Tiny robot drones learn to navigate the world like honeybees

"A honeybee leaving the hive first takes a short learning flight to memorize nearby landmarks. As a bee flies away, it keeps track of the direction and speed of its movement, de Croon says, in a process called path integration. Because path integration is prone to accumulating tiny measurement errors over time, the insect relies on the memorized landmarks to correct its course as it gets back home."

"First, a drone performs a beelike learning flight around its starting point using a minuscule omnidirectional camera to capture the surrounding scenery. Midflight, it trains a tiny onboard neural network to map these images to home vectors, basically invisible arrows pointing back to the launchpad. This establishes a safe zone called the Learned Homing Area."

"Once trained, the drone can be sent far away and come back using path integration first, backtracking based on measured speed and direction. If the drone winds up anywhere inside its starting safe zone, the visual neural network guides it the rest of the way home."

"The Bee-Nav does this using an off-the-shelf Raspberry Pi 4 computer the size of a credit card that runs neural nets with between 3.4 and 42.3 kilobytes of memory—thousands of times less than conventional mapping setups use."

A honeybee memorizes nearby landmarks during a short learning flight and then uses path integration to track direction and speed while flying away. Because path integration accumulates small errors over time, the bee uses the memorized landmarks to correct its route back to the hive. Bee-Nav copies this workflow for tiny drones. A drone performs a beelike learning flight around its start point, capturing surrounding scenery with a minuscule omnidirectional camera. During the flight, it trains a small onboard neural network that maps images to home vectors, defining a Learned Homing Area. After training, the drone can travel far using path integration and then rely on the visual neural network to finish homing when it enters the safe zone. The system runs on a Raspberry Pi 4 with very small memory requirements compared with conventional mapping setups.

Read at www.scientificamerican.com