When AI Meets Conservation

Setting up a network of camera traps in the thick jungles of Gabon isn’t easy. Reaching the cameras just to gather SD cards and replace batteries can take days of trekking in the humid, harsh and hazardous terrain of these forests. There’s the threat of contracting malaria or an injury, with no easy access to hospitals. And there's also the risk of stumbling across poachers and wildlife like elephants and gorillas. That’s why Okala, a company leveraging tech for land management, have been developing smart camera traps and testing them out in a forest concession not far from Gabon’s capital.

Their cameras are equipped with a mini-computer and a satellite connection that, with the help of artificial intelligence, can send them a notification about which species passed by. That system is powered by Machine Learning (ML), a branch of AI that's really good at doing repetitive tasks and doesn't need sleep or rest like humans. Okala’s team of researchers has spent months training their AI to identify elephants, gorillas, guineafowl, duikers and more. But it’s not just the real-time biodiversity data that’s making an impact for these researchers. Their system also tells them when a camera is down, or running low on battery, taking out a lot of the guess-work that is conventionally required when researchers deploy camera traps. These efforts can help minimize the amount of time spent in the field, and increase the amount of time spent on scientific analysis. It can also be used in anti-poaching and anti-logging efforts.

Our documentary follows the story of conservationists using AI for their work observing endangered animal species in Gabon. They use modified camera traps that have AI on-board to gather and analyze information.