$, which are intermediate variables. The ultimate purpose of control is to adjust $pitch$ from the current angle value (including 0) to 0; adjust $altitude$ to the same value as $altitud{e}_{t\arg et}$, immediately, stop adjusting, depth control also allows for setting a reasonable error value, and in order to prevent the loss of control, it is necessary to limit $h_{out}$ to a reasonable range.

b.

Constant-Depth Mode

This technique is designed to maintain the AUV at a specific depth beneath the sea surface, relying on dedicated depth sensors for depth measurement. Its performance is significantly affected by tidal fluctuations, with limited susceptibility to the influence of ocean waves. In the context of vertical motion, an emergency protection mechanism is in place. It is engineered to trigger the submersible’s jettison system, causing it to ascend to the surface when the vehicle–seafloor distance falls below a preset alarm threshold or when an obstacle is detected within a distance less than the prescribed alarm range. Its principle is as follows: The height measurement equipment measures the depth of the seabed at a fixed working frequency, typically set at 10 Hz for emergency mechanism devices. An alarm is triggered when the measured depth remains below the predetermined safety threshold for a specific duration, usually 1 s (equivalent to continuous measurements for 10 times). The safety threshold is determined based on actual conditions. The principle of depth control aligns with that of height control. For visual reference, please refer to Figure 5, which illustrates the constant-altitude and constant-depth motion control modes.

2.1.2. Principles of Side-Scan Sonar Operations

Following the standards outlined in International Hydrographic Organization Standards for Hydrographic Surveys (S-44, Edition 6.0.0) [29] and in consideration of the operational requirements, the fundamental principles governing side-scan sonar operations include, but are not limited to:

(a)

“For Order 1a, a 100% feature search may be achieved with a survey system that does not measure depth. Under those circumstances, the least depth measurements from an independent bathymetric system will be required for any detected significant feature. It is recommended to conduct a 100% feature search in conjunction with 100% bathymetric coverage whenever possible. A feature search greater than or equal to 100% must be planned and conducted to detect all features of the sizes specified in this standard. Where more than 100% feature search is required, including 200% for Exclusive Order, it may be accomplished by adequately overlapping collection or by acquiring more than one independent dataset within a survey.” The recommended degree of overlapping for feature search is 100% and 200%, respectively.

(b)

Based on different detection levels, the edge length of detected cubic objects shall be “>10% of depth”, “>2 m”, “>1 m”, and “>0.5 m”, respectively.

(c)

The survey lines shall be parallel and, in optional scenarios, to the water-flow direction, the long edge of the survey area, the isobath, or shipping channels.

(d)

The depth above the seafloor shall be 10% to 15% of the measuring range.

(e)

The operating speed shall be less than 6 knots (kt).

(f)

The length of survey lines shall reasonably exceed the survey area’s boundaries.

2.1.3. Differences between Shipborne and AUV-Mounted Side-Scan Sonar Operations

Shipborne side-scan sonar systems are typically classified into two installation modes: hull-mounted and towed. The hull-mounted configuration offers enhanced positional accuracy by aligning the sonar system’s attitude with the vessel’s. However, its practicality is limited to shallow waters, with the attainable depth range depending on the device’s measuring capabilities. Conversely, the ship-towed sonar system can operate in deeper waters, with the attainable operational depth dependent on the length of the tow cable. Nonetheless, this configuration has limitations, including limited cable length, uncertainties arising from imprecise depth control, and reduced positional accuracy caused by the simplified reduction model required to locate the system. Furthermore, underwater attitude hinges on a separately built-in attitude sensor, which is not a standard feature in side-scan sonar equipment. The sonar system’s altitude or depth is influenced by velocity, a parameter that does not always align with the vessel’s speed, consequently leading to inertial delays.

When deploying an AUV for side-scan sonar payload operations, it is imperative to consider both the fundamental principles of these operations and the vehicle’s performance characteristics. The submersible has the potential to outperform shipborne side-scan sonar while remaining aligned with the principles of side-scan sonar operations [30]. Unlike shipborne side-scan sonar, its AUV-based counterpart is incorporated with its carrier, ensuring remarkable consistency in speed and attitude. As the AUV system combines satellite navigation, inertial navigation, and bottom-tracking navigation, reinforced by high-precision attitude sensors, it exhibits precise satellite positioning capabilities on the sea surface, along with robust attitude measurements underwater, complemented by altitude and depth measurement functions. Beyond that, the vehicle’s ability to maintain a consistent attitude and vertical position enables it to perform vertical motion, a capability absent in shipborne operating modes. Nevertheless, given its inability to receive satellite signals underwater, the AUV primarily navigates through underwater regions via a combination of inertial and bottom-tracking navigation. Underwater acoustic communication-based positioning can also be an option, though it is infrequently employed due to precision limitations.

In essence, the AUV-mounted side-scan sonar differs from its conventional shipborne equivalent. These two systems exhibit distinct operational characteristics, differing significantly in many aspects such as localization, navigation, operational depth, sea area coverage, underwater attitude, survey line deployment, and data characteristics. These disparities underscore the divergence in their operational requirements and mission planning considerations.

2.2. Technical Design for AUV-Mounted Side-Scan Sonar Operations

The technical design of AUV-based side-scan sonar operations can be summarized by the workflow presented in Figure 6. Initially, existing data are analyzed, and based on the analysis, specific tasks are prepared. The key aspects of preparation include line planning and equipment inspection to ensure that the AUV operates smoothly according to the plan. Subsequently, the operation is executed, with troubleshooting measures taken in case of any interruptions. If the operation progresses without any issues, it continues as planned. Once the operation is completed, the equipment is organized and inspected, and the collected data are exported.

2.2.1. Survey Line Deployment

According to the principles governing side-scan sonar operations, the layout of survey lines should be parallel and optionally aligned with the direction of water flow, the long edge of the survey area, the isobath, or shipping channels. Therefore, before the strategic deployment of survey lines, a preliminary scan of the survey sea area from the surface is essential to map the seafloor features. The planning of survey lines follows this assessment. Typically, the preliminary extensive scanning is performed in parallel with the long edge of the survey area, maintaining the same line spacing as in official operations. However, it is essential to note that data collected during the preliminary scan should only serve as a reference dataset due to potentially varying surface water velocities and greater seafloor depths. Integrating other empirical data, such as nautical charts, is required for a comprehensive seafloor assessment in areas where water depths exceed the instrument’s measuring range.

The deployment of survey lines can be categorized into two scenarios based on seabed topographic characteristics, as Figure 7.

(1)

In cases where the seafloor is relatively flat or displays gradual depth variations with a slope gradient of less than 14°, maintaining image clarity and resolution while optimizing equipment performance necessitates the adherence to the following principles during survey line deployment:

• The constant-altitude mode shall be employed as the preferred vertical motion mode, setting the altitude within the range from 15 to 30 m. In sea areas with depths below 15 m, consideration shall be given to increasing the submersion depth while prioritizing safety to maintain a submersion depth of at least 5 m.
• Given the minor fluctuations in water flow velocity and depth after submersion, and taking into account the AUV’s customary operation in non-harbor shipping channels, it is recommended that survey lines be oriented parallel to the longer side of the survey area.

(2)

In scenarios featuring complex seabed topography and significant depth variations, significantly when the slope gradient exceeds 14°, the following principles guide the layout of survey lines:

• In such water environments, the constant-altitude mode would trigger multiple protection mechanisms, resulting in operational interruptions. Given that, the mode of vertical motion shall be depth-keeping. At the same time, submersion depth changes shall correspond to the varying water depths, ensuring that the distance from the seafloor consistently falls between 15 and 30 m.
• The survey lines shall be parallel with the local directions of isobaths.

The survey line deployment strategies mentioned earlier are tailored to the unique characteristics of the survey area. When determining the survey line spacing, it is crucial to arrange the lines in a way that minimizes the number of lines needed to adequately cover the survey area. The range should be set based on the selected line spacing and the desired level of overlap, aiming for a 100% or more overlap on one side of the range. However, considering the AUV’s design for submersion, steering, and positioning, each survey line must extend beyond the confines of the survey area. Extending the initial survey line facilitates the submersible to descend to greater depths, enabling it to cover a distance that allows the vehicle to reach the designated depth at its maximum speed. Moreover, extending the remaining survey lines eliminates any invalid side-scan sonar data generated during pivot steering. For fixed-point motion control, a guiding line should be added to the initial survey line to ensure the submersion depth remains within the measuring range of the DVL. The top view of the survey line deployment is illustrated in Figure 8.

Because of the design structure and performance constraints of autonomous underwater vehicles (AUVs), it is crucial to maintain the equipment’s optimal horizontal motion state while underwater, with communication equipment positioned above the waterline. Generally, the maximum allowable pitch angle for the equipment should not exceed 20°. Exceeding this maximum pitch angle can result in a weakening or loss of control over the equipment.

When an AUV transitions into a vertical state, the equipment gradually loses its ability to move within a plane. Vertical motion requires specific structural strength of the propeller and the state structure of the carrier, making it necessary to limit the pitch angle. However, the pitch angle limitations may vary based on the specific equipment’s performance.

During hill climbing maneuvers, the equipment’s slope limitations primarily depend on the maximum allowable pitch angle. If the slope angle surpasses the pitch angle limit, the AUV will trigger a near-bottom alarm and eventually activate floatation after climbing for a certain duration.

In constant-altitude mode, we establish a set of assumed AUV constant-altitude navigation parameters, as illustrated in

Table 1. AUV assumed parameters.

Altimeter operating frequency	2 Hz	Two measurements per second
Response time	1 s	At least two measurements are required to initiate response
Maximum diving speed	0.4 m/s
Height tolerance error	0.3 m	The closer the plan is to the bottom, the smaller the allowable error
AUV speed	3 kn	Comprehensive efficiency and resolution requirements
AUV height from bottom	20 m	Comprehensive safety and resolution
Duration required for alarm	30 s

. This set of parameters is based on the working principle of constant-altitude mode and the continuous altitude failure alarm principle.

As shown in Figure 9, this analysis examines the limiting situation where the AUV does not activate a continuous height-holding failure alarm during fixed-altitude mode. Under the assumed parameter conditions, the maximum slope angle for the AUV to maintain a constant height over the seabed is 14.068°. This indicates that under the assumed parameter performance, the fixed-altitude mode of the AUV has limited applicability to large slope seabed terrains.

Additionally, based on the collision avoidance principle of the safety system, the assumed parameters are presented in

Table 2. AUV assumed parameters.

Installation angle of collision avoidance device	Horizontal downward 45 °
Operating frequency of collision avoidance device	10 Hz
Collision avoidance response time	1 s	Start responding after 10 consecutive measurements
Safe distance	2 m
Response completion time	0 s	Response and reverse loading
AUV speed	3 kn

.

As depicted in Figure 10, when analyzing the critical conditions triggering a collision avoidance alarm while ensuring equipment safety, it becomes evident that equipment collisions occur when the slope angle exceeds 84.797°.

Considering the above analysis under assumed parameter performance conditions, it is important to note that the applicability of the constant-altitude mode for certain types of terrains may vary due to differences in equipment performance. For instance, the installation of oil pockets and sliders can enhance the equipment’s ability to maintain non-horizontal positions and relax the maximum pitch angle limitations. Consequently, the choice of the vertical motion mode should be made based on the specific performance characteristics of the AUV.

2.2.2. Equipment Inspection and Operational Execution

A regular AUV system is equipped with various modules, including propulsion, power, communications, positioning, attitude, deployment, and recovery, each requiring thorough inspection. Initially, an overall inspection should be conducted before initiating the AUV. During this phase, the AUV should exhibit no visible damage, and the integrity of essential components such as propellers, rudders, and flotation materials must be verified, ensuring that backup components are readily available. Additionally, the internal pressure must exceed the maximum pressure expected at the most profound operational depth while maintaining stable pressure. Second, thoroughly examining the propulsion, power, and communications modules is necessary. This assessment involves using remote sensing techniques to adjust the speed and direction of the AUV’s propellers and rudders, inflate the oil bladder, and move the sliders to test the three modules simultaneously. The remote-sensing part involves multiple communication methods, including radio, mobile communications, short-range Wi-Fi hotspots, satellite communications, and underwater acoustics. Third, the positioning module is calibrated, with the DVL subjected to inspection after submersion and before decoupling. During the calibration process, which usually involves a stationary period of 10 min, the AUV needs to be rotated to assess the attitude module functionality. Fourth, verifying the full functionality of emergency recovery tools and auxiliary deployment and recovery mechanisms is essential when assessing the deployment and recovery module. Examining side-scan sonar payloads follows the inspection of the AUV’s primary modules. First, it is essential to verify the integrity of the transducers on both exterior sides and ensure the functional connectivity of the cables. Second, upon submersion, the performance of several key aspects, which encompass verifying adjustments to the side-scan sonar’s measuring range, assessing the interface between data logging and real-time data feedback, confirming the seamless connection of positioning and attitude information with side-scan sonar data, and evaluating the quality of the waterfall plot data for both port and starboard sides, across various frequency ranges must be observed. Also, a test should be conducted to determine if all relevant parameters are optimized to facilitate precise feature detection by the side-scan sonar.

The emphasis of AUV operations lies primarily in the deployment and recovery of equipment and the ability to address potential operational disruptions. Due to increased collision risks during these phases, a stable vessel and optimal sea conditions are essential. Specifically, it is crucial to ensure that the swing angle of the mother vessel during deployment and recovery does not exceed 15° and that the sea state remains at or below Level 4, indicating moderate or calmer wind-wave conditions. In the event of an operational interruption, it is crucial to promptly receive notifications of the disruption, determine the equipment’s surface location, and define the cause of the interruption to address the specific issue and assess the AUV’s fitness for ongoing operations. This evaluation informs decisions regarding whether to resume operations or initiate recovery for maintenance. Before surfacing in the recovery process, all acoustical instruments of the AUV should be switched off. Upon recovery, inspecting the vehicle’s watertight enclosures, assessing payload status, and evaluating remaining power reserves should be performed while resetting the recovery module and cleansing the equipment with fresh water. Once the data are retrieved, the AUV should be powered down or recharged, if necessary. Moreover, job logging can be conducted while AUV operations are in progress, focusing on operational modes, statuses, timelines, and encountered challenges. It is also advisable to cross-reference and document the items inspected both before deployment and after recovery.

3. Experimental Results and Analyses

The experiment involved the deployment of a mid-sized AUV and a small-sized AUV, as depicted in Figure 11. These AUVs operate within complex underwater terrain with depths ranging from 20 to 500 m, as well as a flat seafloor area at a depth of 40 m. Notably, these two AUVs displayed differing maximum cruising speeds, with one achieving a top speed of 4 kt while the other boasted an impressive speed of 20 kt. Each AUV was equipped with distinct side-scan sonar models: the Lcocean-developed Shark-S450D and the ES1000 from Hydro-Tech Marine. Both vehicles exhibited the capacity to operate at 1000 kHz and 500 kHz frequencies, offering a maximum operational range of 100 m for high-frequency mode and 200 m for low-frequency mode.

In the complex underwater terrain, experiments were conducted employing the constant-altitude, fixed-direction approach and the constant-depth, fixed-point approach. The goal was to evaluate the suitability of survey line deployment methods in complex seabed topography. To conduct feature detection tests in the flat underwater terrain, a combination of constant-altitude and fixed-point modes, selected for their focus on enhancing positioning precision, and a fusion of constant-depth and fixed-point modes were employed. These efforts were designed to evaluate the suitability of survey line deployment approaches in flat-seafloor underwater settings and to acquire empirical data to analyze side-scan sonar image features during AUV feature detection operations.

3.1. Experiment in Waters with Flat Seabed Topography

In the flat terrain area, the seabed gradually changes from northwest to southeast, with an average depth of 40 m. Additionally, this region features fewer underwater features or obstacles, as shown in Figure 12. During this experiment, a 1-meter-diameter object model was submerged underwater to enhance the texture richness of the seafloor.

3.1.1. Experiment in Constant-Altitude and Fixed-Point Modes

The objective of the experiment is to assess the applicability of constant-altitude and height-varying fixed-point modes in flat terrain and to validate the directional principles of survey line planning. To achieve this, specific requirements must be established for the experimental conditions:

(1)

It is essential to ensure that the equipment maintains a low height above the ground in order to acquire high-resolution image data at both high- and low-frequency rates;

(2)

It is crucial to ensure optimal usage of the effective working range to meet the overlap coverage requirements according to the survey line planning principles;

(3)

The survey line coverage range must ensure complete coverage of the target area;

(4)

The survey line direction must adhere to the established survey line planning principles;

(5)

The flight speed setting must ensure data clarity and quality.

In this experiment, the AUV consistently maintained an altitude of 20 m while using low-frequency side-scan sonar with a 150 m measuring range. Four survey lines were employed, each measuring 500 m long and spaced 75 m apart. The AUV was configured to operate at a cruising speed of 2 kt, and the survey direction alternated between ±45° from the horizontal plane for two sets of repeated experiments. Moreover, an experiment was conducted using a mid-sized AUV, while another set of repeated experiments was carried out with a small-sized AUV equipped with a 75-m range capability and cruising at a speed of 3 kt. The specific parameters are shown in

Table 3. Flat terrain constant-altitude fixed-point experiment parameter settings.

Repeat Group Number	2
Altitude (m)	20
Range (m)	75
Survey Line Spacing (m)	75
Survey Line Length (m)	500
Survey Line Quantity	4
Navigation Speed (kn)	3
Survey Line Direction (°)	45°
Experimental Equipment Model	Small AUV
Main Purpose of the Experiment	Verify the applicability of constant-altitude mode

. Figure 13 displays the AUV’s depth, course, and image data.

Several sets of experiments conducted under constant-altitude mode have validated the applicability of the constant-altitude and fixed-point mode in flat waters, in addition to the planning principles for survey line direction, spacing, and overlap.

3.1.2. Experiment in Constant-Depth and Fixed-Point Modes

The purpose of this experiment is to test the applicability of fixed-depth and variable-depth fixed-point modes on flat terrain, as well as to verify the planning principles for survey line direction, spacing, and overlap. The differences between this experiment and the constant-altitude experiment are:

(1)

The survey line direction follows the principle of being parallel to the edge of the survey area to analyze the effect of the fixed-depth mode on violating the principle of parallel isobaths on flat terrain;

(2)

Controlled experiments with different navigation speeds are established to observe the impact of speed on images;

(3)

A dense control group is established to analyze the applicability of survey line layout methods in terms of spacing and overlap.

The conditions for this experiment are twofold. Firstly, the AUV maintained a constant depth of 20 m and operated at a low frequency, equipped with a measuring range of 150 m. Four survey lines were deployed, each spanning 500 m and spaced 75 m apart. The vehicle cruised at a speed of 2 kt, with the experiment conducted in horizontal (0°) and vertical (90°) survey directions, resulting in one experimental group for each direction. Furthermore, the experiment involved a mid-sized AUV with one experimental repetition, a small-sized AUV equipped with a 75 m range capability, which operated at 3, 4, and 5 kt in the horizontal survey direction. Second, to enhance the depth mode measurements, the AUV maintained a constant depth of 20 m, using a low-frequency setting with a 75 m measuring range. Six survey lines were deployed, each measuring 160 m long and spaced 30 m apart. The AUV maintained a constant speed of 3 kt, and the experiment was performed in both horizontal and vertical directions, with one experimental group designated for each direction. Moreover, an additional experiment was conducted using the AUV equipped with a 50 m measuring range, involving the traversal of four survey lines, each stretching 200 m in length, in addition to one experimental repetition undertaken by a small-sized AUV. Experimental parameter settings are presented in

Table 4. Flat terrain fixed-depth fixed-point experimental parameter settings.

Experiment Number	1				2
Repeat Group Number	2	-	1	-	2	-
Altitude (m)	20	-	-	-	20	-
Range (m)	150	75	-	-	75	50
Survey Line Spacing (m)	75	-	-	-	30	-
Survey Line Length (m)	500	-	-	-	140	200
Survey Line Quantity	4	-	-	-	6	4
Navigation Speed (kn)	2	3	4	5	3	-
Survey Line Direction (°)	0/90	-	0	-	0/90	-
Experimental Equipment Model	Medium AUV	Small AUV			Small AUV
Main Purpose of the Experiment	Verifying the applicability of fixed-depth and fixed-point mode				Verifying the applicability of survey line spacing and overlap

. The vehicle’s depth, course, and image data under the initial experimental conditions are illustrated in Figure 14, whereas the second experimental conditions are presented in Figure 14, Figure 15 and Figure 16.

Through the image analysis of Figure 14a,b, it can be observed that the seabed line in Figure 14c is not parallel. When combined with the survey area’s northwest shallow seabed and southeast deeper seabed, as well as the movement direction of the AUV from north to south, it can be inferred that the seabed line gradually expands from bottom to top, which is correct. However, due to the small depth change, the expansion trend is not particularly evident.

Figure 15 displays the image data captured by the AUV at speeds of 3, 4, and 5 knots while positioned at the same location, The red box shows that the shape of the anchor hook changes with the speed:

Based on the analysis of Figure 15, it can be observed that the change in texture angle in the upper right corner indicates a correlation between speed and image compression/stretching in the direction of travel. This observation is consistent with traditional side-scan sonar operations, where an increase in speed results in image compression, while a decrease in speed leads to image stretching. Therefore, it is essential to maintain speed control during operation to ensure accurate image interpretation.

Figure 17 presents the position deviation of anchor ruts after accounting for the offset distance between tracks. The black lines illustrate the position of anchor ruts in different tracks. It can be observed that all six tracks detected anchor ruts, which redundantly supplies information and thereby compromises efficiency. The overlap degree and occurrence ratio of corresponding targets within each group are listed in

Table 5. Overlap parameter table.

	Experiment 1		Experiment 2
	Medium Group	Small Group	Group 1	Group 2
Overlap (one-sided)	150%	100%	160%	140%
Number of lines required to cover the area	4	6	14	16
Number of occurrences of the same target	3	2	6	4
Proportion of occurrences of the same target	75%	33.3%	42.9%	25%

.

To optimize the efficiency of side-scan sonar operations, the number of lines required to cover the area should be minimized while ensuring that spacing and range settings meet the requirements of not missing any targets and ensuring adjacent strip connections. However, it is also crucial to avoid having too many occurrences of the same target. By satisfying the criterion of having the same target appear twice, co-visibility target-to-strip connections can be applied to enhance operational efficiency. According to the data in the table, increasing overlap does not increase the number and proportion of target occurrences. Hence, having numerous target occurrences is meaningless for underwater terrain exploration. Shrinking both the spacing and range of the lines will increase the number of lines, resulting in a decrease in efficiency. Therefore, the spacing and range of the lines should be set to ensure that the number of lines is as small as possible to improve operational efficiency while still meeting the minimum overlap requirement. The four overlap degrees in the table are relatively high, with a one-sided overlap of 100% for the degree with the target occurring twice. Due to AUV positioning accuracy issues underwater, it is necessary to increase the overlap percentage based on shipboard side-scan sonar operational standards to ensure co-visible target acquisition.

The tests in fixed-depth mode demonstrated that this mode was not superior to constant-altitude mode in flat waters.

By comparing tests conducted in fixed-depth fixed-point mode and constant-altitude fixed-point mode, it was observed that in flat terrain conditions, constant-altitude fixed-point mode is the preferred mode for vertical movement. This mode has strong applicability for establishing lines’ direction, spacing, and overlap degree. In flat waters where depth changes are minimal and unrestricted, there is no need to consider depth changes. However, combining fixed-depth mode with parallel zone edges can yield poor results.

3.2. Experiment in Waters with Complex Seabed Topography

In an underwater environment featured by complex topography, the water depth experiences significant variations, spanning from over 10 to 500 m and covering a distance of 800 m from east to west. Notably, the overall slope of this terrain exceeds 20°, and the presence of densely clustered rocks near the platform reef results in a rugged and uneven landform, as depicted in Figure 18. Given the AUV’s limitations in maintaining a specific altitude under its constant-altitude mode, the experiment experienced frequent interruptions. The AUV repeatedly triggered alarms and initiated premature surfacing during the guidance phase before reaching the specified altitude. Nevertheless, the constant-depth-based experiment was successfully conducted upon several attempts in various modes, including a combination of constant-altitude and fixed-point modes and a blend of constant-altitude and fixed-direction modes. As such, the performance of current AUV models cannot support sustained operations in the constant-altitude mode.

3.2.1. Experiment in Constant-Depth and Fixed-Direction Modes

The fixed-depth directional experiment aims to verify the applicability of the two modes, constant depth and direction, as well as their combinations, in complex terrain sea areas, and to establish the effectiveness of the survey line direction planning principle. This experimental design takes into account the unique characteristics of complex terrains, including variable oceanic conditions, diverse seabed features, and irregular bathymetry. In this experiment, the AUV was stationed on the western boundary of the survey area, and it conducted scanning along a survey line that extended for 2500 m from north to south. The entire survey line was deployed at an approximate depth of 100 m. The AUV operated using a 500 kHz frequency, with a single-sided range of 150 m. Throughout its operations, the AUV consistently maintained a depth of 80 m below the sea surface. The experimental parameters are shown in

Table 6. Parameter settings for the fixed-depth directional experiment in complex terrain.

Repeat Group Number	1
Depth (m)	80
Range (m)	150
Survey Line Spacing (m)	150
Survey Line Length (m)	2500
Survey Line Quantity	1
Navigation Speed (kn)	3
Survey Line Direction (°)	180
Experimental Equipment Model	Small AUV
Main Purpose of the Experiment	Verify the applicability of constant depth and direction

.

Regarding the overall course, it displayed a slight deviation of roughly 2° to the south by east, as illustrated in Figure 19. The AUV submerged from depths that exceeded the operational capabilities of the DVL and adhered to the specified heading and timeline. One notable limitation of this operational mode is its independence from DVL-derived data. The position information within the data acquired by the side-scan sonar system is derived from inertial navigation. In cases where the AUV’s heading angle deviates from the pre-defined course angle, the AUV calibrates its heading angle to align with the predetermined course. In situations with lateral water flows, there may be an angular disparity between the planned course and the actual path taken. Due to inherent inertial navigation errors, disparities can occur between the actual and recorded trajectory, resulting in variations in the position information along survey lines compared to the initially configured parameters.

Figure 20a illustrates the variation in the AUV’s depth relative to the sea surface over time, measured in meters (m). Figure 20b,c depicts waterfall plots produced by the AUV’s side-scan sonar during its submersion. From Figure 20b, it is noticeable that as the AUV descends, the seabed profile lines generated by the side-scan sonar gradually converge, ultimately leveling out as the submersible reaches a specific depth. As the AUV descends further, the gap between these lines progressively increases, as illustrated in Figure 20c. Eventually, the seafloor details become invisible from the diagrams due to surpassing the detection range of the side-scan sonar.

3.2.2. Experiment in Constant-Depth and Fixed-Point Modes

The objective of this experiment is to demonstrate the advantage of the variable depth constant depth mode over the constant depth constant depth mode and to establish the superiority of the fixed-point mode over the directional mode. Additionally, it aims to demonstrate the applicability of the survey line planning principle with regards to survey line direction, spacing, overlap, extension lines, and auxiliary lines. Compared to the experiment under constant-depth and fixed-direction modes, the experiment carried out under constant-depth and fixed-point modes leverages DVL positioning data, which provides more precise position information. The specific requirement for the water depth positioning in the positioning mode fundamentally supports this approach. This necessitates that the distance between the vehicle and the seafloor consistently falls within the DVL’s measurement range, considering the side-scan sonar’s detection range.

In this research, the AUV descended from a water depth of 40 m following a guiding line to the beginning of the first survey line. Then, it covered a distance of 2500 m from north to south along the first survey line, executed a 180° turn and ascended. Following this, it proceeded to the starting point of the second survey line, 75 m away from the first line, and traversed 2500 m from south to north along the second survey line, as illustrated in Figure 21. The experimental parameters are shown in

Table 7. Constant depth fixed-point experiment parameter settings.

Repeat Group Number	1
Depth (m)	90–97
Range (m)	150
Survey Line Spacing (m)	75
Survey Line Length (m)	2500
Survey Line Quantity	2
Navigation Speed (kn)	3
Survey Line Direction (°)	180
Experimental Equipment Model	Small AUV
Main Purpose of the Experiment	Verify the applicability of constant depth fixed point

.

Setting the AUV’s submersion depth is critical in this experiment. To align with the operational ranges of the DVL and side-scan sonar, the depth configuration must be approached on a stepwise basis. It is essential to ensure that any discrepancies in the empirical depth data fall within the prescribed ranges of the side-scan sonar and DVL while avoiding activation of the bottom-off altitude alarm. The stepwise depth profile for each segment was established using empirical data. The experiment was divided into two submersion stages. In the initial stage, referred to as the “guidance stage”, the AUV descended from the sea surface (0 m) to a depth of 90 m. After completing the first survey line, it executed a right-angle turn to transition to the second survey line. In the second stage, the AUV descended from a depth of 90 m at the start of the second survey line to 97 m and then initiated its ascent, as illustrated in Figure 22. The right graph in Figure 23 visually represents the AUV’s surfacing process, accompanied by the gradual expansion of the seabed profile.

The experiment further confirms the strong applicability of the survey line direction, spacing, overlap, extension line, and auxiliary line setting principles for complex seabed terrains. In combination with the flat terrain constant depth experiment, it is evident that the constant depth mode combined with the bathymetric contour direction survey line planning principle significantly outperforms the combined parallel survey zone edge line setting principle. Through the combination of constant depth direction and constant depth fixed point, this set of experiments fully validates the survey line planning principle’s strong applicability in complex terrain sea areas.

Two main challenges were faced in the experiment where the AUV’s depth and point were controlled. Firstly, the reliability of empirical depth data and safety considerations led to the determination of submersion depth without accounting for the side-scan sonar and DVL ranges. Thus, there was a lack of data in certain deeper areas, as illustrated in Figure 24. Second, after submersion, the performance of seabed tracking beyond the DVL range was suboptimal. To complete the experiments, the DVL alarm information was overlooked, reducing the precision of position information. As shown in Figure 25, regions A, C, and E in the left strip image exhibit similarities and share features with regions B, D, and F. This observation indicates that seabed information from the same location appeared twice in the left and right strips, albeit with varying position information.

To address these challenges, it is crucial to provide accurate empirical depth data and improve the planning of submersion depths. Additionally, removing the restrictions of maintaining a constant depth for each survey line is essential to mitigate the risks imposed by the shallowest point along a single survey line. Moreover, it is recommended to adjust the depth of individual survey lines in stages, considering the specific seabed topography. This approach forms the foundation for a robust scanning design along each survey line. By addressing the limitations of the side-scan sonar range, the proposed framework can also rectify inaccuracies in position information.

Based on the comprehensive analysis of the final results data from the experiment, it is evident that the survey line planning method incorporated in the technical design of AUV-mounted side-scan sonar operation is not only feasible but also highly effective. In addition to assessing and analyzing the selection of movement modes, the experiment also carefully compared and evaluated several key factors including the survey line direction, survey line spacing, survey line coverage overlap degree, and the deployment principles of extension lines and auxiliary lines in the survey line planning method. The experiment’s findings strongly support the practicality and effectiveness of the survey line planning method content present in the technical design scheme proposed in this chapter.

4. Discussion

4.1. Experimental Analysis

The data obtained from the experiments described above suggest that the technical framework of AUV-based side-scan sonar operations is not only feasible but highly effective.

In the depth-directed and fixed-point experiments, it was verified that complex seabed topography and oceanic hydro-environments have a significant impact on directional movement patterns, making them unsuitable for operations in complex environments. In these experiments, two sets of tests on complex and flat seabed topography verified the stronger applicability of the variable depth depth-directed and fixed-point mode to complex seabed measurements. When using the depth-directed and fixed-point mode on flat seabed terrain, the AUV’s capabilities cannot be fully utilized, resulting in non-parallel seabed profiles between left and right images. This limits the effective area of seabed images and eliminates the advantages of bathymetric line tracking and ranging corrections in data processing. Complex seabed environments require a greater emphasis on avoiding risks of collisions with underwater features and instability resulting from equipment pitching when using the constant-altitude mode.

In the constant-altitude experiment, it was verified that the directional mode has poor controllability. Therefore, it is sufficient to only use the fixed-point mode with the constant-altitude mode for experiments. These experimental image data have high quality and can clearly display seabed topography, possessing a certain target detection ability, achieving the expected goal of survey line planning principles and demonstrating the applicability of the constant-altitude and fixed-point mode operation plan for detecting flat seabed areas.

In vertical movement mode, the constant-altitude mode is effective in flat seabed areas. When the AUV’s performance allows, it should be prioritized for use in complex seabed environments. Under the constant-altitude mode, DVL can fully leverage its advantages, while the directional mode is an unreasonable choice that sacrifices DVL’s strengths. When AUV performance limitations dictate the use of constant-altitude mode, stepwise depth changes in constant depth mode are more effective than constant depth mode. In planar movement mode, the fixed-point mode is more advantageous than the directional mode. Therefore, the preferred order of mode settings in complex seabed environments is: constant-altitude and fixed-point mode → stepwise depth transformation constant depth and fixed point mode → constant depth and directional mode.

Based on the principle of the fixed-point mode track correction algorithm, it can be observed that the survey line planning approach that prioritizes the use of the fixed-point mode possesses robustness against environmental interference such as waves and currents. Nevertheless, a closer analysis of the results from directional experiments and complex terrain constant depth fixed point experiments reveals that when operating in the directional mode, the AUV’s stability against interference is reduced. Additionally, when the DVL bottom tracking is lost in the fixed-point mode, relying solely on inertial navigation data may also diminish its robustness against current interference. Thus, it is evident that for AUVs conducting side-scan sonar operations, the survey line planning method proposed in this article is highly reliant on the performance and accuracy of positioning and navigation hardware. Any malfunctions or issues in the navigation system can significantly impact its robustness. Furthermore, beyond the navigation system’s influence, problems with the manipulation system and side-scan sonar payload can also result in operational failure or interruption. This highlights the crucial importance of meticulous inspections of both the AUV itself and the side-scan sonar payload during operational preparations.

In most existing AUV path planning methods, the main focus is on researching algorithms for obstacle avoidance, anti-current interference, and path control [7,8,9]. In research related to AUVs equipped with side-scan sonar [10,11], the emphasis is more on target detection, which differs from the purpose of using AUVs for seabed terrain detection in this article. Therefore, the proposed method in this article is not directly comparable in quantitative indicators. However, qualitatively comparing and analyzing, the survey line planning method proposed in this article is more suitable for AUVs equipped with side-scan sonar to conduct seabed terrain detection operations. From the perspective of algorithm principles, the survey line planning method proposed in this article, especially the fixed-point mode algorithm principle, is simpler than path planning methods used for target detection. Due to its simplicity, this method’s principle may have lower applicability to more complex environments and path planning. Overall, however, the survey line planning method proposed in this article achieved a result with good applicability.

4.2. Experimental Image Analysis

The main characteristics of the side-scan sonar images obtained by the AUV, as deduced from the experimental findings, can be summarized as follows:

(1)

In the AUV’s constant-altitude mode, within flat topography, the seabed profile typically follows the AUV’s course, resulting in an inability to detect terrain variations. Nevertheless, these variations become significant when the vehicle encounters minor topographical undulations, as presented in Figure 26.

(2)

In the constant-altitude mode of the AUV, the operation becomes challenging in the complex underwater terrain. As bottom tracking requires pitch movements of the vehicle, this leads to a more intricate seabed profile, as shown in Figure 27.

(3)

In the AUV’s constant-depth mode, the seabed profile is displayed to reflect changes in seabed depth. Deeper submersion reduces the distance between the AUV’s course and the seabed profile. Simultaneously, the sea surface line may appear in the seafloor image if not adequately filtered, as illustrated in Figure 28.

(4)

As the AUV alternates between submersion, surfacing, and depth adjustments, the seabed profile exhibits a funnel-like shape, similar to the changes observed in shipborne side-scan sonar systems when vessel speed alters the diving depth, as presented in Figure 29.

(5)

During pivot steering, the outward waterfall plot of side-scan sonar imagery exhibits distortions, while the inner sections exhibit strip images that overlap and contain duplicate scans of the same geographical area, as shown in Figure 30, Figure 31 and Figure 32. Although such images present difficulties in application, their distinct characteristics enable more straightforward interpretation.

(6)

The additional acoustic equipment onboard the AUV introduces notable interference in the side-scan sonar imagery. This interference is characterized by recurring noise patterns, as displayed in Figure 33, making it readily identifiable but challenging to eliminate.

(7)

Upon correction of slope distance, the geocoded strip image generated within the depth range that falls within the side-scan sonar’s measurement capacity takes on an elliptical shape at both ends, as presented in Figure 34.

(8)

As the AUV accelerates, it induces a compression effect in the images along the heading direction (Figure 35). Nonetheless, the AUV generally operates at speeds between 2 to 4 kt during side-scan sonar missions, which are slower than shipborne side-scan sonar systems. Consequently, the compression level is lower and remains below the levels observed during rapid acceleration in shipborne side-scan sonar operations.

In the experiment conducted under the constant-altitude mode, it was theoretically expected that variations in vehicle velocity would lead to a scaling effect in the images along the heading direction. Yet, in practice, this effect is not particularly significant due to the moderate cruising speed of the AUV during side-scan sonar operations. Shipborne side-scan sonar measurements are typically carried out at speeds below 6 kt, whereas AUV-based side-scan sonar systems operate at lower speeds of 2 to 3 kt. This reduced velocity ensures the AUV can maintain image resolution and stability, even when encountering water flows. The AUV’s relatively low cruising speed and minimal speed fluctuations help minimize the impact of cruising speed on the produced image.

5. Conclusions

We have introduced a comprehensive technical framework to address the lack of systematic operational standards for seabed topography surveys employing AUVs equipped with side-scan sonar systems. This framework has been designed to account for numerous variables while integrating the specific characteristics of the AUV’s vertical motion control modes. The experimental results demonstrated the effectiveness of the established approach, ensuring the safety of the submersible and the quality and efficiency of operations in AUV-mounted side-scan sonar surveys. The present study offers two main contributions:

(1)

We established a technical framework for AUV-based side-scan sonar operations. This research considered a wide range of factors, encompassing vehicle safety, equipment performance, operational efficiency, motion control modes of the vehicle, and data quality. Furthermore, the proposed methodology systematically optimizes the procedural framework for conducting underwater surveys using AUV-based side-scan sonar. Therefore, it can serve as a timeless reference for future research in this field.

(2)

We conducted an in-depth analysis to study the properties of image data produced through AUV-based side-scan sonar operations. As such, we successfully identified ten distinct attributes that characterize experimental results acquired through ship-towed side-scan sonar surveys. This research provides the foundation for future work to improve the post-processing of data retrieved from AUV-based side-scan sonar operations.

Based on the findings, the survey line planning method proposed in this article is different from existing AUV path planning research and is more suitable for AUVs equipped with side-scan sonar to conduct underwater terrain detection operations. Additionally, the operational approach proposed in this article holds tremendous engineering significance and can provide a reliable guideline for AUV-based side-scan sonar operations.