Because animal locomotion is a very interesting interplay between the body, the spinal cord and the environment.

But to understand this interaction, we really need a body and that's where the robot is a key tool.

So we designed a salamander robot that very closely mimics the real motion of the real salamander,

and it can both walk and swim exactly like the real salamander.

It's a very old animal, it's an amphibian and it's almost a living fossil of the first terrestrial vertebrates,

so the ancestor of vertebrates like us, mammals.

So by studying the salamander we can study this key moment in evolution, the switch from swimming to walking,

and especially how the spinal cord has reorganized itself to allow a new mode of locomotion.

So the novelty of this work is really the approach we took to try to be as close as possible to the real physics of the body.

The species is called Pleurodeles. We looked at detailed x-ray recordings from the animals

so we could track the bones in 3D space and

that enabled us to create a very detailed model of the movements of the bones of the animal.

Neurobiologists have shown that if you electrically stimulate the spinal cord,

at low level of stimulation, it will induce a walking like gait.

If you stimulate a bit more, this gate accelerates, and at some point there's a threshold and it switches to swimming

so just changing the global drive that the brain sends to the spinal cord makes a complete switch between different modes of the locomotion.

So understanding this is very important, for instance, for neuroprosthetics.

Being able to re-stimulate those circuits in humans in the long term is something very important

and for that you need to understand how the spinal cord works.