Estimating the risk of drifting ships with the use of Bayesian networks

Kurzfassung

It is a common procedure for vessels waiting for mooring in the port or for the next order to anchor at roadsteads close to a harbor. Incidents like the drifting of the MS Glory Amsterdam in 2017 show the risk that can evolve from the roadsteads. In the morning of the 29th of October 2017 started to drift from its anchoring position towards the German coast, even though two anchors were dropped. The vessel stranded close to the island Langeoog, with damage to the vessel itself but not the environment or cargo lost. Therefore, this paper aims to develop a tool to estimate the risk of drifting for a specific ship. To do so, risk influencing factors that a vessel will start to drift were identified. The focus is on vessel, environmental and human & organizational risk influencing factors. Based on that, a Bayesian network has been developed. The Bayesian network determines the risk of the specific vessel on a three-part scale (low, medium, high). The Bayesian network can be used together with a drift estimation of the considered vessel as a decision support system.