Desert ants are true masters of navigation. In their hot and dry environment they reliably find their way home to their nest again and again. However, they cannot rely on chemical signals because, unlike in other habitats, chemical signals on the ground are out of the question as navigation aids: the droplets with the signal molecules would evaporate much too quickly. That is why the desert ants rely on other clues: Polarized UV light from the sun shows them the direction. And they calculate the distance covered from the texture of their surroundings and the number of steps they take. Combined with each other, both information always give the current position, compared with the starting point at the ant’s nest.
Navigation with the simplest means
These are ideal conditions for finding your way in areas without GPS reception. This is why researchers from the Institute of Movement Science – E.J. Marey, jointly operated by the French research institute Centre national de la recherche scientifique (CNRS) and the University of Aix-Marseille, have used this ingenious natural mechanism for their AntBot robot. “Ants are highly efficient navigators who travel many kilometres in search of food and then return to their nest,” explains the biorobotics expert Julien Serres from the University of Aix-Marseille. “Desert ants have very few neurons in their brains, but a visual system with surprising properties. It is precisely these characteristics that the French researchers wanted to exploit – which is why their “AntBot” is directly inspired by the desert ants and can navigate in unknown environments with its six red and black legs and without GPS.
Two photodiodes on top of his head measure the polarization of the UV light of the sun to determine the direction of movement. Even when the sky is overcast, the measurement deviates by a maximum of 0.4 degrees from the true value. Another optical sensor with a resolution of just twelve pixels looks down at the ground and picks up the underground that passes AntBot. From this, the robot calculates the “optical flow”. This in turn – combined with the number of steps – provides the distance covered.
AntBots work amazingly accurately
The accuracy achieved with the ant sensors is astounding: after covering a distance of 14 metres, the AntBot, which weighs 2.3 kilograms, still knows its position with an accuracy of one centimetre. All this does not require a GPS or a complex computer. “We only want to use a minimum of resources to perform very complex tasks“, says Serres. The control center of the AntBot consists of a small Raspberry single board computer.
In the future, the ant navigation system could be used in industrial production and logistics – for example in production halls or warehouses with artificially generated polarized light as a substitute for the sun. For example, in the modular production of cars, where robots take over on-site logistics in a constantly changing production environment. Instead of rigid control systems, a few parameters would then control the routes between the warehouse and production. The result would be the absolute flexibility of a very simple system.
“With the eyes of a desert ant: The #AntBot uses sunlight and the texture of its environment for #navigation. #GPS it doesn’t need.“Tweet WhatsApp
The ant-inspired technology could also become interesting for logistics centres. The location of objects is playing an increasingly important role there. “Autonomous materials handling vehicles must generally orient themselves – but lose their overview as soon as they leave the area mapped for them,” says Thomas Reppahn, Head of Logistics Product and Process Management at DB Schenker. “In order to expand their radius of deployment, it is currently necessary to invest heavily in positioning technology in order to set up a local, artificial positioning system. The AntBot approach is an interesting positioning method borrowed from nature that will continue to be observed.