Application of bayesian network in the maritime industry: Comprehensive literature review

Abstract

Solving uncertainty-related tasks such as accident investigation, risk analysis, reliability prediction, and port sustainability analysis has been a key focus in the maritime industry. Over the years, conventional tools such as fault tree analysis, event tree analysis, and statistical-driven methods have been used for accident, risk, and reliability analyses in the maritime industry, which is considered weak. For this reason, the application of the Bayesian network (BN) is attracting growing interest in the maritime industry. A comprehensive review of the past 20 years on the application of BN in the maritime industry is provided. To conduct the review, relevant publications involving 115 journal articles focusing on BN application in the maritime industry were extracted. Based on the extracted studies, a classification framework involving application areas, data sources, operational waters, geographical locations, and model validation was proposed. The proposed framework was used to analyze the literature in detail to reveal the necessary findings and extract possible future research areas. The findings of the research revealed an increasing trend in the application of BN in the maritime industry in diverse areas of risk assessment, human error analysis, reliability estimation, and accident investigation. Furthermore, issues emanating from the systematic review analysis are discussed.