Search and Rescue
Swiss researchers have managed to teach drones to recognize and follow forest trails—a capability that could be very useful for search and rescue missions in the wild.
The team, composed of members from the Daile Molie Institute for Artificial Intelligence, University of Zurich and NCCR Robotics uses an artificial intelligence software that equips quadcopters with this skill, thus allowing drones to respond to the thousands of calls that come in annually from hikers who have lost their way around trails or have been injured.
“While drones flying at high altitudes are already being used commercially (see Amazon or DHL), 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,” says Prof. Davide Scaramuzza from the University of Zurich.
Check it out in action here:
The drones that Swiss researchers are using rely on AI algorithms instead of sophisticated sensors to maneuver themselves around complex environments, such as forests. It is able to observe its surroundings using small cameras, where the images get interpreted using the AI algorithm so that it recognizes man-made trails.
Using Deep Neural Network, the drones are then able to solve complex tasks based on a set of “training examples”—kind of like how the brain learns from experience.
“In the last eight years we have developed huge Deep Neural Networks (DNN) to solve difficult problems from the fields of biology, automation and document processing. This is our first attempt at creating a small but performant DNN capable of running on a computer on our drone. I am happy to see that the same networks we have used to analyse biological brains, detect cancerous cells and diagnose retinal disorders can also be used to drive autonomous quadcopters,” says Dr. Dan Ciresan at the at the Dalle Molle Institute for Artificial Intelligence.
The drones can be deployed in large numbers to speed up response time and be used in tandem with human teams to promptly locate and respond to emergency situations, but there is still a lot of work ahead.
It will take some time before a fully autonomous fleet will be able to sweep through forests and numerous technological issues will need to be resolved. But with these drones now being able to recognize forest trails, the next step would be to teach them to recognize humans.