According to KMST(Korea Maritime Safety Tribunal)'s marine accidents statistics, 82.8% of the collision accidents occurred in the last 5 years(2010~2014) are caused by not observing the good seamanship, such as maintaining a proper look-out, checking the ship’s position and course regularly.
Especially, most of the collision accidents that occurred during this period in Korea are caused by OoW(Officers of Watch)s' human errors.
In order to decrease these dangers of marine accidents and secure the safety and efficiency of traffic for vessels, VTS(Vessel Traffic Service) has been established all the countries around the world. Fifteen(15) Port VTSs and three(3) Coastal VTSs are operational in Korea.
The VTSO (Vessel Traffic Service Operator) monitors the vessels through the VTS equipment and provides safety information to the ship to prevent collision between vessels.
When VTSO determine the degree of collision risk of two vessels, they consider comprehensive information about each vessel’s course, speed, CPA(Closest Point of Approach), TCPA(Time to the Closest Point of Approach), and encounter situation, among many other considerations.
In this study, we propose a utility function based on risk attitudes of VTSO about the Risk Index(RI) which be calculated with the risk of encounter(Eij), the risk of time(Tij) and the risk of approach(Aij) for two vessels to predict each ship’s collision risks from the VTS viewpoint.
First, we proposed an algorithm in which two relative bearings as information about two vessel’s encountering situation are used to evaluate the risk of encounter(Eij). In order to verify whether or not the proposed algorithm is reasonable, we set up twenty encountering situations and conducted a survey of collision risk on VTSOs and OoWs on those situations. By comparing the degrees of collision risk with the proposed algorithm to the results of the survey, we found that relative bearing at CPA is very important factor for VTSO to recognize the level of collision risk when two vessels are in encountering situation.
And then, we surveyed the degree of risk of VTSO in the spare time for the ship with impending danger of collision by the questionnaire and calculated the risk of time(Tij). Thereafter, we obtained the ship domain with minimum safe distance of the VTSO converted to the length of the ship and calculated the risk of approach(Aij) to move the domains of each vessel at the point of CPA.
Additionally, we confirmed that there is a difference in perception of risk according to individual VTSO even in the same vessel crossing situation. In order to reflect this, the Ship Collision Risk Model(CoRi) associated with different VTSO’s risk attitudes on the value by calculating the RI was obtained by referring to the study which calculated the utility according to the risk attitude of the decision maker.
We obtained each coefficient of the RI and the risk attitude through the survey of collision risks among VTSOs of Korea. In order to verify whether the proposed utility is reasonable, we validated by applying the to some historical cases of accidents in Busan port along with the Ship of ES value(ESs) of ES(Environmental Stress) model.