Ever been on a trek and found out that you have lost your way? At the same time, did you find yourself wishing that you’d left a breadcrumb trail? You may not necessarily have to tap into your inner Hansel due to the efforts of Swiss researchers at University of Zurich, University della Svizzera Italiana and the University of Applied Sciences and Arts of Southern Switzerland. These researchers have developed a software that can be implemented by drones to detect and follow forest paths autonomously. This can prove to be very useful while trying to detect missing persons in forest trails.
The research paper was published in IEEE Robotics and Automation Letters. A group of researchers from the Dalle Molle Institute for Artificial Intelligence has developed the AI for the drones.this AI uses an algorithm known as the Deep Neural Network algorithm. The application of the algorithm proved to be successful. The algorithm was modeled after the human brain and the process of learning.This algorithm learns complex tasks from a set of “training examples”. The team hiked for several hours along different trails on the Swiss Alps and took more than 20 thousand pictures from cameras set up on their helmets. When tested on a new, unseen trail the Deep Neural Network was able to find the correct path in 85% of the cases, whereas humans could find the correct direction in 82% of the cases.
The drones designed by the the Swiss researchers have cameras attached to them and they manage to find the correct path using this powerful AI instead of sensors. The drones observe the environment through the cameras and if a trail is visible, the software will steer the drone in that direction.
Every year there are a lot of people who get lost in the wild throughout the world. In Switzerland alone, 1000 emergency calls come from hikers who have lost their way or have been injured. These drones have proven to be an effective complement to search and rescue teams.
“While drones flying at high altitudes are already being used commercially, drones cannot yet fly autonomously in complex environments, such as dense forests. 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,”
– Prof. Davide Scaramuzza University of Zurich.
Prof. Juergen Schmidhuber, Scientific Director at the Dalle Molle Institute for Artificial Intelligence says:
“Our lab has worked on deep learning in neural networks since the early 1990s when only a few labs were interested in the topic. Today I am happy to find our lab’s methods not only in numerous real-world applications such as speech recognition on your smartphone, but also in lightweight robots such as drones. Robotics will see an explosion of applications of deep neural networks in coming years.”
Watch the Drone in action here: