This study presents the development and validation of active sonar-based perception and wall-following methods for autonomous underwater vehicle (AUV) inspection in enclosed aquatic environments such as reservoirs, dams, and water tanks. A hardware-in-the-loop (HIL) simulation framework was established to reduce development risk, integrating a ray-based rangefinder simulator with a mechanically scanned imaging sonar (MSIS) generator that includes noise and side-lobe modeling. Two complementary methods were implemented: a dual-rangefinder approach estimating lateral offset and heading deviation from paired starboard rangefinders, and an MSIS-based approach extracting wall geometry using median filtering and Hough transform, fused with Doppler velocity log (DVL) and gyro data for state estimation between image updates. Simulation and towing-tank experiments demonstrated consistent performance. In HIL trials, the rangefinder method maintained a 1.2 m stand-off with mean deviations of 0.02 m and 0.69°, while the MSIS method held a 1.5 m offset with errors of 0.06 m and 1.66°. Experimental results confirmed these trends: the rangefinder method ensured precise tracking, whereas the MSIS method maintained 1.45–1.56 m altitude and robust wall following under magnetic disturbances. Overall, the two methods exhibit complementary strengths, collectively demonstrating a validated perception–control framework for AUV wall following and inspection in confined underwater environments.
In cycloidal propulsion system, propulsion and steering are performed by disc rotation and blade oscillation, respectively. The power is transmitted to the disc and blades through gears and linkage mechanisms, resulting in lower efficiency. To overcome this, an electrical cycloidal propeller is proposed, integrating the propeller, ship hull, and the prime mover. This configuration improves overall efficiency. The design features an annular rotating disc with a dual-stator electric motor arrangement, wherein the load torque is shared between the inner and outer stators. The dual-stator design decreases the height of the propulsion motor. The use of an annular disc decreases the wetted surface area, thereby lowering frictional drag, mass, and mass moment of inertia of the propulsion system. In the proposed design, the blades are provided with individual electric motors. An optimal control scheme for dynamically varying the pitching motion of the blade has been devised. This improves the propulsion efficiency and crabbing maneuvering ability of the system, while retaining the original maneuvering ability. The characteristics of the disc and blade DC motors are analyzed for different maneuvers with the proposed optimal control scheme.
We present a hierarchical scalable methodology to analyze the maritime traffic of year 2022 associated with cargo, tanker, and passenger vessels traveling within the West Mediterranean Sea. The methodology is based on data-mining and machine learning techniques to process large Automatic Identification System (AIS) datasets. We present both a high-level representation of network traffic for the global area and low-level networks for local areas. The intermediate steps of the methodology generate relevant results about critical aspects of maritime traffic. We explore a large test matrix to assess the impact of tuning parameters upon the results and to propose practical guidelines for the methodology application in congested areas. The described approach highlights the traffic features of each ship category at different levels of detail and may support maritime traffic planning and management activity.
Maritime navigation safety is the basis for global logistics and marine ecosystems. The increasing availability of Automatic Identification System (AIS) data has opened new avenues for forecasting vessel trajectories with higher precision and robustness. This review presents a comprehensive survey of recent data-driven approaches, including probabilistic models, classical statistical methods, and deep learning (DL) architectures. Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), Recurrent Neural Network (RNN), Transformer-based models, and hybrid systems are evaluated for maritime trajectory prediction. By categorizing studies based on their input features (speed, course, and position) and computational methods, this review highlights methodological trends and performance benchmarks. Additionally, an overview of dataset accessibility and contextual variable use, including weather and wave state, is presented to support future reproducibility. We conclude by identifying gaps and future directions toward explainable and adaptive maritime forecasting systems.
In recent years, underwater cavitation jetting has emerged as a prominent and innovative cleaning technique for deep-sea aquaculture cages, ship hulls, and other marine structures. The submerged cavitation nozzle, serving as the core functional component, enables cavitation generation. However, practical manufacturing processes may introduce coaxiality errors, which can significantly affect the performance of the cavitation jet. This study focuses on angular cavitation nozzles and investigates the effects of coaxiality errors in the throat and diffusion sections on the submerged flow field using large eddy simulation (LES). The numerical simulations were conducted for nozzles with coaxiality errors of 0.05 mm and 0.1 mm under a 10 MPa pressure inlet condition. The results demonstrate that coaxiality errors substantially influence the flow field and cavitation characteristics. Specifically, a 0.1 mm offset in the diffusion section induces an axial flow deflection angle of up to 6.3°. Compared to errors in the throat section, diffusion section misalignment exerts a more pronounced effect on cavitation performance. At a 0.1 mm diffusion section offset, the maximum vapor phase volume fraction reaches 77.4
To address the high-noise issue associated with the high-speed circulating water channel (HS-CWC), an integrated model combining the circulating water channel and axial-flow pump was established in this study. A systematic research was conducted using computational fluid dynamics (CFD) numerical simulations. Multiple noise-monitoring points were arranged in various regions of the HS-CWC, and the direct method is employed to obtain the acoustic field distribution characteristics. The causes of noise were revealed through the analysis of the entire pipeline flow field. Based on the observed inflow non-uniformity of the axial-flow pump, a workbench model with ideal uniform inflow conditions was established for comparative computations. To reduce flow-induced noise, partial optimization design was implemented for the HS-CWC, including increased curvature radii at the No.1 and No.2 corners and the installation of anti-separation plates upstream of the No.1 corner. Comparative analyses of flow and acoustic field characteristics before and after modification are performed. The results indicate that significant flow separation occurs near the No.1 corner, inducing intense flow-induced noise. The axial-flow pump downstream of the No.2 corner is affected by non-uniform inflow from upstream, leading to strong periodic fluctuations in blade loading, which is identified as the primary cause of excessive noise near the axial-flow pump. The calculation result of the sound field is consistent with the on-site auditory perception. Under uniform inflow conditions in the workbench model, the overall sound pressure level at monitoring points near the axial-flow pump is reduced by approximately 4 dB. The flow separation phenomenon of the modified HS-CWC has been significantly reduced, and the uniformity of the inflow of the axial-flow pump has been significantly improved. Overall sound pressure levels are reduced by approximately 2 dB in the No.1 corner region and by more than 2 dB in the axial-flow pump region. This study provides theoretical guidance for the optimized design of HS-CWC, contributing to enhanced flow uniformity and reduced pipeline noise.
This paper discusses current advancements and future trends in automated pipe routing optimization within the maritime industry. Pipe routing is a critical process in the design of systems to transport fluids and gases across various industries, notably in shipbuilding, where efficiency, cost-effectiveness, and safety are paramount. The paper reviews existing methodologies for automated routing, addressing the challenges posed by spatial constraints and the need for optimal solutions. It delves into the role of computer-aided design (CAD) and optimization algorithms in streamlining the routing process. Additionally, it explores future perspectives on artificial intelligence (AI) and machine learning (ML) techniques in enhancing decision-making for pipe routing. The review highlights the integration of multi-objective optimization approaches to balance competing factors, such as cost, safety, and space utilization. Emphasizing the need for robust solutions in maritime industries, the paper suggests that AI-driven technologies will significantly contribute to minimizing human error and improving routing decisions in complex shipbuilding environments. Furthermore, the paper discusses the adoption of simulation tools and real-time data processing for better outcomes. The review concludes that while advancements in automated pipe routing are promising, challenges related to system complexity, technological integration, and scalability must be addressed to fully harness these innovations.
Saudi Arabia’s extensive coastline provides nearshore wave and tidal resources that, while moderate in global terms, are technically suitable for infrastructure-integrated marine energy applications. This study presents a feasibility and performance assessment of nearshore wave and tidal energy systems at eleven representative sites across the Red Sea, Gulf of Aqaba, and Arabian Gulf. The methodology integrates geospatial site screening, long-term metocean resource characterization, and technology-site compatibility analysis, with pilot-scale performance modeling conducted using the National Renewable Energy Laboratory’s System Advisor Model (SAM) and P10–P50–P90 uncertainty envelopes to represent resource and operational variability. Screening results indicate nearshore wave power densities of 0.4–1.2 kW·m⁻1 along the Red Sea and Gulf of Aqaba, supporting infrastructure-integrated concepts including breakwater oscillating water columns, on-structure floaters, and oscillating surge converters. SAM-based modeling yields P50 capacity factors of 13–17.5
Industry-wide simulation platforms are increasingly needed to address complex societal challenges, such as decarbonization. However, their development is often hindered by diverse stakeholder interests, fragmented models, and evolving regulations. The Maritime and Ocean Digital Engineering Laboratory (MODE), an 18-company industry–academia consortium, exemplifies this challenge in developing a shared simulation platform for maritime decarbonization. This study proposes a quantitative system architecting framework to prioritize development pathways, shifting the focus toward architecture-based system integration. The framework employs object-process methodology (OPM) and design structure matrices (DSM) to define a connection coverage metric, which quantifies how effectively a set of models captures critical cross-domain interactions. Applying this framework to the MODE project, we conduct a comparative simulation between ad hoc and architecture-based development strategies. The analysis reveals that the former suffers from significant periods of “synergy lag,” during which development effort accumulates without realizing system-level desirability due to uncoordinated modeling. In contrast, the proposed framework identifies a Pareto-efficient roadmap that minimizes this lag. Monte Carlo simulations show that architecture-based planning generates value both earlier and more efficiently than random bottom-up construction. When aggregated over the entire development horizon, the integrated impact of the architecture-based strategies is more than three times larger than that of the random bottom-up approach, indicating substantially higher cumulative value for the same modeling investment. The proposed framework thus provides systems engineers with a structured, quantitative method to build consensus and maximize the return on modeling investment in complex, cross-industry projects.
This paper proposes an analytic method for calculating the collision area to evaluate collision risk between ships in both static and dynamic encounter situations. The collision area is defined as the region that the own ship will intrude into in the future if it maintains its current speed, given a specific safety domain surrounding a target ship. For static situations, assuming the target ship maintains its course and speed, we derive exact boundaries of the collision area using both implicit function representations and parametric representations. This method accommodates various domain shapes, including circles, symmetric ellipses, and asymmetric ellipses. Furthermore, we extend the framework to dynamic situations. We demonstrate calculations for scenarios where the target ship alters its course and speed, and where the ship domain expands over time to account for prediction uncertainties. Finally, we clarify the theoretical relationship between the collision area and conventional indicators, such as the Dangerous Area of Collision (DAC) and the Obstacle Zone by Target (OZT), showing that the collision courses in OZT correspond to the tangent lines to the collision area. This study offers a robust and versatile analytical foundation for collision avoidance in Maritime Autonomous Surface Ships (MASS).
As the global shipping industry expands, severe maritime accidents such as collisions, strandings, and capsizing remain a critical concern. Because unpredictable waves are a primary catalyst for these incidents, providing navigators with accurate wave field forecasts is essential for ensuring operational safety. This study proposes a predictive framework to evaluate ship navigation safety under forecasted wave conditions. Specifically, the WAVEWATCH-III (WW3) model, driven by wind data from the Global Forecast System (GFS), was utilized to calculate wave parameters for a target sea area. These forecasted parameters were subsequently integrated into a mathematical ship motion model to dynamically assess the maneuverability and seakeeping of an S-175 container ship. The results demonstrate that: (1) the WW3+GFS model provides highly reliable wave predictions, validated by a strong correlation with buoy observation data. (2) While the ship’s overall maneuvering performance under anticipated waves meets standard operational requirements, its notably large tactical diameter during turning indicates limited collision avoidance capabilities, warranting navigational caution. (3) By analyzing the maximum roll angles under various combinations of speeds and rudder angles, this framework offers actionable guidance for navigators to optimize operational parameters and ensure pre-voyage safety.
Machine learning (ML) and deep learning (DL) have significantly advanced maritime engineering and hydrodynamics by enabling data-driven modelling and prediction. In hydrodynamics, high-fidelity surrogates offer precision, but low-fidelity surrogates are often preferred for generating large training sets due to their low computational cost. Time series records are a central form of data across seakeeping, structural analysis, sea-state estimation, ship-response prediction, control, and manoeuvring; they encode both the system’s response to excitation and key vessel characteristics. To improve robustness, coverage, and perturbation resilience, augmentation techniques, such as jittering, scaling, warping, and permutation, are commonly applied, but typically tuned by trial-and-error and rarely grounded in physics. To address this, we introduce a novel feature-engineering methodology that derives a physics-informed jitter variance from experiment–simulation discrepancies in time–frequency features, turning jittering from a heuristic into a principled step. The method is evaluated on a spherical floating model tested in regular and irregular unidirectional waves and compared against numerical simulations. Beyond PDF/PSD-based representational fidelity comparison, the augmented datasets are further assessed through downstream simulation-to-experiment classification across benchmark ML/DL models. Compared with raw simulation and RMS-matched standard Gaussian jittering, the proposed physics-informed augmentation improved experimental test accuracy and macro-F1 across most model families. These results demonstrate that the proposed method is not merely a spectral noise adder, but a physics-guided augmentation strategy that improves the ML usability of low-fidelity hydrodynamic time series.
In this study, we conducted captive model tests using a scale model of a 324 m-long cruise ship equipped with two POD propulsors. The aim was to capture the maneuvering hydrodynamic force characteristics acting on the ship model with the POD propulsors. Based on the test results, we developed a maneuvering simulation method for a POD-driven ship. This simulation method includes motion coupling effect with roll. To validate the developed simulation method, we conducted maneuvering simulations of the ship model and compared their results with those of free-running model tests. The simulation results of turning and zig-zag maneuvers agree with the results of free-running model tests within practical accuracy. This verifies the validity of the simulation method. Furthermore, we conducted maneuvering simulations for the cruise ship in full scale to investigate the effect of the metacenter height GM on the ship maneuverability and rolling characteristics during turning. As GM decreases, the ship’s course stability deteriorates, and although the turning performance is improved, the overshoot angles for zig-zag maneuver increase. Meanwhile, we observed that the roll generated immediately after steering became large as GM decreased. Furthermore, an unreasonably large roll tended to occur depending on the magnitude of GM .
This paper proposes a novel velocity command acquisition method for the dynamic base recovery control of an underwater unmanned vehicle (UUV). First, to overcome the challenge of unavailable real-time desired velocity information during recovery control, an improved fast terminal sliding mode control (FTSMC) strategy is developed, generating virtual velocity commands solely based on the UUV’s kinematic model. Second, accounting for the constraints of low underwater communication bandwidth and limited position-only feedback, a globally convergent differentiator is designed to accurately estimate the six-degree-of-freedom (6-DoF) velocity of the moving mother ship using sparse acousto-optic guidance data, thereby supplying critical inputs for the lower-level controller. Finally, the derived velocity information is integrated into a feedback controller to achieve precise UUV trajectory tracking. The proposed method not only enhances the control accuracy and robustness of UUV dynamic base recovery but also offers a new paradigm for underwater robotic control system design. Extensive simulations and experimental validations confirm the method’s superior performance and practicality, establishing a robust foundation for autonomous UUV recovery.
The present investigation examines the geometric optimisation of a submerged wave energy converter situated near a partially reflecting sloped seawall for the Indian sea-state. This is achieved by utilising the Response Surface Methodology (RSM) in conjunction with Design of Experiments (DOE). The study aims to model various shape configurations, specifically circular, elliptical, rectangular, and square cross-sections of submerged wave energy converters (S-WECs), and conduct a parametric analysis. Frequency domain analysis using the Boundary Element Method (BEM) reveals that the submerged rectangular wave energy converter (SR-WEC) outperforms all considered geometries in total capture width ratio ( C_P ), surpassing the circular, square, and elliptical configurations by approximately 33% , 15.8% , and 4.5% , respectively. Moreover, the study aims to elucidate the influence of the reflection wall coefficient of the sloped seawall, submergence depth, oblique wave attack, and relative distance from the sloped seawall on the energy capture efficiency of the SR-WEC. The findings show that the performance of an SR-WEC placed in the vicinity of a fully reflecting sloped seawall is approximately 73% higher than in open sea conditions. The quadratic polynomial derived from the RSM predicted model of maximum WEC power (P_WEC)_max , along with the total C_P , constitutes a two-objective problem. The optimised buoy configuration is used to estimate the annual power extraction potential of the WEC at Chavara, Kerala, India (latitude 9.00^∘ N, longitude 76.420^∘ E), yielding an estimated annual energy production (AEP) of 26.9 MWh m^-1 year^-1 . The study underscores the effectiveness of combining BEM with RSM to customise WEC for specific regional wave conditions, thereby enhancing cost-effective wave energy extraction.
Reducing wave resistance is an important consideration in the design of high-speed displacement ships. The optimal sectional area curve for minimizing wave resistance has been studied theoretically, but the optimal waterline half-breadth curve at the design draft remains unresolved, and hull forms are often determined empirically. This study presents a rational method for determining the hull forms that minimize the wave resistance under imposed design constraints without prescribing either curve in advance. This is achieved by minimizing the sum of the wave and frictional resistance coefficients. The hull form is represented using a double trigonometric series, which enables systematic optimization by expressing the resistance coefficients as functions of the series coefficients. The optimization is conducted using the Lagrange multiplier method under the imposed constraints. Using this method, we optimized the Series 64 hull form, a high-speed displacement hull form researched by the US Navy. The optimization resulted in a hull form characterized by a U-shaped forebody and V-shaped afterbody. Model tests confirmed reductions in the total and residual resistance coefficients near the speed used for the optimization, demonstrating the effectiveness of the optimization in designing hull forms that reduce the resistance.
A marine biomass utilization system is developed in this study to address nutrient imbalances in semi-enclosed coastal seas such as Osaka Bay. Osaka Bay's nutrient imbalances pose significant environmental challenges, with eutrophication occurring in the northeastern part and nutrient deficiencies occurring in the southwestern part. Using biofertilizer derived from nitrified methane fermentation residue, this study examines the potential for enhancing phytoplankton and seaweed growth. The phytoplankton and seaweed culture experiments were conducted at Marine Fisheries Research Center, Research Institute of Environment, Agriculture and Fisheries, Osaka Prefecture. The seaweed photosynthesis experiments were carried out at Nakamozu campus of Osaka Metropolitan University. The biofertilizers used for both experiments were provided by the university’s laboratory as a part of nitrification experiments using methane fermentation residue. The results of the phytoplankton culture experiments showed that the addition of biofertilizer significantly enhances Chaetoceros neogracilis growth rate. The results of the seaweed photosynthesis experiments noticed that the relative growth rates of Ulva sp. with biofertilizer is much higher than that without biofertilizer. The results of seaweed culture experiments suggested that the addition of biofertilizer enhances growth rate and reduces discoloring of Undaria pinnatifida. These results indicate that the addition of nitrified methane fermentation residue effectively promotes growth of phytoplankton and seaweed.
Data assimilation (DA) is useful to improve simulation accuracy and tune model parameters. The improvement of the accuracy of the estimation of flows around a ship hull based on the DA method contributes to design ship hull form or energy-saving device, and we construct a novel framework for DA-based parameter tuning of RANS turbulence models carrying out the experiment at the towing tank. Particle filter (PF) is employed to estimate the likelihood of two parameters ( C_2 , C_4 ) of the explicit algebraic stress model (EASM) with measured data from 40 fiber Bragg grating (FBG) pressure sensors near the ship stern. This study presents the first application of data assimilation for parameter tuning of the EASM using measured data of the pressure distribution on a three-dimensional ship model. In the analysis using a 749 gross-ton general cargo ship model, it is confirmed that the parameter tuning using measured pressure distribution significantly affects the flow field near the stern. Moreover, the reconstructed pressure and wake distributions show good agreement with observations, and pressure recovery after flow separation is particularly improved. These findings imply that turbulence model parameters can be effectively tuned, and unobserved pressure and wake can be estimated by the present DA method combining RANS simulation and FBG measurements. Also, present work has the possibility to contribute to the improvement of the accuracy at a full-scale sea trials in the future.
A novel adaptive fault-tolerant control scheme is presented for underactuated surface vessels (USVs) tracking control under actuator faults, input saturation, dynamic uncertainties and unknown disturbances. First, to overcome the design difficulties caused by the underactuated problem, we convert the original mathematical model into the novel USV dynamics in the state space. Second, to address actuator faults and input saturation, an indirect neural approximation with virtual parameter technique and the smoothness of the hyperbolic tangent function is proposed respectively. Third, compared with other adaptive strategies, the advantage of the proposed adaptive control strategy is that control gains require adjustment of a few parameters. In addition, a Lyapunov method is used to obtain the control law and adaptive law, and the stability of the system is analyzed. Finally, the effectiveness of the proposed control method is proved by simulation.
With the rapid development of maritime autonomous surface ships (MASS), human factors remain critical in shaping accident causation. This study develops a multi-layered causal analysis model by integrating the Human Factors Analysis and Classification System (HFACS) with Bayesian networks (BN), aiming to identify key contributing factors and their transmission pathways. Utilizing data primarily drawn from 9 publicly available MASS accident investigation reports (published between 2010 and 2025) and 45 simulation-based human reliability studies reported in the literature, the model captures the probabilistic dependencies across management, supervisory, precondition, and operational behavior levels. The results show that a weak organizational climate, inadequate operational planning, and impaired operator condition together form the core causal chain of maritime accidents. The most influential behavioral error pathway, linking operational errors (R1) through operator condition (P2) and task and authority allocation (S2) to safety culture and policies (M2), exerts the strongest effect on accident probability. Sensitivity analysis indicates that upper-level management factors indirectly shape behavioral risk through supervisory and precondition layers, while multi-level interactions amplify the impact of operator conditions on unsafe acts. These findings validate the HFACS-BN framework for quantitative human factor analysis and reveal that remote control and human–automation interaction amplify multi-level risk propagation, offering theoretical and preliminary empirical insights for improving intervention strategies and intelligent maritime safety management.
