Solar-powered edge devices in Costa Rican rainforests now identify monkeys and other wildlife in real time, replacing tracking systems that often left researchers waiting 30 days to retrieve camera data.
Computer-vision models discard empty frames locally, while acoustic AI isolates primate calls and other sounds to map animal activity without capturing or tagging them.
Real-time alerts let rangers respond faster to threats: sensors can flag chainsaws or vehicle engines alongside primate distress calls, helping curb illegal logging and poaching in protected reserves.
On the Osa Peninsula, the same systems show how monkey troops use rope wildlife bridges, giving conservation groups evidence on where to direct habitat-restoration funding.
The technology still faces rainforest-specific hurdles—humidity, power failures, moss-covered solar panels and model bias from zoo-trained data—so researchers must tune systems with messy field footage.