杏吧原创

IT鈥橲 a long way from Santa Monica to Mars鈥�56 million kilometres at
least, more if you circle Los Angeles to avoid the traffic. But it is here, in a
coastal town famous for its pier and for hosting the second O. J. Simpson trial,
that scientists and engineers have decided to begin the journey.

But the travellers won鈥檛 be human, of course. Humans are not scheduled to
explore the Red Planet until the year 2015 or later. No, the first Martian
adventurers will be robots currently strutting their stuff on Santa Monica鈥檚
beach.

Last month, the town hosted the 1997 International Conference on Mobile
Planetary Robots and Rovers. For the dozens of humans and machines gathered
here, the goal of exploring Mars by machine is almost a given. 鈥淩overs are
absolutely essential to any Mars programme,鈥� says Chuck Weisbin, head of NASA鈥檚
Robotics and Mars Exploration Technology Office. Robots won鈥檛 make the task
easy, he admits. They won鈥檛 do the job quickly. And building a reliable one is
not exactly a breeze. But you鈥檝e got to start somewhere.

And somewhere is a property behind the Loews Santa Monica Beach Hotel. It
doesn鈥檛 much look like Mars, despite the efforts of conference organisers, who
have laboriously scattered chunks of red volcanic rock over a quarter-acre of
white sand. 鈥淲e鈥檝e tried to recreate a little bit of Martian landscape,鈥� says
Louis Friedman, executive director of the Planetary Society, a California-based
organisation hosting the conference. 鈥淚 think it works pretty well, if you
ignore the Pacific Ocean in the background.鈥�

Blazing a trail

For the moment, that鈥檚 what everyone does, as all eyes focus on Rocky 7.
Rolling spasmodically over the rock-strewn sand, Rocky 7 is a more advanced
version of a rover currently en route to Mars. On July 4, the American
Pathfinder mission is due to land on Mars and use Sojourner, a stripped-down
version of Rocky 7, to prowl the terrain around the lander.

Neither Sojourner nor Rocky 7 appear very impressive. Superficially, both
look like oversized metal shoeboxes riding on six small wheels studded with
steel spikes that will help them carry out simple experiments, to test the
Martian soil and atmosphere. But these rovers are more complex than they look.
Sojourner must work in freezing Martian temperatures, so any moving parts must
be carefully lubricated with a substance that does not freeze. And because its
electronics cannot work at temperatures below 鈭�80 掳C , it must be
insulated and even battery-heated during the cold Martian nights when
temperatures drop to 鈭�110 掳C. The rover carries solar cells efficient
enough to convert even the weak Martian sunlight into usable power and, if these
break down, the batteries take over. Sojourner has a seven-day mission but once
the batteries wear out, its hours are numbered.

During this time, the rover must also be able to negotiate whatever terrain
it encounters. Its six wheels have a complex active suspension system with no
axles or springs, which allows the vehicle to negotiate obstacles twice the
diameter of the wheel. Anything larger and the rover must drive around.

Most important of all, Sojourner will analyse the surface of Mars and send
its signals back to the lander for transmission to Earth. The analysis will be
done by bombarding the surface with alpha particles and monitoring the
back-scattered radiation for the telltale signs of certain chemicals. Finally,
the entire machine must be no bigger than a milk crate and weigh only 11.5
kilograms, small and light enough to hitch a ride with the Mars Pathfinder
mission.

Rocky 7 is even more sophisticated. Tucked beneath its solar panel it carries
a folding 2-metre mast equipped with stereoscopic microcameras and other
sensors, such as laser-finders. These cameras snap pictures of the terrain
ahead, then feed the data to internal processors that produce three-dimensional
images and work out how best to proceed.

鈥淚t remembers all the moves it has made so that it can compensate and get to
where it wants to go,鈥� says Samad Hayati, one of Rocky鈥檚 designers at NASA鈥檚 Jet
Propulsion Laboratory (JPL). 鈥淭he human operator has only to give it one
command. Everything else it does itself.鈥�

By contrast, Sojourner is laboriously controlled from Earth. At the end of
each day, mission controllers examine photographs of its position taken by the
lander, and then radio back their instructions for the following day.

The successors of Rocky 7 will supposedly avoid this sort of labour by
working out at least some of the simpler problems by themselves鈥攕uch as
how to get around a big rock or when not to plunge over a precipice. 鈥淲e鈥檙e
going to need clever rovers,鈥� said Carol Stoker, a biologist at NASA鈥檚 Ames
Research Center in northern California.

This is the goal of scientists like Hayati, who reckon they鈥檝e made
tremendous progress over the past few years, particularly in developing ever
smaller processors capable of handling multiple, complex functions. The trick is
building a complete rover that works every time, even when it is millions of
kilometres away. 鈥淭here鈥檒l be no going back to the shop if something breaks,鈥�
admits Hayati.

Helping hand

Alas, such perfection remains out of reach. At the conference, Rocky 7鈥檚
demonstration turns out to be, well, a bit rocky. The rover seems to balk at
commands radioed from a human operator just a few dozen yards away. Told to
scoop up a small sample of sand, Rocky 7 complies using a small extendable arm,
but then refuses to dump the sample. Eventually, a NASA official slinks over and
physically turns the scoop. 鈥淚 don鈥檛 think that鈥檚 going to be an option on
Mars,鈥� mutters one onlooker.

Later, the Russians arrive with Marsokhod (Mars Mobile), a rover the size of
a refrigerator, made up of three jointed sections, each with a pair of
distinctive, ridged titanium wheels shaped like beer kegs.

Marsokhod has been around for a while. Four years ago, the Russians first
showed off Marsokhod in public tests conducted under the blazing sun of
California鈥檚 desolate Death Valley. But there were persistent problems
transmitting radio commands although, eventually, everybody declared the effort
a success and retreated to the shade.

This time, Marsokhod performs fairly well. The wheels grind noisily over the
volcanic rocks, as its flexible sections twist and tilt independently like a
caterpillar with a backache.

Then, inexplicably, Marsokhod halts, its stereoscopic cameras and other
sensors telling it to proceed no further.

鈥淲hat鈥檚 happening?鈥� asks an observer.

鈥淚 am not sure what this machine will do next,鈥� replies Viacheslav Linkin, a
scientist with the Russian Space Research Institute. 鈥淚t is not speaking with
尘别.鈥�

鈥淎re you saying it is out of control?鈥� asks the observer.

鈥淣o, it鈥檚 just not consulting 尘别.鈥�

Eventually, Marsokhod resumes its journey, crunching past a group of young
children on a day trip to the shore. They stop and gawk, then throw stones at
the rover鈥攕omething the Russian engineers hadn鈥檛 prepared for. The
children are politely hurried along.

What a climb-down

The Japanese rover is altogether different. Engineers from the National Space
Development Agency based in Tokyo proudly unveil a three-wheeler that, at this
conference at least, is in a class of its own. After lowering it from a
simulated lander (by rope, suggesting that the Japanese may still be working out
a few bugs) the rover limbers up three retractable legs, each of which ends with
an offset disc making it look like a bar stool on wheels.

But all in all, the Japanese rover scuttles across the landscape well,
something that cannot immediately be said for the Belgian prototype鈥攁
small, six-legged device that in theory scrambles across the terrain like an
ant.

Paul Alexandre, a researcher at the Free University of Brussels, hooks up the
insect-like rover, with its protruding metal legs and exposed circuitry. Pairs
of miniature servomotors in each leg whir into life, but the result is less than
satisfactory. The legs swing and twitch but the rover merely churns up the sand.
The effect is like watching a tiny chorus line suffer massive, simultaneous
hamstring failure.

Alexandre is mortified. He mutters softly, flipping the rover onto its back
for some quick diagnostic work. Legs up, it looks like a dead bug. But a day
later, Alexandre solves the problem and successfully displays his creeping,
crawling creation.

For the most part, the rovers are not much bigger than large suitcases. In
fact, says Friedman, the trend is towards ever-smaller rovers, partly because
they must fit into smaller rocket payloads, partly because improved electronics
have allowed them to shrink naturally. This is a mixed blessing.

Small rovers can go into places big ones cannot. 鈥淎 microrover, defined as
anything weighing less than 50 kilograms, doesn鈥檛 have to crawl over big rocks.
It can go around them,鈥� says Brian Wilcox of JPL鈥檚 Robotic Vehicles Group. Some
designers have proposed a large rover that can carry one or more smaller ones on
its back. These microrovers would be deployed like spiderlings on their mother,
to carry out specific tasks.

But being small also means being able to do less. 鈥淚n general, when you make
things smaller, you make them less sensitive. A smaller maximum aperture on a
camera, for example, means it collects less light,鈥� says Carl Kukkonen, director
of the Center for Space Microelectronics Technology at JPL.

Ultimately, rover designers want to build vehicles capable of ranging over
hundreds of kilometres for up to 1000 days. These semiautonomous rovers would
conduct multiple experiments, analysing the Martian soil and atmosphere at
various sites, or collecting samples and bringing them back to the lander where
larger, more sophisticated equipment would do the tests.

鈥淩over technology has come a long way in the last few years,鈥� says Donna
Shirley, who manages the Mars Pathfinder mission at JPL. 鈥淔or a long time,
scientists weren鈥檛 much interested in rovers. They didn鈥檛 think they would
amount to much. But that鈥檚 changing. Everybody wants one now.鈥�

Back on the beach, Ben Abbott of Utah State University is directing ARC-III,
a toaster-sized rover that rockets about on six wheels, each containing its own
motor and power source. ARC-III isn鈥檛 really a Mars rover, as it has rubber
wheels which would freeze and crack in the cold, but it proves nonetheless to be
one of the conference鈥檚 biggest hits.

Zipping over the boardwalk, directed by Abbott and signals from a Global
Positioning System satellite navigation unit, ARC-III races up to passers-by,
spins, then dashes off. 鈥淧eople do one of three things when they see a rover,鈥�
says Abbott. 鈥淭hey ignore it. They run away. Or they kick it.鈥� The big question
is: what will Martians do?