Researchers at Durham University have developed a new drone navigation system that could help autonomous aircraft travel faster and more safely through complex environments.
The system, called CORTO-Planner, is designed to identify safer and more efficient routes for drones flying at high speed through spaces filled with obstacles. The university said the technology could eventually support search-and-rescue operations, infrastructure inspections and environmental monitoring.
More room to manoeuvre
Autonomous drones face a particular challenge when they need to move quickly through narrow or cluttered areas without hitting anything. Existing navigation systems can force aircraft to slow down because they use rigid safety zones that limit how freely a drone can move.
Durham’s research team has developed a flexible safety corridor that changes according to the surrounding environment. The approach creates flight paths with up to six times more usable space and offers up to 59 per cent wider clearances around obstacles than previous methods.
By creating more room for manoeuvre, CORTO-Planner is intended to help drones maintain higher speeds while making smoother movements in difficult surroundings. The system is aimed at improving both the speed and safety of autonomous flight, rather than simply increasing the aircraft’s top speed.
Tests included narrow passages and maze-like courses
The researchers also developed a faster way for the drone to calculate its route. The new method removes the need for slow and complex mathematical processes, allowing the aircraft to make navigation decisions in real time as it flies.
The technology was tested in computer simulations and on real drones. The trials included narrow passages, sharp bends and maze-like courses. In those tests, CORTO-Planner recorded faster flight times and higher average speeds than several leading drone-navigation systems while continuing to avoid obstacles safely.
The findings could be relevant to time-critical operations where reaching a location quickly matters. The team identified search and rescue, forest exploration, industrial inspection and warehouse automation as possible uses for the technology.
Software released for other researchers
Durham’s researchers have made the software openly available, allowing other researchers to build on the work. The university said the approach could also be adapted for other autonomous technologies that need to move safely through confined spaces.
Potential wider applications identified by the team include self-driving vehicles, robotic arms and underwater robots. The study was carried out by Dr Junyan Hu, Professor Farshad Arvin, Hang Wang and Honghao Pan.
The full paper has been published in IEEE Robotics and Automation Letters. The university’s Department of Computer Science is ranked eighth in the UK in the Complete University Guide 2027, according to the university’s release.