Swiss researchers have developed software enabling drones to autonomously detect and follow paths in forests with the aim of finding people lost in forests.
Professor Davide Scaramuzza is Director of the Robotics and Perception Group at the University of Zurich. He said the biggest challenge in using drones in heavily forested areas is teaching them how to follow a path.
“While drones flying at high altitudes are already being used commercially, drones cannot yet fly autonomously in complex environments, such as dense forests," he said.
"In these environments, any little error may result in a crash, and robots need a powerful brain in order to make sense of the complex world around them.”
The drone they used observes the environment through a pair of small cameras, similar to those used in smartphones. It then uses powerful artificial-intelligence algorithms to interpret the images to recognize man-made trails and steer in the corresponding direction.
They achieved this using a deep neural network, an algorithm that learns to solve complex tasks from a set of 'training examples'.
In order to gather enough data to train their algorithms, the team hiked several hours along different trails in the Swiss Alps and took more than 20 thousand images of these trails using cameras attached to a helmet.
When tested on a new, previously unseen trail, the deep neural network was able to find the correct direction in 85% of cases; in comparison, humans faced with the same task guessed correctly 82% of the time.
Scaramazza says there is still a lot of work to be done before a fully autonomous fleet will be able to swarm forests in search of missing people.
“Now that our drones have learned to recognise and follow forest trails, we must teach them to recognise humans,” he said.
However, he feels with the pace of development in this field, it is only a matter of time before real-world application such as this are a reality.
The research was conducted jointly by the University of Zurich, the Università della Svizzera italiana, and the University of Applied Sciences and Arts of Southern Switzerland.
A drone detects a forest path and follows it autonomously. Photo: UZH; USI; SUPSI