Bee-Inspired Navigation Helps Drones Fly Without GPS
European researchers have developed Bee-Nav, a navigation system that allows drones to return autonomously to their starting point using visual memory and minimal computing power.
The technology could support lighter, lower-energy agricultural drones capable of operating in areas where GPS access is limited or unavailable.
Scientists in the Netherlands have replicated aspects of how bees navigate in nature to improve drone flight.
Bee-Nav enables drones to fly without GPS or powerful onboard computers. The drone first performs a short familiarization flight to memorize the surrounding terrain. It can then return autonomously to its base after flying several hundred meters.
Compact Neural Network
The system combines visual memory with odometry, a method used to estimate distance and direction based on visual signals.
Image processing is handled by a compact neural network, making the technology suitable for lightweight drones with low energy consumption.
In tests, the drone successfully returned to its starting point in both indoor and outdoor environments.
Key figures include:
More than 600 meters covered during tests
Autonomous return to the starting point
Neural network capacity of only 42 kilobytes
Operation without GPS
Researchers believe the technology could eventually support agricultural applications such as detecting plant diseases and pests.
