Revolutionizing Road Safety: How Artificial Intelligence is Helping to Save Endangered Species in Brazil

A new project in Rio de Janeiro is harnessing artificial intelligence technology to reduce wildlife collisions on roads in Brazil. According to the Brazilian Center for Studies in Road Ecology (CBEE), over 475 million vertebrate animals die each year due to collisions with vehicles in the country, with species such as the capybara, armadillo, and opossum being among the most affected.

Gabriel Souto Ferrante, a student at the University of Sao Paulo, has developed a computer vision system that detects animal presence on roads and alerts drivers. The project is a collaboration between USP’s Institute of Mathematical and Computer Sciences and aims to protect endangered species such as the puma, giant anteater, tapir, maned wolf, and jaguarundi.

To make this technology effective, it requires support from companies managing roads. Access to cameras along routes and safety equipment will enable real-time alerts to drivers and road management companies. Additionally, it could improve road safety for humans who are also at risk from animal crossings.

While Brazil has implemented strategies such as ecological corridors and green bridges to reduce habitat fragmentation caused by roads, these measures are not enough given the country’s vast size. Traditional road signs warning of animal presence have limited impact on driver behavior. More extensive efforts are needed to protect wildlife while ensuring safe movement for both animals and humans.

The Urubú System, a citizen science social network created by CBEE coordinator Alex Bager, has played an essential role in collecting data on wildlife roadkill in Brazil. This information has been used to create public policies and legislation protecting animals despite financial challenges. Efforts continue to reactivate the platform address growing concerns about roadkill in the country.