A group of tribes in Brazil finding ways to put barricade on deforestation with old mobile devices.
The Amazon in the largest tropical rain forest in the world, also known in English as Amazonia or the Amazon Jungle, home to a quarter of the world’s bio-diversity.
According to the source 20% of the Amazon Jungle has been destroyed by deforestation from 1970.
In northern Brazil indigenous Tembé people live on 2,800 km2 of rain forest. Over 30% of their territory has been deforested by cattle ranching, fires, and illegal logging.
The Tembé people are well organized, well educated; they are not afraid to use new tools and new technologies. They are looking for collaboration. They’re not looking for help.
In 2014, Chief Naldo reached out to Topher White, founder of environmental nonprofit Rain forest Connection, and together they embarked on an ambitious project using recycled Android phones and TensorFlow, Google’s open-source machine learning model, to track the sounds of illegal logging in real time.
How recycled cell phones and machine learning help the Tembé protect their homeland.
An Android phone is affixed to a solar power adapter and external microphone. These devices, nicknamed Guardians, can hear the sounds of illegal logging up to 1 kilometer away.
The Guardians are hidden high up in trees for better cell service and access to sunlight for power. They listen to all the sounds of the forest around the clock.
Rainforest Connection’s TensorFlow model uses machine learning to analyze the audio recorded by the Guardians, and learns to identify the sounds of chainsaws and logging trucks.
Within minutes of an identification, a real-time alert is sent to the Tembé rangers, a select security force of villagers who can intervene or report the logging activity to the authorities.
Rainforest Connection is putting this same acoustic monitoring system into the hands of other partners fighting deforestation in five different countries, including Peru, Ecuador, and Romania.