Aug 2026· Journal of Marine Science and Engineering· Vol 14, pp. 1496· 0 citations· 31 references
TL;DR
These findings demonstrate that H3-indexed context, structured at multiple geographic resolutions and integrated through a selective mechanism, serves as transferable spatial context for vessel trajectory prediction.
Abstract
Deep learning-based vessel trajectory prediction using Automatic Identification System (AIS) has become a hot topic in the fields of maritime traffic monitoring, situational awareness, and navigational decision support. However, most previous studies have focused primarily on end-to-end model training using trajectory data from a single water area, which limits the resulting models’ ability to generalize to regions with different traffic patterns. To address this issue, this study proposes a method that constructs traffic context from historical AIS records at multiple geographic resolutions using H3, a hexagonal hierarchical spatial indexing system, and integrates this context with a Transformer-based trajectory predictor. A reliability-aware selector determines the contribution of the context to the final prediction, conditioning this decision on the vessel’s motion state and the retrieved historical patterns. Experiments on AIS data from three distinct water areas demonstrated that H3-indexed context improved cross-water prediction accuracy without requiring model retraining on the target area. These findings demonstrate that H3-indexed context, structured at multiple geographic resolutions and integrated through a selective mechanism, serves as transferable spatial context for vessel trajectory prediction.
It is suggested that multi-scale local motion modelling can stably improve the accuracy of AIS-based vessel trajectory prediction and is evaluated under a unified data preprocessing, resampling and multi-step autoregressive prediction framework.
Qi Xu, Hua-Sheng Nong, Tianwei Ma et al.· International Conference on...· 0 citations
Vessel trajectory prediction is critical for maritime safety and accident prevention. While most existing trajectory prediction models rely on Automatic Identification System (AIS) data due to its precision and availability, small vessels mostly operate without AIS, resulting in a significant monitoring gap. To address this, we propose Mix&Fix-Net, a dual-stage mixer-based trajectory prediction model designed to handle vessel trajectory time-series data derived from both AIS and (non-AIS) vision data. Our architecture integrates a Primary Trajectory Predictor with a Residual Trajectory Adjuster, enabling more refined trajectory prediction. Additionally, we introduce a new video-based dataset derived from webcam streams, from which vessel trajectories are extracted to represent non-AIS data. Extensive evaluations on both AIS and non-AIS datasets across six metrics (mean squared error, mean absolute error, symmetric mean absolute percentage error, final displacement error, Frechet distance, and average Euclidean distance) demonstrate that Mix&Fix-Net consistently outperforms existing baselines across most metrics and datasets.
M. Murad, Bora San Turgut, Yasin Yilmaz· Ocean Engineering· 0 citations
Accurate short-term vessel trajectory prediction is important for traffic monitoring and collision-risk screening in port-approach waters, where vessels frequently turn, accelerate, decelerate, and merge into traffic lanes. This study develops a maneuver-aware residual multi-scale long short-term memory (LSTM) framework for Automatic Identification System (AIS)-based 10–30 min trajectory prediction. The method predicts residual displacement relative to the last observed position, constructs maneuver-aware features from local displacement, course variation, speed variation, turning rate, and acceleration-like terms, and compares fixed and maneuver-guided multi-scale fusion strategies. Experiments are conducted on public AIS data from San Francisco Bay and adjacent approach waters using Maritime Mobile Service Identity (MMSI)-level train/validation/test splits and three random seeds. The largest observed gains come from residual prediction and maneuver-aware features. In the three-seed main evaluation, the fixed multi-scale LSTM (Fixed-MS-LSTM) provides the strongest 10 min accuracy, while the maneuver-guided multi-scale residual LSTM (MGMS-RLSTM) achieves lower average displacement error (ADE) at 20 and 30 min and learns distinct temporal-scale preferences across straight, turning, and speed-changing samples. Encounter-oriented closest point of approach (CPA) and time to closest point of approach (TCPA) evaluation further shows that the residual multi-scale models support more accurate CPA/TCPA-based high-risk screening under the evaluated benchmark. These findings indicate that maneuver-guided fusion can be characterized as a horizon-dependent scale-selection mechanism that complements the fixed multi-scale counterpart.
Xinyue Lin, Xiao-Han Zhang, Wendong Bao et al.· Journal of Marine Science an...· 0 citations
With the growing scale and number of vessels, inland waterway traffic environments have become more complex, especially in bridge waterways where vessel-bridge collisions occur frequently. To enhance navigational safety, this paper proposes a reliable early warning framework based on spatiotemporal trajectory prediction and anomaly detection. We utilize Automatic Identification System (AIS) data and construct a trajectory prediction model that integrates a Multi-Head Attention mechanism with a Long Short-Term Memory (LSTM) network. A collaborative optimization strategy is adopted to fine-tune the model’s hyperparameters, significantly enhancing its performance on complex spatiotemporal sequences. To address the challenge of identifying abnormal vessel trajectories, we design an enhanced autoencoder network that couples spatial encoding with dynamic behavior modeling to effectively extract latent trajectory features. By incorporating Dynamic Time Warping (DTW) and time series clustering, the model further enables unsupervised anomaly detection and classification. Furthermore, typical abnormal navigation patterns in bridge waterways are simulated using the full-mission ship maneuvering simulator, generating high-quality abnormal trajectory data to improve the model’s generalization and early warning capability. Experimental results demonstrate that the proposed method achieves excellent performance in both trajectory prediction and anomaly detection. It offers a reliable and practical solution for intelligent vessel navigation and proactive risk warning in complex inland bridge environments.
Jing-Xin Cao, Yuan-Zhou Zheng, Long Qian et al.· Scientific Reports· 0 citations
Predicting the future trajectories of surrounding vehicles in autonomous driving is important for collision risk assessment and safe ego-vehicle path planning. Conventional neural network-based trajectory predictors typically achieve strong prediction performance by exploiting agent history, dynamic scene graphs, and semantic maps. However, in specific motion regimes such as acceleration, deceleration, and turning, these predictors may fail to reflect physically feasible trajectories. To address this issue, this study proposes a framework that fuses the output of Trajectron++, a neural network-based trajectory predictor, with extended Kalman filter (EKF)-based multiple trajectory candidates at a late stage. On the nuScenes dataset, the proposed method reduces the average displacement error and final displacement error of the Trajectron++ robot baseline by 13.7% and 14.6%, respectively, without modifying the baseline architecture. These results indicate that EKF-based trajectory candidates can effectively complement neural trajectory prediction through learned fusion.
Seong-Jun Kim, Seung-Hyun Kong· Journal of Institute of Cont...· 0 citations
Abnormal ship behavior detection is important for maritime traffic surveillance, navigation safety, and risk prevention. However, existing methods often depend on handcrafted features or a single reconstruction or prediction signal, which limits their ability to detect both sustained trajectory abnormalities and abrupt vessel movement changes. This paper proposes a Dual-Error Fusion LSTM–Transformer framework, referred to as DEFLT, for AIS-based abnormal ship behavior detection. A motion-aware vessel representation was first constructed by combining the geographical position, speed over ground, course over ground, and their temporal variations. An LSTM autoencoder reconstructs historical trajectory windows, while a Transformer prediction module estimates subsequent vessel states. The standardized reconstruction and prediction errors are fused into a unified anomaly score to capture complementary evidence from historical trajectory inconsistency and unexpected future motion. Experiments were conducted using real-world AIS data collected during September 2019 from four representative Danish waters. The study considers four abnormal behaviors: speed anomalies, course anomalies, loitering, and route deviations. Compared with KNN, LOF, Isolation Forest, Random Forest, the LSTM-AE, and the Transformer, DEFLT achieves F1-scores of 0.96, 0.97, 0.88, and 0.93 across the four study areas. For type-specific detection, the Macro-F1 values range from 0.61 to 0.86, while Macro-Recall remains between 0.88 and 0.96. Friedman and post hoc Wilcoxon signed-rank tests further demonstrate that DEFLT provides a significant and consistent improvement over all baseline methods. These results verify the effectiveness of dual-error fusion for detecting heterogeneous abnormal ship behaviors from AIS trajectories. In operational settings, DEFLT can serve as an alert-prioritization tool for vessel traffic services and port authorities by directing attention to atypical trajectories that require timely review, thereby supporting safer and more resource-efficient maritime traffic coordination.
