Maritime Traffic Risk Analysis in the Northern Baltic Sea from AIS Data

Thumbnail Image
Journal Title
Journal ISSN
Volume Title
School of Engineering | Doctoral thesis (article-based) | Defence date: 2021-10-01
Degree programme
Aalto University publication series DOCTORAL DISSERTATIONS, 120/2021
Non-accident critical events, such as near misses, detected from AIS data have been widely utilized as a basis to assess the maritime traffic risk, thereby helping to achieve a safe maritime transport system. However, this method originated from other fields, such as road and aviation transportation, and its validity has not been fully verified. Therefore, the direct utilization of non-accident critical events as a basis to correctly understand the maritime traffic risk remains challenging. Further, increasingly crowded ship traffic also poses new challenges to these non-accident critical event-based methods. More attention needs to be paid to violations of Inter-national Regulations for Preventing Collisions at Sea (COLREGs) and multi-vessel encounters. The primary aim of this thesis is to improve maritime traffic risk analysis to support decision making for the prevention of and response to collision risk from the traffic management perspective, with a focus on advancing the latest methodology of utilizing non-accident critical events detected from AIS data as the basis to assess traffic risk. To this end, a review and analysis of these related works is firstly conducted to understand the state-of-the-art of this non-accident critical events-based methodology. The analysis shows the feasibility and challenges of using it as a basis for maritime traffic risk analysis. This thesis seeks to improve response to these identified challenges, including the inadequate consideration paid to the dynamic nature of ship manoeuvres, the utilization of the only ship attributes to detect near misses and the inadequate attention paid to multi-vessel encounters.Therefore, this thesis proposes a framework for near miss detection from AIS data based on ship manoeuvre characteristics. The implementation of this framework is based on the understanding of the process of a navigator formulating and executing her manoeuvre strategy for collision avoidance. This framework enables the consideration of the dynamic nature of ship manoeuvres and the impact of four contributing factors on ship manoeuvres. The framework is applicable for both ship-pair encounters and multi-vessel encounters and is well integrated with the rules in the COLREGs. Under the guidance of this framework, methods for analysing the collision risk represented by ship manoeuvres are developed, and then applied to assess traffic risk in the Northern Baltic Sea. The dangerous waters with high risks of ship collisions in the Northern Baltic Sea are identified and found to be highly consistent with those dealt with in other similar works in terms of the occurrence position and frequency/possibility of dangerous encounters. The results also call for more attention to be paid to the waterway crossing between Stockholm and Turku as many serious encounters were detected there. In addition, the results identify traffic complexity as one of the direct causes of serious encounters. The findings of this thesis contribute to the development of an intelligent traffic supervision system and decision support system to enhance maritime traffic safety.
Defence is held on 1.10.2021 12:00 – 14:30 Via online at Zoom
Supervising professor
Valdez Banda, Osiris A., Prof., Aalto University, Department of Mechanical Engineering, Finland
Thesis advisor
Kujala, Pentti, Prof., Aalto University, Finland
Goerlandt, Floris, Prof., Dalhousie University, Canada

near miss, ship manoeuvre, AIS data, COLREGs, Northern Baltic Sea
Other note
  • [Publication 1]: Du, Lei; Goerlandt, Floris; Kujala, Pentti. 2020. Review and analysis of methods for assessing maritime waterway risk based on non-accident critical events detected from AIS data. Reliability Engineering & System Safety, 200, 106933.
    Full text in Acris/Aaltodoc:
    DOI: 10.1016/j.ress.2020.106933 View at publisher
  • [Publication 2]: Du, Lei; Valdez Banda, Osiris A; Goerlandt, Floris; Kujala, Pentti; Zhang, Weibin. 2021. Improving Near Miss Detection in Maritime Traffic in the Northern Baltic Sea from AIS Data. Journal of Marine Science and En- gineering, 9(2), 180.
    Full text in Acris/Aaltodoc:
    DOI: 10.3390/jmse9020180 View at publisher
  • [Publication 3]: Du, Lei; Goerlandt, Floris; Valdez Banda, Osiris A; Huang, Yamin; Wen, Yuanqiao; Kujala, Pentti. 2020. Improving stand-on ship's situational awareness by estimating the intention of the give-way ship. Ocean Engi- neering, 201, 107110.
    Full text in Acris/Aaltodoc:
    DOI: 10.1016/j.oceaneng.2020.107110 View at publisher
  • [Publication 4]: Du, Lei; Valdez Banda, Osiris A; Goerlandt, Floris; Huang, Yamin; Kujala, Pentti. 2020. A COLREG-compliant ship collision alert system for stand-on vessels. Ocean Engineering, 218, 107866.
    Full text in Acris/Aaltodoc:
    DOI: 10.1016/j.oceaneng.2020.107866 View at publisher
  • [Publication 5]: Du, Lei; Valdez Banda, Osiris A; Huang, Yamin; Goerlandt, Floris; Kujala, Pentti; Zhang, Weibin. 2021. An empirical ship domain based on evasive maneuver and perceived collision risk. Reliability Engineering & System Safety, 107752.
    Full text in Acris/Aaltodoc:
    DOI: 10.1016/j.ress.2021.107752 View at publisher