?Underwater senses － Navy seals? （The Economist August 11th 2018 海洋生物(アザラシ)の知覚と潜水艦探知）
?
?
Pinniped whiskers provide a model for a new generation of submarine detectors
Vision is useless in murky water. To deal with that deficiency dolphins have evolved sonar. They emit clicks and interpret the echoes to find their prey. But not all marine mammals are so equipped. Seals, for instance, have no sonar, yet that?does not stop them finding distant meals as effectively as dolphins can. This puzzled researchers for years, until they discovered that the secret lies in the animals’ whiskers － which they are now trying to copy, to develop novel underwater sensors.
An object moving through water leaves a series of miniature whirlpools in its wake. This trail is called a Karman vortex street. And that is what seals, using their whiskers, follow. As Michael Triantafyllou of the Massachusetts Institute of Technology (MIT) observes, “You can set a harbour seal loose to follow a towed fish, and even 30 seconds later they will be able to follow the exact track, whether it’s straight or zigzag or circular.”
Dr Triantafyllou and his colleges at MIT’s Centre for Ocean Engineering are one of several groups studying how seals do this. A rival team, led by Ben Calhoun of the University of Virginia, and involving the University of California, Santa Cruz; the Naval Undersea Warfare Centre Division at Newport, Rhode Island; and Woods Hole Oceanographic Institution, has recently completed a three-year investigation of the matter. Other projects are underway at Jeju National University in South Korea and at Cleveland State University.
Seals can pick up the trail of fish such as herring when blindfolded and wearing earmuffs. Cover their whiskers, though, and supper eludes them. The bases of seal whiskers are rich in nerve cells, making them as sensitives as human fingertips. But that is not all there is to it. Under a microscope, seal whiskers are not circular when sliced through, as might naively be expected. Instead, they have an oval cross-section. Moreover, those whiskers’ surfaces have an elaborate undulating geometry.
Oh my ears and whiskers!
This complex shape looked familiar to Dr Triantafyllou and his team. They had arrived at something similar when working on mooring lines for off shore gas rigs. Their purpose was to stop those lines vibrating as water flowed past, a phenomenon similar to telephone wires or power cables humming in the wind. Dr Triantafyllou confirmed his suspicions about the similarity when he tested a scaled-up 3D-printed model of an artificial seal whisker. This, too, failed to vibrate in what are known as laminar-flow currents （that is, those without eddies in them）.
Paradoxically, this insensitivity to laminar flow increases sensitivity to vortices, as Dr Triantafyllou proved. He and his team towed an enlarged artificial whisker through water, to find out how to vortex street laid down in front of it affect its behaviour. They aligned their model whisker so that it was edge-on to the direction of travel, just as real one would be. This caused it to cut through the water like a knife blade. Currents at right angles to its direction of travel, such as those created by vortices, exerted forces on its flat surfaces. These caused it to skirt around those vortices a skier negotiating moguls, thus vibrating in a way that it would not when presented with a laminar current.
Bigger moving objects generate bigger vortices, so the amplitude of this vibration changed with the size of the object being followed. The frequency of the vibration changed with the size of the object’s speed. Assuming that this also happens with real whiskers, it would permit a seal to assess its target’s bearing, size and velocity.
In 2016, with help from researchers at Singapore University of Technology and design, Dr Triantafyllou built on these discoveries by attaching an artificial whisker to a membrane that, when distorted, generated a pulse of electricity. This arrangement proved sensitive to the slightest of water movements. The next stage is to understand what the pulses mean.
That is a challenge Dr Calhoun, at Virginia, has already taken up. Recruiting a trained seal for the task, he and his colleagues attached a recording device to one of the animal’s whiskers. They found that even following the simplest object generates several types of vibration in this whisker. And a seal has dozens of whiskers.
Seals’ brains can make sense of all this input. Dr Triantafyllou hopes to do likewise using artificial intelligence (AI). He and his colleagues will employ a form of AI called deep learning. The work involves training appropriate software on thousands of different inputs from an array of artificial whiskers. Once trained, such software should be able to pick out patterns in the data and so learn to recognise the trails left by objects of different types and sizes, travelling at different speeds.
Dr Triantafyllou’s team’s purpose is to create a whisker-based sensor for underwater robots. This will detect the wakes of natural objects, such as fish and marine mammals, and artificial ones, such as other robots, surface ships and submarines.
Not surprisingly America’s armed forces are taking keen interest in all this. As well as the Naval Undersea Warfare Centre’s involvement, some of the research has been supported by the Office of Naval Research. The navy hopes that vortex sensors may meet the challenge of spotting submarines, which are getting ever quieter and harder to find with sonar. A submarine leaves far bigger trail than a fish, and that trail can persist for hours, even days. Fleets of small, bewhiskered robots might thus be able to track otherwise undetectable submarines as easily as seals find shoals and herring.
A vortex detector would have civilian applications too. A static detector would be able to measure turbulent currents flowing past it. Such a system, Dr Triantafyllou observes, might have been useful during the Deepwater Horizon oil spill in the Gulf of Mexico, in 2010. Then, vortex sensors could have helped map the plumes of oil released, helping predict the spread of the spill. Nor are applications restricted to marine settings. Appropriate sensors might measure liquid flowing turbulently through pipes and air flowing similarly over aircraft wings.
Extending vortex sensors into the air brings to mind another possible zoological analogy. Nightjars （which are, as their name suggests, nocturnal）, prey on flying insects. especially moths. The assumption has always been that these birds have particularly good night vision. What they definitely have, though, are arrays of whiskers around their beaks. Time, perhaps, to get the microscopes out again, to see exactly what shape these whiskers are.
