It sounds like science fiction, but also strangely familiar: drones buzzing around, inspecting tomatoes in greenhouses, delivering your package or inspecting an industrial site. With all the talk about drone-swarms, development in drones seems to move fast. But their navigation still requires a lot of computing power and memory, making them heavy, expensive and energy-hungry. Scientist led by Delft University of Technology (the Netherlands) have copied Mother Nature's homework on how honeybees find their way to finally solve this problem. They've published their results in Nature.
The scientific team, consisted of roboticists from Delft University of Technology and biologists from Wageningen University (both The Netherlands) and Carl von Ossietzky University of Oldenburg (Germany) presents "Bee-Nav": a robot navigation strategy inspired by honeybees. It allows even very small robots to travel far away from home and return successfully, using a neural memory of only 42 kilobytes. In a new environment, the robot first performs a short learning flight near home, just as honeybees do. After that, it can travel away for hundreds of meters and still find its way back. Bee-Nav enables lightweight, safe robots to navigate on their own, opening the door to applications such as butterfly-like drones monitoring greenhouses. The research also offers new insight into how flying insects may find their way home.
Challenges of navigating
Many future robots will need to navigate on their own, even where GPS is unavailable. Most current systems do this by building detailed maps of the environment. But that requires a lot of computing power and memory, making such systems expensive and energy-hungry.
Honeybees show that there may be a much more efficient solution. Despite their tiny brains, they can travel long distances and still return home. They do this in part through odometry: they estimate how far and in which direction they have moved using visual motion cues. Think of it like forms of counting steps, so to speak.
Unfortunately, odometry alone drifts over time, so it becomes less and less accurate. That is why insects also rely on visual memory. They remember what the world looks like around important places such as their home. Scientists understand insect odometry well, even up to the neural level, but visual memory is still much harder to explain. It was also not yet clear how the two can be combined to help very small robots navigate autonomously.
The "Bee-Nav" robot navigation strategy
The research team from The Netherlands and Germany were inspired by what honeybees do when they first leave the hive. These tiny creatures begin with short learning flights close to their hive. After that, they can travel much farther away and still return successfully. A bit like stepping outside your own house and walking through the first couple of streets around it. You now recognise your neighbourhood however you approach it on the way back.
"We were fascinated by the fact that honeybees can fly far away from home along winding paths, yet return almost straight back," says Guido de Croon, Professor of Bio-inspired AI for drones at Delft University of Technology (The Netherlands). "Biologists have shown that bees rely on odometry for the return journey, and use visual memory more as they get closer to home. But exactly what and how they learn for their visual memory is still not fully understood. That was the gap we needed to bridge to create a practical navigation strategy for robots."
In Bee-Nav, the robot also first makes a short learning flight near home. During that flight, it collects panoramic images of the environment. A small neural network then learns to process those images for estimating the direction and distance back home.
"Like an insect, the robot may not always know exactly where home is," says Dequan Ou, PhD candidate at Delft University of Technology and first author of the paper. "Home may be too small to see, or hidden behind some trees. So we trained the neural network using odometry estimates of the direction and distance home, even though these become less accurate over time. The key question was whether that would still be enough for the robot to learn to return home."
Turns out, it was. Odometry drift did not prevent successful visual homing. The image titled 'Cyberzoo' shows four robot flights starting from different points within the learned area. Using a neural network of just 3.4 kilobytes, the robot interpreted panoramic images of its surroundings and estimated which way to move and how far it still was from home. The estimated distance allowed the robot to move faster when farther away and slower as it approached home. In all flights, the robot successfully returned home.
Towards real-world applications
After succeeding in small indoor homing experiments, the researchers tested the full navigation strategy in larger indoor and outdoor environments. In one outdoor test at the Dutch drone research field-lab Unmanned Valley in Valkenburg, the drone flew more than 600 meters and still returned home, using a neural network of just 42 kilobytes. In large indoor spaces such as hangars, the system was successful in every test. In windy outdoor conditions, success dropped to 70 percent. One key reason was that wind forced the drone to tilt, making its images harder to use for navigation.
"The experiments are very encouraging," says Dequan Ou. "But they also show that our current system needs to become more robust in real-world conditions."
One promising application is greenhouse monitoring. Lightweight drones could inspect crops and detect diseases or pests at an early stage, helping growers increase yield while reducing waste. Bee-Nav is especially suitable for such drones, because they need to be lightweight and safe for people working nearby.
Finally, the research also provides a new perspective on how honeybees return to their hive, and how visual learning may shape that journey.
