According to international law, large vessels must broadcast their locations using an onboard automatic identification system (AIS), which can be used to monitor illegal activity. A recent study used AIS to track reefers, to try and detect IUU fishing. When fishing watchdog see that large vessels turn off their AIS, this might be a signal of suspicious activity.


Now Jessica Ford, a research statistician at Australia’s Commonwealth Scientific and Industrial Research Organisation and lead author of the Australian study, proposed a new way to detect areas where IUU fishing might be occurring. This will be done by tracking the movements of bunker vessels. Ms. Ford noted that bunkers are more connected to other ships than reefers, and as they are also less, they are easier to follow.

The idea to track bunker vessels came up when the statistician was examining AIS signals in the Arafura Sea, between Indonesia, Papua New Guinea, and northern Australia. She saw that some of the bunkers there were moving strangely. Some months after that, reports were published regarding illegal fishing in the locations the ships Ms. Ford had seen.

The, the team chose an area in the southern Indian Ocean and analyzed all AIS signals from ships traffic during six months. There were 181 vessels in the data, but it was difficult to notice interactions between ships.

Ford then used a method called social network analysis to discover patterns within the AIS tracks. This created a picture of how different individuals are connected to one another. The more connections someone has, the more central they are to the network. In fact, Ford’s analysis concluded that one bunker vessel scored highest across various different ways of measuring centrality, and then discovered the most connected vessel in the network.

Social network analysis can also detect which fishing ships are related with certain tankers or reefers, which might be of assistance for authorities to track down the IUU criminal networks.

See more information on the full report below