Human Factors Analysis for Maritime Accidents Based on a Dynamic Fuzzy Bayesian Network

Area of Science:

• Maritime safety
• Human factors analysis
• Risk assessment

Background:

• Human factors are critical in maritime accident prevention system failures.
• Conventional quantitative methods for human factor assessment are static and struggle with uncertainty.
• Existing models limit comprehensive human factors risk analysis in maritime contexts.

Purpose of the Study:

• Introduce a novel dynamic human factor analysis framework: the Multidimensional Analysis Model of Accident Causes (MAMAC).
• Integrate Human Factors Analysis and Classification System (HFACS) and Business Process Management (BPM) with intuitionistic fuzzy set theory and Bayesian Networks.
• Develop a flexible model capable of handling uncertainty in human factors risk analysis for maritime accidents.

Main Methods:

• Developed the Multidimensional Analysis Model of Accident Causes (MAMAC).
• Integrated Human Factors Analysis and Classification System (HFACS), Business Process Management (BPM), intuitionistic fuzzy set theory, and Bayesian Networks.
• Applied the MAMAC model to a Chinese sand carrier accident database.

Main Results:

• Identified the top 10 primary events contributing to sand carrier accidents due to human factors.
• Found that direct human factors (unsafe acts) receive less focus than unsafe preconditions and supervision failures.
• Highlighted supervision failures by shipping companies and ship owners as critical.

Conclusions:

• The proposed MAMAC model offers a flexible and effective approach to dynamic human factors analysis in maritime safety.
• Unsafe preconditions and supervision failures are more significant contributors to maritime accidents than previously emphasized.
• The model successfully calculates accident probabilities, demonstrating its practical utility for safety countermeasures.