Bee-inspired AI reduces computing power needed for autonomous drones

Researchers developed Bee-Nav, a bee-inspired navigation system that helps lightweight drones return home without GPS.

The method uses a tiny 42-kilobyte neural network and minimal computing power, drawing on honeybee homing behavior and visual memory.

It performed well in large indoor trials and a 600-meter outdoor flight, though reliability dropped in windy conditions.

The work could make small drones more energy efficient for tasks like inspections and crop monitoring, while shedding light on bee navigation.