Technology Innovation Institute Achieves Fastest Speeds with Vision-based AI Drone Racing

Oct 28, 2024
Vision-based AI Drone Racing

 

  • • TII’s autonomous drones outperforms human pilots
  • • TII’s autonomous drones hit a top speed of 22.3 m/s (80 km/h), without prior localization,  or residuals, with no external localization, and no external localization during training
  • • Top speeds were achieved with the added complexity of a curved split S shaped track
     

Abu Dhabi, UAE – 17 October 2024 – The Technology Innovation Institute (TII), a leading global scientific research center and the applied research pillar of Abu Dhabi’s Advanced Technology Research Council (ATRC), is revolutionizing the future of autonomous drones through its cutting-edge work in autonomous fast mobility. Leveraging vision-based artificial intelligence (AI), TII is designing and testing autonomous drones capable of navigating complex environments at high speeds – an advancement with the potential to transform multiple industries.

TII's latest innovation uses drone racing as a platform to design and test autonomous flights, demonstrating the possibility of a fast mobility ecosystem for aerial, marine, and ground vehicles. These advancements have the potential to soon extend to human transportation, reshaping the future of autonomous transit.

AI-Powered Autonomous Racing Drones: A Glimpse into the Future of Fast Mobility

Unlike conventional drones, TII’s unmanned aerial vehicles (UAVs) operate without human pilots or external localization systems, moreover, no external localization systems are used for training. Equipped with only a camera and an inertial measurement unit (IMU) sensor, the drones are steered entirely by AI and can reach speeds of up to 80 km/h. Additionally, top speeds were achieved with the added complexity of a curved split S shaped track.

The autonomous UAVs use onboard AI-driven visual odometry and reinforcement learning to adapt to their environment in real time, this means they receive no external support. The drones are designed to operate by trial and error, mastering flight in simulation before applying their knowledge in real-world scenarios. These advanced algorithms allow the UAVs to optimize trajectory planning and perform agile flights even on complex racing tracks.

TII has been testing these innovative approaches by competing against professional human pilots drafted from the Drone Championship League showcasing the potential of AI-powered UAVs to perform on par with and even surpass human-operated drones in agility and precision. In the direct competition in the TII arena, the TII drone set a best lap time of 4.38s, compared to the best lap times of the human pilots of 6.32s and 5.34s respectively.

This research is an integral step in developing a fast air-mobility ecosystem. The technology not only enhances safety and efficiency but also offers cost-effective and environmentally sustainable solutions for the transport of goods – and eventually people.

“TII is at the forefront of a new era in transportation,” said Dr. Najwa Aaraj, CEO of Technology Innovation Institute. “By harnessing the power of AI and autonomous systems, we are pushing the boundaries of what’s possible in fast mobility. Our work with vision-based drones is a significant leap forward, and we see vast potential for applying this technology across various industries in the near future.”

To learn more, visit: https://www.youtube.com/watch?v=J0l0KmDe_V0