Survey on Multi-Source Data Based Application and Exploitation Toward Smart Ship Navigation

Xiaodong Tang, Jie Zhou, Shujie Hou, Yang Sun, Kai LuoJournal of Marine Science and Engineering 2025 · DOI

TL;DR

This paper surveys the rapidly evolving field of multi-source data applications for smart ship navigation. It provides a comprehensive framework for leveraging diverse data streams (AIS, radar, imagery, etc.) from acquisition to application, aiming to enhance maritime safety, efficiency, and intelligence. The review highlights key technologies and challenges in areas like trajectory analysis, target recognition, and behavior detection, paving the way for future intelligent shipping systems.

What Problem Does This Paper Address?

Maritime transportation is crucial for global logistics and national security, but it's plagued by accidents, largely attributed to human error. While advancements in sensing and data collection have enriched maritime data, effectively integrating and utilizing this multi-source information for intelligent navigation remains a significant challenge. This paper addresses the need for a systematic review and a unified framework to understand how diverse data streams can be leveraged to improve ship operations, enhance situational awareness, and ultimately reduce risks in maritime navigation.

Methods

The authors conducted a systematic literature review, focusing on research from 2020-2025, drawing from major academic databases. Their methodology involved defining research questions, identifying keywords, constructing search strings, and then meticulously selecting, appraising, and synthesizing relevant papers. The review categorizes approaches based on the data layer, application layer, and system layer, analyzing techniques for data acquisition, preprocessing, fusion, and application. Key areas covered include:

  • Data Acquisition & Pre-processing: Examining mainstream maritime data types like AIS, optical/infrared/SAR imagery, and radar, along with their processing techniques.
  • Data Fusion: Reviewing methods for integrating heterogeneous data sources to enhance navigation capabilities.
  • Application Layer Analysis: Delving into typical scenarios such as:
  • Trajectory Extraction & Prediction: Methods for understanding and forecasting vessel movements.
  • Target Recognition: Identifying ships and other objects using various sensors.
  • Vessel Behavior Detection: Recognizing normal and abnormal ship behaviors.
  • Path Planning & Collision Avoidance: Developing strategies for safe and efficient navigation.

Datasets

The paper references several datasets commonly used in maritime research, including:

  • SSDD (Ship Detection Dataset)
  • HRSID (High-Resolution SAR Images Dataset)
  • SeaShips
  • SAR-Ship Dataset

The authors also note that self-collected data is frequently used. The review compares these datasets based on modality, source, resolution, preprocessing, and split, highlighting HRSID as a popular choice for SAR imagery in ship recognition.

Key Insights and Ideas

  • The "Human Factor" is Still Key: Despite technological advancements, human error remains the primary driver of maritime accidents, underscoring the need for intelligent systems to augment human decision-making.
  • Multi-Source Data is the Future: The integration of diverse data streams (AIS, radar, imagery, etc.) is not just beneficial but essential for achieving comprehensive maritime situational awareness and enabling smart navigation.
  • Large Models are Emerging: The paper touches upon the growing influence of large neural network models ("large models") in areas like target recognition and trajectory prediction, hinting at their potential to unlock new capabilities in maritime AI.
  • Data Imputation is Crucial: Incomplete or corrupted AIS data is a significant issue, necessitating robust imputation techniques to ensure the reliability of navigation systems.
  • Fusion is Complex but Powerful: Effectively fusing data from different sensors (e.g., AIS with VHF, LiDAR with radar) presents challenges but offers substantial improvements in accuracy and capability for tasks like berthing and surveillance.
  • Deep Learning Dominates: The review consistently points to deep learning methods, particularly LSTMs and CNNs, as the leading approaches for trajectory prediction, target recognition, and behavior analysis.

Key Results

The paper is a survey and framework construction, rather than a presentation of novel experimental results. However, it synthesizes findings from numerous studies, indicating:

  • Improved Accuracy: Various multi-source data fusion and deep learning approaches demonstrate significant improvements in accuracy for tasks like ship detection (e.g., accuracies in the high 90s for some models) and trajectory prediction.
  • Enhanced Capabilities: The integration of multi-source data enables more sophisticated applications, including real-time traffic situation awareness, abnormal behavior detection, and more reliable collision avoidance.
  • Efficiency Gains: While some advanced methods require significant computational resources, there's a clear trend towards developing lightweight and efficient models for edge deployment.

Limitations and Future Work

The authors acknowledge several limitations and suggest future research directions:

  • Parameter and Density Sensitivity: Many clustering-based methods (e.g., DBSCAN) are sensitive to parameter choices and data density, impacting their robustness.
  • Cross-Domain Transferability: Models often require significant retraining or reconstruction when applied to different maritime regions or environments.
  • Inconsistent Benchmarks: A lack of standardized benchmarks and metrics across studies makes direct comparison of performance challenging.
  • Computational Cost: Target recognition and some deep learning models still face high computational costs and long pre-training times.
  • Lightweight and Real-Time Models: There's a strong need for developing easy-to-deploy models with low computational cost and fast response times, especially for shipboard embedded platforms.
  • Multi-Sensor Fusion: Further research is needed to improve the processing speed and accuracy of individual data sources while promoting cooperative use across different data types.
  • Maritime Environment Perception: Achieving a more comprehensive perception of the maritime environment and ship status through the joint analysis of multiple heterogeneous sensors is a key future direction.
  • Data-Driven Knowledge: Exploring how to express water traffic situation knowledge based on AIS data is also an area for further study.

Summary updated

Abstract

Maritime ship transportation is not only the core infrastructure of the global logistics system but also is closely related to national security and sustainable development. However, the human factor remains the primary source of risk leading to maritime accidents during ship navigation. In recent years, multi-source data has been recognized as an important means to improve the efficiency of ship operations and navigation safety. In this paper, the major research methods and technical pathways of maritime multi-source data in recent years have been systematically reviewed, and a comprehensive technical framework from data acquisition and preprocessing to practical application has been constructed. Focusing on the data layer, application layer, and system layer, this paper comprehensively analyzes the key technologies of maritime navigation based on multi-source data. At the same time, this paper also highlights the advantages and cutting-edge methods of multi-source data in typical application scenarios—such as track extraction, target recognition, behavior detection, path planning, and collision avoidance—and analyzes their performance and adaptation strategies in different usage contexts. Through the combination of theory and engineering practice, this paper looks forward to the future development of ship intelligence and water transportation systems, providing a theoretical basis and technical support for the construction of intelligent shipping systems.
BibTeX
@article{tang2025survey,
  author = {Xiaodong Tang and Jie Zhou and Shujie Hou and Yang Sun and Kai Luo},
  title = {{Survey on Multi-Source Data Based Application and Exploitation Toward Smart Ship Navigation}},
  journal = {Journal of Marine Science and Engineering},
  year = {2025},
  doi = {10.3390/jmse13101852},
}
Figure