Spanish researchers are implementing ever-more sophisticated designs for robotic locomotion thanks to some inspiration from nature.
By deeply investigating the neural networks that control the legs of invertebrates such as crabs, lobsters and spiders, the scientists have discovered a rhythmic nerve impulse that dictates the gait of the animal.
The researchers then took these insights and applied them to the control systems of a robotic worm, which in turn more accurately mimics natural movement.
Fernando Herrero-Carrón, Francisco de Borja Rodríguez Ortiz and Pablo Varona say the rhythmic impulses, which they call “central pattern generators,” or CPGs, help the body automate repetitive tasks.
Simply, CPGs are the body’s physical “autopilot,” allowing you to walk and talk at the same time.
The CPGs of invertebrates and how they generate rhythmic impulses are well understood by scientists because the animals’ neural systems are relatively simple compared to other species.
But applying that knowledge to robotics is a different story.
Traditionally, engineers have defined a series of rules that dictate how the legs should operate. But the scientists say recording these impulses allow them to generate control sequences for artificial limbs without rules, allowing them to better cope with unexpected situations.
In other words: intelligence, at least on a basic level.
What’s more, CPGs function as discrete circuits that can be strung together to create more complex systems — allowing scientists to concentrate on individual limbs, rather than set rules that apply to a robot’s overall movement.
The researchers applied the theory on a worm-like robot with eight sections. They developed the movement in a simulator, then downloaded the final product to the now-autonomous robot, which demonstrated in real life their bio-inspired approach.
Now, all they’ll have to do is figure out a way for it to retain knowledge and continue learning once it’s downloaded. I think there’s a word for that: “evolution.”
Watch the robotic worm in action:
Their research was published in the journal Bioinspiration and Biomimetics.
