New navigation system helps drones fly faster through obstacles
Researchers at Durham University have developed a navigation system that enables autonomous drones to fly faster through crowded environments while maintaining safe distances from obstacles.
The system, named CORTO-Planner, is designed to help drones select and continuously adjust safe flight paths in real time. Rather than simply calculating the shortest route, it identifies paths that balance safety, efficiency and speed as the surrounding environment changes.
The research, published in IEEE Robotics and Automation Letters, addresses a challenge in autonomous flight: allowing drones to move quickly through confined or cluttered spaces without compromising obstacle avoidance.
Existing navigation systems can force drones to slow down sharply near obstacles because they use rigid or restrictive safety zones. Durham researchers have instead developed what they describe as a ‘piecewise parametric safe corridor’, a flexible three-dimensional flight space that adapts to its surroundings while remaining sufficiently smooth for high-speed flight.
The approach creates flight corridors with up to six times more usable space than previous methods and provides up to 59% wider clearances around obstacles. This gives autonomous aircraft greater freedom to manoeuvre while maintaining safe separation from their surroundings.
The researchers also developed a computational approach intended to reduce the mathematical processing required to generate and update the corridors. This allows CORTO-Planner to operate in real time, enabling a drone to respond to changes in its environment while maintaining its planned trajectory.
The system was tested in computer simulations and on real drones navigating courses containing narrow gaps, sharp turns and maze-like layouts.
Across the tested environments, CORTO-Planner achieved faster flight times and higher average speeds than several existing drone planning systems while continuing to avoid obstacles.
“This work addresses a long-standing challenge in autonomous navigation: how to combine safety, speed and smooth motion in complex environments,” said Dr Junyan Hu, co-author of the study at Durham University.
“We've developed a new way of representing safe flight space that gives drones much greater freedom to manoeuvre while still maintaining reliable obstacle avoidance.”
The researchers say the approach addresses a trade-off in existing navigation methods. Flight corridors that closely represent complex environments can result in abrupt or inefficient drone movements, while smoother corridors can be more restrictive and force aircraft to reduce their speed.
CORTO-Planner is designed to combine the two approaches, producing flight paths that are smooth and adaptable while allowing autonomous aircraft to operate at higher speeds in confined environments.
Smoother flight could also reduce unnecessary manoeuvring, potentially improving battery efficiency and reducing wear on drone components.
Potential applications include search and rescue, environmental monitoring, forest exploration, infrastructure inspection and warehouse automation. The underlying navigation approach could also be applicable to other autonomous systems operating in confined spaces, including ground vehicles, robotic arms and underwater robots.
The research team has made the software openly available to support further research.
Related Content
Glasgow and Dublin to host world’s leading mathematics congress in 2030
Glasgow and Dublin have been selected to jointly host the International Congress of Mathematicians (ICM) in 2030, marking the first …
