Traffic-conflict-based modeling of collision risk in port waters

Abstract

This research aims at exploring the use of non-collision data in modeling collision risk in port waters. Traffic conflicts are innovatively proposed as an alternative to the collisions, which are used by the traditional models. The use of conflicts in risk modeling is explored by developing mathematical models of risk measurement and prediction. A risk measurement model is developed that quantitatively measures collision risk in individual interactions, statistically characterizes the risks collectively and obtains risks of collision in waterways by identifying the interactions with high potential of collision. For prediction of risks, a binomial logistic model is formulated that explains the relationships between the risks and waterway characteristics by accounting for the correlations in risks at different time periods in waterways. The modeling technique is illustrated for Singapore port waterways. This research proves that collision risk can be evaluated in a fast, reliable and proactive manner by using conflicts.