Marine accident learning with Fuzzy Cognitive Maps : a method to model and weight human-related contributing factors into maritime accidents

Navas de Maya, Beatriz and Kurt, R. E. (2022) Marine accident learning with Fuzzy Cognitive Maps : a method to model and weight human-related contributing factors into maritime accidents. Ships and Offshore Structures, 17 (3). pp. 555-563. ISSN 1754-212X (https://doi.org/10.1080/17445302.2020.1843843)

[thumbnail of Navas-de-Maya-Kurt-SOS-2020-Marine-accident-learning-with-Fuzzy-Cognitive-Maps]
Preview
 Text. Filename: Navas_de_Maya_Kurt_SOS_2020_Marine_accident_learning_with_Fuzzy_Cognitive_Maps.pdf
Accepted Author Manuscript
Download (1MB)| Preview

Abstract

Previous statistical maritime accident studies are focused on identifying human factors. However, the previous studies were not capable of modelling the complex interrelations that exist between these factors. As accidents are complex processes, researchers fail to agree on the contribution of each human factor. Therefore, in this research study, a new Fuzzy Cognitive Map (FCM)-based technique known as MALFCMs has been introduced and applied. Its novelty is the application of FCM concepts to model the relationships of accident contributors by combining historic accident data with expert opinion. Our approach is capable of integrating information obtained from real occurrences, therefore, the results can be considered more objective. Thus, in this paper, MALFCMs was applied to grounding/stranding accidents in general-cargo vessels, revealing that ‘unprofessional behavior’, ‘lack of training’, and ‘inadequate leadership and supervision’ are the most critical factors, with a normalised importance weighting of 13.25%, 13.24%, and 13.24% respectively.

ORCID iDs

Navas de Maya, Beatriz ORCID logoORCID: https://orcid.org/0000-0002-3595-9401 and Kurt, R. E. ORCID logoORCID: https://orcid.org/0000-0002-5923-0703;
CORE (COnnecting REpositories)