Follow Slashdot stories on Twitter

 



Forgot your password?
typodupeerror
×
Transportation Technology

Robotic Audi To Brave Pikes Peak Without a Driver 197

Scifi83 writes "A team of researchers at the Center for Automotive Research at Stanford (CARS) has filled the trunk of an Audi TTS with computers and GPS receivers, transforming it into a vehicle that drives itself. The car will attempt Pikes Peak without a driver at race speeds, something that's never been done."
This discussion has been archived. No new comments can be posted.

Robotic Audi To Brave Pikes Peak Without a Driver

Comments Filter:
  • I, for one (Score:3, Funny)

    by electricbern ( 1222632 ) on Friday February 05, 2010 @01:16PM (#31037060)
    would like to welcome our robotic driver overlords.
  • Explanation (Score:5, Funny)

    by OpenSourced ( 323149 ) on Friday February 05, 2010 @01:18PM (#31037080) Journal

    The researchers have programmed Shelley to handle like a racecar by using a set of computer calculations called algorithms

    Ha! So that's how they did it! Quite simple, really, once you know the trick.

    • Re: (Score:3, Funny)

      by Anonymous Coward
      damn it! why didn't anyone tell me about these "algorithms" before? They sound so useful!
    • by whopub ( 1100981 ) on Friday February 05, 2010 @01:43PM (#31037444)

      Truth is, it's probable just The Stig in the trunk, with a laptop.

    • by tool462 ( 677306 ) on Friday February 05, 2010 @01:44PM (#31037454)

      Bah. "Algorithm" is just doublespeak for "Mechanical Turk."

      Some say he can steer a car just by thinking evasive thoughts.
      And if he turns the wheel, the road will slide easily underneath his car like a waitress with Tiger Woods.
      All I know is that it'll be driven by The Stig.

      • >Some say he can steer a car just by thinking evasive thoughts.

        Works great, if you can think in German, that is.

    • by mcgrew ( 92797 ) *

      The researchers have programmed Shelley

      God, I hope that's a typo. Does no one read Asimov any more? The car's name should be Sally [wikipedia.org]. It would fit perfectly; iirc the story tales place in 2020, and Sally is an antique.

      As Asimov coined the word "robotics" and this car is a robot, it's a damned shame if they didn't pay homage.

      • Re:Explanation (Score:5, Informative)

        by MaWeiTao ( 908546 ) on Friday February 05, 2010 @02:08PM (#31037710)

        The car is named after Michèle Mouton, her nickname apparently was Shelley, the most successful female rally driver to date. Apparently she's considered the most successful female driver in all of motorsports. And it just so happens she drove Audi's and she was the first woman to win the Pikes Peak hill climb. I'd say that's a far more appropriate reference than anything from Asimov.

        • Top Female Driver?

          Where does that put her overall?

          • Re: (Score:3, Informative)

            Top Female Driver?

            Where does that put her overall?

            Well, she won something, came second in something else and then quit due to rule changes. But please, read the article [wikipedia.org] already.

            • I'm not much of a race fan - so while her Wikipedia does tell what races she's won, I don't have a frame of reference.

              I know that the Pikes Peak race has been going on for nearly 100 years...and that this was the first girl to win.

              I don't know how she relates to male drivers.

              Is racing a male-dominated sport? Or is gender insignificant?

              • Racing is very male dominated. While there aren't any rules (anymore) preventing female drivers from competing, there just hasn't been a lot of female participation in auto racing of any sort.

            • Re: (Score:2, Funny)

              by mrdoogee ( 1179081 )

              Pfft. Call me when she's making sexually suggestive domain registrar commercials.

          • by Goonie ( 8651 )

            The World Rally Championship is the highest level of competition in that sport. So winning rounds of it, and coming second in the championship, indicates that you're pretty bloody good.

            I'd hazard a guess that such results rank her in the top 30 rally drivers, all-time.

        • "I'd say that's a far more appropriate reference than anything from Asimov"

          Yes, Shelley or Walter Rohrl! :-D _O_

        • Re: (Score:2, Informative)

          by yachius ( 1348219 )
    • While the algorithms that make it handle like a racecar seem interesting I'm more concerned about the algorithms for emotion, I mean the headline says this driver-less car will "brave" Pikes Peak. How exactly is the bravery implemented? What if next it decides to "brave" global domination?
      • by Chris Burke ( 6130 ) on Friday February 05, 2010 @02:12PM (#31037784) Homepage

        How exactly is the bravery implemented? What if next it decides to "brave" global domination?

        Don't worry! The scientists have that covered.

        To have bravery, you first must have fear. So the first and most difficult step was to program the car to be afraid all the time. Then, to get bravery, they simply program it to ignore its fear when it's driving up Pike's Peak.

        The rest of the time it's a total scaredy-car. If you think it's trying to dominate the globe, just shout "boo!" at it and it'll drive off to cower in the corner and cry.

      • How exactly is the bravery implemented?

        /s/fear/courage/g

      • by bareman ( 60518 )

        They did mention that the car was going to be equipped with a "Kill Switch".

  • Note to Self (Score:5, Insightful)

    by Kozar_The_Malignant ( 738483 ) on Friday February 05, 2010 @01:19PM (#31037102)
    Observe trials from uphill side of road.
  • It may not be as tall as Pike's Peak, but it's a more challenging road, IMHO. http://www.concierge.com/cntraveler/blogs/perrinpost/2007/08/this-car-climbe.html [concierge.com]
  • by smellsofbikes ( 890263 ) on Friday February 05, 2010 @01:21PM (#31037134) Journal
    Having raced up Pikes Peak myself, I think this is pretty impressive. The road surface varies from almost-pavement-quality treated packed dirt, to completely loose gravel on rut-filled rock, with (as I recall) an average 10% climb. It'd be a great test ground for offroad stuff at a slow speed with nobody else on the road, but doing it at full speed requires a *lot* more than just the ability to see where the roadway is: giving a robot the ability to keep a fast car from skidding/sliding on loose gravel on an off-camber turn appears to me to be a wholly different type of challenge than previous autonomous driving projects.
    • by Tiger4 ( 840741 ) on Friday February 05, 2010 @01:33PM (#31037306)

      Rapid reaction time, good slide slip sensors, and some great counterintuitive steering routines are all going to be essential if this is going to be "at race speeds". Good trained drivers screw this kind of thing up all the time. A robot can be programmed to be repeatable, but is this one flexible enough to conditions to be fast AND safe?

      Look out below!

      • by nomel ( 244635 )

        Rapid reaction time, good slide slip sensors, and some great counterintuitive steering routines are all going to be essential if this is going to be "at race speeds".

        Unfortunately, *realtime* reaction time has nothing to do with staying on the road if physics wont allow it. If you go into a corner too fast, then you're quick reaction could put you into a beautiful sideways controlled drift, right past the point where the surface of the road stops existing. This means the car must also have some very impressive "look ahead". If it doesn't have a map of the road, it'll have to predict a safe speed for any blind turns (I imagine nearly all of them are) while considering th

        • Definitely, I just don't see them letting the car drive up there without a map. It's difficult enough for human drivers to turn in competitive times without pace notes, but it's going to be even more difficult for a computer to deal with.

          The most interesting part here, IMO, is seeing how the car deals with driving at racing speed vs the more leisurely pace of the previous AI challenge race, where, IIRC, the cars averaged below 30km/h. At those sorts of speeds the car just goes where the wheels are pointing

    • Not many humans [youtube.com] can drive up the Pike's Peak at racing speed.

      If they can create a robotic car to do this then one major criterion for a "human level" AI has been reached.

    • "giving a robot the ability to keep a fast car from skidding/sliding on loose gravel on an off-camber turn appears to me to be a wholly different type of challenge than previous autonomous driving projects."

      Actually, that's the easy part. It's called Traction Control [audiusa.com], and it's standard on pretty much all cars sold in the US these days. Given that the article mentions how they're using a lot of Audi's electronics and sensor systems, I expect they'll leave these systems in place.

      It is true that a good
      • I don't think Traction Control handles the problem I'm pointing out. It's one thing to actively control the torque going to the wheels to prevent them sliding. It's a completely different thing to be on a long straight, going fast, and decide that the approaching hairpin turn is filled with gravel and sand and you need to slow down *now*. I've seen people slide off curves with their ABS brakes on full because they misjudged the road conditions and went into a turn too fast. No amount of traction control
        • by Andy Dodd ( 701 )

          Most drivers I've talked to with far more experience driving on gravel than I do seem to be of the opinion that traction control = BAD NEWS on gravel and that even moderately experienced drivers can do far better than what any TCS can do/allows you to do in terms of maintaining control on gravel.

          Most TC systems are apparently tuned more towards pavement or ice/snow, not towards gravel.

          There's also the fact that, as you state, taking a turn properly requires prior knowledge of the conditions of that turn - t

    • by G-Man ( 79561 )

      I'm curious when you raced Pikes Peak - I watched the race a couple of years ago from lower on the course, and the road was tarmac at that point. It's my understanding that every year the road's owners pave a little more of it, with the goal of eventually paving all the way to the top - they want to make it easier for the tourists to reach the summit. Changes the nature of the PPHC, but economics wins out. Now if 'Shelley' has to handle both tarmac and gravel in the same run, that's actually fairly impress

      • Re: (Score:3, Interesting)

        I did it once on a motorcycle back in the mid-1980's and then a couple more times on a bicycle in the mid-90's. (Not that it's worth anything to anyone else but I managed to beat a bunch of pro racers in one of the bike races. The only way you can legally bike up PP is during the race, since it's closed to bicycles the rest of the year.)

        It might not just be economics: there is an enormous amount of environmental damage done by maintaining a high-altitude, high-traffic dirt road. The city of Colorado Spr

  • meh (Score:5, Funny)

    by OglinTatas ( 710589 ) on Friday February 05, 2010 @01:23PM (#31037150)

    I already saw this on Speed Racer.
    Spoiler Alert!
    Speed wins the race anyway, and helps Inspector Detector catch the nefarious people behind the robot car

  • by S-4'N3 ( 1232394 ) on Friday February 05, 2010 @01:25PM (#31037192)
    I am not certain what will be more astonishing: watching this succeed or watching it fail.
    • Re:W1N vs. FA1L (Score:5, Insightful)

      by Carik ( 205890 ) on Friday February 05, 2010 @01:41PM (#31037422)

      Easy. Watching it succeed will be more astonishing.

      Watching it fail will be more entertaining (assuming some safety precautions preventing anyone from being killed).

      • by Chris Burke ( 6130 ) on Friday February 05, 2010 @02:09PM (#31037734) Homepage

        (assuming some safety precautions preventing anyone from being killed).

        Good point!

        The summary only says that the car won't have a driver. I hope they remember to have the passengers get out, too!

        • by Carik ( 205890 )

          I was mostly thinking "make sure there aren't any pedestrians in the way" and "make sure that when it falls off the mountain, it doesn't land on anything important," but getting the passengers out would be valuable too.

          • Also, don't forget the hooker in the trunk.

            • by Carik ( 205890 )

              Didn't you look at the pictures? That's not a hooker, that's a computer! (Gotta wonder about some of these /. types... though I suppose it could be a robot hooker...)

              • Re: (Score:3, Funny)

                by Chris Burke ( 6130 )

                That's a computer?!

                Then what's the thing that balances my checkbook and snorts all my coke?!

        • This projects strikes me a extremely elitist. I want to see a robot driving a huge bus full of robot passengers doing their regular robot things, like optimizing their energy efficiency, updating their belief systems from a trusted repository, and compressing old files.
  • Yikes! (Score:4, Funny)

    by neiras ( 723124 ) on Friday February 05, 2010 @01:27PM (#31037232)

    The researchers have programmed Shelley to handle like a racecar by using a set of computer calculations called algorithms

    See what happens when you let Liberal Arts majors playing journalist direct the public's understanding of technical things?

    Soon: "John's car rolled out of his driveway all by itself and hit a fire hydrant, honey! He should sue General Motors for faulty algorithms!"

    • Soon: "John's car rolled out of his driveway all by itself and hit a fire hydrant, honey! He should sue General Motors for faulty algorithms!"

      If someone's car did that, couldn't "faulty algorithms" actually be the problem?

    • So you mean, “algorithms” becomes the new “gene”, as in “genetically modified”, meaning “It can’t possibly also have good applications.”?

      Let’s see them become Amish then. ;)
      *must... wait...until... then... to tell them that our own bodies are basically based genetic algorithms* ;)

    • by jdgeorge ( 18767 )

      The researchers have programmed Shelley to handle like a racecar by using a set of computer calculations called algorithms

      See what happens when you let Liberal Arts majors playing journalist direct the public's understanding of technical things?

      Soon: "John's car rolled out of his driveway all by itself and hit a fire hydrant, honey! He should sue General Motors for faulty algorithms!"

      You're ridiculing the author for clearly and correctly defining the terminology? If the intended audience [stanford.edu] of the article [stanford.edu] is unlikely to know what "algorithm" means, don't you think a concise definition is in order? Now, I'm sure that writers assuming an Ivy League audience [harvard.edu] would reasonably expect people to know what the big words mean.

      Not to mention the fact that there are many Computer Science and Engineering majors who are also capable of effective communication; this is not the sole domain of "liberal a

  • I have an image in my mind of the Price is Right Cliffhangers game, where the guy goes right over the top when you lose. Oh, and the song [televisiontunes.com] is now firmly implanted in my brain for the rest of the day.
  • It seems like it would be braver to have no driver but still have a passenger.
    • Re: (Score:3, Insightful)

      by Fnord666 ( 889225 )

      It seems like it would be braver to have no driver but still have a passenger.

      Maybe they could require the development team to ride along. I'll bet the quality of the code would go way up.

  • Shelley has reached speeds of 130 miles per hour without a driver on testing grounds at the Bonneville Salt Flats in Utah.

    I could do much better with a brick.

    • What kind of brick?

      • A common red brick has a mass of 1922 Kg/m^3 (ref [simetric.co.uk])
      • A standard brick has dimensions of 3 5/8" x 2 ¼" x 8" (ref [mc2-ice.com])
      • The smallest side (and therefore ideal side for max terminal velocity calculations) is 3 5/8 x 2 1/4, and therefore 8.15625 square inches
      • The mass of the brick would be 2.05510989 Kg (Google calculator ref [google.com]
      • The coefficient of drag would be 2.1, according to this calculator site [calctool.org] (which, incidentally, gives a nonsensical result of 3mph - which is why I did the math myself below)
      • The
      • Re: (Score:3, Funny)

        by BoppreH ( 1520463 )
        You know, you could just put the brick ON THE PEDAL OF THE CAR.

        Given a straight and long enough track, of course.
        • Doh. But isn't it at least *mildly* interesting that a brick's terminal velocity is just about the same 130 MPH? Something needs to make me feel better about completely missing the point (and spending the time doing the math). :)

        • by Barny ( 103770 )

          Huh, what?

          Oh yeah.

          Fucking WOOOOOOOOSSSSSHHHHHH.

  • If you were going to build a robot car, why not build it out of something you can get real cheap. Like, say, you know, your Grandma's Plymouth Aries K.

    • Re:Audi?! (Score:5, Insightful)

      by Tiger4 ( 840741 ) on Friday February 05, 2010 @02:00PM (#31037620)

      Audi, specifically Audi Quattros, have been rally car favorites for years. Big engine, good tranny, four wheel drive. There are a lot of people that know how to get them running well, and the cars are built well to do the job.

      • Re: (Score:2, Funny)

        by Anonymous Coward

        Big engine, good tranny, four wheel drive.

        I'll thank you to leave my mother/father out of this.

      • Specifically Audi Quattros? That's anything made by Audi with AWD. It's the exact same system as used by Volkswagen and called 4Motion, where they use torsen LSDs with a 5:1 maximum slip ratio, with the maximum slip then limited by automatic application of the ABS. Porsche uses basically the same thing. The most likely explanation is that Subaru and Mitsubishi have enough WRC wins to not need the publicity, but the post-VW-acquisition Audi could use some positive press, and it's worth it to provide a car an

        • Re: (Score:3, Informative)

          by moosesocks ( 264553 )

          From what I remember, 4Motion does not necessarily refer to an AWD system with a torsen differential, although this has been the case in several instances (VW have also used a viscous coupling for AWD). Quattro, on the other hand, refers to a very specific set of technologies.

          As far as the engines go, Audi make one of the better (if not the best) 2 Liter engines out there. They put most V6s to shame.

    • To paraphrase the brilliant (but eccentric) Dr. Brown: The way I see it, if you're gonna build an automatic driver into a car, why not do it with some style?

    • If you were going to build a robot car, why not build it out of something you can get real cheap. Like, say, you know, your Grandma's Plymouth Aries K.

      They specifically state in the article that they're shooting for real racing speeds. While I don't know what a Plymouth Aries is, I think my grandma has a Reliant. I'm pretty sure that isn't going up Pike's Peak at anything like racing speeds.

      The Audi is a car that can let drivers shine, and any flaw in the driving algorithm will be quickly apparent at high

  • by imaswinger ( 592216 ) on Friday February 05, 2010 @02:04PM (#31037680)
    Ari Vatanen racing up pike's peak: http://www.youtube.com/watch?v=TKgeCQGu_ug [youtube.com]
    • Once you get past the boring intro... damn! That is some fancy drivin'! Drifting through turns at such high speeds on unpaved roads inches away from a cliff-top drop to certain death... Makes me wonder how he can drive comfortably with such enormous balls.
  • The good news about this is it can now drive only itself off a cliff when the gas pedal gets stuck.

  • Rod Millen's time

  • It's going to be interesting to see how the car detects and handles the drainage culverts. The last few miles have corrugated metal drainage culverts crossing the road periodically to carry off the snowmelt. These often get overloaded and instead you have a big mud puddle with a hard metal culvert under the mud. Humans can use a little intelligence and slow down for these. It will be interesting to see how the automated Audi handles these and other unexpected situations!

  • My understanding is that this [wikipedia.org] is generally solved but doesn't really lead to anything interesting unless you know the right formulas for the for the domain. Any news on whether they got passed formula 1?

To thine own self be true. (If not that, at least make some money.)

Working...