(03)

Science

↓ (W30 ‘26 · Jul 20–26)
01openalex2026-07-24Shuling Zhao, Shengtao XieDOI ↗ (opens in a new tab)

Survey on ship route planning towards autonomous ship era (opens in a new tab)

This paper reviews the evolution and current state of ship route planning algorithms for Maritime Autonomous Surface Ships (MASS), highlighting human error as a key cause of maritime accidents. It categorizes algorithms into three groups: baseline trajectory methods, environment-adaptive methods, and multi-objective intelligent decision-making methods, analyzing their applications, limitations, and COLREGs compliance. The study identifies gaps in existing algorithms, suggesting future trends include integrating multiple approaches, interdisciplinary collaboration, and addressing uncertainty. Refining COLREGs integration is noted as critical for advancing autonomous shipping and global trade.

02openalex2026-07-26Joseph Mino, Mukesh KrishnanDOI ↗ (opens in a new tab)

Cyber-Attack Detection on Maritime Autonomous Surface Ships: Python-based Predictive Analysis (opens in a new tab)

This paper explores cyber-attack detection for Maritime Autonomous Surface Ships using Python-based predictive analysis. It analyzes dynamic data like geo-location, timestamps, and navigation status with machine learning and deep learning techniques to identify anomalies in real-time and Automatic Identification System datasets. The proposed hybrid model combines LSTM and random forests to improve detection accuracy, reliability, and scalability. The system aims to support proactive maritime monitoring and risk mitigation for autonomous operations.

03openalex2026-07-24Ke Zhang, Xiongfei Geng, Jie Wen, Liu X +1 moreDOI ↗ (opens in a new tab)

Adaptive collision-avoidance timing recognition for autonomous ship encounters using multi-evidence Bayesian backward evidence detection (opens in a new tab)

This study introduces a Bayesian framework to improve collision-avoidance timing for autonomous ships by analyzing historical AIS data. It integrates maneuvering, motion, and risk evolution to detect adaptive trigger thresholds for different encounter types (overtaking, head-on, crossing). Tests on 6,186 AIS encounters showed avoidance actions began 6.5–8.6 minutes before peak collision risk, with adaptive thresholds ranging from 0.245 to 0.368. The method outperformed fixed-threshold approaches in stability and interpretability.

04openalex2026-07-22Zixuan Qiu, Shaosong Min, Cong LiuDOI ↗ (opens in a new tab)

Autonomous Approach and Stable Tracking of Dynamic Target for Intelligent Ship Using Recurrent Soft Actor-Critic (opens in a new tab)

This paper introduces a reinforcement learning method called SAC-LSTM for autonomous ships to track moving targets. It uses LSTM networks to analyze past motion patterns and improves tracking by balancing distance, heading, speed, and smooth control. Tests show SAC-LSTM outperforms other methods, increasing success rates by 7.5% and reducing heading errors by 35.5%. The approach proves effective under various motion conditions and disturbances.

05openalex2026-07-20Nizirwan Anwar Tomy RonaldiDOI ↗ (opens in a new tab)

LONG-TERM OPERATIONAL MODEL OF AUTONOMOUS UNDERWATER VEHICLES (AUV) FOR MARITIME DEFENSE AND SECURITY (opens in a new tab)

This study proposes a long-term operational model for Autonomous Underwater Vehicles (AUVs) to improve maritime defense and security. It focuses on optimizing deployment, endurance, and navigation by integrating AI-based systems, energy-efficient propulsion, and acoustic communication. Findings show these technologies enhance AUV performance but face challenges like high costs, environmental limits, and cybersecurity risks. The research recommends local manufacturing and system interoperability to support Indonesia’s maritime surveillance.

06openalex2026-07-22Francesco Maione, Seher Suendam Arici, Agus Hasan, Modestas Sukarevicius +2 moreDOI ↗ (opens in a new tab)

Agentic AI for autonomous condition monitoring and predictive maintenance of marine vessels (opens in a new tab)

This paper introduces an Agentic AI system for marine vessel condition monitoring and predictive maintenance. It combines a Transformer-based Autoencoder for anomaly detection with a Large Language Model (LLM) to provide explainable, interactive analysis. The system processes high-frequency operational data to identify issues like abnormal loading or equipment degradation, using structured tool-calling to reduce errors. Tested on 91 days of vessel data, it successfully detected four fault scenarios with lead times of 97 to 637 minutes.

07openalex2026-07-23Jing Zhu, Zhang Zibei, Zijing LiDOI ↗ (opens in a new tab)

Advanced path-following control of unmanned surface vessels using fully actuated system theory (opens in a new tab)

This paper introduces a new control method for unmanned surface vessels using fully actuated system theory to simplify underactuated dynamics into a decoupled form. The approach combines an extended line-of-sight guidance law with time-varying look-ahead distance to handle unknown drift, while Lyapunov analysis confirms system stability. Simulations show improved tracking accuracy, faster convergence, and better robustness than existing nonlinear methods. Future work will address parameter-dependent compensation and actuator constraints.

08openalex2026-07-22An Li, Fan Gao, Zhenqiang Song, Yuqing Li +3 moreDOI ↗ (opens in a new tab)

An Original Field Experiment for Positioning of the Seafloor Acoustic Beacon Based on the Ranging Measurement from Surface-Navigation AUV (opens in a new tab)

This study tested a method to precisely locate seafloor acoustic beacons using an autonomous underwater vehicle (AUV) instead of costly surface vessel surveys. The experiment, conducted at 2,400 meters depth in the South China Sea, used two-way travel-time data and a robust estimation technique to determine beacon coordinates. Missing data (about 40%) affected accuracy, but simulated complete data improved 3D positioning precision from 1.817 meters to 1.127 meters. Results suggest AUVs offer a cost-effective alternative for underwater positioning in marine surveys.

09openalex2026-07-22NOAA Ocean ExplorationDOI ↗ (opens in a new tab)

Okeanos Explorer (EX2502): 2025 ROV and Mapping Shakedown (opens in a new tab)

NOAA Ocean Exploration conducted the 2025 ROV and Mapping Shakedown (EX2502) from March 17–27, 2025, near Hawai'i to test ROV, mapping, and sampling systems aboard the *Okeanos Explorer*. The expedition completed 8 ROV dives at depths of 322–4,683 meters, totaling 32.5 hours of bottom time and collecting 41 samples. Taxonomic identifications were provided by the University of Hawai'i, while biological, geological, and eDNA samples were archived at the Smithsonian and the University of Rhode Island. Data and sample records are publicly available through NOAA’s National Centers for Environmental Information and other repositories.

10openalex2026-07-21Pierluigi Carbonara, Carmen Ferrà, L. Marsaglia, Rigers Bakiu +8 moreDOI ↗ (opens in a new tab)

Mapping vulnerable marine ecosystems to support conservation measures for Isidella elongata in the Southern Adriatic Sea (opens in a new tab)

This study mapped the distribution and biomass of the bamboo coral *Isidella elongata* in the Southern Adriatic Sea to support conservation efforts. Using trawl surveys, ROV observations, and environmental data, researchers found the coral thrives at depths of 500–650 meters, with its presence influenced by seafloor slope and fishing pressure. Areas with high suitability but low biomass were linked to degradation from bottom trawling. The results guided the design of a Fisheries Restricted Area in the Otranto Channel to protect these vulnerable ecosystems.

11openalex2026-07-21Yuan Gui, Hongchen Luo, Liqi Qu, Longyue Fu +1 more

NaviAIS: A Scenario-Level Vessel Trajectory Prediction Dataset withVectorized Lane Priors and the NaviLane Forecasting Framework (opens in a new tab)

The paper introduces **NaviAIS**, a standardized dataset for vessel trajectory prediction, addressing limitations in existing AIS datasets like inconsistent sampling, noise, and lack of structured navigational data. It provides unified temporal windows, local coordinate systems, vectorized lane priors, and rasterized navigable maps to improve reproducibility and environment-aware forecasting. The authors also propose **NaviLane**, a hierarchical framework that encodes trajectory-map interactions, generates multimodal predictions, and refines them using a consequence-aware evaluator. Experiments demonstrate NaviLane’s superior performance over baseline methods in both single-modal and multimodal prediction tasks.

© 2026 Ship Autonomy · Summaries are AI-generated; articles belong to their publishers · Privacy · Contact