2024 RT Amination Banner.gif

China Daily

News> Business> Content
Thursday, October 29, 2020, 10:42
How fish-recognition tech is assisting demand for canned tuna
By Bloomberg
Thursday, October 29, 2020, 10:42 By Bloomberg

This undated file photo shows an electronic monitoring camera on a fishing vessel. (PHOTO / BLOOMBERG)

The pandemic is forcing marine protection observers to adopt technology that monitors fishing boats remotely instead of getting on the vessels and risking infection.

Commercial fishing fleets are facing a jump in demand for canned tuna, but the coronavirus outbreak has prevented industry watchdogs and environmental groups from sending people onto boats to monitor whether the catches are sustainable. Traditionally, those observers spend months on vessels collecting data and watching for illegal activity.

Some vessels are installing video cameras, sensors and systems that use algorithms to detect different types of fish and marine life, similar to the way Facebook Inc identifies people tagged in photos

Instead, some vessels are installing video cameras, sensors and systems that use algorithms to detect different types of fish and marine life, similar to the way Facebook Inc identifies people tagged in photos, said Mark Zimring, large scale fisheries program director at The Nature Conservancy, a US-based environmental nonprofit organization.

The goal is to make sure boats don’t misreport the contents and volumes of their catches and ensure at-risk species like turtles and sharks are safely released when they’re caught by accident. Satellite imagery, machine-learning tools and artificial intelligence are also used to detect practices such as illegal shark-finning and labor abuses.

Ship owners must agree to install the equipment and use it to show they are adhering to rules designed to stop illegal, unreported, or unregulated fishing. Allowing the extra scrutiny means vessels can show they are meeting environmental requirements of governments and major buyers, whose customers are increasingly demanding proof that their food is sustainably produced. Access to the data also helps fishing fleets become more efficient.

Much of the remote monitoring requires cooperation from the crew. For example, fishers may be required to bring sharks into the view of the camera before releasing them. The systems have alarms to prevent tampering.

“We’re really starting to see the momentum shift around electronic monitoring and Covid is a key contributor to that,” Zimring said in a video interview from California. “Electronic monitoring can’t get sick, it can’t get your crew sick, and it can’t be corrupted or threatened.”

This undated photo shows unloading frozen yellowfin tuna in Victoria on Mahe Island, Seychelles. (PHOTO / BLOOMBERG)

Monitoring systems cost between US$14,000 to US$24,000 a year per vessel, with about 40 percent to 60 percent of the money spent on analyzing the data and footage collected. That puts them well out of reach of the hundreds of thousands of artisanal fishers in developing nations like Indonesia, which together account for a major share of the global catch. Currently, fewer than 2,000 out of 100,000 large-scale vessels globally are equipped with the systems.

Satellite imagery, machine-learning tools and artificial intelligence are also used to detect practices such as illegal shark-finning and labor abuses

“We are starting in large-scale fisheries because they can better bare the financial and transaction costs,” said Zimring at the Nature Conservancy. As consumer demand for sustainable fishing grows and more large-scale fleets adopt the technology, the costs will fall, making it more widely available. For local artisanal fishing boats, cheaper, phone-based apps are being tried.

In Indonesia, the group is working with fishers using “FishFace,” a project that combines cheap satellite tracking and smartphone cameras to photograph the catch.

Governments are playing a part in pushing for sustainable fishing. The Federated States of Micronesia, which has the world’s 14th-largest exclusive economic zone in the seas around its Pacific islands, committed to 100 percent on-the-water monitoring in 2018. The Seychelles made a similar promise this year, joining countries such as Australia, Chile and New Zealand.

The first generation of machine learning and AI tools can identify the presence and absence of fish on deck. The next stage will be to identify catch volumes and fish species.

The pandemic is forcing regulators to limit the number of people on boats and at ports. The Western and Central Pacific Fisheries Commission, which oversees waters where two-thirds of the world’s tuna are caught, in April suspended its requirement for observers to be on board some ships that engage in “purse-seine fishing,” a method that involves hanging nets from the boat.

About a third of global fisheries are over-fished, while 60 percent are fully-fished, United Nations Food and Agriculture Organization data show. That’s threatening marine ecosystems that are already grappling with global warming, pollution and acidification. Global Fishing Watch estimates that more than US$23 billion worth of seafood is taken from the ocean each year through illegal, unreported and unregulated fishing.

The lack of comprehensive monitoring data means that over 95 percent of illegal activity in fisheries takes place in licensed fleets, Zimring said. In the Pacific, where about a billion hooks are dropped into the ocean by long-line fishers each year, only 2 percent of tuna fishing is monitored, he said.

“You need much more granular information on what’s happening on these boats to have confidence that seafood products coming off from these boats were harvested legally, sustainably and without labor issues,” Zimring said. “The conversation around electronic monitoring is starting to shift from ‘this is impossible,’ to ‘this is inevitable.’”

Share this story

CHINA DAILY
HONG KONG NEWS
OPEN
Please click in the upper right corner to open it in your browser !