Today@Dal

» Go to news main

Media Highlight: Dalhousie scientists play big role in Leonardo DiCaprio‑funded app

Posted by Communications and Marketing on September 22, 2016 in Media Highlights

Dalhousie University researchers have played an important part in bringing to life a 'revolutionary' new app that allows anyone with internet to monitor and track commercial fishing activity around the world, as well as potentially identify illegal fishing.

Unveiled last week, Global Fishing Watch is a joint project between Google, digital mapping non-profit SkyTruth and ocean conservation group Oceana. It's funded by the Leonardo DiCaprio Foundation.

The app allows users to view a world map with over 35,000 major fishing vessels moving in real time.

Engineers behind Global Fishing Watch collaborated with researchers at Dalhousie University in the process of developing new ways to identify and hone in on fishing vessel activity, said Jacqueline Savitz, vice-president for U.S. and Global Fishing Watch at Oceana.

Global Fishing Watch will also eventually use algorithms developed at Dalhousie, to provide a more complete picture of fishing activity on the high seas.

Dalhousie scientists have spent about two years developing technology that uses satellite data to monitor and track commercial fishing.

International maritime law requires modern ships around the world have automated tracking systems that send satellites data on identification, position, course, and speed.

Biology PhD student Kristina Boerder recognized the potential of using that satellite data to see how fishing vessels behave around marine protected areas and approached her advisors with the idea.

"You can track your favourite football player, your spouse or your kids with your smart phone, so there's a lot of tracking happening, but that's kind of ending at the water line," Boerder said.

"So it's hard to regulate and it's hard to put it in a context, for example, for conservation actions and for fish stock management."

Dal biologists worked with computer scientists at the university, including Canada Research Chair Stan Matwin.

Matwin said he and team member Erico N. de Souza figured out how to feed satellite data on fishing activity into newly developed machine-learning algorithms.

Machine learning is a subfield of computer science that pools existing data to make predictions.

He said those algorithms can determine which of three types of fishing a ship is involved in (trawling, long-lining or purse seining), whether a ship is fishing or not, and if a ship is moving in an abnormal way or not.

Using that information, it would be possible to distinguish illegal, unreported and unregulated fishing activity.

Read full story (http://www.cbc.ca/news/canada/nova-scotia/dalhousie-scientists-fishing-tracking-technology-1.3768578)