How AI win Over three World Champion Drone Racers, AI piloting system power

An advanced AI pilot has once again demonstrated its dominance over human competitors, raising questions about the future of human-AI competition. The University of Zurich and Intel collaborated on the "Swift" AI piloting system, which outperformed three reigning world champion drone racers.

In a remarkable development that raises questions about the future of human competition against AI, an advanced AI pilot has once again demonstrated its superiority. Researchers at the University of Zurich, in collaboration with Intel, pitted their cutting-edge "Swift" AI piloting system against three reigning world champion drone racers – and the results were astonishing.

Swift represents the culmination of years of intensive research in the field of AI and machine learning by the University of Zurich. Back in 2021, an earlier version of the flight control algorithm faced off against amateur human pilots and consistently outperformed them in every race. This achievement marked a significant milestone because self-guided drones had previously relied on simplified physics models, limiting their top speed.

Subscribe our WhatsApp channel Click to join

The latest breakthrough is even more impressive as Swift achieved victory without the need for the cumbersome external camera arrays used in its predecessor. This system responds in real-time to data from an onboard camera, similar to the way human racers operate. It incorporates an integrated inertial measurement unit to track acceleration and speed while utilizing an onboard neural network to determine its position in space via front-facing cameras. All this data is then processed by a central control unit, which operates as a deep neural network, calculating the shortest and fastest path around the racing track.

Read Also: Google Expands AI-Powered Search To India And Japan Despite Growing Concerns About AI Misuse

According to Davide Scaramuzza, head of the Robotics and Perception Group at the University of Zurich, "Physical sports are more challenging for AI because they are less predictable than board or video games. We don't have a perfect knowledge of the drone and environment models, so the AI needs to learn them by interacting with the physical world."

Rather than letting the AI-controlled quadcopter physically learn the intricacies of the racing circuit over an extended period, the research team simulated the learning process virtually, achieving proficiency in just one hour. Swift then went head-to-head with champion drone racers Alex Vanover, Thomas Bitmatta, and Marvin Schaepper, ultimately posting the fastest lap overall, surpassing human racers by half a second.

While human pilots demonstrated greater adaptability to changing conditions during races, the potential of AI in drone racing is undeniable. Scaramuzza noted, "Drones have a limited battery capacity; they need most of their energy just to stay airborne. Thus, by flying faster we increase their utility." This achievement opens up exciting possibilities for AI applications in Search and Rescue operations, forest monitoring, space exploration, and film production, showcasing the transformative potential of AI in various fields.

Most Read: Snapchat's Dreams Feature: Transform Your Selfies With AI Magic Is A New Trend