Autoblog

After the DARPA dust settled, only 40 minutes separated the first, second and third place contenders for this year's $2 million bounty. The Carnegie Mellon team, behind the virtual wheel of a tech'd-out Tahoe dubbed the "Boss," won the DARPA Urban Challenge, the first event held in a mock city environment.

The Mellon crew beat out Stanford, Virginia Tech and MIT, and according to Wired's report, none of the top competitors were cited for traffic violations – something that most manned vehicles can't claim. Average speeds ranged from 13 to 14 mph over the course of the 55-mile trek. There was no mention of top speed and we're still uncertain about what specific obstacles were set out to befuddled the ghosts inside the machine.

There's no word yet on whether the DARPA event will continue, as many maintain that since the technical end of things have been sorted, it's time for private firms to take the lead. We'll have more on DARPA when Mr. McElroy reports on the event later in the week.

[Sources: GM (Press Release after the jump), Wired]

The Defense Advanced Research Projects Agency (DARPA) is part of the US Defense Department, and it provides funding to all kinds of researchers for projects that might at some point have a useful military application. Over the last several decades they have funded many projects, some of which have made it into the civilian mainstream, like a little network you might have heard of called the internet (which was originally born as ArpaNet). One of the better known recent projects, is the DARPA Grand Challenge, the first running of which in 2004 turned into a complete fiasco. The $1 million prize for the first running went unclaimed, because none of the vehicles completed the 142 mile course through the California desert. In fact, the farthest anyone got was seven miles, with most of the teams not even getting off the starting line.

Eighteen months later, with the prize money doubled, they tried again, and this time a team from Stanford University finally succeeded (along with three other teams), and completed a 132-mile course from Barstow, California to Primm, Nevada. The Stanford team equipped a Red-Bull sponsored Volkswagen Touareg, nick-named Stanley, with more computer power than the Apollo moon missions, and the SUV ran the distance in a little over 7 seven hours with no human intervention once the race began. In November, DARPA will try again, this time with an Urban Challenge. Instead of a desert course, the teams will traverse a sixty-mile mock urban environment. For this year's event, Stanford has created Junior, a VW Passat wagon packed with laser rangefinders, radar, GPS, and enough Intel Core 2 Duo processors to power a small data center. The car is not allowed to crash and must obey all California traffic laws while it's running, which would likely make it the only vehicle on the road to do so. You can find out more about Junior at the Stanford Racing Team web-site, and learn about the event at the DARPA Urban Challenge site.

[Source: Stanford Racing Team via Engadget]

Named in honor of Herbie the Love Bug, the famous number 53, Volkswagen's GTI '53+1' is car that requires no human nanny. It's a fully automatic car that can negotiate twists, turns and straights at the edge of its performance envelope, all without the aid of a driver's touch. The '53+1' uses radar and laser sensors as electronic eyes, much like many of the entrants in the DARPA Grand Challenge, to read the road ahead, and a satellite navigation system to pinpoint its position to within an inch.

The '53+1' was let loose on a race circuit and reportedly can reach its top speed of 150 mph with no driver on board. To prove the system's abilities, Volkswagen allowed guests at an event to design different road courses using cones and then watched as the driverless GTI ran through them. It even ran through courses faster than the company's own engineers could manage.

While the linked article is sparse on the down-and-dirty details of how the system works, it appears that it's set up to recognize only cones, or at least simple objects on an uncluttered road course. Still, the next DARPA challenge takes place in an urban setting and Stanford, winners of the last challenge with an autonomous Touareg, are planning to use a Passat this time around. If these two ever team up, in the near future we'll be able to order a Rabbit with a Twincharger engine, DSG and Auto Pilot. Now that's innovation.

[Source: Daily Mail]

We had the pleasure of catching Dr. Sebastian Thrun's keynote presentation at the 2006 Sensors Expo on Tuesday, where he spoke to the audience about his experience leading Stanford's DARPA Challenge entry to victory last year and how the project's success may affect the way we drive in the not-so-distant future.

The technical content of "Stanley" is fascinating, of course. Due to the relative inaccuracy of GPS (2 meters of error doesn't cut it on a 2.5-meter-wide mountain pass), a variety of sensors were used for environmental recognition, allowing the vehicle to discern "good" terrain from non-drivable areas. A array of five lasers scan the area in front of the vehicle for obstacles, and their downward trajectory combines with the forward movement of the vehicle to allow for the return of 3D data. Unfortunately, the 20 meter range of the lasers and the 70 Hz scan rate limits the maximum practical speed of the vehicle to 35 MPH or so when relying solely on this navigational device.

To allow the faster travel speeds thought necessary to secure a victory, an optical camera was implemented to learn from the laser system how drivable terrain appears, eventually allowing the vehicle to see much further down the intended path and thus facilitating higher speeds. Doing so isn't as easy as it may seem - color can't be used due to the wide variety in coloration of drivable surfaces (roads may be brown, black, red, tan, or any number of other shades), and texture detection is nearly worthless as Thrun noted that the sky is smoothly textured but yet not a practical path. Radar is also used to provide ranging information.

[Click through for much more on this vehicle and the future of (not) driving...]

 

We all know the basic facts about the DARPA Grand Challenge that pitted over 20 autonomous vehicles against each other and the elements out in the Mojave Desert last October. After watching the NOVA special The Great Robot Race on PBS last night the entire enterprise has taken on a whole new dynamic thanks to the excellent backstory provided by the program that reaches all the way back to the first DARPA Grand Challenge in 2004.

While The Great Robot Race features background on many contenders, including a “nobody” darkhorse team from New Orleans that finished fourth overall just weeks after Katrina blew through its hometown, the main push of the program centers around the two teams from Stanford and Carnegie Mellon. Sebastian Thrun heads the Stanford team, which is actually comprised of Carnegie Mellon defectors including Thrun himself, and they enter Stanley, a Touareg donated by Volkswagen, in the contest. Red Whittaker, an ex-marine turned college prof who presides over an army of graduate students and a pair of autonomous H1 Hummers named “H1ghlander” and “Sandstorm”, leads the Carnegie Mellon team.

Read on for more of the gritty details that led up to a showdown of these two teams in the desert.