×

Announcing: Slashdot Deals - Explore geek apps, games, gadgets and more. (what is this?)

Thank you!

We are sorry to see you leave - Beta is different and we value the time you took to try it out. Before you decide to go, please take a look at some value-adds for Beta and learn more about it. Thank you for reading Slashdot, and for making the site better!

Google Cars Drive Themselves, In Traffic

CmdrTaco posted more than 3 years ago | from the invite-me-to-ted-darn-it dept.

Google 293

An anonymous reader noted that "At the TED 2011 conference this week, Google has been giving extremely rare demos of its self-driving cars. TED attendees have even been allowed to travel inside them, on a closed course. The car is a project of Google, which has been working in secret but in plain view on vehicles that can drive themselves, using artificial-intelligence software that can sense anything near the car and mimic the decisions made by a human driver."

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

Road Rage (1)

Anonymous Coward | more than 3 years ago | (#35386426)

Will it cut off the slow car parked in the fast lane too? That'll really be mimicking human behavior then.

Amazing (-1, Troll)

ubuntufan6 (2007438) | more than 3 years ago | (#35386428)

A link to the video of driving in traffic here [tinyurl.com]

Re:Amazing (0)

Shikaku (1129753) | more than 3 years ago | (#35386432)

Mod down, goatse link.

Re:Amazing (0)

ebs16 (1069862) | more than 3 years ago | (#35386444)

GOATSE. shikaku, i wish i saw your post in my rss reader.

Re:Amazing (0)

MikeDataLink (536925) | more than 3 years ago | (#35386456)

Asshat.

Re:Amazing (0)

hoytak (1148181) | more than 3 years ago | (#35386544)

Well, I don't have the Barney song stuck in my head anymore.

Re:NOT AMAZING (1)

masterwit (1800118) | more than 3 years ago | (#35386642)

You may be have a name with "ubuntu" in it...but what you did was far from "humanity to others"

What is seen cannot be unseen.

One Question (0)

stms (1132653) | more than 3 years ago | (#35386450)

Does it track you everywhere you go?

Real question is does it advertise, not track (2)

perpenso (1613749) | more than 3 years ago | (#35386538)

Does it track you everywhere you go?

Its from Google, of course it harvests data to better deliver targeted advertising. The real question is will it deliver targeted ads while driving. A pleasant voice telling you of the sponsored sites you are driving near.

Re:Real question is does it advertise, not track (2)

Cinder6 (894572) | more than 3 years ago | (#35386610)

Nah, it'll just take you on "detours" and park in front of businesses similar to the one you were going to...

Re:Real question is does it advertise, not track (1)

FatdogHaiku (978357) | more than 3 years ago | (#35386638)

...The real question is will it deliver targeted ads while driving....

This could be a real boon for the billboard industry! The car plans the route to expose you to the client's advertising product most relevant to your selected destination. And, if you are willing to give up information on your friends personal preferences, you might travel the most direct path along with the message:

As our way of thanking you for your
positive contributions to Google Transit,
you are eligible to disable advertising...

Re:Real question is does it advertise, not track (2)

perpenso (1613749) | more than 3 years ago | (#35386750)

This could be a real boon for the billboard industry! The car plans the route to expose you to the client's advertising product most relevant to your selected destination.

You are on to something but I see a different implementation. No need to change the route, since you are not driving the system can project targeted ads onto the windshield. Billboards are obsolete. Advertising is how we will finally get heads-up-displays into our cars. [/satire]

Re:Real question is does it advertise, not track (1)

FatdogHaiku (978357) | more than 3 years ago | (#35386768)

Billboards are obsolete.

I would agree but sadly old business models never die the just buy legislators. HUD in a car questions the very need for a windshield at all, and we can't have that for many many contributors, I mean "reasons"....

Sad day (1)

NateOsit (1990192) | more than 3 years ago | (#35386454)

This is a sad day for the robot chauffeur industry ...

Video of the card in traffic (-1, Troll)

ubuntufan6 (2007438) | more than 3 years ago | (#35386476)

Not shown in summary, but here is the video [bit.ly] of google car merging in the traffic.

Bettin' your life on the state of the art (-1)

Anonymous Coward | more than 3 years ago | (#35386496)

Drive!
Push it to the floor 'til the engine screams
Drive!
Drivin' like the demon that drives your dreams

Drive!

Re:Bettin' your life on the state of the art (1)

iluvcapra (782887) | more than 3 years ago | (#35387084)

Hardcastle and McCormick [wikipedia.org] . Well played, sir. Well played.

More info on the AI behind this (-1, Troll)

ubuntufan7 (2007440) | more than 3 years ago | (#35386498)


I just found a page on how their car AI works
its just amazing. [goo.gl]

Re:More info on the AI behind this (1)

tagno25 (1518033) | more than 3 years ago | (#35386506)

Slashdot needs a URL extender, and auto block/ban goatse.*

Re:More info on the AI behind this (3, Funny)

sco08y (615665) | more than 3 years ago | (#35386520)

It's pretty amazing how they've stretched the limits of technology!

On US 101? Irresponsible (0, Flamebait)

Anonymous Coward | more than 3 years ago | (#35386532)

The referenced article describes Google taking a car out into traffic on US 101.

In other words Google turned all the other drivers on the road into involuntary, unpaid, unknowing guinea pigs.

There's a word for that - irresponsible.

Re:On US 101? Irresponsible (0)

Anonymous Coward | more than 3 years ago | (#35386560)

If they have a cutoff where the driver can take over, not so much. Driving instructors do it all the time.

Re:On US 101? Irresponsible (5, Informative)

Nailer235 (1822054) | more than 3 years ago | (#35386570)

Classic fear mongering. The car always had a driver in it (with override capabilities) while on public roads.

Re:On US 101? Irresponsible (0, Flamebait)

Anonymous Coward | more than 3 years ago | (#35386652)

It really doesn't matter how safe it is, there are regulations for testing of devices with human subjects, those standards are there for a reason. They are included in HIPAA, to give a concrete starting point. A few decades ago it was a lot more lax, and a lot of really awful experiments took place, the bar now is set really high, and its a lot more strict than just being "safe". The FAA created the Class 3 license for experimental space vehicles, the NHTSA should be doing something similar if this kind of thing is going to proceed legitimately.

I for one am surprised that Google legal department would let this one slide, then again maybe they were never even told about what was happening, typical arrogant kids (I mean that in the best possible sense, of course).

Re:On US 101? Irresponsible (0)

Anonymous Coward | more than 3 years ago | (#35386656)

So you'd trust Average Joe to do the same with something he built?

Re:On US 101? Irresponsible (1)

Ravon Rodriguez (1074038) | more than 3 years ago | (#35386884)

Do you trust student drivers with instructors in the passenger seat? At least in this case a capable, trained driver can take over at a moments notice. Being cautious isn't necessarily a bad thing, but there comes a point with every new technology where it has to be tested in the real world (or are suggesting that it never be allowed off the track? I can see it now: fully automated stock car races... And you thought NASCAR was boring before).

Re:On US 101? Irresponsible (4, Interesting)

BrightSpark (1578977) | more than 3 years ago | (#35386600)

Have you seen the lunatics out there? Give me a robot any day! We are given a licence (one test only) in our youth and then out you go, rain, hail or shine, fit or unfit, tired or not, drunk or senile or both. That's ignoring the meatheads who want to deliberately drive dangerously and those not paying attention on a mobile phone texting "RORL" (roll off road laughing). I see your point but lets move on.

Re:On US 101? Irresponsible (0)

Anonymous Coward | more than 3 years ago | (#35386626)

Seems to me they need an override and learning mode where a person can see when the car is starting to do something wrong and tell it what to do, same as you would to a sixteen-year old learning to drive.

Re:On US 101? Irresponsible (0)

Anonymous Coward | more than 3 years ago | (#35386694)

What, are you saying every time someone wants to drive for the first time every existing driver should sign a form? Explain how this is less responsible than when Drivers Ed kids hit the public streets for the first time.

google has deep pockets (1)

Joe The Dragon (967727) | more than 3 years ago | (#35386722)

google has deep pockets.

so you get hit google pays you alot to shut up.

Re:google has deep pockets (1)

Luckyo (1726890) | more than 3 years ago | (#35386846)

Last time google accidentally tried to sidestep a law, they landed themselves in a very big and very public pile of shit. It cost them a lot too, and not just money.

This is no longer "just" privacy issue. Here they are playing with people's lives. A car packs a serious amount of energy when in motion on a road, and it doesn't matter how good of a software or a driver watching software perform is - there are LAWS that specifically require extreme measures for such testing. Reason being that human testing which is what it's doing no matter how you spin it - we're risking people's lives by putting a high energy kinetic "projectile" on wheels in machine's hands and releasing it onto a busy street.

Get a permit and a properly controlled environment, and test away. At least in EU, car manufacturers have been doing it for at least a decade.

awful, awful awful awful (-1, Flamebait)

decora (1710862) | more than 3 years ago | (#35386536)

there is no way a @#$ robot can judge what to do about oncoming accidents, like a pedestrian, a deer, a squirrel, a semi jackknifing, an ambulance passing, a crash ahead of you, a gigantic pothole, a box full of dishes that fell off a truck, a big tree branch, a patch of black ice, a tire blowing out, a semi weaving in a strong wind, etc etc etc.

the solution to the 'i dont want to waste my time driving' problem is to build more trains, and make cities more walkable. to do that, you first have to win over the 'trains = communism' crowd using some kind of distributed jobs program (like the military does for its socialist money wasting mega-projects) and through targeting conservative locations for building up the rail infrastructure so they will get pork from it.

Re:awful, awful awful awful (2)

maxume (22995) | more than 3 years ago | (#35386542)

So how do humans do it?

Re:awful, awful awful awful (3, Insightful)

perpenso (1613749) | more than 3 years ago | (#35386634)

So how do humans do it?

They start with a small number of basic "rules" and acquire the majority of their learning from experience.

In general it seems like an expert systems AI project. You have a domain that has an incomplete definition, many variables and many inputs. You hand code some basic rules as a starting point. You rig up sensors so that a computer can observe the environment and the human and it generates new rules based on its observations(*). And/Or you let the computer loose in a simulated environment and it learns through trial and error.

(*) In the movie Starman the alien learns to drive via observation. Red means stop, green means go, and yellow means go faster. Selection of the expert to observe is a critical step.

Re:awful, awful awful awful (2)

TheLink (130905) | more than 3 years ago | (#35386654)

Most can't. The big difference is we hold them personally responsible for it. And they often die in their own screw-ups which is about as accountable as you get.

The responsibility/accountability is the main problem: if Toyota released a self driving car and it crashed and killed people in "corner cases" where even a skilled human might, Toyota would still get in trouble.

Using the elevator is generally much safer than using the stairs. Fortunately for elevator manufacturers and suppliers an elevator shaft is a more predictable environment than a network of roads.

Re:awful, awful awful awful (1)

Ravon Rodriguez (1074038) | more than 3 years ago | (#35387036)

I couldn't agree more. Many don't realize how much computers run our lives, mostly because of how seldom they fail. On the other hand, most peoples' direct experience with computers are PC's running badly written software. The paranoia is understandable, but people need to realize that these are two different animals. A system designed for a specialized task running on standardized hardware is more reliable than any person on the best day of their life... Which is why we have anti-lock brakes on cars and autopilot on airplanes.

I'm not suggesting that an AI be given full, unchecked reign of the road (at least not right now, anyway), but a type of "smart" cruise control with a human backup wouldn't be a bad thing.

Re:awful, awful awful awful (1)

khraz (979373) | more than 3 years ago | (#35386554)

there is no way a @#$ robot can judge what to do about oncoming accidents, like a pedestrian, a deer, a squirrel, a semi jackknifing(...)

The same holds true for many people. And robots probably have a much easier time driving and talking on their cellphones.

Re:awful, awful awful awful (5, Insightful)

causality (777677) | more than 3 years ago | (#35386720)

there is no way a @#$ robot can judge what to do about oncoming accidents, like a pedestrian, a deer, a squirrel, a semi jackknifing, an ambulance passing, a crash ahead of you, a gigantic pothole, a box full of dishes that fell off a truck, a big tree branch, a patch of black ice, a tire blowing out, a semi weaving in a strong wind, etc etc etc.

Perhaps not, but it's likely to be a hell of a lot better at not doing the idiotic things that cause the overwhelming majority of accidents in the first place.

Almost all accidents other than collisions with animals that run out in front of you are due to human stupidity. Black ice may be an exception, maybe, except that if the conditions allow it to happen a prudent driver accounts for the possibility (note that if you hit a patch of black ice the accident is considered your fault esp. for purposes of determining liability). Everything from:

  • Tailgating
  • Running red lights
  • People who think the purpose of the left lane is to drive the exact same speed as the car to your right so other drivers are tempted to perform dangerous maneuvers just to get around your inconsiderate punk ass, rather than submit to your roadblock
  • General failure to yield
  • A belief that your text message is more important than the lives of others
  • A sudden urge to make a right turn from the left lane because proper planning of your route is too much to ask from a puny intellect and you're far too self-important to go a little up the road and find somewhere to turn around and go back
  • Drunk driving

You name it. It's plain old human stupidity. It's a particularly egregious kind of imbecility too, the kind that fails to recognize that other people exist and can be harmed by your poor decision-making. If "robots" can be programmed not to do these things I'm all for it. Alternatively, if robots can be programmed to beat the living shit out of people who do these things, I'm all for that too.

to do that, you first have to win over the 'trains = communism' crowd using some kind of distributed jobs program

That's a new one to me. I have heard complaints that many train systems would be uneconomical, in the sense that they'd never survive without some kind of subsidy. I haven't heard anyone actually refer to alternate transportation as a tenet of Communism, however.

Re:awful, awful awful awful (1, Interesting)

gr3y (549124) | more than 3 years ago | (#35386878)

People who think the purpose of the left lane is to drive the exact same speed as the car to your right so other drivers are tempted to perform dangerous maneuvers just to get around your inconsiderate punk ass, rather than submit to your roadblock

Spoken like someone who thinks the purpose of the left lane is to allow someone to exceed the legal speed limit. The left lane is for passing traffic not traveling the legal speed limit.

Someone traveling the legal speed limit is not an obstacle, they are a responsible, law-abiding citizen. What lane they are in is irrelevant, since there is no legal justification for exceeding the legal speed limit to pass them.

The root cause of the problem you cite is poor impulse control. There is no compelling reason for society to tolerate an inability to exercise self restraint and drive the legal speed limit. It does not matter if you agree that this is true because it is true; it is self-evident from even the most casual appraisal of the facts.

Re:awful, awful awful awful (1)

CosmeticLobotamy (155360) | more than 3 years ago | (#35387026)

Everyone assumes the speed limit will be exceeded by 5 - 10 mph, including those who set it. Its main purpose is to generate municipal funding through what is essentially a random tax, and to ensure that traffic doesn't go much more than 5 - 10 over the number posted on the side of the road, 'cause that number plus 10 is usually about what's actually safe.

Adhering to the speed limit as though it's set by God is not virtuous, it's just annoying. Please move over for people that want to get past you. If they're creating a life-threatening situation, you'll know it no matter what the sign on the side of the road says. Feel free to call 911.

Train systems (1)

Beryllium Sphere(tm) (193358) | more than 3 years ago | (#35386948)

"the real reason for progressivesâ(TM) passion for trains is their goal of diminishing Americansâ(TM) individualism in order to make them more amenable to collectivism."

Not quite calling trains Communism, but in the same league.

The quote is from George Will,http://www.newsweek.com/2011/02/27/high-speed-to-insolvency.html

Re:awful, awful awful awful (2, Insightful)

Anonymous Coward | more than 3 years ago | (#35386730)

there is no way a @#$ robot can judge what to do about oncoming accidents, like a pedestrian, a deer, a squirrel, a semi jackknifing, an ambulance passing, a crash ahead of you, a gigantic pothole, a box full of dishes that fell off a truck, a big tree branch, a patch of black ice, a tire blowing out, a semi weaving in a strong wind, etc etc etc.

A bunch of your example are redundant, object in lane will suffice.

If the computer maintains safe following distances and brakes when there is sudden deceleration ahead or an object ahead then it will probably do better than many drivers I've seen on the road. If it has sufficient sensors to be situationally aware, ie is it safe to change lanes, and is able to change lanes to avoid an object then it will probably do better than most of the drivers out there. Keep in mind that the sensors may have better perception than the human visual system. It may be able to detect that deer beyond headlight range. It certainly would not suffer from the most common human driving failure, inattention.

These complaints sound similar to the arguments made when antilock brakes and traction control systems were introduced. How will the computer know what kind of surface I am on (paved, gravel, dirt, wet, snow, etc) so it can break accordingly. While perhaps a good question in theory when compared to an expert driver (as in professional racing/pursuit instructor) but when compared to the average driver on the road not very relevant. The computers only need to outperform the average driver.

That said. I wouldn't own/operate such a car until laws are passed to shield operators from lawsuits. Cars with automated driving are going to be law suit magnets regardless of who is at fault.

Re:awful, awful awful awful (1, Funny)

Jeremi (14640) | more than 3 years ago | (#35386932)

How will the computer know what kind of surface I am on [...] so it can break accordingly.

Ah, a WinCE-based system, I see.

Re:awful, awful awful awful (1, Funny)

Anonymous Coward | more than 3 years ago | (#35386734)

Well, if it had to make a choice, I would program it to hit the softest target and to avoid pissing off the animal rights groups. That would leave the pedestrian.

Re:awful, awful awful awful (4, Insightful)

Mr. Freeman (933986) | more than 3 years ago | (#35386740)

Absolutely false. The car is completely capable of detecting pedestrians, deer, stopped cars, etc. This thing knows how to stop in the event that some shit goes down (see link below). You're just making up a lot of bullshit based on literally no research.

SOURCE: http://www.youtube.com/watch?v=Atmk07Otu9U Skip to 2:10 to see where the ABC reporter makes a move like she's going to run out in front of the car. The thing slams on the brakes.

Re:awful, awful awful awful (2, Informative)

The boojum (70419) | more than 3 years ago | (#35386876)

TFA even mentions one of these avoiding a deer: "You could see the cars avoiding things like a deer that dashed in front of one or another making it carefully around a small hillside road, as a large truck came toward it."

There was also a story here a year ago [slashdot.org] about Stanford's efforts in this area with a computer-controlled car doing a 180 spin into a tight parking space. "That means Junior could have an entire language of extreme driving maneuvers it could unleash when called upon ... itâ(TM)s also a sign that the cars of the future will be able to respond to any adverse condition with remarkable driving talent."

Put them together and we're well on our way to computerized control that's safer than the average driver.

i am not "making up a lot of bullshit" (1)

decora (1710862) | more than 3 years ago | (#35386936)

i am thinking about every unforseen situation that car drivers encounter, that could never possibly be predicted. then i am trying to imagine a robot taking over those decisions. it doesn't work.

deer and pedestrians jumping out in front of a car are one thing.

i mentioned several other things that the robot has no clue how to deal with. pulling over for a fire truck that is trying to get down the road? not going to work. there are a ton of other things that can happen that there is no way you can ever predict.

what if the car is going up an icy road, and then it starts sliding backwards because it couldnt make it? how is the robot going to slide back down the hill? it doesnt even know where the road is, the road is covered in ice.

how about your wheel falls off? it happens. what does the robot do?

what if the robots sensors go out? get jammed by some unforseen event?

most of all what it lacks is judgement. the robot can hit the brakes, but maybe it should have been going slower in the first place. drivers know in certain neighborhoods you slow down, they know when its dark to slow down, when its wet you slow down, they know where blind corners are, they know about the semi trucks that cut too close on the sharp turn on the way to work every day (and how you might have to back up) , people understand how to deal with a 4 way stop if the lights go out, they know the highway has a bad bump here and you might lose traction for a half of a second, they know there are drug dealers on this block so dont go this way,

they know that if you have two choices, one goes by an accident up ahead and the other goes through the parking lot of the church, you go through the church parking lot so as to not be a rubbernecker. but you slow down because its sunday and old people are walking through on walkers and, they are driving and their reaction times are bad.

etc etc etc.

Re:i am not "making up a lot of bullshit" (1)

Sarten-X (1102295) | more than 3 years ago | (#35387016)

you go through the church parking lot so as to not be a rubbernecker.

...which may be illegal (depending on your jurisdiction), and doesn't account for the fact that a robot has no interest in staring at carnage. It can't be a rubbernecker, regardless of its lack of a neck.

The other situations you describe can be handled easily. When something goes wrong, the car says "oh shit" and alerts the person inside, who can assume manual control as needed. I'm a roboticist myself. We do this routinely. It's not hard for a sensor to notice a tire's missing. It's not hard to notice that the car's stopped moving in the direction is's supposed to be. It's not hard to notice that the road under the back bumper isn't the same as it was under the front. There's even sensors to verify that sensors are working properly.

Prediction isn't the idea. Instead, there are a set number of rules, including all of those pesky laws like "no cutting through private property". The AI just has to determine which rules are applicable to a given situation, and that's a pretty easy job. That deer looks a lot like a wolf, which looks a lot like a box, which looks a lot like a tire tread, which looks a lot like a child. Whatever it is, it's in the middle of the road, and I should change lanes.

Re:awful, awful awful awful (2)

Zancarius (414244) | more than 3 years ago | (#35386848)

there is no way a @#$ robot can judge what to do about oncoming accidents, like a pedestrian, a deer, a squirrel, a semi jackknifing, an ambulance passing, a crash ahead of you, a gigantic pothole, a box full of dishes that fell off a truck, a big tree branch, a patch of black ice, a tire blowing out, a semi weaving in a strong wind, etc etc etc.

the solution to the 'i dont want to waste my time driving' problem is to build more trains, and make cities more walkable. to do that, you first have to win over the 'trains = communism' crowd using some kind of distributed jobs program (like the military does for its socialist money wasting mega-projects) and through targeting conservative locations for building up the rail infrastructure so they will get pork from it.

Good grief, man! You were making a really good point until you turned half of your post into a rant against the right-wing military industrial complex or whatever the current groupthink buzzword is for defense spending and generic angsty hatred toward individuals who don't happen to agree with you.

But yes, I certainly agree. AI platforms for driving vehicles have a long ways to go until they're able to interact with any imaginable situation (though I suspect it'll mostly involve stopping/slowing down and letting the human take over). That still doesn't negate the point that it's fairly impressive!

For future reference, though, save the political rant for a more appropriate discussion like those you might find on politics.slashdot.org. Otherwise you come off as an angry disaffected hipster with an axe to grind.

I, for one, find Google's efforts in this area to be very interesting even if it's not commercialized for decades. Worthless for now? Maybe, but it's a demonstration of technology that could potentially make roads safer in the future even if only parts of the system are implemented. Think collision warning/avoidance for the masses since not everyone has $80,000 to blow.

if you want to build trains (1)

decora (1710862) | more than 3 years ago | (#35386882)

you have to deal with the people who have been stopping the building of trains, for the past 100 years. there are many different groups that oppose trains, but they are in general, on the conservative side of the spectrum model of political opinions.

the 'individualists' would

Re:awful, awful awful awful (2)

Sarten-X (1102295) | more than 3 years ago | (#35386950)

the solution... is to build more trains

Trains don't come to my front door, which is rather necessary when I'm carrying a load of groceries to feed my household. Trains don't go to my friend's front door, who's effectively wheelchair-bound. Trains don't go from my grandmother's house to the post office, where she has her mailbox. Trains require tracks, which require a bigger initial investment than roads, and simply can't reach withing reasonable walking distance of everywhere people need to go. Then there's the noise, the difficulty in meeting their schedules, the limited per-person carrying capacity, and they also can't stop fast enough to avoid all those hazards you mentioned, even with a human driver!

Buses are a bit better in most of these regards, but just not enough to make shared transit an effective alternative to personal cars. If your house and your job are close to a train station, then go ahead and use it. Don't expect your situation to apply to anyone else, though.

Re:awful, awful awful awful (1)

shutdown -p now (807394) | more than 3 years ago | (#35386964)

there is no way a @#$ robot can judge what to do about oncoming accidents, like a pedestrian, a deer, a squirrel, a semi jackknifing, an ambulance passing, a crash ahead of you, a gigantic pothole, a box full of dishes that fell off a truck, a big tree branch, a patch of black ice, a tire blowing out, a semi weaving in a strong wind, etc etc etc.

I highly suspect that majority of accidents on the roads happen due to human recklessness and/or inattentiveness, not because of any of the above.

Car runs Linux! (-1, Troll)

ubuntufan8 (2007442) | more than 3 years ago | (#35386540)

The code [goo.gl] they use for navigation actually runs on Linux.
And they plan to open source it! and hardware design too!
(They use 8 cameras and few dozens of sensors)

Re:Car runs Linux! (0)

Anonymous Coward | more than 3 years ago | (#35386604)

Parent link is goatse.

Re:Car runs Linux! (0)

Anonymous Coward | more than 3 years ago | (#35386614)

Mod parent up.

It's too late for me, it doesn't have to be for you.

Re:Car runs Linux! (0)

Anonymous Coward | more than 3 years ago | (#35386628)

Oh GOD MY EYES!

Re:Car runs Linux! (0)

by (1706743) (1706744) | more than 3 years ago | (#35386616)

The code [goo.gl] they use for navigation actually runs on Linux. And they plan to open source it! and hardware design too! (They use 8 cameras and few dozens of sensors)

$ wget http://goo.gl/zjJOI [goo.gl] -O /dev/null 2>&1 | grep -i goatse
Location: http://goatse.ru/ [goatse.ru] [following]
--2011-03-04 19:55:36-- http://goatse.ru/ [goatse.ru]
Resolving goatse.ru... 78.47.200.67
Connecting to goatse.ru|78.47.200.67|:80... connected.

To the writer of awful, awful, awful (0)

Anonymous Coward | more than 3 years ago | (#35386568)

It will seach the internet for what to do about oncoming accidents!! DUH! its Google

do they get 1 million miles per gallon (0)

Anonymous Coward | more than 3 years ago | (#35386580)

... the way Bill Gates allegedly said they would by now if they followed Moore's Law?

Notice, I did say "allegedly".

Can't wait ... (3, Interesting)

antdude (79039) | more than 3 years ago | (#35386620)

I can't drive due to my disabilities. This would be useful. Of course, it has to be bug free (OK almost). It probably won't be ready until after I am dead though. I always wanted KITT type of car! :(

Re:Can't wait ... (2)

Gorimek (61128) | more than 3 years ago | (#35386788)

It only has to be better than the average driver.

At least from a rational standpoint. From a legal standpoint it probably has to be 100 times better and wait 10 years "just in case".

Re:Can't wait ... (0)

Anonymous Coward | more than 3 years ago | (#35386866)

It has to be better than a really good driver. The concern with robot cars is not that they have a tendency to crash. Average drivers have a tendency to crash, and people will still ride without concern.

The concern is that one cannot control the variables. I can choose drivers who are sober, and experienced, while I cannot discern a robot.

This is the same concern people have with flying, despite it's vastly safer statistical margins. People are bothered when they are not in control of their own fate.

Re:Can't wait ... (1)

vlueboy (1799360) | more than 3 years ago | (#35386914)

It only has to be better than the average driver

No, drivers can be delicensed too. It only needs to be able to *get* a license, voice-assisted by whatever "disabled" person is on the driver's seat --think disabled men, or backseat children whose parent driver just had a heart attack at 70mph. Good plan, unless lawmakers actively kill this automation effort in its infancy on grounds of "automated terror attacks that leave no video-trackable humans to chase... like we caught that Times Square guy in 2010" or something.

IBM's recently lauded "Watson technology" will be miniaturized and eventually the car will be like the Star Trek enterprise's "computer" listening and responding to its Picard. To pass a license test, it has to understand the driver w/in milliseconds. That's to hear, parse, evaluate for command flaws (like asking to brake while stationary), compare to its realtime input and then act.

Re:Can't wait ... (1)

magarity (164372) | more than 3 years ago | (#35386944)

This is the same concern people have with flying, despite it's vastly safer statistical margins. People are bothered when they are not in control of their own fate.

Flight is an excellent example of how human control is still far better than any computer; would an autopilot have safely landed that US Air flight in the middle of the Hudson river? Would even the best modern AI have brought UA232 in as well as the human crew did?

Re:Can't wait ... (3, Insightful)

Lachrymite (115440) | more than 3 years ago | (#35387014)

That's just anecdotal. What about instances where clear pilot error has caused fatal crashes? You can't just pick out particular instances, one way or the other, and made your judgment on the issue based on that. I honestly don't know whether it's true or not, not having looked at any data on it myself, but I think it's a huge mistake to jump to conclusions like that.

Re:Can't wait ... (1)

magarity (164372) | more than 3 years ago | (#35386898)

I can't drive due to my disabilities. This would be useful. Of course, it has to be bug free (OK almost). It probably won't be ready until after I am dead though. I always wanted KITT type of car! :(

Keep in mind the guy with the robotic arms whose car swerved off the road and crashed for no apparent reason when they offer to give out these cars to the handicapped.

2nd order effects (4, Interesting)

hajus (990255) | more than 3 years ago | (#35386622)

The 2nd order effects from this are going to be interesting. If you only have robot drivers (and you will, cause with lower accident rates, you'll have lower insurance rates if you always let the computer drive), you won't need visible signs or traffic lights. How would this affect pedestrian crossings? Would pedestrians feel irrationally unsafe crossing a road with robot drivers on it? Will we remove speed limits as computer reaction and cognitive ability gets faster?

Re:2nd order effects (1)

Anonymous Coward | more than 3 years ago | (#35386684)

No, we will see speed limits changed randomly and capriciously in order to try to trap robot cars which haven't updated their data in the last several minutes.

Re:2nd order effects (0)

Anonymous Coward | more than 3 years ago | (#35386718)

Assuming they work the bugs out (and they will), I would feel much safer crossing in front of a car with a computer driving. Hardware doesn't fall asleep or get distracted by talking about their latest sexual exploits on cellular phones. If designed well, they can detect when they are not operating correctly due to hardware failure and take steps to correct the problem. Plus, dealing with failed sensors in a car scenario is fairly easy: worst case you can have the computer just slam on the brakes. Airplanes and spaceships are harder because it's not obvious what you should do when failures are detected.

Re:2nd order effects (1)

Jeremi (14640) | more than 3 years ago | (#35386900)

If you only have robot drivers (and you will, cause with lower accident rates, you'll have lower insurance rates if you always let the computer drive), you won't need visible signs or traffic lights.

Until humans driving cars is actually made illegal, there will always be some cars still driven by humans (if only because some people like doing the driving themselves, or want to use a "classic" car that doesn't support auto-drive, or -- most likely -- don't trust a computer to drive them safely). And as long as there are any humans driving, visible signs and traffic lights will still be required.

So maybe in a number of decades, practically no humans will be driving anymore... but all of the world's robot-cars will be designed to work with the existing traffic lights and visible signs. So even then, the traffic lights and visible signs will still be necessary, for backwards compatibility with all of the existing robot cars on the roads.

Re:2nd order effects (0)

Anonymous Coward | more than 3 years ago | (#35386930)

Give the cars a way to continually check how good the brakes are and adjust distance behind vehicle in front accordingly AND cruise control for fuel economy and relaxed speed limits would make sense.

If this is transferred to public transport and haulage, how long before there's strikes and sabotage? I don't see this in anything other than private cars / pickups / whatever

Re:2nd order effects (1)

shutdown -p now (807394) | more than 3 years ago | (#35386980)

Would pedestrians feel irrationally unsafe crossing a road with robot drivers on it?

Maybe, but I don't think it'd last for long. Do you feel irrationally unsafe whenever you're walking past an automatic door? I mean, it could close right on you if it malfunctions. And don't even get me started on what could happen if it's an elevator door!

Re:2nd order effects (1)

Sarten-X (1102295) | more than 3 years ago | (#35387054)

I'd assume there will always be visible signs. It doesn't make much sense to equip every car with automated systems, because each one would likely require custom programming. Also, in the event of a computer malfunction, you don't want the human backup drivers, who are already not used to driving, trying to work without direction.

Speed limits are an interesting question. Assuming that cars actually did reach that point, I'd expect to see speed limits become raised or lowered to adjust traffic rates as needed. Higher-capacity routes get higher speed limits, to make them more appealing to the route-planning software. Areas where the government would like to simply avoid having cars could be slower, but have a faster route pointing away from them, so the shorter path takes longer. It's an interesting avenue of thought.

Re:2nd order effects (0)

Anonymous Coward | more than 3 years ago | (#35387066)

I gave up driving after getting into a really bad car accident. Both that experience, and my long time as a pedestrian since then, has made me someone who would be overjoyed at seeing robot drivers. Because they could be counted on to consistently follow logic and the various laws. Between the people who are crappy drivers because they're too nice, and the people who are crappy drivers because they just don't care, I never quite know what to expect when crossing the street.

obligatory (1)

CohibaVancouver (864662) | more than 3 years ago | (#35386670)

In Soviet Russia, car drive you.

Re:obligatory (1)

MickLinux (579158) | more than 3 years ago | (#35387012)

Well, in Google Van, Aliens abduct you. http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=Harrisonburg,+VA&aq=0&sll=37.0625,-95.677068&sspn=46.495626,67.763672&ie=UTF8&hq=&hnear=Harrisonburg,+Virginia&ll=38.407894,-78.900323&spn=0.001421,0.002068&z=19&layer=c&cbll=38.407894,-78.900323&panoid=ruQjeIA46xTLhu6UG8YFMA&cbp=12,234.51,,0,0 [google.com] Take a look -- everything goes dark, except for a beam of light coming straight down... you look up, and there's a flying saucer with its tractor beam above you... I don't know what's going on here, but until I do, I'm staying out of Google cars.

Whats the worst that could happen? (1)

gmuslera (3436) | more than 3 years ago | (#35386674)

well... an average human driver could take the wheel.

Re:Whats the worst that could happen? (2)

SuperQ (431) | more than 3 years ago | (#35386890)

Absolutely true. Every time I watch traffic from my commuter bus on the highway I feel very very glad I'm not out there driving.

It gets worse when I'm out cycling. I was cycling on a nice quiet park road in SF (Washington+Arguello Blvd) the other weekend and some tourists were driving 15mph swerving in and out of the shoulder/bike lane. I ended up picking a safe time to pull around and pass them on the left since I wasn't interested in waiting for them to do something stupid like slam on the brakes.

If this comes to market... (1)

DJ Jones (997846) | more than 3 years ago | (#35386682)

I hate to give credence to hollywood scripts like IRobot but there could be potential downfalls to having a pure AI computer controlling a potentially fatal machine on public streets. Suppose for whatever reason, an accident occurs in which a computer going 70 mph must choose between hitting a 7-year old girl or killing the driver. The human driver might sacrifice himself at all costs. A purely logical computer might not.

All in all it would probably be better than normal drivers out there but food for thought.

Re:If this comes to market... (0)

Anonymous Coward | more than 3 years ago | (#35386728)

What is a 7 year old doing on the freeway?

Re:If this comes to market... (0)

Anonymous Coward | more than 3 years ago | (#35386744)

Rationally speaking, the driver is more valuable than the 7-year-old girl, unless it's her 7-year-old girl. A driver represents many more years of sunk costs in educating, her, feeding her etc. By contrast a 7 year-old girl can be replaced in a little under 8 years. But it's good of you to point out this advantage of the computer driver.

Re:If this comes to market... (1)

Nimatek (1836530) | more than 3 years ago | (#35386746)

Likely human reaction in this hypothetical situation: frantically trying to do something and getting everyone killed. Another likely reaction: "Screw the girl and the charges, at least I will live." It's not their girl, after all. Human driver sacrificing himself at all costs? You watch too much Hollywood indeed.

Re:If this comes to market... (1)

Dun Kick The Noob (904001) | more than 3 years ago | (#35386780)

That can be mitigated by programming with bias for children, its just a matter of how long it takes and whether governments are willing to enact legislation for it. I dont have the statistics for this but I would think the proportion of accidents attributed to driver error and driver being a dick is significantly higher than other causes. If we can get rid of this with a programmable system why not. Besides having a whole city run by centralized programming reduces gridlock I would think because the system can optimize the whole driving population at a given time and automatically reoute areas experiencing congestion. On a nonsensical note, I think its cool if they park my car in a "elevator thingy"hahha

Re:If this comes to market... (1)

davester666 (731373) | more than 3 years ago | (#35386894)

Great. A car that aims for old people.

Things to factor in:
-age
-how much of a drain on society are they [ie a old person also uses more healthcare and gets more benefits, similar with disabled person]
-how famous are they

Re:If this comes to market... (1)

Raptoer (984438) | more than 3 years ago | (#35386832)

You're giving the computer too much credit. The choice that the computer makes is dependent on how it was programmed. No computer at the application level is purely logical.
An AI isn't going to be driving your car, an automated driving program is. The driving program won't even consider your worth compared to the 7-year old. It's likely that it will be programmed to chose paths which will cause the least damage while prioritizing some types of damage.

Having an AI driving your car is silly, I don't want my car to have an existential crisis while on the freeway. Having an automated driving program driving a car would be fantastic.

Re:If this comes to market... (2)

Jeremi (14640) | more than 3 years ago | (#35386910)

I don't want my car to have an existential crisis while on the freeway.

Here I am, brain the size of a planet, and they ask me to drive to the AM/PM to pick up beer. Call that job satisfaction? 'Cause I don't.

Re:If this comes to market... (0)

Anonymous Coward | more than 3 years ago | (#35386874)

Unfortunately, your post demonstrates why the US will be the last place in the world to have this technology, and it comes down to one word: lawyers. What company is going to take on the liability of programming a system that will almost certainly maim and kill people on a regular basis. Yes, it would reduce injuries overall. Yes, it would improve traffic. Yes, it would increase vehicle efficiency. But it would only take a few million dollar court judgements to make it largely unprofitable.

Re:If this comes to market... (0)

Anonymous Coward | more than 3 years ago | (#35386998)

A robot may not injure a human being or, through inaction, allow a human being to come to harm. If one of the two people must die, that doesn't help
A robot must obey any orders given to it by human beings, except where such orders would conflict with the First Law. Either way will conflict, so that doesn't help either
A robot must protect its own existence as long as such protection does not conflict with the First or Second Law. Disobey first law or first AND third law? The girl gets it, with laws written like that. They don't have to be, though - make the first law something like "A robot may not injure a person outside the car....." and the second "A robot may not injure a person inside the car..." and the driver is sacrificed. The media also get a major scare story about cars that kill drivers, but if the girl got hit, they'd have a field day over that too. I'd guess about 10-20% of people would say "so auto related deaths has gone down 99% with these things" and the rest would be against them because of the headlines over a single event

How can Google afford to provide this free service (0)

Anonymous Coward | more than 3 years ago | (#35386748)

But why are we stopping at J.C. Penny's? Hey car!!!

Legal Issues (1)

kombipom (1274672) | more than 3 years ago | (#35386770)

It'll be interesting how the governments of the world choose to deal with self driving cars. They are bound to be statistically much better than humans but I bet it will be a long time before you don't need a license holding, sober, able-bodied person behind the wheel even though most of them will be checking their Facebook updates on the dashboard screen rather than "supervising" the car.

Re:Legal Issues (0)

Anonymous Coward | more than 3 years ago | (#35386828)

You mean pretty much what some dumbasses are doing nowadays w/o the benefit of a self-driving car?

Months ago, I had the pleasure of seeing a ~15 y/o kid look up from his bike (texting) in total surprise to see me driving on the road, I slowed down but he freaked out and with his one free hand, applied his front brake full stop and flipped over. He'll make a mighty fine driver one day.

Not the first (1)

artor3 (1344997) | more than 3 years ago | (#35386836)

Google isn't the first to do this, not by a long shot. Last year, there was a story about an autonomous car driving from Italy to China [ieee.org] . There were humans on board to take over in the event of a problem, of course. I'm sure there are other examples as well.

It is cool tech, but I think it'll be a long time before it's mature, and an even longer time before it gains acceptance. People want to be in control of their lives, even if they're better off relinquishing control. Whether it's long road trips due to fear of flying or keeping a gun in their nightstand, people often choose to do something that is statistically more dangerous rather than put their lives in someone else's hands. I'm not saying that that's a bad thing or a stupid thing, just that it's human nature. Not many people will be willing to trust a computer to drive them, even if it's safer.

Re:Not the first (1)

xnpu (963139) | more than 3 years ago | (#35386942)

Just a matter of time really. My parents will think it's ridiculous, I'll think it's "scary" for the reasons you mentioned, but my kids (or their kids if this needs a lot more time) will know no better.

Re:Not the first (1)

internettoughguy (1478741) | more than 3 years ago | (#35387000)

People want to be in control of their lives, even if they're better off relinquishing control.

That's true only until the first iCar is released.

Mimic human decisions? (0)

Anonymous Coward | more than 3 years ago | (#35386840)

I would prefer that it makes better decisions than human drivers.

"Car turning left in front of me...better change lanes, I'm sure its clear, no need to look or signal."

"Motorcycle coming, should I wait before pulling out right in front of it? Naw"

A 25%+ reduction in fuel use (1)

h4x354x0r (1367733) | more than 3 years ago | (#35386922)

which would easily be attained by an automated transportation system, is nothing to sneeze at. Or 30,000 fewer US highway deaths per year. Or a reduction of 3-5 Million ER Trauma admittances per year (if you want to see medical costs go down). There's no technical reason we couldn't have a completely automated transportation system in the next 15 years, except for the fact that the US couldn't even switch to Metric. But yeah, because there would probably still be a few thousand (instead of 36,000) deaths per year, it simply wouldn't be good enough. Go figure.

Tech for the handicapped (0)

Anonymous Coward | more than 3 years ago | (#35387024)

There are a lot of great tech coming out for the handicapped. I live in a house with 3 other drivers that I feel would be less capable than most automated cars, and I'm afraid to get in the car with them after dark. I'm by far the better driver, and have better eyesight and faster reflexes but can't have a drivers license now due to unpredictable seizures. I say bring on the handicapped tech.

Daemon anybody? (0)

Anonymous Coward | more than 3 years ago | (#35387062)

I always pictured Matthew Sobol exactly like this

http://img.timeinc.net/time/2010/poy_2010/poy_mz/poy_cover_z_1215.jpg

made the book a lot better.

Quick analysis (0)

Anonymous Coward | more than 3 years ago | (#35387068)

This video has three important features. 1: the car drives really fast. 2: the wheels make a noise that shows the whole steering/speeding control scheme is badly implemented (good driving will never make your wheels cry like this, here you can predict a high risk of losing adherence). 3: the people engaged in the test are not stressed, it seems to be quite a usual task to them, which may prove to be a high level of confidence and/or a low level of security measures.
The fact that the article was published in a very bad website (here bad means full of bullshit) makes me wonder if any information given in it is reliable. Given the loose structure of the video, I would say "hell no".

Load More Comments
Slashdot Login

Need an Account?

Forgot your password?