A group of American computer engineers and biologists developed the EEAGER algorithm to find beaver dams using satellite images of wild areas. This information could help better manage water resources in drought-affected areas and even stop wildfires. reports Wired version.
The author of the idea to create EEAGER was former Google employee Eddie Corwin, who was interested in sustainable real estate development. While working on water regulation, he became interested in beavers and their dams, which can hold millions of liters of water and prevent floods.
Corwin, along with his friend and colleague Dan Akerstein, began designing a system that would automatically map the infrastructure created by beavers.
During their research, the experts turned to beaver expert Emily Fairfax from the University of Minnesota. Fairfax has studied how beaver dams help other animals escape during wildfires and sometimes even stop flames by soaking the area.
A team of scientists and programmers trained the EEAGER model on 13,000 satellite images with beaver dams and 56,000 landscape images without beaver ponds. The neural network learned to detect rodent structures with 98.5% accuracy.
Conservation officials and Google plan to test EEAGER in California in 2024 and use it to locate all beaver dams and ponds. This will give state officials information on exactly where beavers live, what their numbers are, and what areas they could be introduced to to improve local wildlife suffering from droughts and fires.
A beaver before managed We leave 14 cities in Canada without internet.