April 24, 2024

How AI is being used to prevent illegal fishing

  • By Emma Woollacott
  • Business reporter

Image source, Getty Images

Image subtitle,

One in every five fish caught worldwide is done through illegal or unregulated fishing

At the end of last year, the Spanish government fined 25 Spanish-flagged fishing vessels operating near Argentine territorial waters.

The financial sanctions were imposed because the vessels illegally turned off the GPS-based automatic identification systems (AIS) that transmitted their positions. This is often a sign that a vessel is involved in illegal fishing.

“Illegal or unregulated fishing is estimated to account for up to 20% of what is caught,” says David Kroodsma, head of the Global Fishing Watch research team.

This illegal, unreported and unregulated (IUU) fishing can capture up to 26 million tons of fish every year, or one in every five fish, according to the Food and Agriculture Organization of the United Nations. He says that in financial terms this black market is worth up to 23 billion dollars (£18 billion).

This contributes significantly to overfishing, with the UN adding that a third of global fish stocks are now being fished beyond biologically sustainable levels. For example, it is estimated that the bluefin tuna population represents only 2.6% of its historical unfished size.

Image source, David Kroodsma

Image subtitle,

David Kroodsma says illegal fishing needs to be tackled more seriously

“Sharks and rays are on the brink of extinction,” says Kroodsma. “There are threats of extinction of seabirds and turtles due to accidental capture. It’s really sad, because we could manage the oceans much better.”

Global Fishing Watch is a partnership between Google, marine conservation body Oceana and environmental group Skywatch. The latter studies satellite images to detect environmental damage.

To try to better monitor and quantify the problem of overfishing, Global Fishing Watch is now using increasingly sophisticated AI software and satellite imagery to globally map the movements of more than 65,000 commercial fishing vessels, both with and without – AIS.

AI analyzes millions of gigabytes of satellite images to detect vessels and offshore infrastructure. It then analyzes publicly accessible data from ships’ AIS signals and combines it with radar and optical images to identify ships that are unable to transmit their positions.

And while not all vessels are legally required to use AIS, the AI ​​and its “fishing detection algorithm” aims to discover which of these “dark” vessels are most likely to be involved in fishing.

“We use information like the length of the vessel, as well as environmental information about where the vessel is located, the image of the area, the density of vessel traffic in the area, the state of the ocean such as the temperature – a lot of information about where the vessel is operating,” says Fernando Paolo, senior remote sensing machine learning engineer at Global Fishing Watch.

“And this model infers whether the ship is likely to be a fishing vessel or not, like passenger ships, oil tankers, transport ships and that kind of thing.”

Researchers found that three-quarters of the world’s industrial fishing vessels are not publicly monitored, with specific hotspots in Africa and South Asia.

Global Fishing Watch is now working to introduce higher resolution images to help detect more smaller fishing vessels.

Image source, Global Fisheries Observation

Image subtitle,

Global Fishing Watch creates a map where the darker the orange, the more fishing in an area is not publicly tracked

However, eradicating illegal fishing means getting data as quickly as possible – and a project from the University of Southampton and local company RS Aqua aims to do just that.

The team is building an underwater robot that uses underwater sensors and AI to identify the sound of fishing and transmit the information in real time. AI can already differentiate the natural sounds of the ocean and is now being trained to identify the noise of trawlers operating in protected waters.

“Part of the motivation for the project is to try to implement something that helps monitor fishing activity within protected marine areas [MPAs]”, says Paul White, professor of statistical signal processing at the University of Southampton.

“There is a definite concern that by creating these ecosystems with higher densities of fish, they will be attractive to fishermen.”

There are currently around 15,000 MPAs around the world, representing 8% of the world’s oceans. However, one study says that less than half are actually “fully protected” from fishing.

White says the goal is that if the robot detects fishing, it will automatically contact authorities who can send a patrol vessel.

He adds: “We haven’t yet decided exactly what form of AI we are going to use. Part of the problem is that for this to work, you have to stay in the ocean for many months. launch something in a week and then have to go back and change the batteries.

“It has to be very energy efficient, which means you can’t have very complicated algorithms running, but they need to be powerful enough to distinguish something like the noise from a trawler from other sources of noise in the ocean.”

Image source, University of Southampton

Image subtitle,

The University of Southampton’s underwater robot, pictured, is being trained to recognize the sound of fishing

Paul Lansbergen, president of the International Fisheries Coalition, says illegal, unreported and unregulated (IUU) fishing “is a scourge of the industry.”

“The sustainability impacts and economic harms are very real, but they are also no longer an issue that lingers in the shadows. Industry leaders, policymakers and associated stakeholders are all focused on the challenge.

“Emerging technologies like AI complement traditional law enforcement and put IUU fishing in the crosshairs. But consumers also need to make sustainable choices and make things tougher for IUU fishermen. We all play a role.”

Leave a Reply

Your email address will not be published. Required fields are marked *