Stanford's Stanley wins DARPA Grand Challenge

tonyquan writes "DARPA has just announced that Stanford's "Stanley" autonomous ground vehicle has won the Grand Challenge, a $2 million contest for driverless vehicles over a 132 mile course in California's Mohave Desert. Stanley's winning time over the course was 6 hours, 53 minutes and 58 seconds, for an average speed of 19.1 mph. Second was Carnegie Mellon's Sandstorm (7:04:50), third went to another CMU vehicle "H1ghlander" (7:14:00) and fourth to the Gray Team's KAT-5 (7:30:16) More info from DARPA."

so wait..

[molo (94384)]

Last year they had NO vehicles even make it out of the obstacle course.. and this year they had several vehicles actually complete the desert course?? What gives?

-molo

Re:so wait..

[Zathrus (232140)]

There are several factors here. First and formost, the vehicles are more capable. The software is vastly better and the hardware is somewhat improved. Did you think that they've been sitting around doing nothing for the past 17 months? They've been working on improvements since the last challenge, and they've spent a lot more time actually testing their vehicles in desert terrain.

There are some people who say that this year's course is far easier than last year's. I don't know myself -- I'm not involved with any of the teams and I don't have detailed knowledge of the courses. But there has been some commentary by those involved to this effect, as well as from bystanders. One huge difference is that the most difficult part of the course (Beer Bottle Pass, a narrow road with a steep drop off on one side) was at the end of the course this year, while the equivalent part was near the start last year. Stanford's leader is quoted as saying something to the effect that if they'd inverted last year's course then a lot of cars would've gone much further, even if none of them finished. The complete lack of media attention last year may have been one reason why DARPA swapped the course around -- it's rather anti-climactic to write about a race where the best racer hardly even got off the finish line and leads to the kind of stories DARPA really doesn't want to see (waste of taxpayer money, will never work, etc.).

In any case, given that less than 25% of the vehicles finished, I'd hardly say that it was a trivial thing to do. It's still amazing. Congrats to those who did, and to all of those who participated for that matter -- it's quite an accomplishment, even if there's a long way to go still before this is really usable in a real world environment.

Re:so wait..

[EEJD (901217)]

It's not so much an improvement in the AI as it is an improvement in the sensors. These vehicles look ahead about 30 feet and plot their course based on very simple logic. If there is a negative obstacle (a hole), it is more difficult for sensors to detect than if there is a rock sticking up in the path. Last race, the only thing that stopped red team was a hairpin turn. Their sensors looked straight ahead and only a little to the sides, but when faced with the hairpin turn, the vehicle almost fell off the side of the mountain! But the rules of the AI haven't changed much- just the sensors. If you're driving through jungle, for example, you have to have sensors that don't see leaves as obstacles. Otherwise the path will look totally impassable.

Re:Luckily...

[CreatureComfort (741652)]

Actually, planetary rovers are just a tiny, tiny portion of the reason for this challenge, otherwise NASA would be sponsoring this, not DARPA. The primary reason for this challenge is for troop supply and support vehicles that can accompany troops into a battlefield, or be sent in autonomously. Which means the jungle scenario is non-trivial. One of the reasons the challenge is being held where it is, is due to the development lifetime projected force deployments being in mainly desert regions. Another major projected use for these kinds of vehicles is for deployment in a bio-hazardous area for testing and sampling in an autonomous measure. But once again, the is a DARPA challenge, not a CDC one.

Re:so wait..

[by Anonymous Coward]

The universities competing in this competition know perfectly well they're helping the armed forces kill people.

You're making the common mistake of assuming that the purpose of the military is to kill people. It's not. The purpose of the military is primarily to defend your country, and secondarily to defend other people where this is deemed beneficial to your country's interests. Killing people is one of the ways this is done, but the primary goal in a war is to persuade the enemy to surrender, not to kill as many of them as possible. If you can use smart weapons and special forces to take out their infrastructure or their commanders, you can get the majority of the opposing forces to give up. Similarly, the average soldier, faced with an enemy that knows no fear, feels no pain, and has nothing to lose but money - in other words, an unmanned assault vehicle - is not going to go out and fight it if he can help it.

Oh, and I'll just add at this point that the most recent thing I heard in the media about the US army was this: that they just sent eight military helicopters to help survivors of the earthquake in central Asia. That's not "killing people". That's your army spending a heckuvalot of money to help people who are not only foreigners, but, by and large, actually hate America. This is called "doing good", and I speak for much of the world when I say that we admire America when it does good. And it doesn't take much imagination to think of other ways America could do good, if it had better AI and robotics technology at its command: think of small autonomous reconnaisance robots, being used to locate survivors in the rubble.

Re:so wait..

[Dun Malg (230075)]

I'm sorry, but there's no pretense about this: the competition is designed to help the defense department deliver on its promise to congress to get most of its ground assault vehicles unmanned in the 2010 to 2015 time frame. They state it explicitly, and all over the place. The universities competing in this competition know perfectly well they're helping the armed forces kill people.

By all means, don't allow facts to get in the way of your hysterical editorializing. The DoD is not developing an unmanned ground assault vehicle, and they do not state that explicitly at all. They are looking to procure an unmanned cargo carrying vehicle by 2015. You will, of course, probably point to how everything in the military is designed to support operations and is therefore contributing to killing people, but that'd just be weaseling. You clearly thought they were developing killer robots. Let's hear it for reading comprehension! Moron.

Re:so wait..

[Dun Malg (230075)]

So yeah, it seems utterly clear that the DOD has no plans to incorporate technologies for ground navigation into assault vehicles.

Nice attempt to braoden and generalize, weasel. You said:

"the competition is designed to help the defense department deliver on its promise to congress to get most of its ground assault vehicles unmanned in the 2010 to 2015 time frame."

Clearly you're talking specifically about the Grand Challenge project, for which the time frame of "2010 to 2015" is relevant, and this is what I addressed. Stands to reason, as that's what this /. topic is about. You added in to word assault between "ground" and "vehicle" based on some internal editorial bias. I never claimed the DoD isn't interested in autonomous armed combat vehicles. I am simply stating the fact that the DARPA Grand Challenge is the preliminary step in developing an unmanned, unarmed, cargo carrier to be fielded in 2010-2015. This is what the linked DARPA site says. This is what all the articles say. You flasely claimed otherwise.

Re:so wait..

[cloudmaster (10662)]

It'd probably be better if our armed forces stopped inovating, and just waited for the rest of the world to advance beyond our own capabiilty to respond, eh? Because, if we stop concerning ourselves with war, the whole world will instantly fall into a state of peace and be covered with pretty flowers.

Obviously, any students who learn more about effectively automate vehicles will *never* find a way to apply that technology in a non-lethal environment...

http://www.edmunds.com/insideline/do/News/articleI d=107011 [edmunds.com]
http://www.jaguarusa.com/us/en/vehicles/s-type/pri ces_and_specs/opt_equipment.htm [jaguarusa.com] (look at the first available option)

Yes, anything that can potentially be used to kill someone [256.com] should be off-limits for research, regardless of its usefulness in other arenas. Especially if, heaven forbid, the *military* encourages development!

Good news

[Data Link Layer (743774)]

I honestly didn't think this contest would ever be won. Maybe in 20 years we can have auto driving cars that can make it so there is next to 0 car accidents.

Re:Good news

[cerberusss (660701)]

Maybe in 20 years we can have auto driving cars

I can already imagine the following scene:
You: *steps in car* "Drive me to uncle George?"
Car: "Why do you want me to drive to uncle George"
You: "Because it's his birthday, dammit. Now start driving!"
Car: "You seem to be a bit angry. Where does this anger come from?"
You: "Start DRIVING you gas-guzzling piece of shit!"
Car: *accelerates to 100mph* *dumps core*

Re:Good news

[lowrydr310 (830514)]

Is your dune buggy street legal?

It's funny how our laws are written. I know a guy who built a Lamborghini Countach kit car out of some steel tubing, a pre-made fiberglass body, and an engine that isn't even close to passing federal emission laws. He had no problems getting it inspected, registered, and getting a license plate for it. Custom choppers are the same: It's easy to weld some tubing together and slap on wheels, an engine, and a transmission and you're out on the streets in no time!

I want to import a new Toyota Hilux diesel pickup because a compact diesel pickup truck isn't available in the USA. Unfortunately I'm not allowed to do this because it hasn't been tested against US crash standards and the engine isn't EPA certified (despite being less polluting than just about any diesel engine currently offered in the USA).

"MOVE OVER BUDDY"

[MosesJones (55544)]

Less than 20mph in an SUV through the desert. These Robot control cars are worse than my Grandmother on an interstate.

Quite clearly these Robot controlled cars are part of a sophisticated plot to increase the amount of road rage in the US to enable the Robots to take over the country... and then the world.

It is not too late to stop them, we must insist that the next competition involves only Ford Broncos and takes place on the Freeways of Los Angeles during rush hour.

Re:"MOVE OVER BUDDY"

[earthforce_1 (454968)]

Yeah but according to Moore's law, the robot cars should be able to break Mach 1 sometime around 2010.

How few remain

[necro81 (917438)]

Looking at the final stats on the Grand Challenge website [grandchallenge.org], it would seem that only five teams, out of the 23 that made the finals, were able to finish the course. The team that got the farthest before calling it quits managed about 80 miles, which means that the cut between those who made it and those who didn't was still pretty big. Another interesting thing about the final results is that, if you look at the pretty red and blue graph lines, they describe what looks like a sort of decaying function...

Or perhaps I'm just a dork.

Re:How few remain

[zurmikopa (460568)]

They blew a tire and were somewhere around 60 feet off course when they were eliminated, if I remember correctly.

I know that a good number of the teams were actually still moving when they were eliminated; they had generally just wandered far enough off course that it was determined that they would be unable to finish.

There were a number of reasons why people did so much better this year than last year.

The biggest reason I think is that people knew a little better what to expect this year, and focused development on more important items for the race. For instance, for the first race I had done work on using a terrain database for path planning, but it turned out that the waypoints are so close together that it ends up just being a waste of CPU cycles for the most part.

Another important reason is there was a rather large jump in the quality of the software running on the bots, and a moderate jump in the quality of the hardware. The integration was much more refined.

Finally, the course was easier overall this year and the difficult part was put near the end. There was nothing in the course really comparable to Daggett ridge from the first race. Also, pretty much the entire course was graded along with the edges of the road often had banks. We had cliff detection that pretty much went unused this year due to this.

Overall, it was a pretty good race this year. Stanford did an awesome job and really deserved the win. Not that you guys have that much interest, but we (Axion) ended up in 7th place (right after Ensco) with about 66 miles. We ended up getting stuck in some sand. The current candidate for the cause is a broken sway arm bracket that caused us to pull to the right a bit. Further analysis will be required to determine if that's actually the case.

The downside to this

[Elrac (314784)]

While I'm happy that these hard-working academics were successful, I can't help but note the downside to this development.

Forget military applications. What I foresee is that, for computer scientists who've lost their jobs to outsourcing, this will deprive them of one more alternative, namely a career as a taxi/truck/bus/etc driver.

Stanford + Volkswagen

[SamuraiMike (768946)]

"Drivers unnecessary"

More info

[Zathrus (232140)]

For far better info than the anemic (and completely flash based) gc.org site:

http://www.darpa.mil/grandchallenge/discussion.htm l [darpa.mil] -- DARPA's GC message boards
http://www.tgdaily.com/2005/10/08/darpagrandchalle nge2005/ [tgdaily.com] -- Was updated throughout the actual event. Best coverage I've seen yet.
http://www.popsci.com/popsci/darpachallenge/ [popsci.com] -- Popular Science's rather disorganized site

I'm still looking for "highlight" video myself... or pretty much any non-bland video (seeing them cross the finish line is nifty and all, but that was not a challenging part of the race). I particularly want video of Alice trying to take out some reporters!

Quote from the winner

[VeganBob (888165)]

01001001001000000111000001110111011011100110010101 10010000100000011110010110111101110101001000000110 0001011011000110110000101110

cmu won all three

[by Anonymous Coward]

sort of.

the stanford leader (thrun) and their lead software developer
(mike montelermo (sp?)) were originally from cmu.
they only recently moved to stanford. although thrun claims it's coz of his wife, some people think it was coz of too much competition and bad blood at cmu which has lots of people working in mobile robots (wittaker, simmons, nourbaksh, choset, ...) while i think palo alto has much better weather than pittsburgh :)

the particle filter based localizer and mapper was developed while at CMU. Frank Dellaert (now at georgia tech) first introduced that to mobile robotics after reading about the
condensation algorithm in computer vision (i like to believe that i had a part in that last bit :) I would'nt be surprised if they also use large parts of the basic control and command software infrastructure (TCX) written by thrun and others while at cmu. if it is, no wonder they required
7 PCs for redundancy, that is some of the worst spaghetti code i've ever had the displeasure of working with. it's easier to make it fault-tolerant by just throwing more hardware at it.

i'm not trying to belittle stanford in any way, but i just thought people might be interested in knowing that the real story in this case is a lot more complicated. the relationship between the winning teams were a lot more incestuous :)

thrun BTW is an amazing all-round guy with an infectious smile all the time.

Why isn't this a bigger deal?

[prozac79 (651102)]

Why is the fact that 5 autonomous vehicles where able to traverse 132 miles not a bigger deal? I hardly saw any media coverage on this (not even mentioned in those closing "isn't this interesting" segments on local news). IMHO, this is another great "first" for mankind on par with Lindberg crossing the Atlantic or Rutan winning the X Prize. In the future, when automnomous vehicles are more ubiquitous, we will see that the pioneers were vehicles like Stanley. These engineers solved (or at least furthered our understanding) some very difficult problems of computer vision and perception. However, whenever I mention the Grand Challenge to people, they just give me a blank look. One person asked me if the Grand Challenge was some sort of football event.

Oh well, from what I heard no one was too excited about the Wright brothers' achievement at the very beginning either.

Team Grey is the Real Winner

[humankind (704050)]

When you look at the results, and you see two colleges with virtually unlimited resources and millions of dollars spent on their vehicles, huge corporate sponsors and engineers at their beck and call from Boeing to Catepillar, who finished, and then this dinky little Team Grey from a suburb of New Orleans, with a splintered development team as a result of the Hurricane Katrina disaster, and they FINISHED just behind the big guys, leaving other heavily-funded vehicles in the dust.

Relatively speaking, a small indy group, even if their time was a tad slower than CMU or Stanford, essentially put those three teams to shame when you compare the resources they had available to them.

The real story here is who is behind the Grey team's car. It must be a far superior design than either CMU or Stanford's considering the limited resources and experience they had in addressing the challenge.

Re:19.1?

[Schweg (730121)]

Almost 20 miles per hour on unpaved roads with an autonomous vehicle? That's not the same as driving on paved roads in the city or on the highway. I think that's pretty good, actually.

Re:19.1?

[necro81 (917438)]

The course did have a fair number of twists and turns in it. There were some places, like dried lake beds, where the cars could open up a bit, but for the most part it was bumpy dirt tracks one which even you or I couldn't do more than, say, 40 mph. There were also, intentionally, a fair number of obstacles designed to throw the computer systems off. You and I wouldn't have much difficulty in recognizing a cattle gate on a road, but imagine trying to teach a computer vision system to distinguish that. In other cases, the robots had to drive through tunnels that would not only be dark (making vision systems less accurate) but also lack any GPS signal.

So, yes, it did average out to a pretty slow "race." But, on the other hand, it is a marked improvement over last time, when no one even came close to finishing. I think that, in the interests of trying to ensure that they safely finished the course, let alone win, the various teams were playing it a little conservatively, and not trying to go for pedal-to-the-metal performance. Maybe next year, now that they have some confidence.

Re:Patriotism... sigh

[reed (19777)]

Well, so were Einstein, Werner von Braun... etc. :)