New frontiers in the risk assessment of ship collision. (15th April 2023)

• Record Type: Journal Article
• Authors: Marino, Massimiliano; Cavallaro, Luca; Castro, Elisa; Musumeci, Rosaria Ester; Martignoni, Matteo; Roman, Federico; Foti, Enrico
• Is Part Of: Ocean engineering. Volume 274(2023)
• Publication Date: 2023-04-15
• Subjects: Maritime safety -- Navigation risk -- Collision avoidance -- Neural network -- Bayesian network -- Swarm intelligence; Ocean engineering -- Periodicals; Ocean engineering; Periodicals; 620.4162
• DOI: 10.1016/j.oceaneng.2023.113999
• Languages: English
• ISSNs: 0029-8018

Abstract

Abstract: Navigation is becoming more and more complex over the years. The increase in maritime traffic and vessel size is inducing a global escalation of ship collision accidents, with consequent losses of human lives and economic assets worth billions. This is particularly true for port basins, with maritime authorities struggling worldwide to keep up with the ever-increasing ship traffic. In this respect, the demand for advanced methods to assess and mitigate ship collision risk has never been higher. The interdependency between physical failures, weather conditions, logistics, governance and human factors requires sophisticated frameworks to effectively assist maritime authorities and navigators in decision-making. The present work reviews the most recent advancements in the risk assessment of ship collision. The article focuses on new, rising technologies, identifying the current main trends and discussing future perspectives and challenges. The review revealed a wide and diversified range of methods, including machine learning, clustering techniques, swarm intelligence algorithms and others. To frame the methods in the current literature and compare them with previous efforts, they are categorized according to literature classifications. Advancements of well-established approaches and new promising tools are discussed, considering methods that allow the inclusion of quantitative and qualitative variables in the assessment. Furthermore, a comprehensive analysis of a database of maritime accidents in port areas is carried out to investigate prevailing trends in both worldwide and Mediterranean Sea contexts. Results indicate that ship collision accidents constitute the majority compared with other types of accidents, especially in the Mediterranean. Highlights: A literature review of new methodologies to assess risk of ship collision is carried out. The review discusses a wide and diversified range of technologies, including machine learning, clustering techniques, swarm intelligence algorithms and others. An analysis on a maritime accidents database is carried out, highlighting worldwide and Mediterranean trends.