Normalization of maritime accident data using AIS

Abstract

The Automatic Identification System (AIS), fitted on most ships today, is primarily used for real time ship monitoring. This paper illustrates how stored AIS data can be used to construct activity data that can be used to normalize accident statistics, to turn recorded numbers of accidents into accident rates. We show, by way of some examples, the potential in using AIS to construct different types of activity data, and discuss the advantage of combining measures based on different activity data when monitoring accident trends or trying to identify accident prone types of vessels. The analysis and discussion are based on a combination of the Norwegian database of maritime accidents and 6 years of recorded AIS data. The paper addresses methodological issues regarding the construction and use of these activity measures, demonstrates how they can provide new knowledge both for researchers and authorities and outlines some directions for further research.