Beta
×

Welcome to the Slashdot Beta site -- learn more here. Use the link in the footer or click here to return to the Classic version of Slashdot.

Thank you!

Before you choose to head back to the Classic look of the site, we'd appreciate it if you share your thoughts on the Beta; your feedback is what drives our ongoing development.

Beta is different and we value you taking the time to try it out. Please take a look at the changes we've made in Beta and  learn more about it. Thanks for reading, and for making the site better!

Autonomous Audi TT Conquers Pike's Peak

samzenpus posted more than 3 years ago | from the mega-turbo-boost dept.

AI 187

fergus07 writes "After a year long research program, this week Audi revealed that its Autonomous TTS car had completed the 12.42-mile Pike's Peak mountain course in 27 minutes. An expert driver in the same car would take around 17 minutes — now we have a benchmark, the race is on, and it's almost inevitable that a computer will one day outdrive the best of our species, and it may be sooner than you think."

cancel ×

187 comments

Sorry! There are no comments related to the filter you selected.

in shaa'a Allah (-1, Troll)

mapkinase (958129) | more than 3 years ago | (#34304902)

Then I won't be driving behind an old lady on the left lane.

Do androids get electronic points?

Re:in shaa'a Allah (1)

OneSmartFellow (716217) | more than 3 years ago | (#34304976)

No, but they might dream of electric sheep

Re:in shaa'a Allah (0)

Anonymous Coward | more than 3 years ago | (#34305492)

They've seen things you people wouldn't believe.

This is how it's done (3, Informative)

Anonymous Coward | more than 3 years ago | (#34304910)

http://www.youtube.com/watch?v=TKgeCQGu_ug (Ari Vatanen with peugeot 405 T16, Pikes-Peak 1990)

Re:This is how it's done (0)

Anonymous Coward | more than 3 years ago | (#34305354)

I immediately thought of that video when I read the title of this article. Too bad the article doesn't have a video :(

Re:This is how it's done (1)

dave420 (699308) | more than 3 years ago | (#34305432)

Awesome film. Kind of like a more rustic version of C'etait un Rendezvous [youtube.com] , but without the urban legends :)

Bah! Stupid "the narrative" (5, Insightful)

DNS-and-BIND (461968) | more than 3 years ago | (#34304922)

This is Slashdot, dammit, we're supposed to be talking about the tech the car uses, the sensor fields, blind spots, known bugs, and so forth. What do we get? A typical journalist "the narrative" story, where humanity is in a race against robots which will surely supplant us. Guh, it's like a rejected 1950s sci-fi manuscript. Bonus points for using the tech-y "benchmark" phrase like the car is some sort of Crysis.

Re:Bah! Stupid "the narrative" (1, Funny)

Anonymous Coward | more than 3 years ago | (#34305068)

This is Slashdot, dammit, we're supposed to be talking about the tech the car uses, the sensor fields, blind spots, known bugs, and so forth. What do we get? A typical journalist "the narrative" story, where humanity is in a race against robots which will surely supplant us. Guh, it's like a rejected 1950s sci-fi manuscript.

Well, an automated journalist could be programmed to write about the facts, the technology, like you said. Maybe we should face off a /. writer with a computer to see who writes the most compelling story? As a prize, the winner could enslave the loser's race.

Re:Bah! Stupid "the narrative" (1)

Urkki (668283) | more than 3 years ago | (#34305328)

This is Slashdot, dammit, we're supposed to be talking about the tech the car uses, the sensor fields, blind spots, known bugs, and so forth. What do we get? A typical journalist "the narrative" story, where humanity is in a race against robots which will surely supplant us. Guh, it's like a rejected 1950s sci-fi manuscript.

Well, an automated journalist could be programmed to write about the facts, the technology, like you said. Maybe we should face off a /. writer with a computer to see who writes the most compelling story? As a prize, the winner could enslave the loser's race.

As a member of the same race as current /. editors, I strongly object to this idea. But if it comes to pass, I'll strive to be the first to personally welcome our new automated journalist overlords, of course.

Re:Bah! Stupid "the narrative" (0)

Anonymous Coward | more than 3 years ago | (#34305542)

As a member of the same race as current /. editors,

What race is that, dyslexic morons?

Re:Bah! Stupid "the narrative" (1)

Shakrai (717556) | more than 3 years ago | (#34305410)

This is Slashdot, dammit, we're supposed to be talking about the tech the car uses, the sensor fields, blind spots, known bugs, and so forth.

It uses magic pixie dust. There are no bugs. It will eventually replace humanity. All our base are belong to it. I for one welcome our new machine overlords.

See, that wasn't so hard, was it?

Re:Bah! Stupid "the narrative" (1)

smitty97 (995791) | more than 3 years ago | (#34306206)

you forgot "what could possibly go wrong?" and "in Soviet Colorado, car drive you!"

Re:Bah! Stupid "the narrative" (1)

Urkki (668283) | more than 3 years ago | (#34305446)

This is Slashdot, dammit, we're supposed to be talking about the tech the car uses, the sensor fields, blind spots, known bugs, and so forth. What do we get? A typical journalist "the narrative" story, where humanity is in a race against robots which will surely supplant us. Guh, it's like a rejected 1950s sci-fi manuscript. Bonus points for using the tech-y "benchmark" phrase like the car is some sort of Crysis.

Yeah, like that's news. It's a certainty, that humanity will either exploit itself to extinction, or be surpassed by AI creatures of our own making.

All we can hope for, is that those AI creatures will find us amusing, and perhaps, wether out of pity, curiosity or boredom,will guide us in our otherwise futile attempt to keep this planet habitable for human-like life forms. Of course we must remember, that those AI creatures will be the "humanity" of that era, and if anybody will carry human/Earth legacy to to the stars, it will be them. For them, preserving us will be roughly analogous to us preserving chimpanzees and gorillas and their habitats.

I think this probably will not happen in our lifetime, unless there are major advances in life-prolonging technologies, but it will happen unless humanity goes extinct first.

Re:Bah! Stupid "the narrative" (1)

Shakrai (717556) | more than 3 years ago | (#34305552)

It's a certainty, that humanity will either exploit itself to extinction, or be surpassed by AI creatures of our own making.

Neither one of those are a "certainty".

Re:Bah! Stupid "the narrative" (1)

Urkki (668283) | more than 3 years ago | (#34305780)

It's a certainty, that humanity will either exploit itself to extinction, or be surpassed by AI creatures of our own making.

Neither one of those are a "certainty".

There may be a few oscillations (like a global war, but so mild that human race survives it), but I don't really see any other endpoints.

Maybe a world that is so starved of resources that developing AI just can't happen is a possiblity, but I doubt even that. As long as there's energy (even if it's just sun & short-delay derivatives like wind and hydro), there's potential for recycling ancient waste if nothing else is available. It's not like matter disappears when we use it (except in nuclear fission).

I find the alternative that humans evolve so much dumber that computers/AI will be out of reach very unlikely. Evolving smaller brain is about as hard as evolving a bigger brain, there's the same barrier of having too big head to be practical, yet still too small to be technologically smart. Just look how long it took until natural evolution got over that barrier (like a quarter of a billion years, depending when you start counting). And there were probably many potentially smart species that went extinct before humans made it, but there's just one species to evolve "stupider", so it's virtually certain that result is extinction.

Conclusion: either we eventually develop the AI that leaves us to eat dust, or we go extinct very very soon in evolutionary time-scale.

Re:Bah! Stupid "the narrative" (1)

TheTurtlesMoves (1442727) | more than 3 years ago | (#34306220)

..but I don't really see any other endpoints.

This is a classic failure of the imagination on your part.

Its also devoid of facts, or historic perspective.

Re:Bah! Stupid "the narrative" (1)

Urkki (668283) | more than 3 years ago | (#34307186)

..but I don't really see any other endpoints.

This is a classic failure of the imagination on your part.

Its also devoid of facts, or historic perspective.

I'm unable to imagine what I'm unable to imagene? Well, duh.

However, you appear to be even more devoid of imagination, because you replied, but only complained about lack of imagination, facts or historic perspective in my post, thus actually contributing nothing useful. Good job.

Re:Bah! Stupid "the narrative" (2, Informative)

Cryonix (1234264) | more than 3 years ago | (#34305914)

A quick google search for "Autonomous Audi TTS hardware" turned up these articles, offering various details on the hardware used:
http://sun.systemnews.com/articles/143/2/feature/22601 [systemnews.com]
http://www.audiusanews.com/newsrelease.do;jsessionid=2CB13CF6B9E286A75E8E2B1663E63318?id=1589 [audiusanews.com]
http://www.topspeed.com/cars/audi/2010-autonomous-audi-tts-pikes-peak-ar92542.html [topspeed.com]

like robot == steel man, and he's taking our jerbs (1)

Thud457 (234763) | more than 3 years ago | (#34306090)

I was born a steel driving man [wikipedia.org] , parse that as you will... ;-)

Re:Bah! Stupid "the narrative" (2, Interesting)

Maxo-Texas (864189) | more than 3 years ago | (#34306118)

I'm not sure how long it will be for truly intelligent machines (could never happen / be impossible without wetware).

But machines able to do our jobs are here now. And for a lot of jobs the annual cost is $15,000. Compare that to $32,000 (after benefits) for even minimum wage jobs and you have to think things get ugly soon.

Already diapers.com has "hundreds of robotic warehouse workers" (business week) and some hospital has "hired" 19 robotic workers *instead* of humans. It seems great as a cost savings measure at first-- but then you have to ask, long term, how do people get even a single dollar to afford the less expensive hospital?

Seriously, we could be looking at 50% unemployment in 20 years. It will be concentrated on the low end. How do we handle that as a society? If we don't, it is going to get violent.

What currently produces more taxes? A robotic factory making a billion a year with 3 human owner/managers or a human factory with 500 workers and 3 owner/managers that makes the same amount?
Substantially lower sales tax, use tax, home property taxes, school taxes, etc. from the first. States will be hurting unless they institute income taxes.

Re:Bah! Stupid "the narrative" (1)

hitmark (640295) | more than 3 years ago | (#34306774)

It could be that what is missing is for software that can reason. So that when it is told to grab something from a shelf, it can work out how to do so using internal simulations and observational data.

As for your employment/tax worry, i agree. Except that right now, China have moved that somewhat backwards as there is a long line of workers willing to work for pennies rather then live the life of subsistence farming. But even there the high water mark is rising, and China is sending out "feelers" to the last source of cheap human labor, Africa.

Thing is, the robotic labor issue may well propel the world into a post-scarcity scenario. That is, if we at the same time adopt local "just in time" production rather then the legacy centralized bulk production we have right now. So that when someone wants a new shirt, a 3d scan is fed into a self contained mini-factory that will cut and construct the shirt from a roll of cloth.

This rather then attempting what sunk soviet russia, and what the multinationals are attempting to a lesser degree, predicting or directing the wants and needs of the population at large and producing to cover those. The multinationals need to have a certain level of predictability for their mass production, and so they employ marketing to direct the wants of the public. Russia attempted to predict the needs via massive data collections and computer models, tho i suspect predicting the weather had better success rate.

It is kinda funny how similar the soviet communist system and the corporate capitalist system is at that level, as both need a level of predictability to function.

Re:Bah! Stupid "the narrative" (1)

mini me (132455) | more than 3 years ago | (#34307018)

It wasn't many years ago that 75% of the population were employed as farmers. A new invention, the tractor, replaced the majority of farm workers with machines. Today, only 2% of the population still work on farms, yet we do not have a 73% unemployment rate.

Twenty years ago people were claiming that robots would be the death of the employment and that we'd all be out of work in twenty years. Here we are. The employment rate has not changed significantly over that time.

We have been replacing jobs with machines for a long, long time. Will the result in employment rates for the next twenty years really be any different than the trends of the past?

Re:Bah! Stupid "the narrative" (1)

need4mospd (1146215) | more than 3 years ago | (#34306514)

Sounds like this would be a perfect article for the new Top Gear US then...

Re:Bah! Stupid "the narrative" (1)

timeOday (582209) | more than 3 years ago | (#34307140)

What do we get? A typical journalist "the narrative" story, where humanity is in a race against robots which will surely supplant us.

It's especially bad because the robot didn't even come close to breaking the record! Let's celebrate this technical triumph for what it is, but save the robot overlords concession speech for when it actually applies.

Re:Bah! Stupid "the narrative" (0)

Anonymous Coward | more than 3 years ago | (#34307142)

typical ./ elitist comment from an atypical ./ socially challenged pud with an expressively lame ./ nick... save your super smart comments for the super smart posts hero... let the rest of us enjoy this very interesting story that completely belongs on the best user submitted and evaluated current affairs technology-related news website in the world... jackass

The most important benchmark would still be... (0)

Anonymous Coward | more than 3 years ago | (#34304930)

...security. Once you could objectively demonstrate that it is actually safer to use automatic driving system it will be easy ride...

Re:The most important benchmark would still be... (3, Interesting)

Anonymous Coward | more than 3 years ago | (#34304970)

I imagine the big test comes a few years down the line, when all the major manufacturers let their cars run along Pike's Peak, with varying traffic patterns (i.e., try to hold 80KPH for three minutes, 30KPH for two, et.c.,) randomized for all the cars to see how well the systems handle unexpected events.

In fact, I would probably insist on not buying an automagic car that hasn't been through a multi-car safety test along those lines.

Re:The most important benchmark would still be... (4, Insightful)

fuzzyfuzzyfungus (1223518) | more than 3 years ago | (#34304980)

Not that easy. Liability allocation.

Something like a car, in a natural environment(not some empty closed test-track stuff), will suffer an accident from time to time. Human error, mechanical issues, sensor faults, algorithmic fuckups, whatever.

With a human driver(who basically all juries trivially recognize as an autonomous agent, since they think of all reasonably functional humans as such), the liability for accidents typically falls on one of the operators, unless a mechanical fault, braking issue, or something of that sort can be proven.

With an autonomous control computer, a jury will be much more likely to see the "driver" as an extension of the company who built the car, just like the brakes or the steering column, and assign liability accordingly.

Even if, lets say, computer-controlled cars delivered a 10-fold reduction in morbidity and mortality(which would save something like 35,000 Americans a year from death, plus an unknown but even larger number from serious injury, just to put things in perspective), that would likely be a net increase in liability for the vehicle manufacturers.

Until autonomous vehicles prove superior safety and insurers and/or legislators recognize the new state of affairs, it'll be strictly test tracks, tech demos, and internal use....

Re:The most important benchmark would still be... (0)

Anonymous Coward | more than 3 years ago | (#34305066)

Adding a black box that records the 'other parties' wrongdoing by keeping camera data,echolocation etc that can tell what happened is only a small step.

Liability? Is a solved problem (1)

Alwin Henseler (640539) | more than 3 years ago | (#34305080)

You're seeking a problem where none exists. There's many fields already where there's someone operating potentially deadly equipment (the user), and manufacturer that might face charges if their product was faulty. Existing cars for one, industrial robots, household appliances, etc, etc.

Once the tech has proven itself, surely some laws could be passed to make it clear who's responsible and/or liable for what. In the mean while, there's nothing stopping a manufacturer from having customers signing a waiver that says "use at your own risk". Besides: if the technology proves safer than cars are currently, it might be a net positive for a manufacturer even if they'd hold a larger part of the risk.

Re:The most important benchmark would still be... (1)

TheRaven64 (641858) | more than 3 years ago | (#34306364)

In the UK, you are legally required to have third-party insurance if you drive on the public highway. I'd expect the situation to be the same with robotic cars. The owner assumes responsibility for it and is required to have insurance. If the car is demonstrably safer than a human-driven one, the insurance will be cheaper. If an accident is caused by a fault in the car, then the owner (or their insurer) will be able to sue the manufacturer to recover the damages, just as they can with failures in any other automatic system.

Re:The most important benchmark would still be... (1)

dwandy (907337) | more than 3 years ago | (#34306760)

Until autonomous vehicles prove superior safety and insurers and/or legislators recognize the new state of affairs

Far more likely is that we will still pay insurance to allow the vehicle on the public road. The sale will include a disclaimer of liability and an agreement from the purchaser that they will buy insurance to drive it on the public road. We will all sign it, and continue to buy insurance. The combined interest of the car manufacturer and the insurance company won't let this go any other way.

Re:The most important benchmark would still be... (0)

Anonymous Coward | more than 3 years ago | (#34307000)

So do what we do with airplanes. If the autopilot is on the pilot is still in command, and responsible for the plane. A car must still have a licensed driver who is responsible for any liability issues, but now he is free to take a phone call or send a text message since overseeing the autopilot generally takes less attention than driving the car.

Re:The most important benchmark would still be... (1)

TheTurtlesMoves (1442727) | more than 3 years ago | (#34306256)

Given how bad many drivers are, this will be trivial to show in the not so distant future. However liability will be an issue. Also irrational thinking along the lines of "I know i am a better driver than the machine".

Robo-Thelma&Louise (5, Funny)

digitaldc (879047) | more than 3 years ago | (#34304932)

"It's an old racing adage that it's a lot easier to make a fast driver who crashes safe than to make a slow driver faster. The penalty for error on Pikes Peak is massive as the edge of the circuit is often a massive cliff.
Audi is logically taking a cautious and considered approach because the negative publicity of a car plunging over a fatal drop would hinder the development.".

Actually, the dash-cam video of the Autonomous Audi speeding off the road, going over a cliff and crashing in a fiery explosion would be pretty damn awesome.

Re:Robo-Thelma&Louise (4, Funny)

kieran (20691) | more than 3 years ago | (#34305148)

"All autonomous vehicles are fitted with extra gasoline tanks made out of cheap plastic for explosive awesomeness".

Re:Robo-Thelma&Louise (1)

digitaldc (879047) | more than 3 years ago | (#34305256)

Yes! Extra large explosions for posterity.

Re:Robo-Thelma&Louise (2, Informative)

Shakrai (717556) | more than 3 years ago | (#34305484)

Audi is logically taking a cautious and considered approach because the negative publicity of a car plunging over a fatal drop would hinder the development.

You mean like this [youtube.com] ? Lockheed Martin invited local media out to test drive their new vehicle. One of their guys started the day off by stating "You can't flip this vehicle." Care to guess what one of the reporters managed to do?

Whoops....

Re:Robo-Thelma&Louise (1)

rwa2 (4391) | more than 3 years ago | (#34305936)

"It's an old racing adage that it's a lot easier to make a fast driver who crashes safe than to make a slow driver faster.

Is there any source for this old adage? Google only turns up this slashdot article :-P

Would have thought I'd at least have run into it in the course of Gran Turismo 4, or even a parody of it in GTA training :-P
That, and I /always/ do much better in races by taking the latter approach... and that's when there's not even a penalty for risking dangerous crashes in the video games...

Re:Robo-Thelma&Louise (1)

14erCleaner (745600) | more than 3 years ago | (#34306336)

Local rumor has it that the car did run off the course at least once, requiring extraction by a tow truck. Audi was extra-tight with any information about the tests while they were going on - they didn't make the news in Colorado Springs until their film crew helicopter crashed [kktv.com] .

I suspect... (5, Insightful)

fuzzyfuzzyfungus (1223518) | more than 3 years ago | (#34304936)

I'd be inclined to guess that the easier(and for many purposes more important) area of computer supremacy won't be in absolute speed(outside of carefully-controlled-for-robotic advantage environments like pick-and-place machines and closed rail tracks); but in sheer endurance.

Assuming that you don't totally cheap out on the fault tolerance or get horribly unlucky, the autonomous car should be able to complete the course every 27 minutes, with occasional pauses for refueling, and longer; but even more occasional pauses for hardware service on the car, virtually forever. That expert human driver, though, will do 17 minutes a number of times; but will be a downright danger to himself and others within 24 hours or so.

For many applications(municipal bus service and low-priority-low-cost mail delivery and commodity trucking/train deliver come to mind), it is virtually irrelevant what a top-notch human in fresh condition can do. What matters is either how many of those you can afford as spares, or what an exhausted, bored, hopped-up-on-stimulants-just-to-stay-awake human can do. Computers, on the other hand, may take longer than one would expect to catch up with best of breed humans in anything resembling natural conditions; but will be able to catch up with real world, performance-degraded humans considerably faster...

Re:I suspect... (4, Insightful)

Rosco P. Coltrane (209368) | more than 3 years ago | (#34305008)

but will be able to catch up with real world, performance-degraded humans considerably faster...

If you'd ever watched a semi truck driver cross the country non-stop running on speed (the driver, not the truck), you'd know it's possible to extend the number of hours a human can perform without significantly degraded performances. In my youth, I've always preferred hitching a ride with a truck driver than ride the bus, as I would invariably get there faster, and I never really felt unsafe.

As for the economics of autonomous vehicles, they'll become commonplace when

- a human behind the wheel is massively more expensive than the computer solution,
- people get over their fear of runaway machines,
- drivers unions are squashed

In short, it's not gonna happen anytime soon. Heck, even trains, the one kind of vehicle that could drive itself completely safely today, are still manned by "drivers" who spend their time pushing a button to tell the computer they're still alive, because passengers would be scared without drivers and unions prevent their removal from the trains.

Re:I suspect... (1)

Tar-Alcarin (1325441) | more than 3 years ago | (#34305210)

Heck, even trains, the one kind of vehicle that could drive itself completely safely today, are still manned by "drivers" who spend their time pushing a button to tell the computer they're still alive, because passengers would be scared without drivers and unions prevent their removal from the trains.

Not true. At least not everywhere.
The AnsaldoBreda Driverless Metro [wikipedia.org] is already in operation in Copenhagen, Denmark.
It feels strange to sit in the front seat of the first car, with a completely undisturbed view of the tracks, but (at least for my part) it still feels completely safe.

Re:I suspect... (1)

dgatwood (11270) | more than 3 years ago | (#34306630)

And most of the elevated people-movers at airports, some (not all) of the Disney parks, etc. are driverless, or if there's a human driver, the driver isn't physically present in the vehicle. IIRC, even BART was designed to be driverless, though the trains are not operated in that mode, last I checked.

Re:I suspect... (0)

Anonymous Coward | more than 3 years ago | (#34306994)

There are lots of these systems, the DLR (Docklands Light Railway) in London is also driverless. But the staff member riding the train who helps passengers and makes sure they get on and off safely is actually trained how to drive it (albeit at low speed because there are no signals a human can understand).

Re:I suspect... (2, Interesting)

moonbender (547943) | more than 3 years ago | (#34305214)

Not sure if you're right about trains. I guess that in a modern rail network, the driver could be made superfluous as long as things are running normal. Normal might even include most everyday delays, minor hardware malfunctions, speed limits (e.g. right now due to lots of leafy mush on the tracks). But you need the driver for extraordinary occurrences.

If you were dealing with an isolated vehicle, having an automated system that simply failed safe might be ok, but I have a feeling that the interconnected nature of rail makes that a lot more problematic both in terms of safety and in terms of spreading the effects of a single incidents. The characteristics of those railway systems that do use automated drivers (we've got one at my university) seem to confirm this -- comparatively small scale, controlled access to the tracks, few possibilities of an incident spreading.

Re:I suspect... (1)

hitmark (640295) | more than 3 years ago | (#34306844)

Air travel is mostly automated, with the pilot spending most of the time setting and adjusting the autopilot based on orders from area controllers on the ground (or even just confirming the corrections delivered via data link when crossing something like the pacific).

Expensive humans (0)

Anonymous Coward | more than 3 years ago | (#34305240)

The relative cost depends on the environment: when trucking in hazardous environments, the squishy tragedy of a dead driver is more of a problem (and the premium for delivery is higher, too). Which is why DARPA is so gung-ho for this technology as a means of automating logistics deliveries in war zones.

Re:I suspect... (2, Informative)

sirlatrom (1162081) | more than 3 years ago | (#34305806)

Heck, even trains, the one kind of vehicle that could drive itself completely safely today, are still manned by "drivers" who spend their time pushing a button to tell the computer they're still alive, because passengers would be scared without drivers and unions prevent their removal from the trains.

Well except in at least Copenhagen, Denmark, where our metro is without in-train operators [railway-technology.com] . As far as I know there is no union for the operating computers, as they have yet to gain sentience.

Re:I suspect... (1)

chrb (1083577) | more than 3 years ago | (#34305950)

even trains, the one kind of vehicle that could drive itself completely safely today, are still manned by "drivers"

List of driverless trains [wikipedia.org]

Re:I suspect... (1)

misexistentialist (1537887) | more than 3 years ago | (#34305980)

Drivers using illegal and imperfect chemical enhancement is not much of an answer to machine superiority, though I guess your other reasons show why it can be practical enough.

Re:I suspect... (1)

TheTurtlesMoves (1442727) | more than 3 years ago | (#34306344)

Trains use a driver because its still one of the best "camera" identification system. It can spot the difference between a pattern of leaves vers a person/tree/cow much better than machine vision systems can. For now.

Of course trains like the TGV that can't stop fast enough anyway, its really a bit moot. Some train systems are becoming driver-less. But its pretty limited.

Re:I suspect... (2, Insightful)

drinkypoo (153816) | more than 3 years ago | (#34305120)

The problem with using long-haul trucking as an example is that like so many jobs it consists of long stretches of boredom punctuated by moments of terror. You're either driving for long, boring periods of time, or you're having an "oh shit" moment due to equipment failure or driver error, usually the latter, and not necessarily yours. Why people like to cut off semis I'll never know. Even just hooking a travel trailer to the back of a 7 or 8 thousand pound pickup gives you lots of opportunities to have the same experience without getting paid. There is an answer to the problem that can be solved by machine, but it involves rails. Having robots drive trucks at this point would be dumb, but rebuilding the rail network where it's been neglected or dismantled and having robots drive trains makes perfect sense.

Re:I suspect... (0)

Anonymous Coward | more than 3 years ago | (#34306184)

Agreed on indurance, but some people are very good at keeping concentration for extended time.

Some human drivers are very close on the physical limits of driving, and often use each tiny aspect of car, surface etc. behavior to be faster (in specific corners). What's important, computer should know how to adapt the the car (setup) to it's own driving model, or adapt itself to a vehicle it gets. Some top drivers are very quick in adapting to a car behavior, therefore creative in extracting speed from a setup that computer could have problems with, and which could be faster than a setup that computer is able drive to the limit (as it's trained for it).

Also, In F1 the car is more important than the driver. I think a machine couldn't be significantly faster, not to mention that overtaking, reacting to uneven road conditions (rain), tyre wear adaptation etc. could be a problem. All of it can be solved, but in some cases you still need improvisation and intelligence, which computers won't have for a long time.

Re:I suspect... (1)

renoX (11677) | more than 3 years ago | (#34307090)

You forgot a little detail: normal driving don't happen in empty roads!
Going fast on a road is only a small part of what human drivers have to do: they also have to monitor the other cars, the pedestrian, etc.

And here I was just thinking... (1)

OneSmartFellow (716217) | more than 3 years ago | (#34304964)

...that NASCAR will be the first auto-race to be fully computerized, and maybe then my dream will be answered. Once computers completely control NASCAR, Billy Nochin and his spawn will completely lose interest (one would suppose), and maybe it will just disappear. One can hope.

Re:And here I was just thinking... (1)

Mr Thinly Sliced (73041) | more than 3 years ago | (#34305072)

Actually NASCAR are going the other way FYI.

They've taken the physical car out of the loop and are using technology to replace it.

Together with iRacing they've endorsed the first official NASCAR series champion for virtual racing:

http://www.iracing.com/partners/ [iracing.com]

Richard Towler has become the first non-resident crowned champion of their online world drivers championship:

http://www.nascar.com/news/headlines/official/iracing.standings/ [nascar.com] .-)

Re:And here I was just thinking... (2, Funny)

dave420 (699308) | more than 3 years ago | (#34305456)

Huh? You can fully automate NASCAR with a steering lock and a brick on the gas pedal.

Re:And here I was just thinking... (1)

ledow (319597) | more than 3 years ago | (#34305726)

Yeah, NASCAR isn't really racing now is it? It's a practice track. I get bored by Formula 1 cars because they all finish within tenths of a second of each other but at least there's something interesting to navigate round. NASCAR is like watching a giant dragster race - interesting for the first few seconds and then it's just more of the same forever.

Tip to Americans: The classification of a "sport" does not entail the spectators wishing everyone would just start fighting to actually provide some entertainment... ice hockey, NASCAR fall under this rule.

Seriously, do NASCAR drivers have one arm's muscles grow longer than the other? Can they even steer right?

It's not a sport, it's a practice track, the sort of thing FIAT had on the roof of its factory to make sure cars ran okay. Short of making a major motion picture (The Italian Job, and no, not that recent heap of shit) which features the track, it's not interesting.

Re:And here I was just thinking... (1)

puto (533470) | more than 3 years ago | (#34306460)

Because "football" and rugby fans are never privy to fights.

Re:And here I was just thinking... (1)

Vectormatic (1759674) | more than 3 years ago | (#34306578)

AFAIK NASCAR cars actually cant steer right, the entire suspension setup is built to only either drive straight or turn left, maybe a very small bit of right turning to exit the pit, but those things are built asymmetrical

And yeah F1 can get boring, but the recent season was quite good, with a rather interesting finale.

The best of us? (2, Insightful)

chill (34294) | more than 3 years ago | (#34304992)

I don't care about autonomous cars out-driving the best of us. I want to see common cars that can out-drive the morons on the freeways! Out-drive the mediocre and worst of us and I'd be happy.

Question though... (2, Interesting)

TheCarp (96830) | more than 3 years ago | (#34304996)

While I do not doubt that these e-drivers will quickly come to out drive most of us, doesn't it come down to a question of how close to the physical capabilities of the car the driver can go or is going already? I expect them to be more reliable overall, more attentive (obviously) and more able to repeat their own performance. However, I am not necessarily so sure that there really is that much more capability for them to squeeze out of the cars than a trained pro driver on a test course is already able to squeeze out.

Driving maneuvers are a constant trade off, the closer to the physical capabilities of the car you come, the faster you can traverse a course, however, it also means having less ability to make adjustments and corrections. It is a crude example but, If 99% of my available traction is being used to make this turn, at this speed,I only have 1% more to add if I need to make an adjustment to my course, or speed.

Admittedly cars can then be redesigned to push those limits.....but thats another issue.

-Steve

Re:Question though... (0)

Anonymous Coward | more than 3 years ago | (#34305070)

So will we be able to text while the car is driving itself? ;)

Re:Question though... (2, Insightful)

drinkypoo (153816) | more than 3 years ago | (#34305146)

The driver has to perform a mystical brain computation in order to integrate everything they're learning about the car through their five senses. The car can have as many senses as you have processing time and I/O to handle. The car can [theoretically] make more decisions per second as well. Ultimately the car is going to be faster... someday.

Re:Question though... (1)

umberleigh (793964) | more than 3 years ago | (#34306920)

The driver has to perform a mystical brain computation in order to integrate everything they're learning about the car through their five senses.

What the hell does taste have to do with driving?

Re:Question though... (0)

Anonymous Coward | more than 3 years ago | (#34305330)

You aren't taking into account that it makes no sense to drive a car at its limit for long. No car, economy or high performance can handle the rigors of being close to its maximum for long. You'd burn up tires, tear up control arm bushings (always a costly thing to replace) blow shocks, eat up rotors and brake pads... waaay before they were close to their expected service life. Ask any race team (especially the non professionals who like to race street cars on the track) and you'll see how often things break when you push a car to its limits.

Proper application of car automation will take into account wear and tear on the car and will drive somewhere between sensibly and what most people are accustomed to/expect.

Don't forget that even after autonomous cars become the norm it will be a long time before most of them are, which means you can have chains of autonomous cars traveling at 100 mph following within 100 feet of each other... not that a inter-car computer communication system couldn't make that work, but because the redneck farmer in the other lane might swerve his 1950's pickup into that chain of cars without a computer to tell the chain of cars it's coming.

Re:Question though... (0)

Anonymous Coward | more than 3 years ago | (#34306778)

Whoops - made a mistake, that should be can't in this: "which means you can't have chains of autonomous cars traveling at 100 mph following within 100 feet of each other..."

My post. I've really got to memorize my /. username

Re:Question though... (1)

TheCarp (96830) | more than 3 years ago | (#34306900)

I wasn't trying to take any of that into account. The original post was about a car that made it through in a time that was approaching that of a skilled human driver.

The time to make it through a course, being the only consideration, since it is not "how many times can it make it through the course at that speed before the car mechanically fails".

Otherwise, why not compare its ability to make it through the course against my grandmother? (who, I seriously believe, is unaware of the fact that she has a gas peddle and could use it)

Re:Question though... (1)

dgatwood (11270) | more than 3 years ago | (#34307166)

You aren't taking into account that it makes no sense to drive a car at its limit for long. No car, economy or high performance can handle the rigors of being close to its maximum for long.

Actually, in terms of car life, vehicles that are mostly driven at 65-80 MPH on the highway last much longer than vehicles that are mostly driven in town, with fewer breakdowns and other problems. What causes the most damage to vehicles is in-town driving at slower speeds. Between the steering, the constant starting and stopping, etc., you have a perfect recipe for wearing cars out.

Ask any race team (especially the non professionals who like to race street cars on the track) and you'll see how often things break when you push a car to its limits.

To be pedantic, race teams push cars beyond their safe limits. They burn up tires because they're peeling out all the way around every curve. They break struts because they're operating at G forces that are beyond the vehicle's true tolerances. And so on. Yes, the vehicles can handle such abuse for a short period of time, but that doesn't mean they are operating at their limit. They're operating way, way past it, and every minute those cars remain functional is more luck than anything else.

...the redneck farmer in the other lane might swerve his 1950's pickup into that chain of cars without a computer to tell the chain of cars it's coming.

This is why the law should require all cars, including vintage cars, to have a transmitter installed that reports steering wheel position, brake activation, and vehicle location. It's not ideal, but it would at least help automated vehicles identify manually driven vehicles and avoid them better.

Re:Question though... (1)

ledow (319597) | more than 3 years ago | (#34305932)

An autonomous car would be incredibly boring. Hell, driving down long motorways is already incredibly boring and it's nowhere NEAR the most efficient way to do that.

And e-drivers can outpace us already in every statistic. They're still shit for driving in general, though, because like voice recognition, image recognition and everything else where you try to get a computer to do a job in a machine designed for human senses, it'll hit a limit that makes it entirely unable to actually *APPLY* any of those abilities to the job in hand. No point having a Formula 1 driver's reaction times (or better) if you still don't recognise that the blob on the camera is an old granny crossing the road until it's too late.

Think that's a "solved" problem? Just the other month two "uncrashable" cars designed with every sensor you can possibly imagine (including "cheating" by having ultrasonic distance sensors as well as video cameras etc.) by one of the largest car manufacturer's in the world crashed, head-on, several times, in a demo to press about how they would avoid all crashes. They couldn't handle their own test-case reliably enough not to make the manufacturer look like an idiot, let alone a real road - and that test case was basically driving them at each other on an test track knowing they were both going to meet somewhere in the middle. They didn't even *TRY* to apply the brakes because the sensor data was not interpreted as the other (identical) car being a threat to them despite heading towards them, front-on at high-speed with identical systems enabled (so no "sun in your eyes" excuses).

Autonomous cars for the next few generations will REQUIRE their own, individual, separate roads. If not technically, then legally. That's an INCREDIBLE expense and requires all sorts of work and income to just get off the ground. And it will take several generations of those roads being low-risk in terms of accidents before you can merge human and autonomous drivers, or have every road autonomous.

Give the idea up. Maybe your grandchildren will see one operating on a separate road and not raise eyebrows at it. Maybe their grandchildren will be chauffeured everywhere. But the problem is not the technology, response times, engines, or anything else. It's how to TELL the computer, using a programming language, to do things in such a way that it's ALWAYS (or even MOSTLY) safe for it to drive on roads that present unknown obstacles (and that includes entirely autonomous roads where a deer strays into their path, or where debris from a natural disaster like a tree collapsing covers their lane). We're not even 1% of the way to solving that problem yet.

When you can literally put an autonomous car in London, give it a destination and TELL people that if they are run over by the car they get a million pounds, and it can do so without anyone being able to claim that "prize" (and not through being dead), then you can start thinking about driving alongside them. Think of the equivalent in whatever your area of expertise is (e.g. a machine that can keep a patient alive and always respond to their heartbeat / breathing rate / blood content / other symptoms properly) and you'll see why it won't happen for a long time.

Re:Question though... (1)

amorsen (7485) | more than 3 years ago | (#34306568)

Formula 1 cars are hampered by restrictions because the drivers can't handle 10G+ in the turns and because there is a limit to how survivable you can make 500km/h crashes. Take away the restrictions and you would immediately see a significant reduction in lap times, but the drivers wouldn't last long.

OK, let's see it do Mt. Everest then (1)

fkx (453233) | more than 3 years ago | (#34305014)

Didn't think so.

Re:OK, let's see it do Mt. Everest then (1)

Vectormatic (1759674) | more than 3 years ago | (#34306668)

i'd love to see you drive up Mt. Everest in an audi TTS

You car has been sued - please leave it. (1)

icebraining (1313345) | more than 3 years ago | (#34305022)

Working together, Audi, Stanford University, the Volkswagen Group Electronics Research Lab and Oracle developed a distinct engineering achievement. The Autonomous Audi TTS Pikes Peak integrates advanced algorithms, the Oracle Java real-Time System (Java RTS), Oracle Solaris and GPS with safety and navigation systems found in stock Audi TTS models to maintain control at a physical performance extreme.

Maybe in 50 years we will take it for granted? (1)

voss (52565) | more than 3 years ago | (#34305030)

"Grandpa you mean you actually had to DRIVE A CAR?? What if mom forgot to pick me up from school...she couldnt just send the auto-cab for me?"

we already have that robot... (1)

mug funky (910186) | more than 3 years ago | (#34305040)

the computer that will outdrive the best human? we've had Michael Schumacher for years. he's a pretty good driving program.

Re:we already have that robot... (1)

Vectormatic (1759674) | more than 3 years ago | (#34306696)

recently the germans produced AutoFahrer 2.0, they called it sebastian vettel, it is already outpacing the old version, but still has some compatibility problems with australian software components

Fatal exception (1)

leptechie (1937384) | more than 3 years ago | (#34305060)

Uh, does anyone else have a problem with this thing running on Java? Java causes death [theserverside.com]

Re:Fatal exception (1)

TheTurtlesMoves (1442727) | more than 3 years ago | (#34306484)

Bad code in *any* language is bad and unsafe code, but java (as a language) has some features that make certain types of errors hard to make. Having said that, to "certify" a java app for say FAA standards, you would need the whole JVM stack(java platform) to be certified. Its a long way from that, that's for sure. The license even says something to that effect (as your link shows).

On the other hand, cars have very lax rules with hardware and software compared to medical and aeronautics. Hell just about any loser can work on a car, including the breaks.

So yes. I think java (the platform) is the wrong tool for the job. I would expect some kind of real time OS and associated tool chain and hardware.

Vapor ware smoke and mirrors as a general case. (0)

Anonymous Coward | more than 3 years ago | (#34305282)

Lets take the driver to an similar yet unfamiliar course and see how both drivers do.

The test conditions allows for a unique to pikes peak success to the bot driver.

If the point of the test is to leverage trust and confidence that due to not seeing the bot driver drive off the cliff, I see smoke an mirrors.

Further if the bot driver completed "pike peak" in 5 minutes it would (be amusing) but would not / should not engender any confidence in the ability of the bot to not dive off any other cliff.

Where an expert driver would have a reproducibly safe experience on the first try without any initial information.

Meh

Should have used a DeLorean (1)

Joce640k (829181) | more than 3 years ago | (#34305364)

It's a question of style...

WRC (1)

Overunderrated (1518503) | more than 3 years ago | (#34305386)

"it's almost inevitable that a computer will one day outdrive the best of our species, and it may be sooner than you think."

Sounds like the author has never seen a WRC race.

I hope it remembers to stop (1)

ishmalius (153450) | more than 3 years ago | (#34305426)

When it gets to the top.

Hurry up, dammit. (1)

Lilith's Heart-shape (1224784) | more than 3 years ago | (#34305742)

The sooner this sort of autopilot technology is refined and made standard equipment, the happier I'll be. I fucking hate driving, and would love to just put the car on autopilot, reach down between my legs, ease the seat back, and take a nap.

The real test (1)

Wonko the Sane (25252) | more than 3 years ago | (#34305764)

I drove my car to the top of Pike's Peak several years when I was on vacation. Driving up is easy. Driving back down, on the other hand, is the real challenge.

It was especially fun considering that my car was sold in a part of the country that was basically at sea level so the computer had trouble dealing with the air at that altitude. How will their automated system deal with the engine stalling out, causing a lose of both power steering and power braking ever couple of minutes?

There's something else I took issue with in the article:

Humans are not very good at driving cars, as is evidenced by our ability to destroy 1.3 million souls on our roads each year.

I don't know where to get good statistics for the rest of the world but you can get information about accident rates and total miles driven in the US and it works out that in the USA the averages are about 1 accident (any severity) for every 500,000 miles driven and 1 death for every 80 million miles driven. That's pretty damn good.

Re:The real test (1)

ledow (319597) | more than 3 years ago | (#34306248)

That means that a driver has an accident (of ANY kind) only once or twice in a driving lifetime, or thereabouts. I call rubbish on that, I think the statistics don't show what you think they show - given that the average "no-claims bonus" is about 3-4 years and there's only one man in the entire UK who hasn't had an single accident (actually insurance claim, which isn't the same thing at all) in over 50 years of driving (according to a news article I read a year or so ago). I think you've looked at *insurance* claims (which isn't the same thing at all) and probably even "fault" insurance claims (i.e. where someone messed up rather than it being unavoidable).

For instance, the UK has about 7 deaths on the road each day, with a population of only 60 million. To fit your statistics there would have to be an enormous driving population and/or an enormous amount of mileage for each. 250,000 are injured on the roads each year. That's 700 a day or thereabouts. I have no reason to believe the UK is significantly statistically different from the US in terms of driving accidents, on average.

I think your statistics are vastly skewed - mainly because most people actually couldn't tell you how far they drove in a year anyway. I also think if you looked at *motorway* deaths it would vastly increase the statistics - almost no-one dies on little side-streets in comparison.

However, I agree that the statistics for cars are probably less scary when expressed as accidents per million miles, but the same would happen for aircraft, trains, ships and anything else and I bet cars are STILL more dangerous in those terms. Large passenger aircraft might kill hundreds or thousands but when they do it makes the news and they mostly do daily trips of 1000's of miles all the time. Hell, I can easily drive 50,000 miles a year if I had a driving job, and that would only see me in an accident every ten years. I have a "near miss" about once or twice a year where only sheer luck and/or psychic ability has prevented an accident.

Re:The real test (1)

Wonko the Sane (25252) | more than 3 years ago | (#34306690)

Here's the data [dot.gov] from the US DOT for miles driven and death rates.

As far as the 1 per 500,000 miles I am using 6 million [car-accidents.com] car accidents per year in the US. If you have better data I'd be glad to see it.

I have a "near miss" about once or twice a year where only sheer luck and/or psychic ability has prevented an accident.

Exactly - near misses happen all the time but rarely result in an accident.

Re:The real test (1)

TheTurtlesMoves (1442727) | more than 3 years ago | (#34306546)

The US has a annual road toll of about 50000. That's pretty high and IIRC quite a bit higher per capita than the EU. Yes Europeans drive less. But we still love our cars, and still drive a lot.

Re:The real test (1)

Wonko the Sane (25252) | more than 3 years ago | (#34306754)

Where did you come up with that number?

Re:The real test (1)

Vectormatic (1759674) | more than 3 years ago | (#34306770)

How will their automated system deal with the engine stalling out, causing a lose of both power steering and power braking ever couple of minutes?

with a turbocharger setup, the TTS has a 2 litre TFSI engine

Oracle Solaris and Java (0)

Anonymous Coward | more than 3 years ago | (#34305846)

Audi's press release says that the car runs on Solaris and uses Java for the GPS. What was that about Oracle killing its open software business units?

HUmans not good at driving? Oh really? (1)

Viol8 (599362) | more than 3 years ago | (#34306114)

FTA:

"Humans are not very good at driving cars, as is evidenced by our ability to destroy 1.3 million souls on our roads each year"

Apart from the fact that you don't just measure driving skill by the number of fatalities, thats just BS anyway. Considering how few accidents there are per mile driven we're EXCEPTIONALLY good at driving. If say 1 billion people in the world drive and they each drive 10000 miles a year
on average then thats 10 TRILLION miles a year, so in other words thats 1 death for roughly every
1 million miles driven.

I don't know about you but I'd call that good driving.

Computer driving faster? So what? (1)

drgregoryhouse (1909704) | more than 3 years ago | (#34306180)

As if I want to watch 20 computer racing each other on a race track. Nobody is putting their health at risk. It's more academic than sports. Like 20 geeks trying to prove who is a better programmer. Get a life and actually drive a car.

Where is 17 minutes from? (1)

TurtleBay (1942166) | more than 3 years ago | (#34306338)

I just watched a video of human Rhys Millen doing Pike's Peak in 12:09. The computer is only at half speed. At those speeds and with a good car like the TT, the computer shouldn't even have to use advanced cornering techniques. If it is a Sunday drive up Pike's Peak, even I could do that. I can see what they are trying to do by useing an autonomous vehicle on one of the world's most dangerous tracks. Just going 30 mph up the course completely misses the point of why the drive is considered to be so dangerous.

Re:Where is 17 minutes from? (1)

Vectormatic (1759674) | more than 3 years ago | (#34306832)

did millen use a TTS for that run? (at work, so no youtube for me)

a TTS is a pretty serious sportscar (by my standards anyway, it outperforms 99,9% of other traffic), but compared to the monstrous racers they use on pikes peak, i wouldnt be surprised if 17 minutes is a realistic goal for a pro-driver in a TTS

*Ahem* (1)

PhxBlue (562201) | more than 3 years ago | (#34306448)

No apostrophe in Pikes Peak. It's a common mistake, because you'd think, "It should have an apostrophe in it." But for whatever reason, it doesn't.

We Need Robot Cars In Ohio (1)

mattwrock (1630159) | more than 3 years ago | (#34307210)

I am sure everyplace would have their share of stupid drivers, but I think we have a higher share of them here. The standard operating procedure for merging here to never let anyone in front of you... ever. Just speed up move and brake hard. Since you don't want to let your adversaries, er... fellow drivers know your intentions, never signal your turn. Given that our fastest drivers drive in the lanes designated for slower drivers because some drivers drive just under the speed limit in our fastest lane to "set an example" for the faster drivers, Ohio is a real mess. The other fun thing is our cops (local and state) will bust you if your even 5 miles over the speed limit. Beside the obvious life saving measures, our traffic flow, and even our wallets improve. On a side note, has anyone done research on the impact of NOT losing 35,000 people every year. I don't want to sound morose, but wow that would have major implications on our resources. I would rather keep people alive and try to fix our resource problems though.
Load More Comments
Slashdot Login

Need an Account?

Forgot your password?