The Tembé tribe from the central Amazon were looking for new ways to curb illegal logging of their rainforest. Loggers enter the Tembé land under the cover of night and can decimate hundreds of acres of forest without being detected, as the ever-present sounds of the rainforest drown out the noise from the loggers’ chainsaws and trucks. So the Tembé teamed up with an environmental non-profit to transform recycled cell phones into solar-powered acoustic monitoring devices that are hidden in trees and listen to sounds of the rainforest 24/7. Google’s open-source machine learning tool TensorFlow is then used to analyze the sound data and identify chainsaws and logging trucks the human ear might otherwise miss. Armed with this new technology, the Tembé hope to save their forest and with it, their entire way of life.

The Tembé tribe are one of approximately 180 indigenous groups that live in and off the Amazon. The fight to protect their forest from illegal loggers, who make millions selling and distributing trees as timber, is a David vs Goliath battle that has broader implications for the entire planet. Illegal logging and deforestation releases more Co2 than all of transportation - planes train, and cars combined. Deforestation is also the second highest contributor to climate change.