Quantitative assessment of ship collision risk influencing factors from worldwide accident and fleet data

The paper assesses the contribution of specific Risk Influencing Factors towards the occurrence of ship collision accidents. The approach adopted is inspired by the Cox proportional hazard regression model and combines a Bayesian rule and the least-squares method to assess the influence of risk factors on the collision probability based on historical data of collision accidents worldwide and information of the world's merchant fleet. For the quantitative assessment, a global sample involving 936 collision events from the Global Integrated Shipping Information System database of the International Maritime Organization for the period 2005-2017 is used. Additionally, world fleet statistics are analysed and compared to collision accidents statistics. Based on these data sets, six Risk Influencing Factors are assessed with a direct impact on the collision probability, namely Ship length, Ship type, age of the ship, Classification Society, flag and geographical area. The results show that the Risk Influ-encing Factors having the greater impact on ship collisions are the ship type and the geographical area.

Automated summary

The paper evaluates the contribution of specific Risk Influencing Factors to ship collision accidents using a combination of Cox proportional hazard regression model, Bayesian rule, and least-squares method. Historical data of collision accidents worldwide and information of the world's merchant fleet are used for the assessment. Based on a global sample of 936 collision events from 2005 to 2017, six Risk Influencing Factors are assessed, including Ship length, Ship type, age of the ship, Classification Society, flag, and geographical area. The results indicate that ship type and geographical area have the greatest impact on ship collisions.