杏吧原创

DON鈥橳 be fooled into thinking that humans are the most successful explorers
on Earth. With our highly effective modes of travel, our amazing new materials
and our hi-tech gadgetry, it would be easy to imagine that we were the only ones
fit to conquer our planet. But humans were not the first to plumb the depths of
the oceans, live at high altitude, or even explore the polar ice. Other animals
beat us to it every time. And when it comes to devising novel ways of living in
and exploring the harshest environments, nature wins hands down.

Which is why NASA is adopting a new approach to space exploration. The agency
plans to study how living things interact with their environment so it can apply
similar principles to explore planets and moons in the Solar System. NASA鈥檚
teachers range from cockroaches and centipedes to seeds and plants. And already,
scientists are glimpsing general principles of locomotion that apply equally to
humans, dogs, cockroaches and centipedes and that could be used to create a new
generation of intelligent robots called 鈥渂iomorphic鈥� explorers that react to the
environment in similar ways to living creatures on Earth. In future, these
robots could wander the Solar System tasting the soil on Mars, searching for
signs of life on Europa and even mapping the magnetic fields near Jupiter.

Small, tough and charged up

The challenges are many. NASA engineers must come up with a way of matching
the form of locomotion to the environment they wish to explore. This will
determine the structure of the biomorphic explorers and how they will move
around. Engineers also need to develop simple autonomous control systems so
their explorers can navigate any obstacle without a helping hand from Earth or
wasting time on number crunching. They have to develop power sources that are
small and light, yet can tolerate long spells in harsh environments. And each
explorer needs a reliable, low-power communications system. Finally, the
explorers must carry useful instruments鈥攃ameras or spectrometers, for
example鈥攖hat are small and efficient, yet powerful enough to carry out the
analyses that scientists need.

Robotic explorers have a mixed history. In 1993, an ill-fated eight-legged
robotic explorer called Dante descended into an active volcanic crater in
Antarctica. The experiment was designed to test the feasibility of robotic
exploration in extreme conditions. But Dante鈥檚 control systems were primitive,
and the robot鈥檚 progress was painfully slow because it couldn鈥檛 take a step
before its on-board computer had calculated the position and movement of each
leg. The mission ended after a journey of only a few metres when the fibre optic
cable linking Dante to the researchers became snagged and broke.

Wheeled rovers have had their fair share of problems too. In 1996, NASA鈥檚
Pathfinder probe released the Sojourner rover onto the surface of Mars.
Controlling a rover from Earth is impractical because of the time it takes for
signals to travel back and forth. Sojourner supposedly had enough autonomy to
navigate small rocks. But the rover could never be allowed to tip over onto its
back, since it couldn鈥檛 get back up. So Sojourner was programmed to stop and
wait for commands from Earth if it tilted beyond a certain angle or if its
wheels slipped.

This caused all kinds of problems. While crossing an area known as the 鈥渞ock
garden鈥�, Sojourner constantly tilted beyond the failsafe point and waited hours
for a message from Earth to set it in motion again. 鈥淲e were stuck in the rock
garden for weeks. Sojourner had trouble with rocks you wouldn鈥檛 even notice if
you were walking over them,鈥� says Dave Crisp, chief scientist of NASA鈥檚 New
Millennium Program, which is testing new technologies for future space
missions.

Few living things would have had trouble navigating the rock garden. This is
why NASA wants to study them further. Last August, Sarita Thakoor, a scientist
at NASA鈥檚 Jet Propulsion Laboratory in Pasadena, California organised a meeting
to discuss building biomorphic explorers and what they could do. But it鈥檚 not as
simple as copying the way a cockroach walks, for instance. That would be a bad
idea, says Robert Full, a professor at the department of integrative biology at
the University of California, Berkeley. Full argues that mimicking one type of
animal or insect locomotion rules out the advantages that have evolved in other
species.

Instead he is looking for general principles that underlie many-legged
locomotion, and he has found at least one. Full has been measuring the forces
generated between the feet and ground when animals and insects run. 鈥淲hen humans
run, each leg acts like a pogo stick propelling the body mass forward. The same
is true of dogs, cockroaches and even centipedes.鈥� Full says it is possible to
model each as if it were a system of springs and masses. You can then work out
very general characteristics of these systems.

One important characteristic is the pattern of forces that a foot creates on
the ground during each step. Full says that when creatures run, they generate
identical patterns with each foot. This is independent of how many other legs
they have. Another important factor is the spring constant, which is a measure
of the stiffness of a spring. It turns out that the spring constant is the same
for all polypedal creatures that Full has measured. All creatures seem to share
the same kind of 鈥渂ounciness鈥�, which is unsurprising since they all survive in
the same gravitational field. 鈥淗umans, dogs, horses, crabs all share the same
spring constant,鈥� he says.

Full believes that engineers should design their biomorphic explorers using
these kinds of principle. 鈥淚t would be very easy to work out the ideal spring
constant for a robot that could run on Mars鈥攜ou just take into account the
lower level of gravity and work out the numbers,鈥� he says.

The differences between these polypedal systems are important too. Humans are
highly manoeuvrable on two legs but not very stable. So a two-legged robot would
have to devote a relatively large amount of processing power to staying upright.
Insects, on the other hand, though less manoeuvrable are highly stable since
they usually have at least three feet on the ground. 鈥淚nsects are passively
stable鈥攖he control for their stability is embedded in the design,鈥�
explains Full.

Crabs, which have eight legs, are even more unusual. Although more legs means
greater stability, they are more likely to get tied up in knots. Crabs solve
this problem by walking sideways, which also allows them to take longer strides
and simplifies the construction of the legs. IS Robotics, a company based in
Massachusetts, has already built crab-like robots for hunting mines in surf.
Similar designs may be suited to exploring other oceans in the Solar System, on
Europa or Saturn鈥檚 moon Titan.

But NASA isn鈥檛 looking just at animals for inspiration. While polypedal
creatures invest a lot of energy in moving around, plants invest very little
energy in dispersing the essential parts of themselves. 鈥淧lants actually do a
nice job of moving around,鈥� says Kenneth Nealson, a senior scientist at JPL. He
believes that we could explore Mars by exploiting the planet鈥檚 natural energy
resources鈥攕uch as wind and gravity鈥攊n the same way that plants use
these resources on Earth. He says it would be relatively straightforward to
design a lander that released tiny sensors when the wind was strong enough to
disperse them. So the lander wouldn鈥檛 have to move and could save its battery
power for making measurements and sending the results back to base.

Batteries are problematical in any case because they do not work well at low
temperatures. Sojourner had to be fitted with radioactive materials keep its
batteries warm. But although the low temperature on Mars could be a
disadvantage, Crisp says it may be possible to exploit the huge daily variation
in temperature as a power source. The idea is to use strips made of two metals
with different rates of thermal expansion that flex as the temperature changes.
Explorers could use this flexing to crawl like snakes or burrow into the soil
like worms. 鈥淚magine these things inching their way across a Martian plain,鈥� he
says.

Giant strides

But just how useful would these devices be when the instruments they can
carry are clearly limited in size? Crisp believes that explorers must be able to
do high-quality experiments. 鈥淲hen you start reducing the size of your
instruments, the first thing you end up sacrificing is quality,鈥� he says.
Nevertheless, giant strides have been made in recent years to miniaturise
important instruments. Crisp points to high-quality cameras, optical
spectrometers and mass spectrometers that, even allowing for electronics and
batteries, each weigh less than 10 kilograms and fit in a milk crate鈥攁nd
the size keeps dropping. Lower quality versions can be made much smaller.

Of course, any results are useless if they can鈥檛 be sent back to Earth, so
communications are also important. Generating enough power to broadcast all the
way back to Earth is impractical. Instead, biomorphic explorers would relay
their data back to some orbiting mother ship or, like Sojourner, to a lander
which relays pictures back to Earth. The Mars Global Surveyor now in orbit
around Mars is designed to act as the relay station for other Mars missions.

Crisp and his colleagues are already thinking about small groups of explorers
that will be able to communicate with each other and map out large areas of
alien terrain. One idea is to map out the Jovian magnetic field using a number
of explorers armed with magnetic sensors. These would have to precisely
control their position relative to one another, rather like a flock of
birds鈥� not an easy task. The technologies now being developed for cellular
phones and wireless computers could help with this task.

Just when could this menagerie of high tech explorers begin their quest?
Researchers at JPL are drawing up a technology roadmap that predicts when the
various challenges will be solved. Much of the work is being done outside
NASA. For example, the Defense Advanced Research Projects Agency, a military
research organisation based in Arlington, Virginia, is funding a range of
projects to build robots that do everything from mine-hunting to security
patrols. So Crisp鈥檚 timetable is aggressive. 鈥淥ur focus is on sample return
missions,鈥� says Crisp. Biomorphic explorers could be ideal for sampling large
numbers of rocks and identifying interesting ones for return.

Crisp expects a biomorphic explorer to fly by 2007, possibly earlier. If he
is right, nature will have once again beaten humans in the race to explore the
Universe.