Identifying Factors of Dynamic Positioning Incidents through Association Rule Mining

Abstract

Accidents in the offshore industry can have severe repercussions for people, cargo, vessels, and the environment, making maritime safety a crucial concern. Dynamic positioning incidents, particularly those involving loss of position, represent a significant risk. This study employs association rule mining to analyze DP incident data, leveraging its strength in discovering robust associations. Using the Apriori algorithm, the analysis identifies strong association rules for loss of position (drift-off, drive-off) and loss of redundancy situations. The findings reveal event-related variables and potential causal relationships, providing insights and guidance for reducing the risk and occurrence of future DP incidents through stringent and targeted safety measures.