Evaluation of human error contribution to oil tanker collision using fault tree analysis and modified fuzzy Bayesian Network based CREAM

Human error is a significant factor triggering oil tanker collisions. Drawbacks jeopardizing the effectiveness of human error analysis from traditional studies are spotted. This study proposes a risk assessment of human error contribution to oil tanker collision. The research evaluates the collision probability of oil tanker using a Fault Tree Analysis (FTA) structure under which a modified Fuzzy Bayesian Network based Cognitive Reliability Error Analysis Method (CREAM) is developed to conduct human error assessment. 39 specialists with substantial seagoing experience are invited to provide expert judgement for oil tanker navigation particularly in vicinity of the Taiwanese waters. Different from traditional studies, the proposed methodology provides a higher degree of result distinguishability meanwhile contemplating the weights and quantitative effects caused by the ambient elements without the loss of information from the experts. It is concluded that lack of Bridge Resource Management Communication, lack of Communication between Ships, Fatigue and Collision Regulation Violations are the elements with higher occurrence rates and would also have great potential contributing to oil tanker collision once the likelihoods deteriorated. In addition, Number of Simultaneous Goals in the bridge is the Common Performance Condition (CPC) that significantly causes the occurrence of all the human error basic events.