Development of anti-siphoning model by automatic identification system for marine security.

Abstract

Illegal ship-to-ship (STS) fuel siphoning is a prevalent yet under-reported phenomenon due to national security confidentiality. Modeling of such scenario lacks the robustness of validating the occurrence for its own simple reason of being illegal. Nevertheless, a framework for siphoning detection is required to assist in curbing these illegal activities to preserve the rights and profit of stakeholders. This research paper is proposing a near proximity framework to identify potential illegal STS fuel siphoning outside the port region using automatic identification system (AIS) data along Straits of Malacca and Singapore. Suggested criteria on the characteristics of siphoning process are based on proximity between two vessels and total duration within the proximity. The proximity model is developed in Python to process the AIS data collected, and filters are applied according to the decide criteria with potential level of illegal siphoning. From 50 ship samples of AIS data analyzed, 18.2, 47.9 and 46.7% of the total cases are simulated with high possibility of illegal STS fuel siphoning for proximity category less than 50 m, 100 m, and 150 m, respectively, with more than 10 h duration. Enhancement techniques to the conceptual framework to identify potential illegal STS activities are also proposed in the event without on-site validation.