Slashdot Mirror
DARPA Grand Challenge 3
Meostro writes "DARPA announced the 3rd "Grand Challenge" today, The DARPA Urban Challenge. "To
succeed, vehicles must autonomously obey traffic laws while merging into moving traffic,
navigating traffic circles, negotiating busy intersections and avoiding obstacles." This year's new twist is two tracks for entry: the first is the same as the previous two challenges (develop on your own without Gov't. funding), but the second involves "submitting a detailed proposal for up to $1 million of technology development funds." Here is the PDF press release ."
Yes, they do research in defense but shouldn't there be a little more than a tiny graphic or blurb about what work they're doing? Couldn't they at least take the time to write an abstract or 1-2 page paper with unclassified information on each project?
Instead, I find the following links in the 'Archives':
- Quantum Computing
- Infrared Focal Plane Array/Uncooled Integrated Sensors
- Advanced Lithography
- Most of the Other Links
My alma mater has produced better papers than this in these fields. I know that a lot of this stuff isn't classified and they list their programs on their sites, why can't they do a better job in showing the American public what they've done with our money?The Grand Challenge Forums are flooded with only vendors. Where are the designs and reports by the teams from older Grand Challenges? Why isn't this structured more like RoboCup where the learning algorithms are released every year so that future contestants can build on this?
The fact that this contest isn't run in a more open way makes it seems like less of a "contest" and more of a "do our research for us!" kind of thing.
My work here is dung.
I'm curious about how valuable experience in the previous two challenges is. What was learned that will be applied to the new challenge?
...how this can be "safe" (as far as it can be anyway) with cars which are automated going on roads for which the system works (I'm assuming most are based on the idea of staying within the white lines) I worry about this quote...
We believe the robotics community is ready to tackle vehicle operation inside city limits. - Dr. Tony Tether, DARPA Director
You can build the safest car in the world but there is always a need to be able to take a very quick decision to avoid some other idiot who might be breaking the rules of the road and not be in an automated car... still, if we all had them...
*''I can't believe it's not a hyperlink.''
"City hall" in German is "Rathaus" Kinda explains a few things......
I for one trust stupid computers over many humans I know.
Its too early to go urban. They should have spent at least another 2-3 years perfecting autonomous navigation in unstructured environments.
I know last year's challange seemed to be won rediculously easily, but I have seen no proof that that dormain has been fully conquered yet. If they wanted a challenge why not move onto wooded or swampy areas.
In this case it seems they are juat setting themselves up to fail.
Surur
Information is the location of things. Computation is moving things around.
Let's make this real, vehicles must be equipped to : Avoid hub-cap thieves. Deal with homeless window-washers. Handle crazy taxi-cab drivers. Deal with traffic detours and malfunctioning signals. Understand parking, loading, no stopping and no standing zones. Not run over pedestrians who jaywalk. Avoid accident scences/traffic congestion by planning on-the-fly alternate routes. Be able to pay Tolls. Stop, Look, Listen at Railroad crossings. Be able to do this in even in Snow conditions. Properly prepared, an urban course will be MORE challenging than the Desert course. Sure terrian won't be such an issue but the environment is much more dynamic, and the weather is still a factor.
I'm not going to speculate as to whether the robotics community is "ready" for this challenge, but what do the two challenges have to do with each other from a technical standpoint? In the previous challenges, vision wasn't good enough to tell a boulder from a bush. Are they going to give the robots the GPS location of all the stop signs and traffic circles? If they do, how well would this apply to a city where not all GPS locations are known? If not, how will it differentiate signs from one another and from random stuff in the background?
I'll be impressed with no crashing into each other, before they worry about compliance with all traffic laws. How will the robots recognize the speed limit in their area, or will they all crawl along at 10 mph, impeding the flow of traffic?
I used to carry a bottle of whiskey for snake bite. And two snakes. -Nefarious Wheel
Any teams in the Detroit area? Non-academic ones? This would be a really cool thing to work on full time.
"It is likely that DARPA has become incapable of inovation because of internal politics so they need to attract new ideas from the out side."
Um, Darpa is a think tank. They don't do actual research. It's been that way since the beginning.
I wonder if you get extra points for having your car honk its horn at
other drivers and extend it's little robotic middle finger at anyone that
gets in its way?
Just 1 team has to "win", and win doesn't have to be a complete success. If they can demonstrate the technology is 70, 80, even 90% there then that technology can be taken from the contestants and worked on in industry.
Sure, 'auto car' is cool, but with the world today why not shift to something that will help us all, like a new energy source, food, etc..
---- Booth was a patriot ----
Ummm, yeah. That's the way they're supposed to work.
Best Slashdot Co
Bagdad. Unmanned Military Ground Vehicles. Primarily for supply runs, but could also be used to troll for IEDs.
Best Slashdot Co
... if they have alert drivers in the other cars, and ... if they select an urban area which is in decent repair, and ... if they allow ideal driving conditions like the last race.
:)
Think about it. City driving is designed to be easy. In fact it is really really easy. You are told exactly where to go with visible lines, lights, signs, etc which are all designed to be noticed and easily intepreted.
The hardest part of GC1 was finding the road! When it's layed out for you nice and easy.... man thats a cakewalk.
DARPA won't risk random people in these tests so all the other drivers will be hyper aware of the robot cars. Thus merging will be greatly reduced in difficulty, and that will be one of the two hardest things to do.
So on face this seems like something that Stanley could do today, but what we need to do is brainstorm what could make this a real challenge. Here's my short list.
1. Unaware human drivers.
2. BAD human drivers.
3. Rain/Darkness/Snow (weather that obscures the road)
4. Potholes, stupid pedestrians, trafic cones (urban obstacles)
5. Human Drivers
As you can see, I believe the greatest obstacle to good driving is stupidity. How to program around that? I have no idea.
2006 -- True Artificial Intelligence
2007 -- AI Landrush2009 -- Human-Level AI
2011 -- Cybernetic Economy
2012 -- Superintelligent AI
2012 -- Joint Stewardship of Earth
2012 -- Technological SingularityA robotic M1A1 Abrams main battle tank. It can merge over moving traffic, bisect traffic circles, evacuate busy intersections, and flatten obstacles.
Humans can't navigate traffic circles (at least here in the U.S.), why should we expect robot to be able to? Then again, robots don't have "a**hole driver" functionality and generally are far better trained than human drivers.
Will the DARPA challengers have to use turning signals?
If so this will be impressive, as at least half the driving population of the United States is unable to utilize that technology consistently.
Oh, right, because you are a troll and couldn't give a rat's ass about the plight of others. This is the oldest troll in the book: some cool new thing comes along and some asshat who has never helped another human being in his life says, "What a waste of time! What about all the starving orphans in Africa?" or some such shit. Go crawl back under your bridge, troll, and let the grown ups talk.
- None can love freedom heartily, but good men; the rest love not freedom, but license. -- John Milton
"Using a turn signal is giving information to the enemy."
- None can love freedom heartily, but good men; the rest love not freedom, but license. -- John Milton
There's got to be at least one team who has considered using a black Pontiac TransAm complete with red scanner on the front. C'mon people, make it happen!
Why is it that every time I hear a story about efforts to improve vehicles, I say to myself, "Gee, hasn't that been done already?" All these efforts seem to have one thing in mind: get a car to act like a train, that way we can continue subsidising the auto/oil/rubber industries with the needless purchase of more individual rail cars.
How about a challenge to develop real public transit in the U.S.?
Sorry, I live in L.A. and I'm bitter.
Can you imagine what would happen if you put an idiot driver in a robot controlled car, and the AI encountered a conundrum, i.e. a blocked one way or something that required the driver take control? That idiot is going to be in a situation where his idiocy will endanger everyone around him and probably not have been paying attention while the AI was driving so will be all that more "lost"
"It's because they're stupid, that's why. That's why everybody does everything." -Homer Simpson
obey traffic laws
Chicago traffic/parking laws are cleverly designed to trip up & get a few bucks out of the uninitiated. Examples: 1. no left turn at some intersections 2. other intersections, no right turn on red between 7am-7pm. The notice is posted on the signal closest to your car, so it's harder to see unless you look for it. 3. No street parking during the afternoon rush hour. Signs poorly visible 4. At some intersections, right turn or left turn only on an arrow signal. 5. No parking after 5pm in some downtown areas. Tow trucks wait around the corner to snatch up cars where the driver didn't see the sign. 6. In some neighborhoods, only permit (windshield sticker) parking at certain times. Signs not always where they should be. 7. No parking on some streets after 2" snowfall between Nov 1 - April 1. I could go on but you get the point. Maybe a robot could navigate a modest size downtown area but a major metro area? No way.
stands for defense, i.e., the organizations whose job it is to blow up things and kill people.
"How will the robots recognize the speed limit in their area, or will they all crawl along at 10 mph, impeding the flow of traffic?"
They are probably not really concerned about this, nor about recognizing STOP signs or other traffic signals, really, even though the article summary includes "[the] vehicles must autonomously obey traffic laws". They probably mean some set of operating parameters that fall under the class of "rules of engagement", e.g., whether or not running over pedesterian non-combatants are a concern.
You can build the safest car in the world but there is always a need to be able to take a very quick decision to avoid some other idiot who might be breaking the rules of the road and not be in an automated car... still, if we all had them...
This is the Grand Challenge I was really waiting for. I believe that the experience gained in the previous Grand Challenges is practically useless for this new one. This new challenge will involve true AI, that is, AI that has true general learning capabilities and the ability to adapt to new situations. A true autonomous vehicle will need to have common sense understanding and the only way it can have this is by learning through trial and error, imitation/observation, the capability of being trained via communication (with a trainer) and the well known principles of operant and classical conditioning. In addition the AI will need robust and sophisticated perceptual (visual and auditory) system in addition to sound motor control/learning mechanism. This AI will forcibly be based on some sort of neural network or a integrated collection of neurla networks.
What Darpa is asking for is none other than the solution of the AI puzzle. I'm afraid this is worth much more than a few million dollars.
Anyway, that's quite impressive for two guys with jobs working part time, on their own, out-of-pocket to be in any way competitive with the CMU and Stanford teams. I hope they're up to this urban challenge and get the money. Although the new challenges presented by an urban environment may well be beyond the realm where amateurs can compete.
Can anyone tell me how to set my sig on Slashdot?
"My other car needs a human"
****
"I'd never want to join a club that would have me as a member" - G. Marx
I saw the last one (Stanley), and I must say that it's skills left me a little under the.... under.....
http://www.theinquirer.net/?article=28756
Either way, my insurance company is still trying to go after the estate of a Mr. Babbage, but are having a hard time tracking him down.
-Charlie
I've met the Team Dad guys and am impressed with them, but bear in mind that they own an electronics company, with experienced people on tap who can build custom electronics. It's not two guys working alone.
Ok, there's this good ole boy driving down the highway when he gets a call from his wife, "Honey, be careful. The radio said that some maniac is driving down the highway in the wrong direction." He replies, "One? Hell, all of them are!"
This sig has absolutely no significance and serves only to take up screen space and waste the time of the reader.
The NOVA made it look like they have a company that makes stereo speakers (probably not a lot of crossover applications to autonomous robotics), and it did make it look like they did this on their own, not that they were paying others to help out. It did look like one of them was devoting most of his time to this project, not to work. And they're probably still well off, but still, compared to Stanford or CMU, with millions in funding, and as many as a hundred nearly full-time people working on the project at CMU, etc, they are two self-funded amateurs working from home.
Can anyone tell me how to set my sig on Slashdot?
There's more crossover than you think. Their company makes subwoofers with closed-loop servoing and positional feedback, run by controllers implemented in FPGAs. The Team DAD vehicle uses very similar technology.
We all know what will happen when the first urban challenge takes place.
1. It'll be run in Los Angeles.
2. The vehicle in front will have a bug in its software that causes it go off course and on to the LA Freeway.
3. It will be chased by 50 police cars and 5 news helicopters.
4. The chase will end when some guy with a samurai sword cuts the vehicle's wheels out from under it and shoots the gas tank when it rolls over.
In Soviet Russia, car drives you!
There are 10 types of people in the world; those who can read binary, and those who can't.
These devices WILL be used against the US population in the future, as well as the populations of
other countries.
Why ?
Because the Pentagon knows that the casualty rate for live soldiers is too high for the American public to stomach.
The Army War College determined some time ago that the casualty rate in urban combat will exceed any casualty rate seen
in normal "war" scenarios. Thus, robots will be used instead of humans.
What these bastards forget is that a robot cannot ( yet ) become any smarter, nor can it adapt.
But humans can adapt, and they will, and these robots, when used against humans, will shortly become expensive piles
of useless scrap, whether they are used in Iran, Syria, or Cleveland.
The best you can do in real life is drive defensively, leave time to react, and be aware of your surroundings. Those behaviors don't end collisions, but they might minimize them.
There are so many variables to keep track of, and so many possible scenarios, that a data miner probably can't handle the complexity of "that guy is swerving and changing lanes without signaling, so I'd better keep my distance from him to the extent that I can". It probably requires some kind of reasoning agent with a specialized vocabulary (like a set of microtheories in Cyc). Needless to say, that's a long way off.
I don't think the Urban Challenge is about that part of driving. I think it's a recognition that the definitions of road and obstacle start to blur when you get into the city. For instance, you can't just drive anywhere you want on blacktop. You might get your agent speeding through a grocery store parking lot.
"You are told exactly where to go with visible lines, lights, signs, etc which are all designed to be noticed and easily intepreted."
I think this will take more effort than you are portraying. Humans interpret signs easily; computers don't. Humans also look out while they're driving and know which parts of the scene to focus on. More than one sign is visible at any one time. Also, the computer has to see the signs, and has to know where to look. The computer vision portion of the vehicles will have to be stellar to work in real time. Finally, like I said, the way the computer reasons about the signs will require some pretty heavy programming to induce proper behavior. Whether it's humans writing the driving policy or it's induced by some kind of simulation (and it would have to be a pretty detailed simulation to carry over into the real world), changing the policy in real time could be one of the hardest parts of this challenge.
That's very different than what Stanley was doing to win the Grand Challenge, which was roughly to find the road and stay on it at speed. The Urban Challenge has many more moving parts.
Q: What did the comedian say to the crowd?
A: If I knew, this joke would be funny.
Personal rapid transit
Automated cars are something that is going to move at a really slow pace. They already are moving really slowly. It was probably 15 years ago that I first remember reading in Popular Mechanics about MIT (I think) instrumenting a big cargo van with the necessary sensors and computers to handle driving around a parking lot. They kept working on it year after year, improving their sensors and probably constantly rewriting their code, until maybe 6-7 years ago, they had it to the point where it all fit into a sedan, which they drove across the country (with a person in the driver's seat to take emergency control if necessary).
Since then, I've heard almost nothing about it. I think they're having trouble getting people interested in it. First of all, people are understandably reluctant to hand their life so directly over to a computer. Secondly, I imagine the auto industry is pretty nearly terrified of the liability. They have to be darn sure the computer isn't going to decide to run someone into a median at 70+ mph, and that an occupant can reliably take over when necessary. There are a lot of features on a car that can not fail absolutely. Your brakes are actuated by a vaccuum cylinder instead of a hydraulic valve so pressure is maintained in the event of a power loss, and even if the booster "runs out" of vaccuum, you can still mash down on the pedal and get some pressure from the master cylinder. I believe all power steering systems all still maintain a direct linkage for the same reasons. I've already heard stories of cruise controls going out of control and taking off at wide open throttle. I'm not sure what's been done about correcting that, but giving the computer control of steering, direction control, and braking adds new possibilities to slap a big lawsuit on the manufacturer.
Getting computer control into cars will take baby steps. I think this new Grand Challenge sounds like a good way to advance the technology (just look how fast the controls improved between 1 and 2), but we aren't going to see the first successes hit the road for years. I think Cadillac or somebody is developing a system that lets a computer perform a parallel park. That seems like a good start. I guess a bigger step would be allowing computer control on certain stretches of low accident rate highways (Wyoming seems like a good place).
Eventually, we could actually see fully automated driving be the norm, like in Minority Report (You were controlling the car!?!). The precision and "play by the rules" nature of automated control offers quite a bit of potential to reduce accidents. Of course, there are plenty of personal freedom concerns to address, too, but I'll leave that for others to talk about if they care to.
Something that will help us all like reducing the amount our fellow citizens in the Army need to expose themselves to the enemy? Or how about improving traffic safety by reducing stupid driver mistakes? Maybe allowing long haul truckers to reduce driving at less efficient speeds in order to minimize the cost of driver labor? That's just scratching the surface of what automated vehicle control might offer in the future.
And by the way, I'm almost positive DARPA funds a lot of energy research, too. Lastly, at this point, the world does not face a problem with food shortages. It is entirely in the distribution, with politics being the biggest factor.
nobody else caught the obvious faked name for a man reviewing projects in autonomous robotics.
We believe the robotics community is ready to tackle vehicle operation inside city limits. - Dr. Tony Tether, DARPA Director
-----------------
Willy Wireless
CTO Viacom Cable
The clear intention of the urban project at this time is to act as a bridge to the more useful and applicable project of mixed-terrain autonomous navigation. That is what real-world defense missions tend to be. For example, when I was in OIF, we staged in Kuwait and convoyed up to Baghdad. During that trip we drove on desert sand, rocky desert, somewhat modernized paved highways, gravel sideroads, bombed-out highways, dense jungle roads near the river, over various levels of bridges and some other stuff that's a mix of all of that, and finally dense city traffic interlaced with crazy Baghdad drivers, no traffic signals or cops, random roadblocks and the occasional bomb. Having soldiers consistently do those operations can be a waste of resources and create unnecessary risks. The aim of these projects is to create a situation where autonomous vehicles, that might be observed through on-board cameras or through an overflight (by an autonomous plane), can become a practical and safer alternative. Relying on things like trains or planes alone is silly - often the current infrastructure will not be able to support either of these (no airports or viable runways, too far to fly, too dangerous to fly, no train tracks, no functioning rail signal system, etc.).
Ultimately, there won't be a lot of this kind of work done in downtown Manhattan. While some third-world cities can be even crazier traffic-wise, we should try to be realistic in how these vehicles might ultimately be deployed. Regardless though, I think it's important that the specific requirements of each of these projects exceed those expectations in order to create vehicles that have the ability to accomodate worst-case scenarios.
I can definitely see the ultimate end of these projects being an autonomous cargo ship that departs from an aircraft carrier carrying autonomous vehicles that will disembark on their own and travel to their intended target on their own with minimal oversight through onboard cameras and overhead surveillance. That would be useful.
If soccer-moms around the country can later drive around town by pressing a button and sitting back to talk on their cell phone and drink a latte because of these projects, great. But that's not their targeted market.