© Craig Taylor
10/21/2020 7:30:15 PM
Early Detection of Polar Bears
By Barbara Nielsen, Director of Communications
People who live and work in the Arctic tend to have polar bears on their personal radar, especially during seasons when the bears are forced ashore by melting ice, away from their seal prey. As the ice-free periods become longer and polar bears spend more time on shore, wildlife managers and scientists expect conflicts with people to increase.
BJ Kirschhoffer, director of field operations for Polar Bears International, talked with us about exciting new technology that could reduce dangerous encounters by detecting polar bears before they arrive in town.
Q: Tell us about the radar system. How did you learn about it and why are you excited about it?
A: The idea came about after a series of conversations with a radar company in Provo, Utah that makes a compact radar surveillance device called the SpotterRF®. Devices like these were designed to be used by the military to warn of nearby threats like drones, people, and vehicles, but they’re starting to have applications for wildlife, too.
The SpotterRF is able to withstand extreme weather, so we decided to test whether it could detect polar bears and serve as an early warning system for Arctic communities. If wildlife managers can spot a polar bear as it approaches a town or workstation, they’ll be able to use less-lethal deterrents to drive the bear away—preventing harm to the bear and to people.
Q: Where are you testing the system?
A: We decided that Churchill, Manitoba, Canada, the polar bear capital of the world, would be the perfect test site for several reasons. First, polar bears gather near Churchill every fall as they wait for the sea ice to return. So, we knew we would reliably have polar bears to test there. Second, we already have vast infrastructure in place in Churchill to power our Polar Bear Cams and Tundra Connections® webcasts. Knowing we’d be able to piggyback on that—both the wireless and bandwidth—was a big plus. And, finally, the town of Churchill supported the idea and welcomed our team.
Q: Churchill already has a robust Polar Bear Alert Program. How does the SpotterRF fit into that?
A: Manitoba’s conservation officers do an amazing job of spotting approaching bears and responding to reports from citizens, but human patrols can be limited by darkness, foggy weather, or whiteout snowstorms. If it proves successful, the SpotterRF could complement the efforts of the Polar Bear Alert officers. And it could help other northern communities that don’t have the resources for human patrols.
Q: Is the radar system up and running yet?
A: Yes, we first installed it in the fall of 2017 in Churchill, working with NMS Security, a Native corporation from Alaska that has a security division with certified installers. They came up along with some folks from Utah’s Hogle Zoo, which is helping to fund the project. Having infrastructure already in place was a huge help, and Polar Bears International staff laid the groundwork ahead of time to make the installation quick and easy.
For that first phase, we mounted the SpotterRF on Churchill’s community center, which includes the school and hospital, along with seasonal installations on the Tundra Buggy Lodge and at Cape Churchill. All overlook known polar bear movement corridors. We were able to monitor the radar stream from Polar Bears International’s office in Bozeman, Montana, and staff from the town of Churchill were able to log in and monitor too.
Since that early pilot phase, we have moved the radar system to Cape Churchill year-round and are planning a second installation this year at Polar Bear Point (where the Tundra Buggy Lodge is usually based) to further train the system to identify bear targets.
Q: Have you detected any polar bears yet?
A: Yes, we have! And we’re fine-tuning the radar’s artificial intelligence (AI) system to help it tell the difference between a vehicle, dog, person, or other wildlife, like caribou—training it to make decisions based on the size, speed, and location of the target. And if an object never moves, the system learns that it’s an inanimate object, like a big rock.
When the radar detects a moving object, it triggers an alert. We look at to see if it’s a non-threatening dog, bird, or human—and tag those different options as we see them.
Our team will work with partners to monitor the system and mark when polar bears are located, feeding this data into the AI system and preparing it for deployment as a fully functional polar bear alert tool. Our goal is to fine-tune the AI to correctly identify polar bears, triggering an alert before they hit the first road or the edge of town.
We will also be using combined radar data and direct observations to document an unusually quiet year in the Wildlife Management Area outside of Churchill, which will see a small fraction of the usual visitation by people this year.
Q: Is there anything else you’d like to add about the project?
A: We’ve had a lot of interest in the radar system from wildlife managers in northern communities. And beyond the community safety aspect, there’s a pretty interesting science story. The radar collects geospatial data, location, and speed—data that could potentially help with other studies.
Special thanks to Utah’s Hogle Zoo for funding the pilot program, to NMS Security for help with installation, and to Milestone Systems, which analyzes the livestreams through its surveillance system. We’re also grateful to the town of Churchill for their cooperation during the test phase of the project, to Frontiers North Adventures for allowing us to mount a device on the Tundra Buggy Lodge, and to Amy Cocksedge of York University for making the system the focus of her master’s thesis.