By Eric Huang.    Photo：Pixabay 

Normal ships will turn on the Automatic Identification System (AIS) when sailing. This technology was invented as a collision avoidance tool. It reads AIS signals from other ships through the marine radar installed on the ship to provide information such as the position, course and speed of nearby ships. However, when the AIS equipment on the ship is unable to send out signals due to natural disasters, equipment damage or human factors, the marine radar will not be able to scan the position information of the ship, which may cause the ship to be in danger and cannot be rescued, or deliberately avoided surveillance by law enforcement authorities for illegal maritime activities. There are sometimes legitimate reasons for ships to turn off AIS, such as to prevent other competitors from knowing where there are good fishing spots; but there are two cases for potentially malicious reasons, one is fishing in unauthorized locations fish, and the other is to cover up unauthorized transshipment of catches.

In this regard, Oceana, a non-profit marine conservation organization, SkyTruth, a non-profit environmental watchdog, and Google have jointly launched a website - Global Fishing Watch, which successfully developed and published the first real-time global ocean map, which contains never-before-seen Sightings of dark vessels that have been spotted or ships whose location has been purposely concealed or not shown on public surveillance systems. Powered by satellite radar imagery and artificial intelligence machine learning, Global Fishing Watch's live ocean map automatically updates major map apps daily. Available free of charge to anyone with a computer network connection worldwide, the portal empowers authorities, researchers, and the public who alike to monitor vessel activity in all coastal waters, identify dark vessel patterns and build the awareness necessary to quantify the threaten from oceans.

The new global real-time ocean map draws relevant information from a massive database and uses artificial intelligence machine learning to process millions of Gigabytes of radar imagery captured by the European Space Agency's Sentinel-1 satellite. By analyzing the entire archive of Sentinel-1 radar imagery, Global Fishing Watch has been able to successfully detect 20 million ocean-going vessels with over 10 meters in length worldwide and compare these detections to 100 billion GPS location points data from AIS of ships around the world. This comparison distinguishes ships that disclose their positions from those that deliberately remain invisible to public surveillance systems, allowing for a more complete picture of the operations of various ships in the world. This information can help authorities to pinpoint areas of suspicious activity and identify vessel trajectories that may represent illegal activity or fishing patterns never quantified before.

Data from Global Fishing Watch shows that in regions with high rates of illegal, unregulated, and unreported (IUU) fishing, transshipment in offshore area is more common than in highly regulated regions such as North America and Europe. The data also show that collective catch transshipment occurs along some countries' Exclusive Economic Zones (EEZs) and within these zones in countries with high levels of corruption and limited monitoring capabilities. These correlations do not provide evidence of any specific illegality, but they raise serious concerns and may lead to more informed efforts by international fisheries management organizations to prevent or better regulate catch trans-shipment.

By analyzing more than 21 billion GPS points in public AIS messages from ocean-going vessels between 2012 and 2016, the artificial intelligence system developed by Global Fishing Watch can identify refrigerated cargo ships based on their movements. They confirmed the identity of 794 refrigerated cargo ships through the results of verification of confirmed fishery registration information and databases publicly available online. According to the CIA's World Factbook, this figure represents 90% of the globally identifiable refrigerated cargo ships in 2010. Through further analysis, they mapped the trajectories of 5,065 instances of voyages in which a refrigerated cargo ship and a fish ship moved alongside each other at same speed and within a certain distance for a period of time. Represents a potential phenomenon of catch transshipment.

To expand the potential of using satellite radar technology, Global Fishing Watch partnered with Defense Innovation Unit to host the xView3 competition in July 2021. The challenge, which invited AI machine learning developers around the world to develop innovative computer algorithms to help spot dark vessels, attracted 1,900 registrants from 67 countries / areas. Global Fishing Watch is using the winning entries announced earlier this year to refine and advance dark ship detection methods continuously around the world, and hopefully reveal even more unknown human activity in the oceans in the near future.

Appreciate if you could share TGL Blog among your friends who are interested in first-hand market information of supply chain and updated economic incidents.