To understand how an animal moves, physicists typically study its mechanical interaction with the environment. They need not study the animal’s specific gait patterns, though. Almost all known vertebrates are equipped with central pattern generators—neural networks that are able to provide a rhythmic output without any external driving signal. Because of those generators, animals rely not only on their five senses to get around, but also on proprioception, their sense of self-movement and body position in a given environment. That sixth sense is usually responsible for adaptation of the rhythmic driving of the animal’s muscles, and locomotion is thought to be sustained through a proprioceptive feedback loop: In short, the animal senses how the deformation of its body affects its position and speed and uses that information to adjust how it moves.
Predicting the swimming gaits of fish has been the subject of research studies for more than half a century. In all those studies, kinematic constraints had to be imposed on the fish, which limited the predictions researchers could make for its tail beat amplitude and frequency. Now physicists led by Médéric Argentina of the Côte d’Azur University in Nice, France, have gotten around that problem by having the fish essentially “choose” (via feedback) how and when it moves. They have demonstrated a new mechanism for driving robotic fish, whose tail movements are controlled by proprioception. They designed and built the robot, shown here, to resemble a fish’s exoskeleton, and they placed it in a water tunnel. The robot’s head, body, and tail were 3D printed from soft polymers. A waterproof servomotor (blue) links the body to the tail. The motor’s wheels are attached by two cables that bend the robot’s body and tail side to side with a vigor that depends on information collected by a force sensor (vertical rod) on the robot’s head.
Using the sensor’s information, the robot moves in a way that matches a real fish. And it does so only when submerged. The authors show that the robot’s activation is triggered by an oscillatory instability. In water, the robot’s tail spontaneously oscillates at an amplitude and frequency determined by the feedback loop between the motor and the force sensor, as shown in the video. The oscillation stops when the robot is raised above the waterline. (J. Sánchez-Rodríguez et al., Phys. Rev. Lett. 126, 234501, 2021.)
