Bee Inspired Drone Navigation

in Popular STEM6 hours ago

Bee Inspired Drone Navigation




Ultra efficient navigation


Cross hundreds of meters autonomously, find your way back home and do all this using less memory than a regular photograph stored on your cell phone. It seems impossible, but it was exactly what European researchers managed to demonstrate when developing the Bee-Nav, a navigation system for drones inspired by one of nature's most efficient navigators, the bee.


Today, autonomous drones depend on complex three-dimensional maps and enormous computational power to locate themselves. The problem is that all this technology consumes energy, increases the weight of the equipment and reduces flight autonomy. To overcome this limitation, researchers from the Delft University of Technology, Wageningen University and the University of Oldenburg decided to copy a strategy used in nature.


Bees managed to fly long distances and return to the hive without GPS, without digital maps and using a tiny brain. The Bee-Nav replicates that principle exactly. The system combines two main mechanisms. The first is visual odometry, where the drone calculates distance and direction by observing only the visual flow of the environment around it, the second is visual memory, before leaving, the drone performs small flights around its base, recording panoramic images that function as reference points for the return.


During the trip, small navigation errors inevitably accumulate, it is at that moment that the second stage comes into action, as you approach the destination, the artificial intelligence compares what it is seeing with the previously stored images and corrects the route automatically, recognizing the area even when approaching from different angles.


The results impressed the researchers using a neural network of just 42 KBytes, the Bee-Nav completed external flights up to 600 m in distance. In controlled internal environments, the return success rate reached 100%. In outdoor tests, under strong wind, the index dropped to about 70%, revealing the main current challenge of the project.


References 1


Coin Marketplace

STEEM 0.04
TRX 0.32
JST 0.075
BTC 64496.87
ETH 1680.70
USDT 1.00
SBD 0.42