Stanford team clinches top spot in robot desert race
By Reed Stevenson
PRIMM, Nevada (Reuters) – A Stanford University team won a
$2 million prize on Sunday for sending a modified Volkswagen
across 132 miles of rugged desert, guided only by sensors and
computers in a race the Pentagon hopes will lead to a
technological breakthrough in warfare.
Twenty-three driverless vehicles were sent into the Mojave
Desert on Saturday in a race sponsored by the Pentagon’s
Defense Advanced Research Projects Agency, known as DARPA.
After extending the race a day to allow a slow-moving
robotic vehicle to finish, the Stanford University vehicle,
known as Stanley, was declared the winner of the Grand
Challenge with the quickest race time of six hours and 54
minutes.
Sebastian Thrun, leader of the Stanford team, said the
victory was a win for the automobile’s future, predicting that
all cars would one day be able to drive themselves.
DARPA sponsored the race to spur development of driverless
vehicles that one day could carry water, fuel and other
supplies for the U.S. military in war zones.
In last year’s inaugural race, called the Grand Challenge,
every machine failed within sight of the starting line. The
Pentagon decided to double the prize and hold the event again
this year.
Organizers designed a more difficult course this year and
an assortment of modified Humvees, sports utility vehicles,
pickup trucks and dune buggies were sent into the mountains and
valleys in the Mojave to navigate man-made obstacles, tunnels
and a dry lake bed.
“These vehicles just haven’t achieved world records,
they’ve made history,” said DARPA director Tony Tether.
One broke down at the starting line and 17 others stopped
moving at various points on the course.
Four made it back to the finish line to complete the race
and one, a huge six-wheeled truck called TerraMax, was stopped
overnight so it could make it to the finish line.
Coming in just after Stanley were a pair of modified
Humvees built by a Carnegie Mellon University team. A modified
sports utility vehicle called GrayBot also finished just after
sunset on Saturday.
The rugged, twisting course, about 40 miles southwest of
Las Vegas on the Nevada-California border, was chosen because
of its similarity to terrain where the U.S. military is
currently most active, Iraq and the Middle East.
At one point, the vehicles had to climb through a steep
valley that organizers said was “reminiscent of a mountain pass
in Afghanistan.”
Using global positioning satellites and inertial
navigation, the vehicles were programmed to follow a
pre-defined course disclosed only hours before the race. Radar,
lasers and cameras mounted on the vehicles guided onboard
computers that steered the vehicles around obstacles.
