Data science project initiated to tackle illegal bunkering
A collaboration between rise-x.io, DNV GL and National University of Singapore’s (NUS) Department of Statistics and Applied Probability is looking to tap on data science to predict illegal bunker activity.
NUS students majoring in Data Science and Analytics will be tasked to create computer models to analyst 10 billion lines of automatic identification system (AIS) data to determine whether illegal bunkering can be detected using vessel pattern analysis.
Global hydrocarbon theft and fraud are estimated to be more than $133bn a year and illegal bunkering is estimated to cost companies and governments up to $3bn a year, according to rise-x.io.
Related: DNV GL develops new integrated online tool for LNG bunkering
Singapore is the world’s largest bunkering port with sales of 47.5m tonnes of bunkers in 2019, down from 49.8m tonnes in 2018.
“The market driving illegal bunkering activities and bunker theft costs the industry billions every year. The quicker we can build solutions to address that issue, the quicker the industry can become cheaper and more sustainable,” said Rowan Fenn, ceo at rise-x.io, a technology start-up firm.
With the implementation of IMO 2020 requiring the use of low sulphur 0.5% fuel on a global basis, the need to clamp down on bunkering malpractices would be higher on the agenda as the price of low sulphur fuels are much higher, with recent low sulphur fuel premiums at slightly above $200 per tonne over high sulphur fuels.
The algorithms produced by the project will be integrated into rise-x.io’s QuayChain platform. “These algorithms will provide users of the platform with unique insights into vessel performance and management that builds trust for vessel owners and operators,” said David Barker, cto at rise-x.io.
Beyond the potential direct integration into QuayChain, the algorithms will be enhanced to deliver alternative outcomes. “The value of this project is how flexible the algorithms can be. For example, modifications will allow us to predict metrics such as fuel consumption and CO2 emissions without installing IoT devices on the vessel’s machinery,” Fenn said.
Providing this information not only gives vessel owners more insights into their vessels’ operational performance, but also open doors for the delivery of carbon neutral voyages.
Nic Sabin, DNV GL’s technical lead on the initiative, commented: “Data science is starting to gain momentum in the maritime industry, but is still relatively nascent compared to other sectors. We therefore see this excellent initiative, driving efficiency improvements while helping to prevent illegal behaviour or honest disputes, as a key driver to improve trust and transparency in the industry.”