Designing an Adorable Penguin Rover is Really Hard

Last week, Yvon Le Maho's robotic penguin chick took the internet by storm. An integrative biologist at the University of Strasbourg in France, Le Maho and his team built the adorable rover so that they could get a closer look at penguin colonies without disturbing the birds in their native habitats. They published the results of their research in Nature Methods.

But the road to designing a weather-resistant (and penguin-resistant) rover was a bumpy one, indeed. We spoke with Le Maho about fiberglass penguins, epic king penguin battles, and the autonomous robots of the future.

MOTHERBOARD: Tell us about your first penguin-friendly rovers.
Yvon Le Maho: We initially used a regular [non-disguised] rover. It was a gift from the French Ministry of the Interior, so it cost us nothing. We built the rover to find out if we could approach the penguin colonies very closely, and use the rover for species identification. We tested the rover and we thought, OK, it's a good strategy.

But there were problems, right?
Right. The rover wasn't sand-proof or weatherproof. There was very bad weather while we were there, rain and wind, and that was bad for the electronic equipment. So we needed to improve every aspect of the rover.

Also, we wanted to make sure that the rover would not have a disturbing effect on the penguin colony. So we tried to make fake penguins. The first fake penguin that I tried before this collaboration [with a team of engineers] was made from fiberglass. It was very well done, but it didn't look like enough like a real penguin.

And so the penguin rover was born.
I think it's the material that really made a difference. Emperor penguin chicks are covered in down—not just feathers, but something that looks a bit like fur. We used a fake material that looks like the real down of a real penguin chick.

What happened when the rover approached the penguins?
We actually tested the rover with two different species of penguin. King penguins are a very territorial species. If they see another penguin approach, they will argue with it and fight for their territory. But if they see a human approach, they retreat. Here, we had no difficulty. The rover approached and they did not retreat. Once the rover was close enough to the king penguins, the birds began arguing with it, the same as they would with other penguins.

But then, when I first wanted to approach emperor penguins, I found it much more difficult. Emperor penguins do not have any territorial defense, because they huddle together instead. I needed to camouflage the rover enough that it would be able to approach [that huddle]. The fake chick that I [initially] used was not good enough to fool the birds. Eventually, we used this much better design, and it worked, and we were able to approach.

Image: Nature Methods

Wait… the king penguins actually fought with the rover?
I think that was one of the most difficult parts of designing the rover. It needed to be robust and resistant. When king penguins fight, they can be very tough, so the rover must be very resistant, especially since it's carrying high-level technology. When you have a penguin that is 80 centimeters (about 2.6 feet) high, using its beak and flipper to fight, the antennae and all of the sophisticated parts of your electronic system must be very resistant.

How did you initially become interested in studying penguins?
As a master student at the University of Paris, where I was studying animal physiology, I was a frequent visitor at the laboratory of ornithology of the Museum of National History. There, I met Jean Prévost, the deputy director who had been the first to describe in detail the breeding cycle of emperor penguins. This fascinated me. Prévost invited me to spend a year in Antarctica, studying the physiology of emperor penguins. This was the beginning of the story, and I am still fascinated!

What are the next steps?
Rovers are remote controlled. In the future, we will have fake animals that are autonomous, which can move on their own. The next step is to get information about how colonies function, using one of these fake animals.

I have learned that the key to developing fake animals is that they must look as much as possible like the real thing. I think we are beginning a new era, where we will use fake animals to help us understand the biology of real animals.

Science Is Really Hard is a new series where Motherboard asks scientists about the real work that goes on behind scientific discoveries and what their jobs actually involve.