Dynamic Modeling of Water Depth and Navigation Decision-Making Methods in Shallow Waters Considering Tidal Effects

A ship autonomous navigation decision-making method based on the OODA loop theory

Ship autonomous navigation plays a vital role in ensuring the safety and efficiency of maritime transportation. However, existing methods are limited in handling complex ship encounter scenarios, and their output collision avoidance decisions suffer from insufficient rationality and practical applicability. To address these challenges, this study proposes a ship autonomous navigation decision-making method based on the Observe–Orient–Decide–Act (OODA) loop theory. This framework establishes a complete closed-loop navigation logic, covering situational awareness, risk assessment, decision generation, and motion control. In the observation module, a quaternion ship domain model is used to determine the dynamic safety boundary of ships, and a ship encounter recognition method based on relative orientation is constructed. A risk assessment module is developed that quantifies collision risk levels and multi-vessel avoidance priorities by integrating ship maneuvering constraints and dynamic maneuvering intervals. In the decision-making module, a feasible ship maneuvering interval model and a ship collision avoidance decision model are constructed by combining a three-degree-of-freedom ship motion model and an improved velocity obstacle algorithm to obtain a safe and feasible maneuvering decision scheme that meets the requirements. Finally, the Act module is constructed to execute decision instructions through the ship control system. Through simulation verification on the OpenCPN platform, the proposed method realizes safe collision avoidance of all target vessels in complex encounter scenarios. The minimum relative distance between the own ship and target ships exceeds the safety distance throughout avoidance, and DCPA is substantially larger than the safety threshold even when TCPA approaches zero. By tuning course and speed, this method yields COLREGs-compliant safe collision avoidance strategies and provides feasible technical support for the practical implementation of autonomous navigation systems.

Three-Dimensional Near-Seabed Path Planning for Autonomous Underwater Vehicles Using Fisher Information and Slope–Curvature–Yaw-Rate Constraints

Autonomous underwater vehicles (AUVs) operating near the seabed must maintain localization reliability while satisfying maneuverability and bottom-clearance constraints. This study proposes a three-dimensional near-seabed path-planning framework integrating likelihood-function-curvature Fisher information, multi-criteria start-point selection, an information-guided Hybrid A* planner, vertical-plane homotopy correction, and closed-loop tracking validation. Terrain adaptability is quantified from the curvature of the local likelihood surface and interpreted through the associated Cramér–Rao lower bound (CRLB). For the test area, a case-specific adaptability threshold of 0.6 was selected by balancing the mean CRLB against the usable planning area. Across three start-point schemes, the proposed planner increased mean terrain adaptability to 0.26–0.31 and reduced traversal through low-adaptability regions by 2.8–16.2 percentage points, with only a 0–1.9% increase in path length relative to standard Hybrid A*. After vertical correction, the generated trajectory satisfied the prescribed yaw-rate, slope, and curvature limits. Dynamic simulations of the HHU-1 AUV yielded horizontal and three-dimensional tracking RMSEs of 1.50 m and 2.51 m, respectively. Paired TAN-EKF Monte Carlo tests further showed a 3.32% nominal reduction in horizontal localization RMSE and reductions of 3.37–4.09% under four disturbance scenarios, indicating a modest but statistically consistent localization benefit.

An Autonomous GeoAI Agent for Arctic Eco-Navigation

Arctic maritime navigation is becoming increasingly important as changing sea-ice conditions expand seasonal accessibility while simultaneously introducing substantial operational, environmental, and community risks. Arctic route planning is inherently a multi-criteria problem: routes that improve vessel safety or efficiency may increase exposure to sea ice, sensitive ecosystems, or nearby communities. Existing routing methods prioritize travel time, fuel use, and navigational risk, often overlooking ecological and community impacts. We introduce a human-in-the-loop, multi-agent GeoAI system for Arctic eco-navigation that integrates operational, physical, ecological, and community-related criteria within a unified routing framework. Multiple specialized agents coordinate geospatial data acquisition and preparation, multi-objective route generation, and skyline-based decision support. The ecological criteria explicitly account for exposure to sensitive areas, including Essential Fish Habitat and seal critical habitat. By considering these ecosystem impacts and potential community burdens while keeping consequential value judgments under human control, the framework supports safer, more transparent, and socially responsible Arctic navigation. Project page and code are publicly available. https://samiraat.github.io/Arctic-Eco-Navigation-Agent/, https://github.com/samiraat/Arctic-Eco-Navigation-Agent

Optimal Forecast of Open-Channel Flow through Sequential Node Selection for Data Assimilation in a Hydrodynamic Model

Predicting river flow is important for flood management, river erosion protection, navigability update, and so on. The traditional approach to predicting the flow using a hydrodynamic model (HM) requires manual calibration and recalibrations of uncertain parameters; it has faced inefficiency and uncertainty issues. This study aims to explore assimilating observational data of water level and velocity into the HM as a new approach to predicting open-channel flow. The background of this study is that ever advancing technologies for flow monitoring have offered or will soon offer good amounts of data in real time, making data assimilation (DA) practical. The new approach’s novelty lies in combining hydrodynamic laws (via the momentum principle) with data during HM run time without the need for calibrations and optimizing the prediction accuracy while minimizing computing costs. Its performance is validated using observational data from flume experiments and hydrometric stations in the Danube River. The results show that the adaptive proportional–integral–derivative–based DA technique is more efficient than the model-predictive-controller-based DA technique, enhancing the computation efficiency by an order of magnitude. Both techniques automate the correction of the channel-bed friction factor itself, which may also alleviate the impact of other uncertain parameters on the prediction. Both techniques have achieved low relative errors ( < 1 % ). The sequential selection method successfully locates optimal DA stations in the channel, supported by open-channel flow theories. The new approach improves the flow prediction for the entire HM channel. This study has contributed to the development of a robust framework for forecasting open-channel flow in real time.

Dual-Agent Hierarchical Reinforcement Learning for Typhoon-Avoidance Route Planning of Ships Under Dynamic Wind–Wave–Current Fields

Ship weather routing in regional seas under severe seasonal weather systems, such as typhoons, poses a critical operational challenge for maritime safety and efficiency. Traditional single-agent reinforcement learning (RL) methods frequently suffer from training instabilities and erratic trajectory adjustments when exposed to the highly non-stationary, multi-scale dynamics of coupled wind–wave–current fields. To address these limitations, this study proposes a dual-agent hierarchical reinforcement learning framework (HRL-MOO) featuring a built-in dynamic risk-weight adaptation mechanism. The proposed global agent discerns large-scale environmental evolutions and adaptively updates the relative weights of wind-, wave-, and current-induced risks at a strategic level, while the local agent translates this macro-level guidance into short-term tactical heading and speed adjustments within realistic vessel motion boundaries. The framework incorporates bathymetric constraints through a high-resolution navigable domain mask derived from ETOPO topography to guarantee practical navigability. Simulated experiments are executed using hourly environmental reanalysis data and best-track records corresponding to the passage of Typhoon Yagi (2024) through the northern South China Sea and Taiwan Strait. The empirical results demonstrate that the cooperative dual-agent structure establishes a superior global Pareto frontier compared to conventional standard single-agent Proximal Policy Optimization (PPO) and metaheuristic baselines. Under extreme typhoon conditions, the HRL-MOO model effectively decreases the cumulative environment-induced risk from approximately 0.17 to 0.12, improves path smoothness by achieving a higher index of 0.842, and accelerates policy convergence to within 3000 training episodes, all while maintaining a highly efficient voyage distance with minimal detour overhead. Ablation studies further validate that the synergy between macro-level stream field recognition and adaptive multi-objective optimization significantly enhances decision-making stability. This framework offers a robust and interpretable computational tool for autonomous ship weather routing in fast-changing, high-risk ocean environments.

Determine the navigation safety of the S-175 ship using predicted wave parameters