Jalopnik: Darpa Grand Challenge

DARPA Winner's Victory Lap To End At Smithsonian

Travis Hudson — Mon, 23 Jun 2008 15:00:00 EDT

The Volkswagen Touareg that was crowned as the first DARPA Grand Challenge winner in 2005 (the first challenge was in 2004, but no winners were crowned) is making a grand tour that will eventually lead the fully autonomous vehicle to the top dog of museums, the Smithsonian. The Touareg was nicknamed "Stanley" since it's so much easier to pronounce.

Stanley was developed by Stanford University's School of Engineering with input from Volkswagen, MDV-Mohn Davidow Ventures and Intel. It's completely autonomous, meaning it requires no driver, making its completion of the 132-mile Mojave Desert course in a little less than seven hours at an average speed of 19.1 mph all the more impressive. Technology developed for Stanley has helped advance technologies being used by vehicles today, including adaptive cruise control and electronic stabilization.
[Gizmag]

Robo-Cars Take Next Step, Standardized Communications With JAUS

Ben Wojdyla — Mon, 14 Apr 2008 17:00:00 EDT

After proving unmanned ground vehicles could do things all on their own in the DARPA Grand Challenge and then following up with the Urban Challenge it was time to start talking the same language. SAE has standardized the communications protocols for umanned vehicles and has given it a sexy name to boot — Joint Architecture for Unmanned Vehicles or JAUS, see, sexy. Anyway, it's basically a method which allows one unmanned vehicle to talk to another about situational conditions, location, direction, payload, and other pertinent information

All that sexiness ends when you realize humans are just making it that much easier for robots to coordinate attacks on soft and hardened targets. You just wait. It'll be all fun and games until an autonomous Toyota Highlander smartass does donuts on your lawn while a Toyota robot plays taps on your doorstep. Then who will be the paranoid one, huh?

Toyota Grand Prix To Feature DARPA Race

Travis Hudson — Tue, 01 Apr 2008 15:30:00 EDT

The Toyota Grand Prix, slated for April 18 - 20, will feature more than F1 ~~ChampCar~~ IRL The Tony George Racing League cars hauling ass around the track. As part of the Long Beach festivities, a Robotic Grand Prix will pit three of the biggest, most badass unmanned vehicles against each other. Junior, Boss and Ben are the three robo-racers that will tackle the 1.97-mile 11-turn circuit course. The three completely autonomous vehicles will have to complete one lap of the course and the fastest time will be awarded first place.

Boss is a 2007 Chevy Tahoe that was the work of student and faculty at Carnegie Mellon University. This is the same Boss that won the DARPA Grand Challenge. Junior is a 2006 VW Passat that was the runner up at the DARPA Grand Challenge and Ben is a Toyota Prius that took sixth at the Grand Challenge.

Don't get too pumped, the three vehicles have an average speed in the teens when completing a course. It may sound like a yawner, but autonomous vehicles are the future. Haven't you seen i Robot? [GPLB via BotJunkie]

DARPA Robot Car Crashes Into Ford Taurus

Ray Wert — Tue, 06 Nov 2007 10:00:00 EST

Although the DARPA Grand Challenge has already declared a winner and we've even done a wrap-up post already — we felt the need to run this week-old video of the Axion "Spirit" racing vehicle taking a left turn directly into one of the human-driven Ford Taurus chase vehicles. Yes, This may be our first non-human Hoon of the Day, but with all our DARPA coverage from the past week or two, we felt it necessary to welcome rather than discriminate against our new robotic driving overlords. [via Gizmodo]

DARPA Grand Challenge: Wrap-Up

Mon, 05 Nov 2007 14:15:01 EST

After criticizing MIT's bot for overtaking Cornell's car when it got confused by a barrier and causing a minor bumper brushing, it's come to my attention no points were deducted from either team. In fact, no teams were penalized for traffic infractions of any sort, and the final scoring came down to finish times. MIT managed to come in first place of the teams that didn't win spots on the podium, and that's pretty impressive for a squad conspicuously absent from previous races. MIT's bot, with its extensive array of sensors and custom-build hardware running around 100,000 lines of code written just for the event, was the most geek-tastic of the event. Even though they had 16 out of 40 cores running vision processing from their optical cameras, they still had a little trouble navigating the dirt road portion of the course. Unfortunately, apart from MIT's fourth place win, DARPA will not be providing final scores for the rest of the pack. Happy future defense contracting to all!

DARPA Grand Challenge Round-up

DARPA Grand Challenge: And the Winner Is...

Sun, 04 Nov 2007 15:17:48 EST

After all points were tallied, Carnegie Mellon's Tartan Racing was the winner of this year's DARPA Grand Challenge. In second place was Stanford, with Virginia Tech in third. As podium finishers, the teams will receive $2 million, $1 million and $500,000, respectively. Currently, there are no plans for a fourth challenge, and murmurings from DARPA staff are that autonomous-vehicle development has reached a point such that it no longer needs seed money to flourish and evolve further. Hopefully they're wrong though, because the Grand Challenge a great event and tremendous motivation for public imagination. Maybe the teams will just bathe in Cristal tonight. – Curtis Walker

DARPA Grand Challenge: Team LUX's Lament

Sat, 03 Nov 2007 18:13:56 EDT

I'd been interested in Team LUX from the first time I saw their car in person. That's why I was really hoping to see them in the final race. Unlike some other teams, they stuck around for the finals after being disqualified a few days ago. Turns out they were eliminated because their car was unable to navigate on the dirt portion of the track. Without lane markers or any obvious edges to follow, the car simply didn't know what to do. They also had some technical issues. Their car died once during qualifying, but nobody knew whether it was a car problem, a computer glitch or the kill relay system. Also, in a very unlucky move, the car went kamikaze on DARPA's Dr. Tether's chase vehicle. While the Team-LUX members are confident a collision would not have happened — after all that's their specialty — the good ol' Doc was less optimistic and ordered the kill command be sent. No team knows their scores or the final nail in the coffin of their disqualification, so it's hard to say for sure, but in the end, the dirt road portion of the course would have killed 'em anyway. – Curtis Walker

DARPA Grand Challenge: Update from the Course

Sat, 03 Nov 2007 18:05:23 EDT

Out here on the urban challenge course in Victorville, California, it's UPenn, MIT, Stanford, Carnegie Mellon, Virginia Tech and Cornell out of the original 11 left in the DARPA Grand Challenge. Of these, all said teams' autonomous vehicles have completed mission two as of 1:05 pm PT, with Cornell the final team to finish. All the cars are racing on the same course, but each has a unique mission plugged into it, meaning they have to interface with each other as well as a team professional drivers behind the wheel of several stunt Ford Tauruses. (More after the jump.)

The monster TerraMax Oshkosh truck fouled out when it hit a defunct base exchange. UCF went out of bounds in a residential area, hitting the side of a house. Honeywell was pulled out of the race for driving outside of its lane consistently. Team AnnieWay got stuck at an intersection. CarOLO was also disqualified.

MIT aggressively attempted to overtake Cornell, who's been having a bit of trouble keeping its vehicle moving. The two collided, with MIT at fault. Scoring isn't publicly discussed, but this could be a pretty big penalty. Stanford's Junior is similarly aggressive when the route is blocked, but so far has avoided collisions.

Stanford and Carnegie Mellon's entries are the two most confident and speedy robots on the course, with Virginia Tech right behind them. Watching the cars drive is almost like watching a person drive. They're quick and take corners well. So far, they've been the most fun to watch.

Stanford was the first to complete all of its missions, but final scoring will determine the ultimate winner. (I'm betting it's Stanford). – Curtis Walker

DARPA Grand Challenge: Corporate vs. Collegiate

Mon, 29 Oct 2007 08:49:01 EDT

Three-and-a-half million in prizes is a pretty big deal, and this year, the DARPA Grand Challenge will be webcast live on Saturday, Nov 3 starting at 6:00 am (PT). The event is free for anyone willing to brave a trip to Victorville, California, a Mojave desert town 81 miles northeast of Los Angeles. With 35 semifinalist teams competing, the easiest way to split them up is by whether they're primarily corporate or collegiate, which gets a little tough because there is a fair amount of crossover. What the corporate vehicles seem to have in common is an interest in actually being able to deliver their product to the end user, whereas the college teams seem more interested in trying things out that may not even be necessary. Installs seem to be pretty clean and sensor arrays are relatively modest. One team breaks away from this stereotype: Team Axion's car features gobs of hot glue, a PS3 to run the rear motion cameras, and windshield wipers for their cameras up front. They're also the only team to have brought the same vehicle back for each challenge. Insight Racing is another team with an interesting to look at vehicle: a Lotus with several MacMini's serving as its brain. It seems like only a few teams
really understand the value of a sexy platform in generating interest.

My heart was briefly warmed by the University of Central Florida pit when I saw their budget minded Subaru, a one-time DARPA Challenge veteran. Rather than dedicating the bulk of their tight budget to a fancier sensor, they adapted their Sick LIDARs to custom wobbler units that rotate the units through 90 degrees of motion in order to double their viewing angle. This is one of the more clever approaches in the field, but time will tell if it actually works.

DARPA Grand Challenge: Million-Dollar Babies

Sun, 28 Oct 2007 12:00:30 EDT

More than a year ago, DARPA awarded 11 teams $1 million each in seed money for vehicle R&D. Of the 60 applicants, Stanford, MIT, Carnegie Mellon, Autonomous Solutions, Virginia Polytechnic, CalTech, Cornell, The Golem Group, Honeywell, Raytheon and OshKosh were the only ones with proposals deemed worthy of the prize. So, what's a million dollars get ya these days? Anything from a mostly stock Toyota Highlander Hybrid to a gigantanormous robotic monster truck that belches Terminator sounds when in autonomous mode. Every team has a unique approach, and while similar hardware shows up on many of the vehicles, nobody seems to have the exact combination to match anyone else. Case in point. Last year's winner thinks it can get the job done without any optical cameras; Autonomous Solutions figures Windows XP is up to the task of commanding an SUV. Things are bound to get interesting on the A Track, and we're not even to the privately funded B Track entrants. [DARPA Grand Challenge]

DARPA Grand Challenge: A Battle of Wits

Sun, 28 Oct 2007 08:24:53 EDT

[Jalopnik's desert stringer Curtis Walker is following the DARPA Grand Challenge, an annual race of autonomous ground vehicles, with a serious case of the techies. We'll have his reports today on the National Qualifying Event semifinals, like this one, as soon as he can type them into his digital audio-visual receipt and transmission unit. - ed.] Among the numerous newcomers to this year's DARPA competition, Massachusetts Institute of Technology seems poised to win the imaginary award for most gizmotastically outfitted. Armed with a cool million in seed money for development, team MIT went all out with the sensors and CPUs. All told, they've got 11 Sick Lidars (Light Detection and Ranging) units, five optical cameras, 15 Delphi long-range radars, GPS and a 40-core supercomputer to run them. All this power comes at a price in the form of electricity required and heat generated. To address this, they installed a 6kW generator and a 2kW roof mounted AC unit. Perhaps even more impressive is that the 100,000 or so lines of code running on the system were all written for this event. [Next: Team Lux, and gallery.]

At the flip side of MIT's absurdly outfitted LandRover is team Lux's Volkswagen Passat, a joint venture between German supplier of automotive sensors Ibeo and parent company Sick. Hands-down winner of the nonexistent prize for stealthiest install, their diesel grocery-getter has a mere three sensors; one on each corner up front, mounted behind black Plexiglas in the bumper and a prototype model mounted in the center of the read bumper. Data from these units is collected and processed by four computers in the trunk. Even the actuators for steering and velocity are stealthily mounted. Apart from the big honkin' kill switches on the read windows, one might miss the fact that it's a robot car at all.

It's hard to pick favorites in an event like this, and looking at this brief comparison shows why. Despite their differences, both vehicles did well in Saturday's qualifying runs. [DARPA Grand Challenge]

Terramax Prepares for Driverless Takeoff

Mike Spinelli — Wed, 03 Oct 2007 13:00:58 EDT

The Oshkosh TerraMax unmanned truck hasn't quite rocked the DARPA Grand Challenge in the past, so the company has a lot to prove this year in Victorville, California, where the grand spectacle of mecha robots takes place later this month (we'll be there). That means the company has a lot to prove, considering it wants a piece of the US defense budget atop a congressional mandate stating one in three ground combat vehicles must be self-driving by 2015. The maker of airport firetrucks and other large industrial vehicles has invested tens of millions of dollars in the project, and says the TerraMax is competitive this year, with lasers, high-tech camera gear and global positioning systems to guide it through the punishing desert course without wandering off into a cactus-filled netherland. We'll see. [Yahoo!]

Stanford Takes the Message to the Streets, DARPA-Style

Davey G. Johnson — Mon, 19 Feb 2007 12:45:00 EST

One may recall Stanley, the robo-Touareg that won the 2nd DARPA Grand Challenge. Well, now the boffins in Paly are working a successor to the SUV, based on a Passat known as Junior. Junior's designed to navigate an urban environment, rather than a desert off-road course. DARPA plans to announce the location of the next race, to be held November 3rd, sometime in October. Meanwhile we're still waiting for information on the atomic-powered Saab 96 entry from Venture Industries. – Davey G. Johnson

Urban road race to test limits of robotic cars [Reuters]

DARPA Robo-Car Challenge Decides Not To Listen To Mom -- Will Play In Traffic

Ray Wert — Tue, 02 May 2006 13:24:50 EDT

We bet DARPA's mom is super upset with them, cause yesterday they decided the next Defense Advanced Research Projects Agency Car Challenge is gettin' a taste of city life. After four vehicles finished the 142-mil3 desert course without human intervention last year, the third Grand Challenge will take place in a mock urban area. Thank G-D it's mock, and so therefore no need to worry about crazed robo-cars running down moms pushing strollers in the middle of downtown — DARPA's leaving that job to the taxi drivers.

The New DARPA Robot Car Challenge? City Driving [PC Magazine]

Related:
More on the DARPA Grand Challenge [internal]

PBS's "Nova" to Feature DARPA Grand Challenge Tonight

Mike Spinelli — Tue, 28 Mar 2006 10:10:19 EST

Prop-heads set your Tivos on stun. PBS's "Nova" tonight will feature a look at last October's DARPA Grand Challenge autonomous vehicle race, on a episode it calls "The Great Robot Race." The program will follow contestants in the lead-up to the race followed by the competition, in which a Volkswagen Touareg named Stanley, built by the team of Stanford University and Volkswagen Electronics Research Laboratory (ERL) won by completing the 132-mile desert course in six hours and 35 minutes. Geek rating: A[sum(X[j]*Y[j], j=1,10)]+.

Related:
Four Entrants Finish DARPA Grand Challenge [internal]

Four Entrants Finish DARPA Grand Challenge

Mike Spinelli — Sun, 09 Oct 2005 10:23:09 EDT

This results of this year's DARPA Grand Challenge robotic (sorry, autonomous) vehicles derby were far different from last year's. For one, an entrant actually finished the course. For two, though most entrants still couldn't figure out how to turn their sensors and algorithms into directional prowess, at least four entrants had finished the course by Sunday morning. The winner was "Stanley," Stanford's autonomous Volkswagen Toureg, which edged out two, Hummer-based competitors from Carnegie Mellon University's Red Team and an entry from the Gray Team. A full report is expected later today.

Four robots finish 132 mile DARPA Grand Challenge course [Gizmag]

Rat Fink Goes DARPA: The A.I. Motorvators Team

Davey G. Johnson — Tue, 26 Jul 2005 11:28:35 EDT

What happens when you put together a crew of hot-roddin', Linux-geekin' guys in search of a cool two mil from Uncle Sam? You get the A.I. Motorvators' creation for the DARPA Grand Challenge. Featuring a Chevy motor, independent front suspension, and bodywork that lies somewhere beyond Thunderdome, the Motorvators' machine uses inertial and DGPS guiding systems with a battery of sensors (including curb feelers!) to wind its way through the desert. Plus, they put a big Rat Fink on the front. Red Team doesn't have a Rat Fink.

A.I. Motorvators [via Technocrat]