Fujitsu in collaboration with the Maritime and Port Authority of Singapore (MPA) has in analyzing marine traffic risks in the Singapore Strait, demonstrated the usefulness of a novel ship collision risk prediction technology that leverages the power of Artificial Intelligence (AI) in predicting near misses between vessels.
Using AI technology “Fujitsu Human Centric AI Zinrai” developed by Fujitsu Laboratories Ltd, this solution can detect ship collision risks and predict areas where collision risks are concentrated as dynamic risk hotspots.
This technology has the potential to be deployed in a Vessel Traffic Service (VTS) system to assist maritime traffic controllers to manage marine traffic proactively with the aim of improving the safety of navigation.
Based on the results of this development, Fujitsu will continue working with Singapore to improve the technology, and deliver a solution supporting safe navigation for maritime traffic control and for ship operators by 2020.
Background
Globally, especially in busy waterways, VTS centers that manage marine traffic rely on anti-collision warning in VTS systems, which uses closest point of approach techniques, to detect and notify ships when vessels get unusually close to one another.
As vessel operations and interactions become more complex, the ability to detect and predict vessel movements in advance, especially in high density vessel traffic areas like Singapore, is key to managing and reducing collision risks.
To this end, Fujitsu and Fujitsu Laboratories have conducted research and development to calculate ship collision risks and predict dynamic hotspots.
About the research and testing
Fujitsu conducted its research and testing with the assistance of about 10 MPA officers from the Vessel Traffic Management Department and Port Systems Division over 24 months. Details of the testing and its results are as follows.
1. Testing details
Drawing on past traffic data for the Singapore Strait provided by MPA, Fujitsu leveraged its collision risk prediction technology to extract information such as examples of collisions or near misses involving multiple vessels, as well as examples of developing dynamic risk hotspots. The prediction technology was benchmarked to human operators.
2. Evaluation
Fujitsu’s risk detection technology demonstrated the possibility of quantifying risk in more detail in advance of the detection by human operators.
For example, based on benchmarking studies, the detection technology was able to flag potential risks approximately 10 minutes before the near-miss, and in doing so, would theoretically provide approximately 5 minutes lead time for human operators to execute the necessary follow on action, e.g. through the provision of advice to the vessel(s).
Fujitsu’s technology was also able to pick out scenarios where the collision risk could be overlooked and raise the alert to a possible collision incident.
Separately, the dynamic risk hotspot detection technology could detect risks up to 15 minutes in advance, enabling staff to take specific measures to avoid an incident.
