Basic Stamp Kits
Books for Kids
Robots at Work
Courtesy of New Scientist Magazine
By Peter Hadfield
MUSEUMS AND AQUARIA could soon be displaying lifelike robotic models of marine
creatures--some of them long extinct--thanks to the efforts of engineers at Mitsubishi
Heavy Industries in Japan. The company has just developed a remote-controlled sea bream
that looks and swims just like the real thing. And since you don't need to feed the
60-centimeter-long fish or clean out its tank, it makes an ideal exhibit.
The robotic sea bream is controlled via a desktop computer. This controls movement of the
tail fin and two pectoral fins. "It wasn't possible to replicate the exact method of
propulsion used by the sea bream," says Mitsubishi robotics engineer Yuuji Terada.
"So we're using an elastic oscillating fin that we first developed for use in our
underwater vehicles." The battery-powered fins oscillate at between 0.2 and 1 hertz,
and will propel the robot at up to 0.25 meters per second. The battery is automatically
recharged by a coil inside the fish that draws power from an electromagnetic field that
permeates the fish tank.
After spending four years getting the bream to swim accurately, Mitsubishi says its next
step will be to build a robotic replica of a coelacanth--the "fossil fish" that
was once thought to be extinct. After that it plans to bring back several extinct fish
from the pre-Cambrian era.
WHY BUILD A ROBOT coelacanth? "Just for fun" is the answer given by Yuuji
Terada. If only life were so simple. But Mitsubishi Heavy Industries also builds
submarines and, like its counterparts elsewhere, has good practical reason to be
interested in building robot fish. Fish are just so much more efficient at moving through
the sea than any submarine or ship and engineers would like to crack their secrets.
Although many people know the old joke that the laws of physics show that bees cannot fly,
fewer know that there is a marine equivalent. According to Gray's paradox, fish just don't
have the muscle power to swim at the speeds that they do. Or to put it another way, they
appear to require only half the power to reach the same speeds as a propeller-driven or
towed model of the same fish.
Once a fish starts swimming, the drag around its body mysteriously drops and it transfers
momentum to its wake with minimum energy loss. For both fish and bees, vortex control
appears to be the secret of efficient propulsion. And, of course, you must add the fish's
ability to hover effortlessly and to turn in less than a body length. Submarine commanders
can only turn green with envy.
In the case of fish, scientists outside of Mitsubishi have made some progress. At MIT
Lycra-clad robot tuna and robot pike are already swimming in tanks. The tuna is a good
model as it is a big fish with a relatively simple swimming pattern in which all the
propulsion comes from the final section of its tail. The pike is particularly interesting
because of its astounding ability to go from 0 to 20 kilometers per hour with acceleration
of up to 12g.
The long-term goal is to understand how fish swim so efficiently and develop autonomous
robot fish that can set off into the oceans and make maps, search for pollutants, check on
fish stocks and become part of "military surveillance". Power sources also
present difficulties, as even the robot tuna which is 2.5 meters long could eventually end
up motionless with its batteries flat. But if robot fish really can swim more efficiently
than robot submarines, expect the military to be very interested in their ability to keep
going that much longer.
Which is why, if you meet an angler who claims to have hooked a metal monster, don't be
surprised if a hoard of gun-toting men in dark sunglasses aren't far behind.
Subscribe to New Scientist