Slashdot Mirror
Google: Our Robot Cars Are Better Drivers Than You
An anonymous reader writes "At a robotics conference in Santa Clara, California, the head of Google's autonomous car project presented results of a study showing that the company's autonomous cars are already safer than human drivers — including trained professionals. 'We're spending less time in near-collision states,' he said. 'In addition to painting a rosy picture of his vehicles' autonomous capabilities, Urmson showed a new dashboard display that his group has developed to help people understand what an autonomous car is doing and when they might want to take over.' This follows another (non-Google) study earlier this week that found the adoption of autonomous cars could save thousands of lives and billions of dollars each year. Urmson also pointed out that determining liability for an accident is much easier using the data collected by the autonomous cars. At one point, a test car was read-ended, and the data showed it smoothly braking to a stop before being struck. 'We don't have to rely on eyewitnesses that can't be trusted as to what happened — we actually have the data. The guy around us wasn't paying enough attention. The data will set you free.'"
Have the Google robot take on the Stig round the top gear test track.
Autonomous cars will more than likely drive at exactly the speed limit. So on that stretch of highway you were used to doing 65mph in a 55 zone... well that slow car (hopefully in the right lane) will be the Google one.
I guess that's when the human takes over?
TODO: create/find/steal funny sig.
We'll soon reach a point where autonomous vehicles are orders of magnitude less likely than human-driven vehicles to have an accident. It won't matter, though; people would rather face a daily one-in-a-million chance of dying due to their own mistake than a daily one-in-a-billion chance of dying due to a machine failure.
Autonomous vehicles will still take over in the end. It's just that this particular rational motive to make it happen won't be contributing very much. So, it'll take longer than it should, and more people will die.
Is it Google? Is it the consumer?
They are right that the data will have a lot of power over you in these situations...
...in the future you are being looked at as being crazy if you tell other people that you are still driving yourself.
"Seriously, how can you live with that - risking the life of others. Robot-Cars are much safer."
Spelling errors were made for your amusement only...
if {collision}
then {arbitrary braking profile}
else {real data}
[Fuck Beta]
o0t!
If you had read TFA, you would have noticed that the robot car operates more safely than humans in the highway infrastructure that is in place today. We don't need to redesign today's infrastructure, if we switch over to autonomous cars.
I'd come to a complete stop once too, and was almost rear-ended. Luckily, I'd been glancing in my rear-view mirror and noticed that the guy pulling up behind me wasn't paying attention. So I rolled forward a few feet, and he ended up just stopping in time. Autonomous and sensors are one thing, but picking up on non-sensor cues and reacting accordingly (going into motion when the vehicle is legally supposed to be stopped) are still beyond the realm of sensors and algorithms.
I came to know that the cost of a Robot car is around $150,000 including a $70,000 LIDAR (laser radar) system. So, what is the use of being better Driver while it is too much costly so that it can not be used by many people even if the Government allows self-driving cars in future?
The car was read-ended? Not sure that is possible.
I get tired after about 5 hours straight driving. But for city and suburban driving, no chance. There is no tech yet that can anticipate a child about to kick a ball out onto a road, or to see that a pedestrian is about to walk out in front of you without looking first.
No they aren't.
Time Bomber the Book coming soon.
I can totally see these cars breaking down all at the same time when in some unforeseen extremely unusual condition. I think it is therefore premature to conclude they are safer than human drivers.
"The data will set you free"
That's what the NSA is saying too.
Why is there some automatic assumption that if A is safer than B, we must permit A and eliminate B?
Are no other factors allowed to be considered? Not wanting everywhere we go to be logged by Google and the Government, maybe? How about just the visceral enjoyment of a perfectly executed heel-toe, nailing the rev match? How about the realization that by far what's most likely to kill me is a heart attack or cancer, not a car accident?
Yes, cars kill people. Yes, I can be killed by the next guy's car. I don't care. Life isn't about cowering from everything in sheer terror at minuscule risks. I also jump out of airplanes from time to time, and climb sheer rock walls. Both can be dangerous. I do them anyway, because I enjoy it.
Why are the "safety before every other thing" people always assumed to be right? We're insulating our children from any perceivable level of risk, and it's leaving them unprepared to deal with real life.
How did Caesar put it? "Cowards die many times before their deaths. The valiant never taste of death but once."
Or imprison you, as the case may be.
the problem isn't that I wouldn't trust a Robot Driver, but, how can you be sure it won't get hacked? or malfunction? Some things should always be left to be in control by a human. Intuition isn't something a robot can acquire.
No way the consumer can control the data. If he could alter it, he could claim innocence while he is liable. So the "carputer" (it's an ugly name so somebody is going to use it eventually) will be closed source or DRM. It's great for public transport, but not for something you want to call My car.
---- MISSING MISCELLANEOUS DATA SEGMENT --- [sigdash] trolololol
Thanks to our dear friends at the NSA, law enforcement will soon have the ability to override the destination selection of autonomous cars and have any driver/passenger they wish promptly delivered to a convenient jail or donut shop.
I love technology!
Scruting the inscrutable for over 50 years.
Two times in the last two days google has blocked forwards from my MX, plus two other times in the last couple of months. That's four times I've had to ask them to unblock MX and whitelist me. (Each time they claim they're getting too much UCE from it – they're not, I have spam assassin and two dnsbls that block way more that what google lets through.)
If their driving software works as well as their ability to keep me whitelisted then I wouldn't trust it to drive me in a golf cart on a golf course.
I commented about insurance and liability a couple days ago when another autonomous vehicle story was posted. This answered my question:
a new dashboard display that his group has developed to help people understand what an autonomous car is doing and when they might want to take over
Well there you have it. As long as a human has the ability to take over, and it's a decision they have to make, then the liability goes from Google to the person sitting in the driver's seat. Subtle but clear as day. Google wants to transfer liability off of their system onto a person in the vehicle. I can see it in court now "Our dashboard clearly indicated to the driver 5 seconds before the accident that it could no longer maintain control of the vehicle given the circumstances involved and that the driver was to disengage the system and take over control."
Better known as 318230.
I'd like to see videos of these autonomous cars maneuvering in almost roadless situations going to various rural cottages, and snowy conditions.
Sure, they will eventually match and surpass us humans, but not yet.
"we actually have the data. The guy around us wasn't paying enough attention. The data will set you free.'"
The old argument we collect all your data and IF you are acused of a crime we can set you free if we have ALL YOUR DATA, but with the data
we can adjust the payment for your insurance even if you have no accident.
This circumvents and undermines the common principles of law: You don't have to proof that you are innocent, an acuser must proof that you are guilty.
These are many steps that will eliminate the freedom in between and a new generation of 1984 conformists are born. Innocent until prooven guilty.
STOP - GOOGLE - NOW!
I really wish that development would progress toward trains than autonomous cars. Trains bring [energy and infrastructure] efficiency to the table, where autonomous cars... not so much of anything.
That's the only way this is going to work with thousands of predatory, fraudulent lawyers waiting to pounce with clients in cahoots. I would also recommend a video log of forward and rear for the inevitable Russian-like heaving yourself onto the car pretending to be hit.
It's also a huge concern for household robots that may clean your house someday.
(-1: Post disagrees with my already-settled worldview) is not a valid mod option.
You'll find jobs doing something else once the machines take over! No really - I read it on slashdot all the time!
Driverless Cars Are Further Away Than You Think: "Most daunting, however, are the remaining computer science and artificial-intelligence challenges. Automated driving will at first be limited to relatively simple situations, mainly highway driving, because the technology still can't respond to uncertainties posed by oncoming traffic, rotaries, and pedestrians. And drivers will also almost certainly be expected to assume some sort of supervisory role, requiring them to be ready to retake control as soon as the system gets outside its comfort zone."
This is all an elaborate scheme to increase the fan base and therfore the profits of auto racing.
That's all well and good until you're given a notice that they will be shutting down the automated driving service because they have Google Taxi now.
Demented But Determined.
You complain about "the expected cost".
Did you ever think about that EVERYTHING anyone earns anywhere is a "cost" for somebody else? Nature and economies are circular systems.
You WANT "costs" to be high - that means incomes are high. Of course, you don't want ANY costs to be high - battle tanks, mines, bridges to nowhere, poison gas, etc. are costs that are bad to have. Paying people to do nothing, by the way, is not on that category - these days A LOT of people would be much better paid to do nothing because what they DO get paid for is actually bad for the majority of people.
So "costs" are over all GOOD, but you have to look at the details, what they stand for. Too much abstraction is bad, comparing apples and oranges ("cost, money" makes everything seem completely equal) has gotten WAY too far.
Our robots drive cars better than your robot cars drive.
- Skynet
Now go and take this out into New York City on 5th avenue at 5pm ET rush hour during the work week.
No, seriously, I want to see how well this car performs in a city where the posted 40mph speed limit oin the Staten Island Expressway is ignored by the vast majority of cops and motorists, the normal speed is about 70mph or so, and people will rear end you out of spite if you go too slow for them.
Then get me the data on how much less it costs to run this car.
Unfortunately, since the organs from car accident victims can be used to save several more lives, driverless cars might cause a net increase in deaths due to fewer donated organs.
1. Will you still be drving drunk if you have your autonomous car drive you home after a night of drinking? 2. What if you are driving link and ass and rear-end someone, will they be able to use that data against you? What if both people are at fault? 3. Who's going to absorb the liability for these cars when something unexpected breaks? The large automotives are going to drag their feet for years on self-driving cars. Their will need to be a lot of testing in real life before they mass produce any cars.
So why is it when I say that computers should run City Hall for 90% of the work that is the same all over the planet, I get strange looks?
People have been living in cities for millennia, you'd think that 90% of the daily operations of a city would be standardized by now. Why does it take such an insane amount of redundant bloated bureaucracy to pick up garbage?
The summary looks like a collection of vaguely-related sentences.
It will do better than you will under those conditions, because it will have GPS, and it can react faster to slips than you can.
Leave my roads alone, you pompous greedy lying, deceitful bhasturds. Stick to collecting data from browsers and phones.
That is all.
FTA: " Urmson showed a new dashboard display that his group has developed to help people understand what an autonomous car is doing and when they might want to take over."
If it was safer than a human driver, this would never happen. What he's saying is that it's safer than a human driver, except when it isn't.
The example accident isn't indicative of anything. If you rear end someone, it's nearly always your fault. There are rare exceptions, but data isn't setting anyone free here, the car in front of you can stop as hard as it wants, at anytime, for any or no reason, and it's your fault if you hit them because you're following too closely and/or not paying attention.
Another bit of his "proof" is that the car accelerated more smoothly than human drivers. Exactly how is this a safety issue?
Also, human drivers braked harder. Did anyone doubt that their algorithms would be conservative? And would never exceed the speed limit? Or maybe the human perception is way more advanced at judging when it's ok to brake later and harder?
So when you're driving today you're in a state of being aware of the situation and are engaged with the surroundings.
If you're letting the car drive, I highly doubt you're paying that much attention. Why wouldn't I let the car drive and I read, do email, surf the web or turn around and talk to the passengers in the rear seats.
In the event where you need to take an emergency action, it's much easy in the first case to go to heightened state than in the second one. Atleast in the first one you aren't completely surprised by the events you're facing before you.
Think of the case of a gravel truck that has a loose load. If I know there's a truck in front of me, I'm not 100% surprised if some gravel comes out, whereby if i'm reading/emailing and I'm forced to take over to avoid gravel, it's more of a surprise and I'm forced to figure quite a bit more out about the situation before I can act. One could also panic because of the amount of elevated emotion or adrenaline dump that would be taking place since you'd go from "reading iPad" to "dodging gravel".
So, what is the use of being better Driver while it is too much costly so that it can not be used by many people even if the Government allows self-driving cars in future?
The cost of ANYTHING is high at first. The main reason for this is fixed costs which are very high on a per unit basis if you haven't produced a lot of units. You need to scale up production to bring the costs down since that allows you to spread the fixed costs over more units. Since we are still in the R&D phase with this technology there is no point in mass producing anything in order to lower the costs. Furthermore as the technology develops we discover cheaper ways to accomplish what was previously expensive.
It took me a while to learn what that meant. Basically, be aware of situations where you could be in an accident and get out of them as soon as you can. Also,try to make it a rule that if there's more than 2 things that could go wrong with a driving maneuver I don't do it :).
:), but I struggle with how to get the point across to my kid that she should periodically be taking stock of her situation.
:)
It's a hard thing to teach though. When I did driver's Ed as a kid they tried to hammer it into us so hard it just came off as a joke ("Blood on the Highway!"). For me, I'm more than a little neurotic, so it came natural
Self driving cars do all that without the messiness of trying to teach them
Hi! I make Firefox Plug-ins. Check 'em out @ https://addons.mozilla.org/en-US/firefox/addon/youtube-mp3-podcaster/
It doesn't have to be closed source, or DRM, which i think is not the term you want anyway. Having some kind of non-repudiation would be nice, but still not completely required. Just take speeding tickets now. The cop catches you on radar speeding and does what? Writes it on a piece of paper. Could he be lying? Absolutely! It's still accepted as evidence.
Time for a contest
For all intensive purposes, "whom" is no longer a word. That begs the question, "who cares"?
EVERYBODY says they are a good driver and better then others. So why would Google do otherwise?
I also like the bit at the end: The data will set you free.
Also : yes, you need eye witnesses. Or at least external experts.
I would not trust a company saying they are innocent in an accident and back it up by THEIR data. "We promise there was no software error in ANY of the cars. All people need to do is sit on the left cheek and hold the doorknob with the right hand. People just are using it wrong. They also signed a waiver when they opened the door. Look it up. It is in the Company-Is-Always-Right law that was passed last week."
Don't fight for your country, if your country does not fight for you.
Yeah, and it will also stop driving as it doesn't believe there's road there. How would it handle parking at a college football game or a county fair where you're directed by a guy in an orange vest with a flash light pointing you to a made up spot in the middle of a grass field?
How will it handle a section of road handled by a flagger where your lane is closed?
How will it handle a flooded roadway? A section of road that has a snow drift where *you* know to slow down?
Before you go out from your daily work, you go to Google maps from your Google desktop using your Google browser (chrome) and you are asked:
(start)
Hello Mr. Smith, do you want to go home or going somewhere else?
- I would like to go to a Chinese restaurant with crab noodles, and on the way pick up my girlfriend.
Checking your girlfriend availability. ... ... ...
She confirmed availability.
These are the restaurants available with tables, please choose one:
- Restaurant one (table available in 35m, 25m to arrive there)
- Restaurant two (table available in 45m, 35m to arrive there)
- Restaurant three (table available in 1h10m, 40m to arrive there)
Thank you for choosing restaurant two. Please await outside at bay 12 for your car. It will arrive in 3m.
(end)
And you will be able to change / track what is going on with your Google glass goggles, your Google clock, your Google belt and your Google wearable computer (the one that was called before a mobile phone in the old days)
They can't come soon enough. We continually build highways wider and wider to solve, really, our own ability to drive properly in rush hour. Rush hour would be very free flowing if we had autonomous cars that automatically made way for merging cars, for instance, instead of those idiots who block them on purpose just to remine 1-2 car lengths in front of them. Rush hour traffic is mainly caused by cars having to force their way in, causing all other cars to have to brake as all those cars are following too closely to begin with. Truly autonomous cars could tailgate each other SAFELY. The density of traffic you could support on our roads would go tremendously.
Also, I'd love a day where I don't have to own a car but I instead say "I need a car in my driveway at 1:15 to take me to work". It arrives at 1:15, and if more people are going to my destination. maybe the car picks them up also. The possibilities are endless, and can truly solve a lot of the problems with the roads today.
However, of course, insurance will be an issue. They won't go down lightly and will find a way to fight it.
Now all we need is Formula 1 driven by robots. This will be truly fascinating.
more so, they will know exactly where everyone is intending to go, the second they get in their car.
that's why Google's pursuing this tech. they know what we're searching for online. now they will know where we are going in real life.
your dash becomes the perfect targeted ad platform.
My God can beat up your God. Just kidding...don't take offense. I know there's no God.
I don't want a driverless car.
I want an automatic chef cook. Because when I go home from work, I still have to sit in the car for 1 hour, and I still have to prepare my food for 45 minutes.
Now, without a driverless car, but having a chef cook, I'd have to sit in the car for 1 hour, and have a meal waiting for me. A net reduction of 45 minutes.
If Pandora's box is destined to be opened, *I* want to be the one to open it.
Where I live there is construction season and winter. Call me when it can deal with those.
I would sincerely like to know how it deals with the this construction situation. Will the car reroute or stop in the middle of the intersection thinking there is a car in front of it.
faked. I'm not sure how admissable it'd be in court but they do accept GPS data so maybe...
google's car may be a great driver in nominal driving conditions but as i've posted before, it cant deal with all situations. if your car cant deal with situations a person can, it's not a safer driver, it's a non-driver.
- bad weather (e.g. heavy snow)
- construction areas
- odd situations
dont believe me? how about the lead engineer?
of course if we added electronic assistance to the road itself (special markers/paint for lines) it could solve the first two but it's the unforeseen that is the biggest issue.
Anons need not reply. Questions end with a question mark.
If you had read TFA, you would have noticed that the robot car operates more safely than humans in the highway infrastructure that is in place today.
None of the three links describe the driving scenario's in which the car operates more safely than humans. Did the article you read mention anything about city driving? That's what i'm interested in.
TFA headline says the cars drive 'better'. How does the Google engineers define 'better'?
lets look in reverse order
> maintaining 'safe' distance from car in front of you...'safe' as defined by the DMV is '2 seconds'...a variable distance that accounts for speed...
so, they programed the Google cars not to get closer than a prescribed ammount...the cars did not significantly get any closer than programmed...so "we win!!"
not so fast...that 'safety' they measure has very rapidly decreasing marginal utility...after awhile a .0000000000000001% gain in 'safety' is not work the extra 20 minutes to your commute b/c the Google control software won't let your car ever get any closer than 2 car lengths...so turning left on that busy intersection takes 3x as long in the morning....
> cars [with drivers] accelerated and braked significantly more sharply....again this is a programmed parameter, not some 'victory' for the car...
they defined 'safe' as smooth stops and starts...programed the machine to do smooth stops and starts...therefore, scientifically "ITS SAFER!"
wrong again...it's ridiculously **slower** which almost by default makes it 'safer'...
humans can estimate distances in real time and **make decisions** that safe TIME
TIME is the key here
anyone can just drive more slowly to be safer...we don't b/c we'd never get where we are going if we didn't cut corners
Thank you Dave Raggett
An autonomous car is a rule based system and can not deal with events outside of its rules. Such as a deer leaping in front of a vehicle, a vehicle with a tire blow out or if a nearby vehicle become uncontrollable. In most instances a Human driver can think outside a set of rules and react in a way the computer cannot to avoid an accident.
I see the biggest benefit to this technology not as saving us from a boring commute but in preventing drunk driving deaths. It's like having a designated driver for everyone. That would save 10's of thousands of lives per year.
Garbage in, Garbage Out
Of course Google's cars are safer, that is until you need to take over -- as evidenced by the new display they are talking about to tell drivers that they need to take over. Of course, that is assuming that the drivers are allert and not doing something else because they haven't had to be paying attention. If the country was interested in highway safety, it has already been proven that lowering the speed limits (and enforcing them) will also produce dramatic results in the reduction of accidents. The main reason being that human drivers will have more time to react.
Simply put, we can already reduce accidents, but the price to pay is safer driving habits. If people aren't willing to exhibit safer driving habits, then they shouldn't be permitted to drive in the first place. In development the old adage is low cost, fast or feature rich, pick any two. The same goes with highway safety. New cars are already out of the price range of the average American, how much will a robotic car cost? Face it, we could have better fuel economy, less highway accidents/injuries/fatalities and lower infrastructure maintenance if all we would do is drop the speed limit from 70 to 60 and enforce it.
But in the future, when the wealthy have their autonomous cars, we will build new prisons to lock up the poor who must be the cause of accidents by driving their antiquated vehicles, or we will just outlaw them all together. Like it or not, at the anticipated price points, autonomous cars are not about highway safety, but maintaining separation of the classes.
If we ever do get flying cars some time in the future, they will almost certainly have to be completely automated.
Last week I had an American friend over and we were talking about driverless cars, and she said she thought they might work in the USA, but having seen what UK roads are like, she was very skeptical they'd work there, so maybe Google should try it!
For example, many roads in tows date back to roman times, and are too narrow for two-lane traffic. You need to look far ahead and work out when exactly you need to duck into a gap behind a parked car to let oncoming traffic through, and when to go for it when you have right of way so as not to block traffic in either direction. And if a block does occur, will it mount the pavement (sidewalk) to free things up, or know when it's time to back up and give in?
The UK has very few towns laid out in a grid, and most roads are twisty, and narrow, other than motorways. Can a driverless car cope with such terrain? If Google really want to prove their technology is better than a human, let them bring their cars over to the UK. If they work here, I'll be impressed.
Before i trust a google software packed car it will be boxed in a faraday cage , dismanteled in components and all of them smashes to bits. .. while you discuss business with a passenger , Google relays it all to the NSA.. bla bla .. or sent to the police for ticketing purposes .. I'd rather trust Big Brother .. of the Devil himself .
Imagine the possibilities
Position is sent to them and your whole travels are nicely tucked away for future uses
Trust a Google Car ? like hell
I propose Google set their cars free in Belgium. If they can prove that they drive safely through Belgium I'll be somewhat convinced.
Why Belgium you ask? Belgium has a very extreme interpretation of yielding right of way at intersections to traffic coming from your right. So extreme that it often extends to blind intersections where you might not even be able to see there is a road intersecting on your right (think alleys in towns). You need to be familiar with the roads in question to know where to yield, otherwise there is no way for you to know what to do.
Well up to a point. The comparison is with 'normal' drivers, American drivers at that. As far as I know driver training in the US is fairly basic compared to many countries, and TFA doesn't actually mention 'professionally trained' drivers at all unlike the /. summary.
I would contest that a 'real' advanced driver, say a UK police Class 1 driving certificate holder, will out-perform a robot any day, not because they are quicker to react but because the skills and experience they have gathered over years of dedicated training and practice would give them the ability to predict problems and prevent incidents, rather than relying on their robotically quick response times to cope with a developing problem the advanced human driver would have already avoided.
Smivs on the intertubes!
If you had read TFA ...
I did. I also read:
Proceed with Caution toward the Self-Driving Car
Completely autonomous vehicles will remain a fantasy for years. Until they're here, we need technology that enhances human drivers' abilities rather than making those abilities increasingly obsolete
http://www.technologyreview.com/review/513531/proceed-with-caution-toward-the-self-driving-car/
and
Driverless Cars Are Further Away Than You Think
http://www.technologyreview.com/featuredstory/520431/driverless-cars-are-further-away-than-you-think/
Personally, I love the idea of autonomous vehicles, but what about privacy? These days, almost all smart phone users must assume that anything they do with the device is being recorded by one or more spy agencies, so will that also be true for self-driving cars? At this point in our history I imagine the answer would have to be yes, which I find depressing.
In the future I hope things will be different, but then we will first have to get money out of politics .
I predict that one day it will be illegal to drive a car manually. I can't predict exactly when, but I bet some of you will live to see it.
So why is Google investing all this money into this project? This is just their excuse to gather more information on you, and it goes without saying that the government will have access to all of it. Google can monitor your health: Airbag arming sensor in the seat measures your weight; Car knows where you shop and where you eat; If you drive to health clubs and how often. Google can determine who your associates are: The car can use biometric cues along with phone data to correlate particular passengers and track meetings even if one of you leaves their phone at home from time to time. Google knows where you are going before you get there and sells that real time information to the businesses at your destination.
I think that is the way to go anyway - why should I drive an vehicle that costs quite a lot and I use twice or 3 times a day? With this part of parking lot problem will go away too and if the companies doing it do it right then we may have shopping minibuses too. The one pity is that the common practice among fired software developers was take to become a taxi driver, now that will be gone.
I cycle to work, so for me it's not much of a big deal (if it's bad in the winter (sub -10C) I take the subway). I could see how commuting could be made better with auto-drive cars. But when I do have occasion to drive during the weekends, I actually enjoy it. Especially in the later evening when the traffic is really thin on the highways and by-ways, where there is pleasure of driving.
Hopefully the option of driving for pleasure, and not simply utilitarian purposes will be possible. (And going to the track really isn't an option for me.)
Similarly with motorcycles. Riding one can be quite enjoyable.
Google's lawyers will have to be better than everyone else's lawyers too, otherwise the very first accident that can't be blamed on "the other guy" will bankrupt even the almighty Google.
And then it is evaluated by a court of law as to validity.
Evidence is not always assumed to be accurate.
While I don't doubt that robotic driver-less cars can drive safer than humans on average, the problem will be what happens when there is an accident. With humans, you can always fault the driver(s). With a driver-less system, the fault will be the vehicle and algorithm. Every accident will be evidence of a flaw in "the system" and will be very difficult to overcome. Every winter we see 5, 10, 20 car pileups in fog and snow and we just accept that as stupid drivers pushing the limits of the weather. If that happens ONCE with driver-less cars they will be pulled off the roads until the system can be fixed, even if the overall accident rate is much lower than humans.
This is the same problem that UAVs are having attempting to integrate with manned aircraft. Manned aircraft always have a pilot to "see and avoid" other aircraft and the ground. You can build a UAV which can demonstrate better or equal "sense and avoid" to a human, but if it fails it is will be seen as a failure of the "system" not a single pilot.
I wonder how these things do under extreme weather conditions, such as ice and snow? Do they still do better than humans?
Years ago, I suddenly hit a patch of black ice and I instinctively steered the wrong way, which only made it worse. I ended up doing a 180 and going over a curb. There wasn't much damage to me or the car, but I was very lucky - a large industrial truck was only about 10 seconds away.
In principle, steering wrong on ice is the sort of mistake that a well-programmed computer wouldn't make. I assume they're trying to account for all situations, but I've never actually seen any coverage about that.
wasnt amy fisher read ended too?
so I made a mistake...its **4** seconds instead of **2** (btw, the DMV rules vary by state in the USA...& some states mandate a time-distance, some don't)
you dodged my greater point...
Google's engineers gamed the test by slowing the cars down...
Of course they are 'safer'
Thank you Dave Raggett
I spend many of my days driving less than 2 mph along the side of the road, usually with the two right tires about 2 or 3 feet off the pavement. Sometimes I'm on the wrong side of the road doing the same thing. All the while looking at an instrument; looking for changes in vegetation; keeping up with how far I've traveled; watching for traffic; reading a map; watching out for holes, broken glass, mailboxes, low-hanging limbs, etc; and trying my best to get near where the gas pipeline is buried. Show me a computer controlled vehicle that can handle that and I'll buy one today. I could sit back, watch the instrument and sip coffee.
Oh, did I mention backing around a mailbox with a 6 foot PVC arm sticking out to the side of the truck? Try that some time; it's a skill not many have. I can back into places where YOU couldn't even drive out of.
I was right all along:
http://en.wikipedia.org/wiki/Two-second_rule
I revoke my apology...
Thank you Dave Raggett
I commute about a thousand miles a week, mostly rural thank goodness and I still enjoy driving. Still work on my own cars too. Robot cars? No thanks.
But for the other screwball drivers out there who seem to have the attention span of a fruit fly, I hope they all adopt the system, and soon. From what I've observed over the past several years, it would seem most cars are already driverless.
/. Dissent will not be tolerated. Think like us or perish.
Why do you think they can't be switched to manual driving if needed? Airplanes have had autopilots for quite some time, and to my knowledge there's not a single one where you can't switch the autopilot off.
The Tao of math: The numbers you can count are not the real numbers.
Here in Europe if you are rear-ended the guilty part is always the car which drives into you. You have to be able to brake to a stop in case of an emergency, it's the fault of the driver behind not maintaining the safety distance or not paying attention.
call me when a robot can get so liquored up it takes its hands of the wheel and is so engaged at
yelling at some other driver that it wraps itself around a tree
Urmson showed a new dashboard display that his group has developed to help people understand what an autonomous car is doing and when they might want to take over.
This is an interesting quote that brings up a few points. If the driver has to take over the vehicle is not autonomous and it is being driven by autopilot. There is a big step between "drive most of the time" and "drive all of the time". I am a big believer in the 80/20 rule. In this case 80% of the work needs to be done to deal with the last 20% of the situations the vehicle gets into. Another issue is that all current tests have been with drivers who's only job is to test the vehicle and monitor it's performance. The average driver will not be that attentive. What happens when the driver is asleep, distracted and/or drunk when the system tells them to take over? What if the driver has not been watching the situation when the system tells him to take over. It will take a few seconds to orient oneself to a dangerous situation and by that time it may be too late to react. This whole "take over when needed" method predicates on drivers doing keeping track of what is going on so they can take over when needed. Most drivers are not attentive when they have to be let alone when they don't have to be.
One of those analyses showed that when a human was behind the wheel, Google’s cars accelerated and braked significantly more sharply than they did when piloting themselves. Another showed that the cars’ software was much better at maintaining a safe distance from the vehicle ahead than the human drivers were.
We’re spending less time in near-collision states
No news here. Computers have better reflexes and measurement techniques than humans and they are infinity more patient. When the car in front slows a bit the computer will instantly know and be able to react which will lead to smoother breaking and better distance control. A human may take a few seconds to react which will mean harder braking and loss of distance. The real question is "is it significant?". Following a vehicle is a simple problem for a computer but there are many, much more difficult issues involved when driving. The Google solution is to have the driver take over which has issues in itself.
Also, who defined what is "near-collision states"? The word "near" is a relative term and is defined by the context where it is used. For example near means something different to an astronomer than it does to a microbiologist. It could be argued that cars are in a "near-collision state" most of the time. All it takes is for the vehicle in the next lane to swerve and there is a collision. If no collision happened is there a problem? is "near-collision state" really a good measure of safety?
When the computer can drive 100% of the time we will have autonomous vehicles. Until then we are trading one problem for another.
Possibly. On the other hand, the robot car has the potential to have hundreds of thousands or millions of years worth of experience in handling such situations (mind you, the current crop of cars don't, but that's because they're still very definitely a prototype design). Regardless, the point of autonomous cars is that they can be used by anyone without such experience: if they're safer than the average driver, they'll save lives. Not to mention potentially eliminating problems like drunk driving entirely.
"None can love freedom heartily, but good men; the rest love not freedom, but license." --John Milton
The fantasy of 'self-driving' cars that will function on ORDINARY roads in a way that matches Human drivers is, exactly that- a fantasy. The real question is exactly what this specific psy-op is all about. Why sell a complete lie about current computing (and sensor) capabilities?
True AI doesn't exists and cannot exist. OK, in this instance, that doesn't matter since pseudo-AI techniques based on culling rules from massive collections of useful data could theoretically drive a car with sufficient sensor capability safely. However, the sensors and datasets most certainly do NOT exist for anything other than the trivial case- simple perfect roads with simple perfect junctions and simple perfect traffic.
You sheeple will mostly say at this point "dribble dribble, I know Google has magic AI technology, dribble, look at their search algorithms and language translation, dribble".
Let's take your two examples. The last first. When language translation (spoken or written) was a task in the hands of traditional AI researchers, the project went no-where, even though computing power had increased by millions of times. Then an individual said "screw AI" and wrote a paper stating that the problem should be attacked using simple pattern matching- using the extraordinary data gathering and storage facilities of modern computers to build simple, straightforward one-to-one translation techniques, then improved by the usual statistical analysis methods like Markov chains. Data from the UN transcripts was used, because by law everything said at the UN has to have first class translations into most Earth languages.
Take the first Google achievement- clever search systems. You enter a few search terms, an Google seems to read your mind. How does it do this? Simple, each day tens of thousands of Googe-employed HUMANS view the common patterns amongst search terms, and they enter hints into the system as to what the user actually meant. No AI, real (impossible) or pseudo- simply Human brains in the bodies of cheap hourly workers looking at what fellow Humans are trying to find,
The task of autonomously driving a car on roads not specifically designed for this task is impossibly hard. You'd have to be a very thick beta indeed not to be able to work this out from first principles. So why does Google lie? Why does Google push the "trust the machine" meme.
Do I really need to ask? We are approaching the age of robotic warfare. America wants to be able to invade future nations with ZERO casualties of US citizens. The creation of a war machine that mass murders, and yet costs not one life of the team doing the mass murdering, is the dream of most evil forces on this planet.
What use are the weapons of self defense in a nation like Iran, for instance, if a future Obama could send wave after wave after wave of un-manned killing machines, by land and air, destroying every town and city that refused to bend the knee to the will of the monsters that rule the USA?
Google is the R+D arm of the NSA, and installations based on Google hardware and software designs are at the centre of intelligence operations in Canada, the UK, France, Germany, Australia, New Zealand, the USA (obviously) and many lesser nations in the UK/USA sphere of control. Of course, no official Google personnel work at these various 'shadow Google' facilities, any more than an employee of IBM operates your home PC, derived from the original IBM design. But Google actually hones its hardware, and software algorithms for the specific purpose of improving the intelligence facilities around the world- the commercial gains for Google are just a very nice side-effect.
The owners/originators of Google are proud racist zionists, and have NEVER sought to hide their allegiance to the most extreme murderous groups in Israel. They see the manipulation of public opinion to accept autonomous killing machines as the greatest gift to zionism. So, you'll continue to have Slashdot push nonsense about robot cars on ordinary roads for the foreseeable future.
Sorry but the only way I would accept giving up the control to a computer is if they upped the speed limits for them too. Surely if a computer controlled car is safer than a human they can go faster right?
There is an ancient saying the roughly goes, "Before the Gods make you fall, they first offer you the gift of Pride." Now how could Google have pissed off the Gods?
That's where this is going.
More and more we are becoming a nation (a world?) where people don't know how to do anything for themselves. Dependence on machines to do everything for us is a disaster waiting to happen. Also, getting your car literally hijacked by a hacker, causing it to take you somewhere you didn't want to go, or intentionally getting you killed. More opportunities for governments and corporations to spy on you and track you. Less privacy. Etcetera.
Are YOU using the TOOL, or is the TOOL using YOU? Think about it!
Indeed. The fact is most drivers are pants, and autonomous cars will undoubtedly save many lives and avoid many crashes. I just wanted to put this in some sort of context.
Smivs on the intertubes!
Yeah, yeah, and before you let that steam drill beat you down, you'll die with a hammer in your hand.
Enjoy your superiority while you can. You've probably got years to do so, but certainly not decades.
...and don't people sometimes malfunction while behind the wheel too?
All robot situations will work great but mixing in humans is just asking for trouble; nobody is so ingenious at finding problems than a foolish human.
I can detect the context of avoiding people on their cell phones and adapt or avoid a mother driving a bunch of kids; these robots can't. That is my "near collision state" to avoid. Now in avoiding these risky situations one could be placed into a worse situation - such as NOT speeding when everybody else is speeding and recklessly passing you - every lane change around you increases the risk.
I wonder what happens when in bad road conditions the google maps indicates to drive off a cliff... Some humans have had troubles in those situations.
Google: Our robot cars have better lawyers than your lawyers.
Democracy Now! - uncensored, anti-establishment news
I will believe them when they can navigate the streets of India :-)
For example, look at https://www.youtube.com/watch?v=RjrEQaG5jPM.
If we're going to talk of safest drivers, then the safest drivers whether human or automatic are the ones that aren't behind a wheel. You can't hurt anyone with a car unless you're driving it. My point here is that Google is ignoring the style of driving, for example, that of the car sitting in the parking lot.
I imagine the automated driver would lose their safety advantage, if human drivers were to drive like the automated ones.
What happens if the GPS is wrong and how do we make the car go to the right location? Just recently, I went to a store and the GPS routed me the wrong way. I also have had many case where the GPS misses the address by several hundred feet. I think before we go robot cars the routing needs to perfect, and that has not been achieved yet.
Huge difference though:
- Both you and the cop have a copy of the ticket, so he can't pull you over for a 10mph ticket and then secretly write it up as 25mph after he's done with you. Nor could you take your copy and do the opposite.
- Police officers in general are in a position of trust (whether its warranted isn't always clear, but its necessary nonetheless for them to do their jobs.) That includes trusting that the officer won't try and do something like this (on the threat of losing their jobs and potentially facing civil or criminal charges themselves if they're caught.)
None of that is really relevant to whether or not its possible to have an open source and still tamper-proof data log. But the answer is "sure why not." Just have the log storage in a "black box" with a tamper-proof seal and an external interface that only allows appends and reads (no modify or delete.)
All of the components could be completely open source / open hardware as long a sufficiently tamper-proof seal can be designed in such a way that it couldn't be re-built without detection should someone attempt to tamper. I'm sure someone clever could come up with such a thing (and probably already has long ago, though I wouldn't know where to look for one, open or not.)
Of course, you probably don't want users being able to muck around with their car's software anyway if safety is a concern (which it is!) Hacking a 1-2ton machine that barrels down the highway at 80mph is a bit more serious of a risk to yourself and others than hacking at Linux kernel on your test box where the worst that happens is you have to hit the power button and try again.
So you could STILL have open everything -- but the implementation would have to be further locked down to prevent unproven hacks from being installed on actual vehicles. Which kind of defeats the purpose to some degree (still leaves the "many eyes" bug-hunting argument for open source but not the "anyone can muck with it" argument.)
Then again, that's kind of needed for closed software as well since there's enough smart people out there who are able to reverse-engineer and start hacking at damned near anything without the source!
Since you can get rid of stop signs and stop lights, I would say that they are better. Well, in non-residential zones anyway.
Their data shows that their autonomous cars don't end up in near crashes as often as human drivers. But it is the same data that is driving the decision-making of the car. Since the decision-making is tailored to what the car considers a near crash (and the sensor data is interpreted conservatively), of course it will get into "near-crashes" according to the metrics of the car much less than human drivers will. Particularly professional drivers.
Their metrics will probably expose human drivers for repeatedly driving recklessly close to tree shadows which they probably avoid hitting by a miracle.
Anybody can make rosy claims like that if collected stats are only for nice weather. If the automated car refuses to drive or function during inclement weather, then it would be easy to boost stats since it takes out all the risk that driving in non-ideal conditions presents. Yet I don't see much if any in regards to that point of data being presented.
Now unlike a robot car, I may have reasons to get from point A to point B during rain, fog, sleet, snow, etc. The conditions during those times are anything but ideal. Also, despite the risks, I have learned to deal with bad conditions to the best extent I can, and am often willing and prepared to drive in them.
Keep in mind that when the weather is bad accident rates go up for human drivers. I may be a good driver even in bad weather, but that person behind me might not be keen enough to slow down and adjust following distance - and that's all it takes.
Show some good evidence that a self-driving car can deal with some awful and crap-tastic weather while maintaining that very low accident rate, and then you will really have my attention.
So, I guess that would be google or google?
I'm all for one of these if google pays when it screws up (and gives me a new spine should serious problems occur.)
We need to redesign some parts of the infrastructure whether or not we switch to autonomous cars.
They are actually remotely controlled by low-paid operators offshore in china and india.
They both control the car, however, both have to apply the gas to accellerate, but only one has to hit the brakes to stop.
If it was cheap enough, this wouldn't necessarily be horrible (well, assuming they had some competition in the market of course..)
If we take an average of say $16k for a low-mid range new vehicle and fully depreciate it over 10 years (ie: equivalent to $1600/yr or a bit over $30/wk or a bit over $4/day.) Add fuel and insurance costs to that $4/day and you could probably have $10-15/day worth of alternative transportation before you start really losing money on the deal.
Of course that's still far far less than current taxi services are willing to offer for any significant trip (at least any I've seen) and also ignores the possibility of picking up a much cheaper used car and additionally ignores the requirement of having enough taxis in service to actually get everyone where they need to go when they need to get there (that last one being the biggest stumbling block by far -- needing thousands upon thousands of vehicles in service for rush hour each day.. all just sitting there idle the rest of the time.)
I don't see how autonomous cars will ever be able to do things that require taking the initiative and forcing your way in, like when you have a stop sign and the perpendicular traffic does not, and because of some obstruction you can't see what's coming, so you essentially just have to stick your nose out and edge into traffic until you can crane your neck forward trying to get a glimpse or just go, hoping for the best, which frequently requires gunning it. I see autonomous cars being much like the Griswolds from European Vacation, trying to merge into the outer circle of that roundabout for hours, to no avail, while appreciating Big Ben every time they go round. Picture yourself merging onto a very crowded fast-moving highway, having to make your own space by wedging yourself in, how can an autonomous car do that, simply because if it goes wrong, it's your liability. "My car did that dangerous move, not me, ticket and sue the engineers" sounds like a future common refrain. Obviously the engineers aren't going to make the car aggressive enough like that, so I forsee a possibly worse problem of these cars coming to a halt when merging, which is even more dangerous, makes traffic worse, and draws the ire of drivers all around you.
Of course the solution is to make all cars autonomous and aware of each other, but can anyone imagine that happening in our car culture?
You can get that from any cell phone now. Self-driving car is an optional accessory.
Having worked on self-driving cars (2005 Grand Challenge), a few points:
The comment about minimizing "near-collision states" is significant. A near-collision state is one where a reasonable variance of the behavior of another vehicle could cause a collision. It's about predicting other-vehicle behavior. That's an important area to study. Aviation people put a lot of effort into minimizing near-misses, and it pays off.
Incidentally, Tesla's announcement that they're starting work on an "autopilot" is them playing catch-up. Audi, BMW, Cadillac, and Ford are already demoing automatic driving systems. It looks like Cadillac will be the first to ship hands-off highway driving, in 2015. All these early systems are highway driving only, although Cadillac includes stop-and-go driving in traffic jams. That's likely to be a very popular feature.
On the sensor side, more progress is needed, and it's coming. That rotating LIDAR contraption on top of Google's self-driving cars is from Velodyne. It's 64 LIDAR units on a spinning turntable. That's a research device, not a production one. There are better ways to do LIDAR, but the cost needs to come down. The approaches used in the Kinect and the XBox One will not work outdoors in bright sunlight. Outdoor LIDAR systems work fine, but they're pulsed, not continuous. For a nanosecond, at one frequency (color) they far outshine the sun. But the total energy per pulse is low, so they're eye-safe. Currently, such devices are very expensive, but that's not for any good reason. It's because some exotic ICs have to be made in tiny quantities.
Radars are getting better, too. A decade ago, in the Grand Challenge, we had to use Eaton VORAD radars, which operate at 24GHz. These could reliably range cars, trucks, and larger bicycles, but not people at long range, or signposts. (Such radars return range, azimuth, and range rate; this isn't a speed gun. I used to have one of these looking out my window at at an intersection, with a display plotting the traffic.) Today's automotive radars are running at 77GHz, with plans to move to 79GHz. There's an effort to standardize on 79GHz internationally. Tripling the frequency, plus applying more compute power to the processing, means that most objects a car might hit are detectable. These radars are getting cheap and small, so a car will have enough of them to provide full-circle data. Long range is needed mostly in front; on the side and in back, much lower power can be used.
A key issue is a high viewpoint. This isn't just about obstacle detection. You also need to profile the road. This was a big deal for the off-road DARPA Grand Challenge, but even on paved roads you need to be able to detect junk on the pavement and potholes. Google has their sensor on top of the roof. This will probably be unacceptable in a production car. I'd go for flash LIDARs at the top corners of the front windows. One possibility is a narrow strip just above the windshield, to contain all the sensors. This is one way to combine vehicle aesthetics and field of view.
Cameras are useful, but computer vision is still kind of dumb. Distance from stereo only works at short ranges, and range rate info from cameras is poor. Digital cameras are so cheap now, so it's tempting to think they can do the whole job. Not yet. Computer vision isn't good enough. Tesla is probably putting too much hope into camera processing. You need cameras to recognize signs, traffic lights, and such. Also, you need multiple sensors because not all objects are visible on all sensors. Radars can't see insulators. Cameras can't see objects with little contrast against the background. LIDARs can't see some materials, such as the charcoal fabric used on many office chairs. Sensor fusion is essential.
Enough for now. This looks quite do-able.
I'll be impressed with Google's solution once they leave the comforts of California and drive day to day on the snow and ice of places like Minnesota, Montana and so on.
I'm sure they will master it someday but until then, Googles solution is incomplete.
Has the Google autonomous car reached the point where it can be trusted on roads it's never sensed before. From my understanding the whole thing is a lot more smoke and mirrors then actual intelligence. Basically, what am saying is that Google drives the road with the car in manual road a few times. Gathers data about how the car would drive. Then fixes any flaws in the cars plane for how to navigate the road. In theory they refine and create new algorithms to better handle the situations.
Try the google car in the chaotic bike filled streets of the Netherlands, show us that it's better than the average driver.... :P
Our Semi-Robot Cars With You As Backup When The Computer Can't Handle It Are Better Drivers Than You Alone
FTFY
They are not comparing with completely robotic systems as there always needs to be a qualified driver to take over when the computer can not deal with the situation. The robots we have are assisting drivers not replacing them.
I realize the article is two years old but it may still be relevant;
Two things seem particularly interesting about Google's approach. First, it relies on very detailed maps of the roads and terrain, something that Urmson said is essential to determine accurately where the car is. Using GPS-based techniques alone, he said, the location could be off by several meters.
Google map data has been inaccurate and/or out of date a significant amount of the time. I would not trust my life to a Google map.
The second thing is that, before sending the self-driving car on a road test, Google engineers drive along the route one or more times to gather data about the environment. When it's the autonomous vehicle's turn to drive itself, it compares the data it is acquiring to the previously recorded data, an approach that is useful to differentiate pedestrians from stationary objects like poles and mailboxes.
So you have a human drive the road, record the path and then the semi-autonomous car does it's best to follow that path. That's cheating. A scan like that is only valid for a few days at most and how much data is needed to store that kind of scan? An everyday vehicle would need a high speed connection to just keep up with the changes.
Standing tough under stars and stripes
We can tell
This dream's in sight
You've got to admit it
At this point in time that it's clear
The future looks bright
On that train all graphite and glitter
Undersea by rail
Ninety minutes from New York to Paris
Well by '76 we'll be A-O.K.
What a beautiful world this will be
What a glorious time to be free
What a beautiful world this will be
What a glorious time to be free
Get your ticket to that wheel in space
While there's time
The fix is in
You'll be a witness to that game of chance in the sky
You know we've got to win
Here at home we'll play in the city
Powered by the sun
Perfect weather for a streamlined world
There'll be spandex jackets one for everyone
What a beautiful world this will be
What a glorious time to be free
What a beautiful world this will be
What a glorious time to be free
On that train all graphite and glitter
Undersea by rail
Ninety minutes from New York to Paris
(more leisure for artists everywhere)
A just machine to make big decisions
Programmed by fellows with compassion and vision
We'll be clean when their work is done
We'll be eternally free yes and eternally young
What a glorious world this will be
What a glorious time to be free
Both are much better for a great many (perhaps a majority of) people when they are part of good PUBLIC TRANSPORTATION!
I'd like to see a self-driving car avoid a "near-collision" state on the Capital Beltway during busy times. Or the MD I-270 of old (they've changed it a lot since I commuted on it) -- I (along with everyone else on the road) used to spend entire commutes in a "near-collision" state, sub-second following distances at ~65mph alternated with panic stops.
That's a classic control theory problem. You want to maintain a following distance as the speed of the car ahead changes, but there's lag due to reaction time. So you can get oscillation. Solutions are known.
Smart cruise controls already do this better than humans. With radar systems you have good range rate information. Vision isn't good at range rate, especially when it's changing. With good range rate info, you can servo on speed difference and range. You can buy such systems now; that's what "smart cruise" systems do. Here's an Audi A7 doing it. 2014 Kia Those demos are in low-speed stop and go traffic. Here's a Porsche 991 doing 90 in a 55 zone on adaptive cruise control. Here's a Subaru Forester in stop and go freeway traffic in Los Angeles with smart cruise control, with speed varying from high speed down to zero and back.
Solved problem. Available now at your car dealer.
Self driving cars aren't meant for longer highway speed driving. They are meant for highly congested roadways which rarely get much over 45 mph anyway. I live in just such an area. The biggest traffic tie ups are always at the merge points. Why? People are terrible at merging, and everyone merges differently. Some follow the driving class' rules and drive to the end of the on ramp. Others find a hole and dive in. Others slowly sneak out. Some only recently started driving or moved into the area from a place that didn't have traffic congestion.
Shows like Top Gear will argue that self driving cars ruin the driving experience. Personally, I don't consider my morning commute a worthwhile experience. It's merely a less time consuming method than riding the bus. If I could let my car handle the daily commute while I grab a nap or read my RSS feeds, I'd leave the driving to Google or whoever. Leave the highway miles and scenic road ways to me. Leave the drudgery to the machines.
Only the dead have seen the end of War. - Plato
Lord have mercy...the most random stuff brings out the trolls
I'm talking about this: http://en.wikipedia.org/wiki/Two-second_rule
It's a common US heuristic taught to help keep a safe following distance
that's all...two second rule exists...
Thank you Dave Raggett
The Google cars that I've seen driving on Highway 101 here in Silicon Valley, are to me, prime examples of how not to drive. They are not safe to have around real drivers. Allowing these things on the road is/will be a mistake. All other cars will be forced off the road because the two cannot coexist safely, and big money is behind the self driving cars.
I'm not into gardening. I'm handy but I don't really care to improve my house any more. Driving's been a fun thing to do ever since I bought the sportscar and joined the car club. In a typical day of running errands, it's the driving between the errands that's the fun part of my day. It's the commute to and from clients that breaks up the work day.
If I stop driving, what's the fun part of my day? Does the google car come with a google hooker? Because there aren't too many recreational dreams a young man has. Driving's probably in the top three. Maybe the google car comes with unlimited google sandwiches. But even if the hooker serves the sandwiches, how many hours of pleasure can hookers and sandwiches provide in a single day? There's a bottle neck, or two, on my side.
And even better, *any* vehicle could become an emergency vehicle by broadcasting an emergency code and all the other vehicles would automatically get out of their way. This would presumably be logged and tracked, to minimize abuse of the system.
I live in the Canadian prairies. The nearest large cities are 5hrs, 7hrs, and 9hrs drive away. If I could sleep or read or watch movies during that time, it'd be *awesome*.
and their deductible is high enough that they'd end up replacing it out of pocket.
The solution is obvious:
Autonomously-driving cars would also have manual controls, activated by a bright purple Special Handle. Pulling the Special Handle enables manual high-speed driving mode and the driver, skills un-degraded by years of not driving, can rush the Special Emergency to the hospital.
Special Emergency Unicorn Mode activates the car-top rotating purple beacon, to alert everyone else (via their cars' Special Unicorn Detector) that there's a Special Emergency nearby.
Unicorn Mode also transmits geolocation and in-car audio/video (copyright waived) to DMV and the History Channel, deploys a trail of glowing purple sparkles behind the car for use by Actionhype News, and marks the driver's hand with permanent purple Tribute Ink.
After the Special Emergency, circumstances can be validated by DMV and hospital staff. Drivers reasonably acting to save a life win a guest appearance on NatGeo's True Unicorns, 10% off Tuesdays at Disneyland, and a $25 Applebees gift card.
Alternately, drivers who Unicorned their very special offspring to the hospital for a split a lip from squabbling over the iPad will get their car's Special Handle removed, their purple-dyed hand surgically attached to their head like that thing on a rooster, and a $25 Applebees gift card.
In theory, here's how it should work.
The highway patrol will have a lot less to do after the adoption of autonomous cars, so its overall expenses are lower. Where all citizens used to pay the patrol's operating costs through a combination of taxes and traffic fines, now we pay for their much lower operating costs through taxes only. The net savings to you and me should be substantial.
That that is is that that that that is not is not.
people would rather face a daily one-in-a-million chance of dying due to their own mistake than a daily one-in-a-billion chance of dying due to a machine failure.
Irrational people would rather face a daily one-in-a-million chance of dying due to their own mistake than a daily one-in-a-billion chance of dying due to a machine failure.
People who are like me would rather take the one-in-a-billion chance.
That that is is that that that that is not is not.
No, YOU got it completely wrong. You are unable to see the difference between micro- and macro-economics. I was talking macro. In micro you only see the individual or firm, in macro you see ALL (at ONCE, if you look at macro again looking at one individual at a time than you are NOT in macro view). You want low costs only in micro view, from the point of view of the individual. In macro costs = earnings (of the next entity in the chain).
It's possible to fly from london to new york in three hours. In fact, you could do for a span of twenty years or so, before concorde was decomissioned. Just because it's possible, doesn't mean it'll become widely available, or even at all available. Ditto driverless cars. The technology will filter down into ordinary cars as extra safety features to augment the driver's slow reactions. For the foreseeable future, anyway.
What I don't want--and what it will enable--is for Google to conveniently route traffic to support people who pay for extra business traffic.
Once we give up control it grants them the ability to do this; and given Google's behavior as of late, they will.
Well i will have to see this for myself i really have no doubt that i can outdrove a robot car
As it currently stands in the US, drivers are already distracted enough, already careless enough, already lack the skill and seriousness that is reasonable when controlling a ~ 4,000lb vehicle.
So lets "solve" this uniquely American problem of what amounts to accepted incompetency by further removing the appropriate level of responsibility that is really merely the bare minimum.
Because technology solves every problem.
Look, you don't own the road, you're sharing it. You don't have a right to drive, a right to be a danger to me, a right to waste my time, a right to BE in your care and BE distracted in ANY way.
Hang up, STFU, and just DRIVE.
I believe I would love the thought and would buy into it, however terminator comes to mind... hopefully the future is not a robot controlled life, at least not in mine or my kids future. Safety at this poi t would be nice... hell I would take a nap on the way to work, I have an hour drive!
Where there are windows... there are doors to get the F out.
Thats easy, no ego involved.
data data data Can someone give me a solid definition of what data is?
I live in Southern California where all of the idiots drive (I guess that includes me). I would LOVE an automated system so I could read in transit. I don't like the idea of Big Brother looking over my shoulder (can anyone say NSA).
We should put railroad wheels on cars (keeping the tires) and slowly build out railroad tracks along highways and maybe even in cities. There'd be a pretty small chance of accidents and you'd probably save a ton of money on gas. The best part would be that it would be like mass transportation with the benefit that you get to still drive regularly after you get off the railroad. You could even chain cars together and probably get even more efficient. As a bonus, it would create useful construction jobs to build out the railroads. And, as another bonus, because it doesn't require some massive investment like high speed rail, you could build it out little by little, adding more only if people like it.
Since autonomous cars would drive perfectly, there would be no more speeding ticket rackets in small town, USA for traveling 2 over the limit. We could theoretically starve these bastards out. Plus I can sleep on the trip into work and back home. No more traffic jams because some arse wiped out while playing with his phone or fiddling with his radio.
Sign me up!
FTFY
How about Google get some of those guys in to teach the system. Then we have the best of both worlds - advanced driving techniques with the fastest reaction times available.
Robots will also reliably use their turn signals to indicate a lane change.
I know it sounds inflammatory but imagine a situation where there is a cyclist pulling onto the road and in order to avoid hitting the cyclist, the only option the car has is to swerve out of lane and into on-coming traffic.
have we imagined that?
Cool. Now, algorithmically, the car needs to decide whether to hit the cyclist or to pull into on-coming traffic. The decision is simple - hitting the cyclist will likely result in serious injury or fatality where as you, inside your car are likely to suffer far less injury in a minor collision. The car intentionally injures you.
Don't get me wrong. When evaluating this kind of thing, the autonomous vehicle should absolutely take the path of least harm - but now tell me how you market a vehicle that is programmed to intentionally harm its owner in certain scenarios?
Google + NSA + autonomous cars: you will get fucked, certainly.
IMHO they meant to say:
Our robot cars are better than human cars, so long as nothing goes wrong and all the human cars are driven by drunks with road rage.
There, fixed it.
-- Tigger warning: This post may contain tiggers! --
If i built my own data logger and was involved in an accident, my data would show that I was the innocent party, too.
They whose government reduces their essential liberties for temporary security, receive neither liberty nor security.
"Self-driving car" is beginning to sound a lot like "horseless carriage". I would suggest "automobile" but that seems to be taken already...
I was right...it's two seconds...which is what I originally said, before I was trolled
trolled 2x now...http://en.wikipedia.org/wiki/Two-second_rule
I knew what I was talking about the whole time...my only mistake was being open to what other /.'ers had to say
Thank you Dave Raggett