The impacts of risk level based on PSC inspection deficiencies on ship accident consequences

Abstract

Global economic growth has stimulated the growth of international trade, leading to more maritime transport activities. However, in a complex and high-risk environment at sea, heavy ship accidents are difficult to eliminate. In order to prevent ship accidents more effectively, this study constructs a Bayesian network model to investigate the impact of various factors on the ship's risk level and accident consequences. To improve the reliability of the model, it proposes two hidden variables, the ship risk level and the PSC inspection level, to the model. The risk level variable represents the overall risk status of a ship. The PSC inspection level variable is used to aggregate complex PSC inspection results. Through analysis, it can not only identify the key factors affecting ship risk level, but also analyze the impact of the risk level on the consequences of the accident. The data used in this study is obtained from three databases, Lloyd's Register of Shipping (LR), International Maritime Organization (IMO) and Tokyo Memorandum (MoU). The model structure is learned through the Greedy thick thinning (GTT) algorithm, and it is evaluated using K-fold cross validation, log-likelihood function (LL), and Akaike Information Criterion (AIC). It also uses an overall sensitivity analysis to verify the validity of the model. The results of the model show that ship age, ship type and the PSC inspection level have the most significant impact on ship risk level. It also validates the effectiveness of PSC inspections on accident prevention.