On the run from predators, giant schools of fish swim in seemingly choreographed motion – remaining together in a group as they try to out maneuver the enemy. This actually works better to save more little fish hides than everyone swimming away in a different direction. But how do the fish manage to swim together in one large body like that? It’s as if they join together – like how all the individual robots come together to form one big one in the Mighty Morphin’ Power Rangers. Yeah, MMPR!! Right guys?! Guys?
But the fish aren’t joined together and their motions aren’t choreographed. So how do they do it?
The basic mechanism behind fishes’ ability to make these synchronized movements has been understood for some time. Fish have what is called a lateral-line system, extending from the fishes mouth to its tail. Along the lateral line are a series of two types of sensors: one which measures changes in the velocity of the surrounding water, and another which detects changes in pressure. The former are called surface neuromasts and the latter are called canal neuromasts. The surface neuromasts are made of sensors that stick up like a mast on a ship, although they are only the size of a short hair. The canal neuromasts are more like little trenches below the scales, and they detect pressure in the surrounding water through two pores positioned on either side of the neuromast.
These neuromasts allow fish (especially those living in murky waters or dark locations) to determine the motion of their neighbors, the presence of obstacles, and the motion of potential predators or prey. The range of this detection method is limited – at most, the body length of the fish. In murky or dark waters, this still beats the use of vision. It also beats techniques like sonar, used by bats, which would require the fish to send out a signal that would also alert predators, or prey, to the fish’s location.
Fish that dwell in water that is mostly still, like goldfish living in quiet ponds, will have more sensors geared toward changes in velocity, good for detecting water motion or wakes created by swimming fish in the area. Fish living in more tumultuous locations, like trout in raging streams, will have more canals that detect changes in pressure.
At the APS March Meeting, Dr. Leo van Hemmen of the University of Technology Munich (TUM) in Germany, discussed his group’s work to understand how this lateral-line system and the ensuing neuronal information processing connect to fishes brain, and help fish create a three-dimensional map of their surroundings. In collaboration with TUM’s electrical engineers, lead by Dr. Sandra Hirche, his group is currently building a water-going robot that operates using a lateral-line system also constructed by the group. Van Hemmen and his group nicknamed the robot Snooki, although there is no connection to the Jersey Shore star (you were so confused by the title of this post! I’m sorry I dragged you along for so many words). The robot Snooki is fully equipped with sensors and additional hardware modeled after the neuronal system.
Besides potentially developing the lateral-line system for use in another robot, the group is using Snooki (shown right, photo courtesy of Dr. van Hemmen) to study studying how the fish manage to integrate the input from the canals into their mental map of their surroundings. Human brains put together sight, sound, and touch into a cohesive picture of our world. This input is transcribed through some complex neuronal activity, and van Hemmen and his group want to try to understand that process of passing the information from the sensors to the brain, and turning it into an integrated local map. Replicating this process in robots, which sometimes use six cameras, is a challenge. The TUM researchers would like to find a way to integrate six cameras so the input is reduced to what you might get from only two cameras (much more manageable).
Their work could also be used to make more sensitive detectors of this type operating in air. Researchers at the University of Twente in the Netherlands have built sensors similar to Snooki’s surface neuromasts velocity detectors, but theirs were patterned after those found in crickets. These detectors can respond to changes in airflow velocity less than 1 mm/sec, but the ideal is to reduce that by a factor of ten so as to attain optimal response. Ideally, Snooki or her descendents will help with this pursuit.