Many ants lay down trails of chemicals wherever they walk. They use these trails to find their way to food and return home with ease.
One species, the desert ant, does not. Instead of following the routes others have laid down, they go out on seemingly random searches and yet always find their way home.
That's because they use a process called path integration, which is a sort of internalised GPS system.
They can also remember how to get home using a memory of their route based on visible landmarks that they store in their tiny brains. This is similar to humans using a familiar route: if we were plonked down at any point on this route, we would probably know where we were.
This extremely effective system ensures that if ants stray slightly away from their usual path, they can use these visual cues to orient themselves.
A team led by Barbara Webb of Edinburgh University in the UK has been trying to understand exactly how the processes in an ant's brain achieve this. Her team has now modelled it to better understand how it works.
"They have a brain that could fit on the head of a pin," says Webb, who presented her team's insights at the 2015 meeting of the Association of Psychological Science in New York. "What we're interested in understanding is what program could be running in the ant's brain."
The team peered into an area called the mushroom bodies, and successfully copied this "neural network" onto a small robot, where each node in this model represents a single brain cell.
The robot is essentially a mobile phone on wheels. Like an ant, it can "see", albeit using a camera, and it has a program that allows it to move around. It also has an omnidirectional lens to mimic an ant's near-360-degree vision.
Finally, it can remember (store) up to 400 images, and use them to identify familiar locations.
The robot was able to follow routes the same way the ant did, and in the same desert environment.
Webb says that confirms that ants really do use a selection of visual images to help them navigate. "If you can recreate the behaviour in the artificial system then you know you've got a good explanation," she says.
The artificial ant's brain also has many applications, says Webb.
"There's potential for having navigating robots particularly in situation where you don't want to have very complex sensors."
The ant's eye – and therefore the robot too – are extremely efficient, yet take up very little memory. Building something small that does not require much power could therefore be widely used.
Ants are known for their striking abilities to operate as one super-organism. This recreation of an ant's brain is yet another example of how processes in nature can be used to inspire technology.
For example, the amazing excavation abilities of invasive ants have been used to help design search and rescue robots. A robot colony that behaves just like an ant colony has been built and robot swarms have been programmed to move into any shape.
Their brains may be small but we clearly still have much to learn from these tiny critters.