Swimming Lamprey Robot Offers Locomotion Insight

Thanks to their swimming robot modeled after a lamprey, EPFL  scientists may have discovered why some vertebrates are able to retain their locomotor capabilities after a spinal cord lesion. The finding could also help improve the performance of swimming robots used for search and rescue missions and for environmental monitoring. AgnathaX is a long, undulating robot designed to mimic a lamprey, which is a primitive eel-like fish. It contains a series of motors that actuate the robot’s ten segments, which replicate the muscles along a lamprey’s body. The robot also has force sensors distributed laterally along its segments that work like the pressure-sensitive cells on a lamprey’s skin and detect the force of the water against the animal.

Transcript

00:00:02 What we did in this paper is to try to solve a  very old debate of neuroscience, which is the role of the peripheral nervous system and the central  nervous system, how these two systems interact to generate locomotion, and to control and sustain locomotion. And for this, we use a robot that is a lamprey, and with this lamprey, we will be able to sense the environment and bring these signals of the sensing of the environment to the central nervous system, and complement each other. AgnathaX is an undulatory swimming robot composed of 10 segments along the body. Each of the segment has a motor that  allows to do side-to-side movements, and additionally, it has a  sensor unit on each of its eyes,

00:00:44 to measure the force between the environment -  which in this case is the water - and the robot. According to the controller we were testing,  we loaded a program into the robot, and made it swim in a pool. Our pool is equipped with a motion tracking system, that allows us to quantify the results of the movement  to later compare them with our expectations. We found that the central and the peripheral  nervous systems together produce robust locomotion. And using both systems at the same time helps to  be robust against neural disruptions, like for example communication failures between segments, muted sensors, or muted neural oscillators. We also found that peripheral force sensors  in the skin of the animal and the robot

00:01:36 that sense how much the water  is pushing against the body were very useful signals to generate locomotion  patterns. It confirms that there are important functions of the peripheral nervous system that are hidden by the central nervous system. So this project has two main contributions; one which  is for neuroscience, one which is for robotics. The contribution to neuroscience is to explore this key question of how do we generate locomotion, and how do we synchronize the rhythms needed for locomotion. The contribution to robotics is really this notion of fault tolerance: being able to resist to big perturbations. Here we have design principles that allow us in  principle to be very robust to many many lesions.

00:02:18 We can provide a potential explanation of why  animals are so robust to do swimming. And this is really a good idea to take inspiration from  animals and bring that robustness to robots.