New Algorithm Beats Humans in Drone Racing for The First Time
Reading Time: 2 minutes

Researchers at the University of Zurich have developed a new algorithm, it calculates time-optimal trajectories that fully consider the drones’ limitations.

Drones need to be quick and with limited battery life, they need to accomplish the task of searching for survivors on a disaster site, inspect buildings, deliver cargo, etc in a minimal period of time. Making their way through obstacles like windows, rooms, or specific locations to inspect. They also have to adopt the best trajectory and the right acceleration or deceleration at each segment.

Human pilots do have the skills to maneuver drones and so far have been at their best when it came to competing with autonomous systems in drone racing. A new algorithm developed by researchers can find the quickest trajectory to guide a quadrotor through a series of waypoints.

Davide Scaramuzza, who heads the Robotics and Perception Group at the University of Zurich and the Rescue Robotics Grand Challenge of the NCCR Robotics, which funded the research said, “Our drone beat the fastest lap of two world-class human pilots on an experimental race track. The novelty of the algorithm is that it is the first to generate time-optimal trajectories that fully consider the drones’ limitations.”

Earlier research relied on simplifications of either the quadrotor system or the description of the flight path, and thus they were sub-optimal.

According to Philipp Foehn, a Ph.D. student and first author of the paper on the work, “The key idea is, rather than assigning sections of the flight path to specific waypoints, that our algorithm just tells the drone to pass through all waypoints, but not how or when to do that.”

During the research two human pilots flew the same quadrotor through the race circuit. External cameras were deployed to capture the precise motion of the drone in the case of autonomous drones to give real-time information to the algorithm on the drone’s current location.

To have a fair competition, human pilots were allowed to train on the circuit prior to the race. The algorithm was an all-out winner, all its laps were faster than the humans and performed more consistently. The algorithm was able to find the best trajectory and repeat it over and over unlike human pilots.

It is a bit early for algorithms to integrate with commercial applications as it needs to be less computationally demanding. Presently it takes a lot of time to compute the optimal time trajectory. The researchers also plan to replace the external cameras with onboard cameras on the drones in the future.

This can prove to be a major factor when integrating the algorithm into huge applications in package delivery with drones, inspection, search and rescue, and more.

Related Articles:

New Antenna Design Will Withstand the Harsh Environment on Europa
New Computer Vision and Deep Learning Approaches for Improving Degraded Images
LED Traffic Signal Lights Designed for the 21ST Century