AI App for Salish Orca Conservation

There are few iconic species in British Columbia as popular or as sacred as the Salish Sea Orca. With their numbers dwindling and only an estimated 73 left in the Salish Sea, which extends from the Southern Coast of British Columbia into Washington’s Puget Sound, Big Tech is joining with the Fisheries and Oceans Canada to apply AI to track, monitor, and observe the orcas remaining in the Salish Sea. This partnership between Google AI, DFO, and Rainforest Connection was formed in 2019, and the nearly 20,000 hours of sub-sea audio kept by the DFO helped teach Google’s deep neural network how to identify the distinct sources of ocean sounds, including orcas themselves. After the neural network learned how to pick up the orca’s sounds, it started to receive audio recordings from 12 hydrophones in real-time located throughout the Salish Sea (north of the U.S. border). Ideally, the neural network will be able to generate information about ocean conditions, nearby prey (like salmon), and track the location of orca pods. An app developed by Rainforest connection is intended to generate alerts about the location of whales and any potential risks, like a speeding boat or oil spill, that officials could act on to deliver help or mitigate harm. It isn’t entirely clear what the status of this project is or if the app is already being used, but the project is relatively new and will hopefully bring a new level of ecological consciousness to regulators, the public, and the people who traverse the waters and value the presence of Salish orcas.


Artificial Intelligence, Biodiversity, Data, Internet of Things, Monitoring, Psychology, Regulation