The Autonomous Racing League, which will take place in the United Arab Emirates on 11 and 12 April 2025, aims to push the boundaries of autonomous flying. Teams must program their drones to fly autonomously through a sequence of gates in the shortest time possible, successfully avoiding all obstacles. They can reach speeds of 150 km/h and experience up to 7 G overload when maneuvering around obstacles.
All this is possible thanks to the use of advanced machine learning algorithms and on-board artificial intelligence. It is in these criteria that the team from FEE CTU represents the absolute world leader. "Of course, we want to fight for the top position in Abu Dhabi, but the primary motivation for entering this competition is to give our PhD students the opportunity to compare themselves with the absolute world leaders in the field of agile flying. Drone races for time, which combine AI and robotics research with extreme sports, have not been our focus and we have not yet participated in a similar competition, so this can push us forward and improve our autonomous flying systems as well, said Associate Professor Martin Saska, head of the Multi-robot Systems Group at FEE CTU.
In addition to the main team competition, the organizers are preparing another robot vs. human competition for greater attractiveness: a fully autonomous drone and a drone piloted by a human will face each other. Martin Saska foresees the dominance of the robotic machine, which is able to make decisions in milliseconds and, thanks to machine learning, choose the optimal trajectory through the gates. The only thing that could derail the robot's victory is the potential difficulty of locating it in an environment without a GPS signal and equipped with only one camera.
Drones prepare for extreme race in Kladno hall
The MRS team operating at the Department of Cybernetics of the FEE CTU will prepare their drones for the competition in the test hall in Kladno, which simulates the conditions of the final circuit in Abu Dhabi by laying out obstacles, until the beginning of April 2025. Thanks to this preparation, students and researchers can work on optimizing algorithms and testing the performance of drones in real conditions.
Among 14 teams from all over the world, often from leading academic institutions, the Czech team is not one of the outright favorites, but it certainly does not intend to be left behind in the finals. "We use advanced trajectory planning and flight control algorithms that allow our drones to fly at maximum speed with extreme precision," adds Dr. Robert Pěnička, the team's chief technical officer from FEE CTU.
Robert Pěnička is researching methods for controlling and planning the flight of autonomous drones in unknown environments with obstacles. His goal is to bring the autonomy of drones to the level of human pilots so that they can fly even in unknown environments with great agility and incredible speed. Dr. Pěnička is also conducting his research thanks to a prestigious grant Junior Star from the GA Czech Republic, which he received in 2022.
For example, agile autonomous drones can be used to inspect critical infrastructure such as energy distribution, monitoring key industrial or strategic facilities, or inspecting solar power plants. Drones are capable of spotting any differences in infrastructure and by flying automatically, the whole process is significantly more efficient than when drones are pilot-controlled. Another promising area for the application of autonomous drones is rescue operations or the autonomous transport of goods or medicines.
For more information about the competition, please visit www.a2rl.io.
Photos and videos from MRS/Fly4Future team preparation link.
Photo credit: Petr Neugebauer
