New AI system detects animals on roads to prevent accidents

Brazilian scientists are developing a system that uses artificial intelligence (AI) to detect and warn drivers about animals ahead, reducing dangerous collisions.

Imagine you’re driving down a scenic highway. Lush rainforests surround you, the sun is shining, and you’re enjoying the road trip. But then, an unexpected obstacle appears – a tapir crosses the road.

Animal-involved car accidents are a serious problem, causing injury and death to humans while also devastating wildlife populations. In Brazil alone, an estimated 475 million animals are killed on roads each year. This research could drastically reduce the numbers.

How does the AI detect animals?

The core of the system relies on strategically placed roadside cameras that provide a continuous video feed. This video feed is then analyzed in real-time by specialized AI software.

The AI “brain”

The AI software has been meticulously trained on a massive dataset of images featuring Brazilian wildlife, specifically species like tapirs, wolves, and anteaters. This training process enables the AI to accurately recognize these animals within the video footage.

Detection and alert

As the AI analyzes the camera feed, it constantly scans for the presence of the target animals. If the AI detects a potential animal match, it triggers an immediate alert.

Driver notification

These alerts are designed to be delivered directly to the driver’s smartphone or integrated into the car’s dashboard display system. This method provides a clear and timely warning, giving the driver valuable seconds to react to the potential hazard.

The importance of speed

The entire process, from image capture to alert delivery, is optimized for speed. This is crucial because even a few extra seconds can make a significant difference in giving the driver enough time to brake, swerve safely, or take other necessary actions to avoid a collision.

Developing the animal detection AI

Brazilian researchers, backed by the São Paulo Research Foundation (FAPESP), are leading this project.

“The species were selected in accordance with the metrics recommended by the Brazilian Center for Road Ecology Studies [CBEE, at the Federal University of Lavras, UFLA]. According to this center’s estimates, some 475 million animals are killed on Brazilian roads every year,” said study lead author Gabriel Souto Ferrante.

“We created a database of Brazilian species and trained a number of computer vision models to detect them.”

Building the system involved collecting over 1,800 publicly available images of the Brazilian mammals most at risk on roads. The researchers used these images to train an AI algorithm called YOLO (“You Only Look Once”).

Promising results

YOLO detects objects fast, making it perfect for quickly spotting animals. The system is designed to run on simple portable computers, eliminating the need for expensive equipment on roads.

“The models correctly detected the species in 80% of images taken during the day with the animal appearing clearly,” said Ferrante.

The researchers acknowledge ongoing challenges like poor nighttime animal detection, but see this as a significant first step.

The future of road safety

The team plans to expand their database of animal images using forest camera traps and actual roadside cameras. They’ve already tested the system with videos of animals in a São Paulo ecological park.

“Drivers also run a significant risk in collisions with large animals. They often have insufficient time to take action to avoid them. Our system uses roadside cameras coupled to a portable computer and is innovative in that respect,” said Professor Rodolfo Ipolito Meneguette, who led the research team.

The experts plan to partner with toll road companies and cities to test the system in real-life settings and potentially integrate it with existing traffic apps.

Potential impact of AI for animal detection

This AI system to detect different animals offers a promising solution to a complex problem – animal-related car accidents. Its success could lead to significant improvements in road safety for both drivers and wildlife populations in Brazil.

Enhanced driver safety

By providing real-time alerts about animals on the road, the system empowers drivers with crucial information. These warnings give them valuable extra reaction time to avoid collisions with potentially dangerous wildlife. This translates to a reduced risk of accidents, injuries, and even fatalities for drivers.

Wildlife protection

Animal-vehicle collisions are a major threat to Brazilian wildlife, causing the deaths of millions of animals each year. This technology has the potential to significantly reduce this number. By giving drivers time to react and avoid collisions, the system can help protect a wide range of animals, from large mammals like tapirs to smaller species like foxes and birds.

Reduced ecological impact

Animal-vehicle collisions can have a ripple effect on ecosystems. The loss of large herbivores, for instance, can disrupt food chains and plant communities. By mitigating these collisions, the system can contribute to a healthier and more balanced ecological environment.

Economic benefits of AI animal detection

Animal-related accidents can also be costly, leading to vehicle damage, injuries, and even fatalities. Reducing the number of these accidents can lead to significant economic benefits for individuals, insurance companies, and society as a a whole.

The overall impact of this technology could be substantial. It has the potential to create safer roads for drivers, protect vulnerable wildlife populations, and contribute to a healthier environment in Brazil.

The work is published in the journal Scientific Reports.

—–

Like what you read? Subscribe to our newsletter for engaging articles, exclusive content, and the latest updates. 

Check us out on EarthSnap, a free app brought to you by Eric Ralls and Earth.com.

—–