Acaté launches AI tool for Indigenous-led biodiversity monitoring in Peru

Acaté Amazon Conservation has developed SentaAI, a machine-learning system that blends Indigenous ecological knowledge with camera traps and acoustic sensors in Peru’s Matsés rainforest. The project aims to fill a major gap in tropical forest monitoring by tracking wildlife and ecosystem health beneath the canopy.

Why it matters: - Tropical forest monitoring has long focused on tree cover loss, while biodiversity changes under the canopy have remained hard to measure. - SentaAI gives Matsés communities a new way to track animal populations and ecological health in their ancestral rainforest. - Nearly half of the world’s tropical forests are protected within Indigenous lands, making locally led monitoring a potentially important conservation model.

What happened: - Acaté Amazon Conservation developed SentaAI, a machine-learning tool for continuous biodiversity monitoring in the Matsés ancestral rainforest in Peru. - The system combines Indigenous ecological expertise with camera traps and bioacoustic monitors. - Senta is the Matsés Indigenous name for the red uakari monkey, an iconic and endangered species. - The launch positions local Indigenous rainforest experts at the center of day-to-day biodiversity monitoring.

The details: - The Matsés identify more than 47 types of rainforest habitats, a level of ecological detail that Western science does not fully match. - Matsés knowledge helps determine where to place camera traps and bioacoustic recorders to capture elusive Amazonian species. - SentaAI processes large volumes of field data to deliver image recognition for Amazonian fauna. - The system combines publicly available recognition systems with a custom machine-learning workflow developed over two years. - SentaAI is now being trained to identify bird, amphibian and mammal vocalizations recorded in the Matsés rainforest. - Satellite platforms can detect forest cover loss, but they provide little information on biodiversity trends or declines in ecological health beneath the canopy. - Rapid species inventories by scientists remain resource-intensive and are usually limited to baseline assessments.

Between the lines: - The project is as much about governance as technology, because it shifts monitoring power toward communities that already understand the forest at a finer scale than outside teams. - The approach suggests AI can extend local expertise rather than replace it, especially in remote ecosystems where conventional surveys are expensive and infrequent. - The initiative also points to a gap in conservation infrastructure: forest protection efforts are growing faster than tools for measuring what happens to wildlife inside protected forests.

What's next: - SentaAI will continue training on vocalizations from birds, amphibians and mammals. - The system is expected to expand the scope and speed of on-the-ground biodiversity monitoring in Matsés territory. - Acaté’s model could be adapted for other Indigenous-led conservation areas across tropical forests.

The bottom line: - SentaAI turns Indigenous knowledge and AI into a practical conservation tool for tracking biodiversity where satellites cannot.