In one episode of the YouTube series “The Age of AI” released last January, Intel visited the Maasai Mara Reserve in Kenya. In this episode, Intel works together with the Mara Elephant Project, and RESOLVE NGO. Their goal was to use Intel’s technology to help catch and stop poachers that kill more than 35 thousand African elephants every year.
The Mara Elephant Project
This non-profit was created with the mission to conserve the Mara ecosystem by protecting elephants. They put devices on the elephants to track them through the Maasai Mara and therefore be able to protect them better. Also, they engage with the community to teach them sustainable economic practices that don’t include elephant poaching.
However, one of the problems they faced was how to detect poacher’s attacks fast enough to stop them. They used traditional cameras that took photos of everything that moves to try and identify the poachers. But It was a common occurrence that, when the park rangers were notified of an attack, the elephant was already dead. Sometimes the carcass was a few days old by the time the rangers identified a poacher attack’s image.
Intel and RESOLVE TrailGuard AI technology
To achieve the goal of protecting the Mara ecosystem, the non-profit organization partnered with Intel and RESOLVE to create the TrailGuard device. The TrailGuard uses AI technology to know when a human is detected with the camera and send a signal to the park rangers.
The TrailGuard device uses a tiny Intel’s Movidius vision processing unit (VPU) along with a very powerful computer chip. But how does it work? The device uses machine learning to analyze all the images caught by the camera. The AI algorithm analyzes body shape, facial geometry, movement, and other features to differentiate between humans and animals.
Then, they put the device in strategic zones (common elephant trails) and test them at night. The device worked perfectly and identified every human that was caught by the camera. And the best part, not only did the device identified all the humans but it sent a signal in mere seconds. This will allow the rangers to act fast enough to save the animals and even catch the poachers.
This technology has so many applications that will improve our world, and this is just the beginning. By saving endangered species to predict earthquakes, it will solve many problems that humans could not in a lifetime. You can check out the video below for other incredible examples of machine learning and AI technology.
YouTube: Saving the world one algorithm at a time
Photo: All the photos belong to Intel Corp. and were provided for press usage. Intel also indicates credit to Walden Kirsch and RESOLVE for the images taken.
Source: The Mara Elephant Project / Intel
