PopSci’s Darpa Grand Challenge Preview

Darpa's blue team making tweaks to the Ghostrider
Howard Chau, Anthony Levandowski and Bryon Majusiak [left to right] make some last-minute tweaks to get the Ghostrider ready for race day. Courtesy Maxime Levandowski

In this, the first of a series, Popular Science profiles one of the favored teams competing to win the Darpa Grand Challenge autonomous vehicle race, which will take place on Saturday (October 8) near Primm, Nevada. Today we look at the Blue Team’s autonomous motorcycle. Stay tuned to popsci.com for more previews throughout the week, and for minute-by-minute videos and updates on race day.

The Blue Team
University of California at Berkeley and Texas A&M

A cultlike group of engineering students and recent grads, including 25-year-old team leader Anthony Levandowski, 25-year-old Charlie Smart, 19-year-old Bryon Majusiak and 23-year-old Howard Chau.

The Ghostrider, a 90cc off-road motorcycle that’s stripped down to its metal skeleton and tricked out with two cameras, 12 optical encoders and a GPS module. These sensors feed information about the surrounding terrain to two on-board 2.2GHz Opteron processors, which synthesize the data and steer the bike accordingly. “Most people are entering trucks and Hummers, but there’s a lot of potential for a lighter unmanned vehicle to be used in the field-one that can go through narrow or tight spaces,” Levandowski says.

Last year’s Ghostrider collapsed at the starting line when Levandowski forgot to turn on a critical balance controller. “It was so frustrating,” Smart says. “It’s like if you call tech support and say your computer’s not working, and they say, ‘Did you turn it on?’ ” Chasing redemption, Blue Team members regrouped, refueled their budget with contributions from companies like Agilent and Raytheon, and set their sights on 2005.

Spotlight tech
Engineering an unmanned motorcycle presents unique challenges-unlike its four-wheeled competitors, the Ghostrider must be able to balance on a single axis. To help stabilize the bike, Chau has designed a pair of robotic arms, one on each side, that muscle it back into a standing position if it tips over. “They’re like glorified kickstands,” he says. Self-adjusting camera mounts ensure that the vehicle’s visual sensors stay level and pointing straight ahead, even as the bike leans into tight turns and rumbles over obstacles.

Working style
Very intense. Most team members put in 12 to 16 hours every day at a corrugated-metal warehouse a few miles from the Berkeley campus. Often they don’t finish tinkering and typing until 2 or 3 a.m.-if they go home at all. “Sometimes I leave at night and come back the next morning, and Anthony’s still there,” Majusiak says.

Piggy bank
The University of California at Berkeley and corporate sponsors have chipped in several thousand dollars apiece, but Levandowski, who moonlights as a successful online entrepreneur, has contributed over $100,000 of his own savings to the effort.

Odds of winning 20:1.
We like Levandowski and co.’s innovative approach, but a wobbly two-wheeler might not make the grade in this year’s race, where reliability and endurance will be the key deciding factors.

For more coverage of this year’s Darpa Grand Challenge, including up-to-the-minute field reports and videos, check out our Darpa Grand Challenge headquarters. You can also visit Darpa’s official site here.