Google's Self-Driving Cars: 300,000 Miles Logged, Not a Single Accident

An anonymous reader writes "The automated cars are slowly building a driving record that's better than that of your average American. From the article: 'Ever since Google began designing its self-driving cars, they've wanted to build cars that go beyond the capabilities of human-piloted vehicles, cars that are much, much safer. When Sebastian Thrun announced the project in 2010, he wrote, "According to the World Health Organization, more than 1.2 million lives are lost every year in road traffic accidents. We believe our technology has the potential to cut that number, perhaps by as much as half." New data indicate that Google's on the right path. Earlier this week the company announced that the self-driving cars have now logged some 300,000 miles and "there hasn't been a single accident under computer control." (The New York Times did note in a 2010 article that a self-driving car was rear-ended while stopped at a traffic light, so Google must not be counting the incidents that were the fault of flawed humans.)'"

Comments

Rear Ended

[sconeu]

The GoogleMobile was behaving properly, and was stopped. It had no possible way to evade the puny human that hit it.

However, after the accident, the GoogleMobile was heard asking another car, "Hey, hot mama, wanna kill all humans?"

Re:Rear Ended

[Anonymous Coward]

We don't know if the Google car could have avoided it. I was in a similar situation one time, and happened to catch a glance of the idiot in my rear view. I cut out ahead of the adjacent lane into the empty crosswalk. The idiot screeched to a stop in what was previously my lane.

There isn't always empty space; but if there is then the Google program should recognize it as available for evasive maneuvers. The Google car will not have a heart to go pitter-patter like mine did; nor a father who turned to me and said, "you're a good driver".

Re:Rear Ended

[norpy]

You are creating a straw man there, 99% of similar situations with human drivers would either have not noticed the exit or not reacted in time.
Additionally you likely broke the law doing what you did and if you caused an accident or ran over a pedestrian because of it you would have been 100% at fault, whereas being shunted by the guy behind you lands 100% of the resposibility on him (unless you stopped too close to a car in front of you).

I would put money on your driving record being way worse than 300k miles accident free. The actual pouplation-wide average is a LOT higher than that, and you are asking for us to give up reducing that number because we can't reduce it to 0.

That's like people saying "Don't build gas power plants to replace coal plants because they still emit CO2", sure it's not perfect but at least it BETTER.

Re:Rear Ended

[Beardydog]

I don't think he's saying we should give up on reducing that number. I think he's saying it may technically have been avoidable in that case. It's not meant to be dismissive. It's meant to be food for thought.

Re:Rear Ended

[Cryacin]

I'm more wondering what it would be like for the driver who actually rear ended a robotic vehicle

Uh, yeah, so I rear ended you. We should exchange insurance details.

I'm sorry, Dave, but I can't do that.

Re:Rear Ended

[Dunbal]

But if everyone went around in automated cars, the point is there would be no "idiot" in your rear view because he also would have been in an automated car; one which would have stopped in time.

Re:Rear Ended

[DemonGenius]

All automated vehicles would have to have some sort of human override on them. Considering that fact, it would be near impossible to rid the road of all idiots, therefore, it would make sense for the automated car to have some sort of collision avoidance algorithm for rear enders that also doesn't put other drivers or pedestrians at risk. Especially since an automated car has several times more visibility than the average human and wouldn't have to react to something in the corner of its "eye".

Re:Rear Ended

[Jeremi]

But if everyone went around in automated cars, the point is there would be no "idiot" in your rear view because he also would have been in an automated car; one which would have stopped in time.

Nonsense. Nature will by then have bred a far superior idiot, who will have downloaded some buggy "mods" into his car's autodriving software, causing it to crash (in one or more senses of the word).

Re:Rear Ended

[Chrisq]

I'm more wondering what it would be like for the driver who actually rear ended a robotic vehicle

I imagine that you exchange details with the human in charge, with the full knowledge that there will be a complete 360 degree video of the accident with measurements of speed of both vehicles.

Re:Rear Ended

[Anonymous Coward]

Considering that fact, it would be near impossible to rid the road of all idiots, therefore, it would make sense for the automated car to have some sort of collision avoidance algorithm for rear enders that also doesn't put other drivers or pedestrians at risk.

With 300,000 miles without "a single accident under computer control." it has a collision avoidance algorithm that is far better than yours. Therefore it would make sense that you wouldn't be allowed to manually override a far superior driver.

Re:Rear Ended

[neyla]

It's way better than *average* but in no way better than -all-, there's plenty of people who've driven for decades and never had any accident whatsoever.

If you don't count accidents where someone else is 100% to blame, I've driven for 18 years and never had an accident of any kind. This is rare, but not unheard-of.

Re:Rear Ended

[LordLimecat]

my lane.

Think I found your problem.

While there are times that someone can make it more difficult to avoid rear ending them, in general you should be aware of your surroundings on the road including who is next to you and if they appear to be preparing to change lanes. Short of someone pulling a "Cops" and aiming for your bumper, you should be able to prevent almost any rear-end collision by simply being aware.

Re:Rear Ended

[Havenwar]

We had a neighbour when I was little who was obscenely proud of the fact that he had been driving for something like 70-80 years without ever being in a car accident. Even then, old, with deteriorating everything (but a sharp mind) he never had an accident!

Incidentally he had also been living on the family farm all those years, and aside from driving the tractor around the farm (and occasionally breaking shit with it, like one time he misjudged his angle going into the barn and tore half a wall down with his rear tire) he actually only drove into town once a month for supplies. A drive of about 15 minutes on a road where you met another car maybe once every five times you drove it, to get to a town where livestock had the right of way and everybody just kind of crawled around in their vehicles around whatever obstacles might appear, be they sheep, pedestrians, or a ninety year old half blind man driving on the wrong side of the road.

I'm just saying, sometimes good drivers have accidents, and bad drivers avoid them, because of whatever outside reasons govern their reality... rear-ended by an idiot, or spending their entire life driving in a very very safe environment. With that being said of course you are right that there are some good drivers who never cause an accident, and of course these are better drivers than an automated car... And there are drivers so bad that they pull the average right back down again.

In short, your argument is invalid - if we replace all cars with robots that have a better than average record, then the average would rise, even if we never let a good driver touch the controls again. And if ALL cars are automated, they can be patched, they can be linked, and they are over all predictable - which is the main risk in traffic today. Unpredictability. Almost every accident happens because of one of two things - 1) something unpredictable happens. 2) the driver failed to predict something obvious due to ignorance/distraction/narcissism/slashdottism/whatever.

Both of those aspects can be near eliminated by letting machines do the driving.

Re:Rear Ended

[Havenwar]

I used to agree with that, then I realized that since we regularly inspect cars and since it would be easy enough to scan the software of every car that is involved in an accident, this is a crime that wouldn't go unnoticed. I like modding, but I'd probably be cautious if I knew that in an accident situation my crime could be upgraded from "manual operation of vehicle outside of my abilities" or whatever to "reckless endangerment" or in cases where someone was actually killed, it could easily be upgraded from "manslaughter" to "murder", or whatever the terms would be in the appropriate jurisdiction.

Sure, some people would still tamper... and the cops would likely have scanners scanning your software on routine stops, maybe even getting feedback from it remotely.

But you know what'll really kill it? Predictability.

If your car is driven by a program that will under no circumstances go above the speed limit, then if you go above the speed limit it's reasonable suspicion that you've tampered with your program. The car is impounded, and if it's found to be tampered with, lost.

Since the kind of people who really love to mod their cars are also the kind of people who really love their cars, this kind of thing is a fairly easily enforceable and not too harsh punishment that would likely make them think twice. And their hobby isn't outlawed, no, just moved to fenced trackways. If they want to race, feel the speed, mod their cars... then that's not for the streets, that's for the raceways.

Will there still be the one in a million idiot that does something insanely stupid, gets someone killed, and then drools in the courtroom saying he didn't know or that the rules somehow magically don't apply to them? Yes. There will also still be lightning strikes. If we got fenderbenders and carcrashes down to the scarcity of lightning strikes however, or lower... man, imagine the billions of dollars and manhours saved... not even mentioning the deaths and injuries avoided.

Re:Rear Ended

[Niedi]

Anyone wants to bet how long it'll take for "Pay 5000$ now or you'll hit that nice tree over there" style rootkits to appear? Scareware indeed...

Re:Rear Ended

[somersault]

Better doesn't imply that, it just means.. better than whatever state you were in before. When people are sick, we often say "I hope you feel better".

Re:Rear Ended

[somersault]

of course these are better drivers than an automated car

I don't think there's really any justification to say that. The automated car can be attentive 100% of the time, whereas humans make mistakes. There probably are bugs in the software right now, but maybe none bad enough to cause an accident. Until the Google car causes an accident, you can't really judge it less safe than any human driver.

Re:Rear Ended

[Havenwar]

Well, the issue here lies in the unexpected. The unexpected does happen, quite a lot, and unless the cars AI can handle this better than a human driver it is not a better driver. I've driven more than a few laps around the planet so far in my life, and I can tell you that unexpected situations are really common. Proving your other point however, the two near accidents I've been is was once when I was young and dumb and showing off, and once when I was really tired and zoned out for a moment. I recovered in time, but of course an automated car wouldn't have had these problems.

On the other hand an automated car would have likely crashed in a few hundred other situations I've encountered, that after having reacted and averted the danger have left me going "huh, what the FUCK just happened?"

So I believe it's possible to give the fair assessment that a good driver is a better driver than an automated one, simply because that's the nature of driving. Unexpected situations. Once we get to the point where all roadway traffic is automated and controlled in unison, then the unexpected situations will be near null, and thus automation will win... But as they taught us in basic physics... "This is only true in a perfect vacuum, without any outside interference."

Re:Rear Ended

[somersault]

I used to think like that, but then after considering the actual problem, I realised it's perhaps not as complex as we think. Basically to be a "safe" driver all you need to do is avoid bashing into anything, or causing anything to bash into you. Now if the thing went over an oil slick or something then I can see that causing difficulty for the AI, but in other situations what with its cameras and radar and whatnot, I think the AI is better placed to tell when someone or something looks to be getting in its way.

But really, most accidents wouldn't fall into the "unexpected" category for someone who has researched into the most common causes of accidents, and generally doesn't trust other drivers to do the right thing. A good driver is always ready for that idiot to pull out without looking, makes sure that they haven't got their turn signal on by mistake, slows down adequately for blind corners (especially when there are buildings nearby, or a known junction on the bend for example). If Google have done their job right, the car probably expects a lot more accidents than any human, and is ready for them.

Re:Rear Ended

[NoZart]

"man, imagine the billions of dollars and manhours saved..."

aggressive lobbying against automated cars activated by mechanics, car makers, hospitals, oil magnates

Re:Rear Ended

[AmiMoJo]

What about cars with no passengers though? Say there is no parking near your work so you send the car home, then tell it to come get you at 5PM. On the way it has an accident. There will have to be some kind of system in place for notifying the owner and allowing the other person involved in the collision to speak with them. Even if it is just a notice placed somewhere on the car it will have to be standardized.

Re:Rear Ended

[Havenwar]

Okay, unexpected things that doesn't include hitting or being hit by things... Hmm.

At one point I was driving down the road, when the road disappeared.

Sure, it didn't literally disappear, it just so happened that it re-located about 30 feet downwards in an instant due to an unfortunate passionate meeting between a boat and a support pillar holding the bridge I was about to cross up. I guess they really liked each other, or whatever, but all the same I was definitely surprised, there was nothing for me to hit but rock bottom and by the time any computer would have seen that it would have been too late, and nothing was about to hit me.

It's possible their car could handle this situation, but that depends largely on how perfectly they can detect every inch of road. From what I've heard they go mostly by road markings and such, which means that in such a situation they'd have to switch to manual control. Hopefully in time to avoid a nosedive into the amorous steel behemoths below.

Children running into the road is childs play - a clear collision avoidance. How about children jumping over you? There I was driving down a narrow "alleyway" between two walls/brick fences that were just barely high enough that I couldn't see over them, and all of a sudden a bunch of teens on bikes fly over the road over and infront of me. I slowed down, and that was pretty lucky because the last guy didn't make it. Granted that's collision avoiding, but would google's car have been slowing down for things flying above it? Can it tell that this is kids on bikes that might crash, or a ball that might be followed (if it's above you and thus not in your direct path), rather than a dove or flying squirrel with a death wish and a meth addiction?

I could go on, but the point I'm trying to make is simple - being a good driver is not just about not bashing into things and avoiding things bashing into you, it's about avoiding putting yourself in a position where you can't stop. It's about avoiding the scene of the potential accident completely, not about behaving competently once you are already there.

Re:Rear Ended

[Havenwar]

Oh that's okay, they'll just stop selling you the parts and start leasing you the rights instead. Sign the EULA, and if you don't pay your fees every time you use your brakes they'll re-possess those brake-lines they fixed.

Hey, it worked for the film industry, the music industry, and so on.

Re:Rear Ended

[Chrisq]

What about cars with no passengers though?

This is currently not legal and I doubt if it will be legal for some time. I imagine that by the time it is legal accidents will be dealt with like product liability. If a blender explodes and pieces cut you - then you have a claim. If you put your hand in it then you probably don't. Of course I am sure that cases will be much more complex, should your car have been reasonably expected to avoid the wheel that fell off the other car? Is it shared liability or all the other car's fault?

Re:Rear Ended

[neyla]

There probably exist situations where creative problem-solving is beneficial, and you're right that humans are better at that for the immediate future.

But this is largely negated by the short reaction-times needed. These are typically split-second decisions, and human beings aren't universally good at being creative and solving complex problems in a few hundred milliseconds.

And often, a excellent solution, where the execution starts after one second, is inferior to a mediocre solution where execution starts in 50ms.

And what matters, is the average performance. If the automated car has a better reaction than that of a good driver in 95% of the dangerous situations, and a worse reaction in 5% of them, then it still comes out ahead, despite the existence of the 5%.

Also, computers and algorithms improve with time -- often rapidly so, whereas human drivers aren't likely to be all that much better in a decade.

Re:Rear Ended

[neyla]

Essentially all dangerous situations are unexpected: if you where *expecting* that (for example) a child might emerge running from behind that corner, you can act so as to make that a non-dangerous thing.

But a huge fraction of traffic-deaths fall into a *tiny* collection of situations, more than two thirds of those who die in Norwegian traffic either leaves their lane and crashes into oncoming traffic, or leave the road alltogether and crash into a hard object. Merely avoiding those accidents, would save 2/3rds of the deaths.

Re:Rear Ended

[Havenwar]

Excellent points, which I've been making all over this discussion over and over again. However in this particular thread we were discussing the very top percent of drivers, the ones that are in fact really good drivers. That includes having a quick reaction time and the experience and skill to choose the best course of action in an unexpected situation.

I was never saying all humans were better than automated cars, in fact I've spent some time arguing the opposite - if all cars were automated then the average driving skill is increased, simply because machines are better at those 95% of use cases.

The fact that a good human driver would be better than an automated one in 100% of the fringe cases is as you said, irrelevant - but since it's brought up often by the opposition to automated transport it has to be acknowledged and admitted.

The real problem with that argument is that study that showed that 80% of drivers considered themselves an above average driver, or whatever the exact numbers were. Most people think they are much better drivers than they are. The vast majority of mediocre drivers will indeed think that they are in that small percent that should be allowed to disable the automatic control because they'd do a better job at it. That's the real danger to automated vehicles in the future.

Re:Rear Ended

[Havenwar]

I disagree, there are plenty of situations which are both expected and dangerous. Like for instance getting into a hunk of metal and speeding down the freeway at over 100kph. Making a left turn on a country road is dangerous. Passing another car is dangerous. Doing anything but crawling around a corner when you don't know what's on the other side is dangerous... Of course people assume the road continues, and most of the time it does. It's that one time in a million that there's a tree over it, or a gay pride parade crossing it, or a kamikaze bumblebee does an elaborate star wars recreation with a wasp and two birds on the hood of your car...

You say two thirds of those who die in Norwegian traffic do so because they stray from their lane into oncoming traffic or off the road. I'm sure numbers are similiar here in Sweden. That proves my point. These are expected dangers, we know these happens all the time, we can be prepared for them... but they still get to us, they still kill some of us. Your argument is that if googe's car stops these 2/3ds of deaths, then that's a good thing. I totally agree.

As long as it doesn't fail to react to those unexpected situations, and causing 2/3ds more deaths there instead.

Re:Rear Ended

[Yvanhoe]

If automated cars can be "driven" by people without the permit, I think most people would end up not taking the class and exam to have the right to override the automatism.

Re:Rear Ended

[somersault]

I seriously had never thought of this angle at all before. Of course you're right, and people are dicks, so it is a big concern. I'm hoping that there will be a hardware switch required to be set before you can modify the software, rather than over the air updates..

Re:Rear Ended

[Dunbal]

there's plenty of people who've driven for decades and never had any accident whatsoever.

Yeah, and there's plenty of people who can't seem to leave the drive-way without having an accident. The point is it's impossible to tell who belongs into which group without actually giving them a car and letting them drive for a while. When considering the individual, of course as much individual freedom as you can get is what your are going to shoot for. When considering public policy, however, your actual individual wants and needs fall second to the needs of the population as a whole, and the only way to reliably assess those needs is through (appropriate) statistics.

Re:Rear Ended

[realityimpaired]

The real problem with that argument is that study that showed that 80% of drivers considered themselves an above average driver, or whatever the exact numbers were. Most people think they are much better drivers than they are.

*shrugs* depends on your definitions. I'm probably more lucky than I am a good driver, but the last time I had an "at fault" accident, I was 17, and the visibility was down to about 100 feet due to a heavy rainstorm... I hit somebody in front of me who was turning without signalling, but because I hit her, I was at fault in the eyes of the law. Even the cop writing the ticket told me, at the time, I should take it to court because I had a good argument that it wasn't actually at fault. I have had one other accident, where somebody hit a pothole on an icy road and spun out in front of me, but that's one of those completely unavoidable situations. Plenty of other avoidable situations that I managed to avoid, but that one I had no choice, the road was too slick and I had about 15 feet to come to a complete stop in order to avoid hitting him (and no, I wasn't speeding at the time, actually I was going 10 under the limit).

So going purely on accident record, yes, I'm a "good" driver. I have had a total of one ticket since I was 17 as well, and it was a parking ticket (stayed too long), so by that measure I'm a good driver too. But how do you define being a "good" driver, really? I don't pay any attention to the speed limit at all, and rely on cruise control for long distance hauls to keep me from speeding too much... I *have* run up against the electronic limiter in my car before, which kicks in at 185km/h. I never drive at a speed that's inappropriate for the conditions, but very often that has nothing to do with the posted speed limit, being either higher or lower than the sign says. Does completely ignoring one of the biggest rules of the road make me a good driver, or just lucky?

Re:Rear Ended

[Havenwar]

I think you're missing the point, it's not about objective skill, but rather the fact that 80% think they are better than average. In reality 50% is better than average, 50% is worse than average. That means 30% of people at the very least, overestimate their skills. It's called illusory superiority, and you can check it out on wikipedia, it's a basic cognitivie bias. It even quoted the studies I referred to.

"For driving skill, 93% of the US sample and 69% of the Swedish sample put themselves in the top 50% (above the median). For safety, 88% of the US group and 77% of the Swedish sample put themselves in the top 50%"
-- http://en.wikipedia.org/wiki/Illusory_superiority#Driving_ability

You can also read the article on the Dunning-Kruger effect.
"The Dunning–Kruger effect is a cognitive bias in which unskilled individuals suffer from illusory superiority, mistakenly rating their ability much higher than average. This bias is attributed to a metacognitive inability of the unskilled to recognize their mistakes."
-- http://en.wikipedia.org/wiki/Dunning%E2%80%93Kruger_effect

My definition of a good driver is one that doesn't get into accidents, and that can handle an unexpected situation quickly and correctly. Whether they obey traffic laws is more about whether they are a good citizen or not, and that's a different argument, however staying within the traffic laws (i.e. lowering speed around schools) prevent accidents, and is as such a trait of a good driver.

Whether you are indeed a good driver is irrelevant to the fact that a lot of people judge themselves as to be better than they are. It's after all not the individual that matters when we're talking about the group as a whole.

Re:Rear Ended

[K. S. Kyosuke]

And if ALL cars are automated, they can be patched, they can be linked, and they are over all predictable - which is the main risk in traffic today.

If all cars were automated and linked, you wouldn't need traffic lights and traffic signs on road intersections and the perpendicular lines of traffic could cross each other at full speed without ever slowing down... Wow, that could easily pass for an adrenaline sport.

Re:Rear Ended

[Havenwar]

Yeah. Except if it's that automated we'll probably be sitting backwards since that's safer from a crash perspective, and the windows will likely be tinted or blacked out as to not give people motion sickness.

Re:Rear Ended

[beelsebob]

I used to think like that, but then after considering the actual problem, I realised it's perhaps not as complex as we think. Basically to be a "safe" driver all you need to do is avoid bashing into anything, or causing anything to bash into you. Now if the thing went over an oil slick or something then I can see that causing difficulty for the AI, but in other situations what with its cameras and radar and whatnot, I think the AI is better placed to tell when someone or something looks to be getting in its way.

I think the AI is better placed to deal with the oil slick too tbh, if it's sanely coded, it will be coded to measure exactly how much wheel slip it's getting at any one point, and to counter steer and slow gently as appropriate to the grip levels available. Compare this with humans where a good 50% panic and do exactly the opposite to what they should do (either they steer more to make the turn and spin it, or jam the breaks on and lose all traction), and the other 50% will do it imperfectly.

Add to that that the AI has a good chance of being coded to read the light readings it gets back from the road and recognise the oil before it even hits it and you're likely to have the AI do a far better job than any human.

Re:Rear Ended

[beelsebob]

I think you're missing the point, it's not about objective skill, but rather the fact that 80% think they are better than average. In reality 50% is better than average, 50% is worse than average.

Actually, it's entirely possible for 80% to be better than average drivers, if they're all only slightly better than average, and the 20% are really terrible. That said... I expect that it's more likely that 90% are slightly worse than average, 5% are significantly worse than average, and 5% are exceptionally excellent drivers.

Re:Rear Ended

[beelsebob]

Uhhh, how would tinted/black windows help motion sickness? The problem with motion sickness is that your ears tell your brain you're moving, while your eyes see the inside of the car and say it isn't moving, making your brain go "uhh, sensors are fucked, time to core dump". Tinting the windows would only make it impossible to see the truth that you are in fact moving, and make the situation worse.

Re:Rear Ended

[somersault]

Well, while the Dunning-Kruger effect is real, maybe those 80% are having a lot less accidents than the other 20%. People don't often get pointed out for being accident free, but accidents are posted all over the news and newspapers, so no wonder people might consider themselves good drivers in this case.

Myself I've been told by a Police driving instructor while doing an advanced/defensive driving course that I was better than most police drivers (as far as the first week of their training goes - we obviously didn't go to the breaking the speed limit stage), so I can safely say it without worrying I'm misjudging ;) I did have accidents where I was at fault when I was younger, but I like to think I learned my lesson, and doing the advanced driving course definitely improved my observational skills and general attitude to driving.

Re:Rear Ended

[Havenwar]

True, thanks for calling me on that. I have a tendency to misuse the word average. Bad me, bad bad. Of course I meant and should have said "in the top 50% of drivers", as it is phrased in the quote right after the paragraph you commented on.

"For driving skill, 93% of the US sample and 69% of the Swedish sample put themselves in the top 50% (above the median). For safety, 88% of the US group and 77% of the Swedish sample put themselves in the top 50%"

Re:Rear Ended

[Jason Levine]

I would think that cars would still require maintenance, so mechanics wouldn't be out of a job. Even if you never got in an accident in your life, you still need to change your oil, rotate your tires, check that the various systems are working properly, etc. Some people do that themselves (at least the first two), but for most people this would mean going to a mechanic. And if there's a car software upgrade or a malfunctioning sensor that needs to be replaced, that would be a job for the mechanic too. (I'm guessing that they wouldn't make it easy for drivers to modify their own car software. Otherwise you are bound to get people who think they can do it but wind up bricking their cars... as the car is doing 70mph on the highway.)

Car makers would stay in business how they do today: Selling new model cars to people who could have otherwise squeezed another decade out of their old car. Oil magnates wouldn't care if the automated cars run on gas and hospitals would rather see less auto accidents coming in. They certainly won't be lobbying to keep accident rates high.

Re:Rear Ended

[TheRaven64]

This is a concern even with 'manual' cars. There have been a couple of instances over the past few years of remotely exploitable vulnerabilities in engine management systems, which would allow an attacker to make you accelerate or stop at very inconvenient moments, although probably not override the steering.

Re:Rear Ended

[HaZardman27]

I think quality is like temperature. You can measure heat but you can't measure cold (it's just low amounts of heat). Bad is just low amounts of quality, so saying "better" is like saying "warmer."

Re:Rear Ended

[kilfarsnar]

It's way better than *average* but in no way better than -all-, there's plenty of people who've driven for decades and never had any accident whatsoever.

If you don't count accidents where someone else is 100% to blame, I've driven for 18 years and never had an accident of any kind. This is rare, but not unheard-of.

Indeed, I have been driving for 24 years and have never been at fault in an accident. I may not have logged 300,000 miles, but I'm probably over 200,000. And personally, they will have to pry the steering wheel from my cold dead hands. I know that for many, driving is a chore or at most a means to an end. I really enjoy driving and don't want a computer to do it for me.

Re:Rear Ended

[bluefoxlucid]

Yeah but I know how to modify data cores so mine broadcasts false information.

Re:Rear Ended

[nolnacs]

In that study, they are calling average the median, but in my experience, people rarely think of average as the median. Instead they think of the mean and it is quite possible for 80% of people to be above the mean. Take the set 6, 6, 6, 6, 6, 6, 6, 6, 1, 1. The mean is 5 and 8 of the 10 people are above the mean.

I also think that people's perceptions of the average driver are perhaps excessively derived from the most memorable drivers - i.e. the terrible ones. I certainly don't remember the 100s of cars that I drove by this morning that stayed in the lane without any trouble.

Re:Rear Ended

[drinkypoo]

Well, the issue here lies in the unexpected. The unexpected does happen, quite a lot, and unless the cars AI can handle this better than a human driver it is not a better driver.

The AI will handle it better than 90% or more of human drivers, who are usually dipshits who can't even keep their lane. All cars now have to have yaw control so the AI can ask for all kinds of great stuff without losing control.

Once we get to the point where all roadway traffic is automated and controlled in unison, then the unexpected situations will be near null

You forgot "inspected" because without regular safety inspections there will still be plenty of equipment failures. Some places have safety inspections, most don't.

Re:Rear Ended

[wisdom_brewing]

Never say never (or always, every, etc...)

It's comments like that that can drive ignorance and mass retardation.

CO2 emissions are NOT bad. Upsetting the carbon cycle is bad. Almost all living (for a given level of living) things emit CO2.

Re:Rear Ended

[Havenwar]

You apparently missed the part where the comments in question - this whole part of the thread actually - is specifically referring to the good drivers, the top few percent if you so wish. To summarize the progression for you:

"It's way better than *average* but in no way better than -all-, there's plenty of people who've driven for decades and never had any accident whatsoever."

"[...] of course you are right that there are some good drivers who never cause an accident, and of course these are better drivers than an automated car."

"I don't think there's really any justification to say that."

"Well, the issue here lies in the unexpected. The unexpected does happen, quite a lot, and unless the cars AI can handle this better than a human driver it is not a better driver."

Does that clear things up for you?

Re:Rear Ended

[drinkypoo]

"It's way better than *average* but in no way better than -all-, there's plenty of people who've driven for decades and never had any accident whatsoever."

Never had an accident, irrelevant. Lots of reasons why that could happen that don't really mean they're a "good" driver. Some people go through life without ever being tested as sorely as some other people.

"[...] of course you are right that there are some good drivers who never cause an accident, and of course these are better drivers than an automated car."

Citation needed. There's no reason to believe that those people are going to be better drivers than an automated car. Let's have some empirical testing involving throwing some trying real-world situations at both before we make that decision.

"Well, the issue here lies in the unexpected. The unexpected does happen, quite a lot, and unless the cars AI can handle this better than a human driver it is not a better driver."

That is, indeed, the point. The car can (potentially) immediately diagnose a number of problems with itself as well as always having a good idea of the road characteristics in the present and the future that could well eclipse all but the tiniest sliver of the driving population.

Does that clear things up for you?

Yes, I can now see the hands waving furiously in the air through the clouds of anal smoke.

Re:Rear Ended

[Havenwar]

You already ceded my point, how nice of you. "that could well eclipse all but the tiniest sliver of the driving population."

Yes. That was, as I pointed out, the sliver we were talking about here.

Did you intentionally misunderstand that the quotes listed were part of a conversation, or are you really that eager to argue that you don't even care? The crudeness at the end would imply the latter, and to that I just say... have fun. As shown above you agree with what I was saying, so I have no need to stay around. Go trolling if you are a glutton for verbal punishment.

Re:Rear Ended

[artemis67]

"Evasive maneuvers, Mr. Google!"

"Aye aye, captain!"

Re:Rear Ended

[crakbone]

Likely when the automated car goes without a human handler it will report the accident and all details to the police dept and its insurance company and share it via bluetooth to your automated car or your phone.

Re:Rear Ended

[Xeno man]

What exactly are you basing that on? I'm so awesome that even a computer couldn't avoid that? Computers are slow and untrustworthy so humans must be better?

I think your suffering from from the common assumption that most people have that your a better driver then you are. I'm not calling you a bad driver but I don't think your as good as you think you are.

The only fair assessment is more miles and more Google cars on the roads. The better the track record, the more confident we can be in Googles cars.

Re:Rear Ended

[artemis67]

Think of the implications of having an automated driving system... the onboard computer is collecting and analyzing data in real time, and it will likely store that information, at least temporarily. So if a Google car is involved in a crash, a full report will be generated, detailing exactly what happened and liability will be very easy to determine in most cases. "Car A has had a faulty motion sensor on the front bumper that the driver failed to have replaced" or "Car B drove through a red light to hit Car A".

I also think that automated cars will observe all safety rules to the letter... like only driving the speed limit (or slightly below), always maintaining a safe distance behind other vehicles, stopping for yellow lights, and having a generous braking distance. Remember, Google could be held liable if the system is reckless, and they aren't going to want that when human lives are at stake.

I think that making automated, passenger-less cars legal will be a very easy decision for legislatures, and will pass quickly. Like I said, I believe automated cars will err far more on the side of caution, like the most grandmotherly of drivers.

Re:Rear Ended

[artemis67]

That's why the "black box" on automated cars will probably be heavily encrypted.

Re:Rear Ended

[crakbone]

Sounds all great unless the conditions change to the point the software cannot handle the situation. Then you would need manual override. Say your roads out and you want to take a dirt road to your destination. Will the googlecar know how best to traverse a partially washed out road? Or that the small tree is easy to drive over in your truck and will shave 2 hours off your time doing so? And what about two feet of snow? or that the road crew has dumped about all the snow in your culdesac in your driveway and you just need to drive over it? Or that you can move through that 2 feet of water to move to high ground during your flood. There are a ton of variables out there that I cannot see programmed in. You will need manual control.

Re:Rear Ended

[artemis67]

The automated car can be attentive 100% of the time, whereas humans make mistakes. There probably are bugs in the software right now, but maybe none bad enough to cause an accident. Until the Google car causes an accident, you can't really judge it less safe than any human driver.

I'm also going to guess that automated cars are going to drive like grandma, observing every safety rule you ever heard about driving.

Most drivers take risks; driving 10 MPH over the speed limit, not braking enough in advance, not leaving enough distance between the car in front, running yellow lights. I imagine that automated cars offer a very boring but safe driving experience.

Re:Rear Ended

[MightyYar]

All automated vehicles would have to have some sort of human override on them.

No. Do you have the ability to override the pilots on an aircraft? Can you take the wheel from a bus driver? At some point you have to accept that the car is the driver.

Unless you mean an e-stop or emergency button - but that would presumably just stop the vehicle.

Re:Rear Ended

[Havenwar]

No I'm basing it on a very thorough understanding of how computers work - within a set of parameters. Computers can not handle the unexpected unless they are programmed for that eventuality, with some exception here for the rising field of machine learning and AI, which as of yet is not at the level of an average human being when it comes to assessing information outside of their pre-defined parameters. It's getting there, sure, but it's not there yet.

And as I informed the other guy below/above, this thread of the conversation was in reference to the computers performance versus the absolute best drivers - a class I'm definitely not included in. Of course had you bothered to actually read the conversation before jumping in you'd have seen I've already discussed that common assumption at length.
( http://slashdot.org/comments.pl?sid=3038597&cid=40944209 )

I think you are totally right that the only fair assessment is more miles, although of course if we truly want a fair assessment then those miles would be without human intervention - which would lead google's car to quickly fail. To date they specifically switch to manual mode in any troubling situations, because they are full aware (and do not hide) that the technology is not quite there yet for unexpected situations. Or to put it differently - they also agree with me.

Re:Rear Ended

[artemis67]

There's also the issue of cost.

Everyone will want an automated car. Hell, who wouldn't want to take a nap during their work commute, or watch TV or read a book?

New technology is always expensive at first. To reduce cost, I can see car manufacturers offering a budget car that has *no* manual controls. I think that first world countries will adopt automated driving very quickly, and that by the turn of the century, manually operated cars won't exist on public roads.

Re:Rear Ended

[MightyYar]

Well, the issue here lies in the unexpected.

You mean like heart attacks, strokes, seizures, or just falling asleep? :)

If the thing maintains a safe distance between it and other cars (what we currently call defensive driving), I don't see how the unexpected couldn't be handled.

We also have to look at cost-benefit. Sure, you might be at a slightly higher risk of driving off the edge of a collapsed bridge. In the last 10 years, we've had one major bridge failure that killed 8 (IIRC) and injured 100. In contrast, 100 people die in car crashes every day - most due to human error. I am willing to "kill" a few hundred people in robot cars due to poorly handled rare events if it saves tens of thousands due to human error.

Re:Rear Ended

[Chrisq]

Think of the implications of having an automated driving system... the onboard computer is collecting and analyzing data in real time, and it will likely store that information, at least temporarily. So if a Google car is involved in a crash, a full report will be generated, detailing exactly what happened and liability will be very easy to determine in most cases. "Car A has had a faulty motion sensor on the front bumper that the driver failed to have replaced" or "Car B drove through a red light to hit Car A".

I also think that automated cars will observe all safety rules to the letter... like only driving the speed limit (or slightly below), always maintaining a safe distance behind other vehicles, stopping for yellow lights, and having a generous braking distance. Remember, Google could be held liable if the system is reckless, and they aren't going to want that when human lives are at stake.

I think that making automated, passenger-less cars legal will be a very easy decision for legislatures, and will pass quickly. Like I said, I believe automated cars will err far more on the side of caution, like the most grandmotherly of drivers.

I'm not so sure. For driving on the highway I think it will be fine, but think about some other conditions. How will it deal with passing a horse and rider on a narrow road? How will it deal with coming across another driver in a country lane at a place where there is no room to pass? How will it drive on a road with a shear drop off on one side and a cliff on another? If this is narrow with passing paces will it know to pull in so that it almost touches the cliff to give extra space to the vehicle near the drop-off? In busy commuter traffic will it adjust the "aggressiveness" of pulling out from a side-road to take into account that if you don't pull out quick and accelerate hard you could be waiting until the end of the rush? On a rutted farm track can it work out that you have to drive with one wheel on the centre of the road and the other on the edge to avoid the tractor ruts?

If you have a driver there is always the option to safely pull over or stop and say "manual intervention required", but once you allow completely automatic use with non-drivers or no driver the car has to do something sensible.

Re:Rear Ended

[somersault]

And that's a great thing! When I'm driving, I always feel a weird pressure to be as efficient as I can, to get there as fast as possible. When I'm a passenger, I really am in no rush. An ebook reader and a good stereo are all you need to make you not care that you're being driven around by robograndma.

Re:Rear Ended

[mooingyak]

Almost every accident happens because of one of two things - 1) something unpredictable happens.

My driving rule #1, above all others (and I think it's good enough that everyone should follow it) is:

"Don't surprise the other drivers."

Re:Rear Ended

[Khashishi]

Good point. We have all seen idiots on the road who are much worse than normal. It's kind of hard to notice drivers who are much better than normal. I'm sure it's a very skewed distribution.
Maybe the survey question should be reworded, "Are you more skilled than the median driver?"

Re:Rear Ended

[Havenwar]

I agree with you to some degree, but I don't agree with how many events of unexpected things happen, and the types. Defensive driving is a really good start, and it'll keep you reasonably safe from the exact sort of thing that a computer would be good at handling. However there are more things than cars on the road, and more unexpected things than collapsed bridges that can happen. At one point I had to stop for what I thought was a traffic accident, turns out a hot air balloon had landed in the middle of the road. Does google's car look up? If it does, could it interpret what it saw enough to predict what was going to happen?

A lot of things are just collision detection, granted, but then a lot of things are predicting areas where collisions might happen. Google's car won't know that your neighbour has five kids and no sense to teach them to not run blindly into the road, for instance. Google's car might spot a ball crossing the road and slow down, but will it spot a kite? A paper airplane? Soap bubbles?

I love the idea of a computerized car, and certainly think that they are smart enough today to handle freeway driving. Hell, the cars we can build today would easily handle all driving if we only changed out all cars at once and rebuilt all streets to keep pedestrians off them or to assigned areas, but of course that won't happen... In my estimation, the unexpected situations as the situation lays today, would mean that a computer would stop people falling asleep causing accidents, but instead cause them in a lot of other cases. Now this is completely just my opinion based on nothing but my experience on the road as a driver, as a pedestrian, and my experience with tech and its benefits and limitations... So if you don't agree then we'll just have to agree to disagree.

Re:Rear Ended

[Havenwar]

Huh. My top two driving rules are

1) Everyone on the road is an idiot.
2) That includes you, sometimes.

I find that mindset has made me slow down and look twice more often, and it's kept me out of a few accidents.

Re:Rear Ended

[DocSavage64109]

I was in a similar situation one time, and happened to catch a glance of the idiot in my rear view. I cut out ahead of the adjacent lane into the empty crosswalk. The idiot screeched to a stop in what was previously my lane.

So you where stopped at an intersection behind a car, noticed a car coming up on you at a rate of speed to fast to stop without hitting you, and you had time and space to accelerate and move into another lane? I don't see how that's possible, and how would you know the fast driver behind wasn't about to change lanes and you just moved into his way?

Re:Rear Ended

[Yvanhoe]

I can understand that. Car accidents killed far more Americans that terrorism did in the last decade. This should be a top national issue.

Re:Rear Ended

[tehcyder]

Additionally you likely broke the law doing what you did and if you caused an accident or ran over a pedestrian because of it you would have been 100% at fault

It is never breaking the law to act in self defence, and even if it was, then your own survival is far more important than the law.

I agree though that it would be morally wrong to kill an innocent bystander just to save your car from damage.

Re:Rear Ended

[tehcyder]

How will it deal with passing a horse and rider on a narrow road? How will it deal with coming across another driver in a country lane at a place where there is no room to pass? How will it drive on a road with a shear drop off on one side and a cliff on another? If this is narrow with passing paces will it know to pull in so that it almost touches the cliff to give extra space to the vehicle near the drop-off? In busy commuter traffic will it adjust the "aggressiveness" of pulling out from a side-road to take into account that if you don't pull out quick and accelerate hard you could be waiting until the end of the rush? On a rutted farm track can it work out that you have to drive with one wheel on the centre of the road and the other on the edge to avoid the tractor ruts?

I see no reason why the car shouldn't be "intelligent" enough to recognise the problems in all these situatins.

As for aggressiveness, there is something wrong with your traffic system if anyone needs to drive aggressively just to keep moving. But presumably a load of rational non-human drivers wouldn't end up in that situation.

Re:Rear Ended

[bluefoxlucid]

"Heavily encrypted" psh. Logic analyzers bypass hardware encryption, son. That's how we got the Sonic ROMs. You try modifying a SOD-CMOS chip in an undetectable way, when the damn thing never alters state but just adds state so you can see all the changes that have ever been made simply by plugging in a UMCP ID and replaying the data core.

Re:Rear Ended

[ThatsMyNick]

Interesting points. Also one should note that Google car is more about being legally in the right, than doing the right thing. It would have been legal for it to run over the child (assuming it did follow speed limit, and tried to react to it, when it did eventually notice the child in front of it), or to just follow the road and fall of the bridge.
 
  This attitude difference is pretty interesting.

Re:Rear Ended

[mug funky]

need an advice dog here:

"accuse parent post of straw-man... ...respond with straw man"

Re:Rear Ended

[dywolf]

No, most accidents are due to simply a poor driver.
But that doesn't mean you should eliminate all drivers.
Millions of people drive to and frm work on long commutes every day without incident. Average 4 serious accidents per day in a city (atlanta) with >4 million commuters, do the math. the number is small as hell.

We have planes that essntially fly themselves, yet we still have a trained pilot and copilot in the front in case the hardware fails. And hardware does fail. You have to really be turning a blind eye to the reality of how often most people actually have their cars worked on.

so now you're also mandating preventive car maintenance, driving up costs of ownership, creating a host of new "crimes", and pricing it out of range of most users.

No, the automated car is a toy for rich people. It will never completely replace the driver, because of safety reasons or costs, or both.

Or you can mandate only driverless cars, and once again the people who can afford it can move around and leave the cities, going/moving where they want, and the lower classes who have no mobility and stay in one place, whether its inner city or rural countryside.

Re:Rear Ended

[thePowerOfGrayskull]

We don't know if the Google car could have avoided it. I was in a similar situation one time, and happened to catch a glance of the idiot in my rear view. I cut out ahead of the adjacent lane into the empty crosswalk. The idiot screeched to a stop in what was previously my lane.

This is the difference between a good driver and a shitty one.

The shitty driver accepts that his responsibility ends with "legally at fault".

The good driver knows that legalities don't stop bones from breaking, and pays attention to avoid even those situations where s/he would not be "at fault".

Re:Rear Ended

[dywolf]

addendum: google's cars had nearly no accidents, but i also garuntee they were continually monitored, constantly maintained, and never let out on the road if there was any inkling of any deficiency in the vehicle.

in other words, they got about a million times more maintenance and TLC than your typical car.

Re:Rear Ended

[dywolf]

exactly. and this is why totally driverless cars will not (or should not) happen. you will need someone, who knows how to drive, to be capable of taking over control when such a situation occurs. otherwise you get this:

"Unknown problem, turning over manual control"
"Huh?, What? I wasn't paying attention"
"Tonight on the 6 O'Clock news, a Ford TaurusDV drove off a bridge that was washed away due to the flooding"
(close enough)

Re:Rear Ended

[darkwing_bmf]

My guess is the cameras used for navigation will also record so that fault can be more easily assessed. Obviously courts will be involved but there should be ample evidence available so that most things can be resolved fairly and efficiently.

Re:Rear Ended

[Jedi Alec]

So I believe it's possible to give the fair assessment that a good driver is a better driver than an automated one, simply because that's the nature of driving. Unexpected situations.

Even if unexpected situations occur 9 times out of 10, they're still a matter of a lot of moving objects with speeds and directions in a 2 dimensional plane.

No matter what the exact reason, if the brakelights of the car(s) in front of you turn on and the distance between yourself and those cars starts to decrease, you should probably decrease speed as well.

If a car next to you starts getting closer (yay for jackasses changing lanes without looking properly) there's only 3 directions to escape, and some situations where you just need to stay where you are, even if it means physical contact between the vehicles.

Honestly, we teach this crap to 18 year olds (16 in some countries). If they can figure it out, so can a computer that has eyes in the back of its head and always pays attention.

Re:Rear Ended

[MightyYar]

So if you don't agree then we'll just have to agree to disagree.

I think we largely do agree.

I'm just pointing out that kids running into the street already get plowed down, and while I'm sure you can come up with scenarios that computers won't be good at, I'd put good money on overall injuries and fatalities going down. So, yeah, there might be the occasional hot air balloon collision, and that would be tragic. But meanwhile probably 100 (a number I completely pulled out of my ass) other lives would be saved. Yes, the kite-flying kid might meet his demise on the hood of a robot car... but 100 other pedestrians won't get smooshed by over-eager cars at busy intersections.

In short, I'm willing to take some "collateral damage". No amount of technology will ever render thousands of pounds of metal safe, but on balance, I'd wager that robot cars would produce less carnage than human-driven cars.

Re:Rear Ended

[HaZardman27]

I think what's really important here is whether or not the total realm of road safety is improved or not. Sure, maybe every once in a while something crazy like kids jumping over the road on bikes will happen and an accident will occur, but if that means that five other accidents did not occur, it's probably worth it.

This is also why a licensed driver will still need to be behind the wheel and sober, at least until the AI advances enough to be able to cover all possible driving scenarios. When you consider the percentage of drivers that probably don't know how to react to sliding or hydroplaning, or would panic in that situation and do the wrong thing, I think I'd feel much safer knowing that vehicles in my proximity are automated.

Inattentive drivers are scary, too. Think about all the times you may have nearly been hit by a driver writing a text message, or simply changing lanes without checking their blind-spot. Even worse, drivers don't even need to be inattentive to be terrible drivers. I used to live in a city where it seemed like people would wait for you to get close to the intersection they wanted to turn right at (this is in the US) before pulling out, rather than waiting for you to pass when there are absolutely no other drivers behind you.

Re:Rear Ended

[mcgrew]

On the other hand an automated car would have likely crashed in a few hundred other situations I've encountered, that after having reacted and averted the danger have left me going "huh, what the FUCK just happened?"

What leads you to think that in those unlikely situations the computer-driven car would crash? After all, it's logged 300,000 miles without a crash, have you? It seems the math kind of makes your hypothesis a bit illogical.

I used to think the same about antilock brakes. I was trained as a driver in the USAF where you learn NOT to lock up the brakes, as locking them up you lose all steering and it will actually take longer to stop than if you don't quite lock them up. But I have to say, after having them, stomping on the antilock brakes is far superior to doing it like the AF taught me in a car without them.

Plus, the computer is going to see the unexpected situation faster than you can, particularly if it's a car running a stop sign. You can only see in one direction at a time, the computer can "see" all around.

Re:Rear Ended

[gorzek]

An automated car is capable of being a better driver simply because it has resources we don't.

It can see everything around it in all directions at all times. We can't.

It can detect exactly how much traction each wheel is getting at any given moment, and react within milliseconds. We can't.

It can instantaneously produce a calculation to determine how to avoid an imminent collision and adjust it in real-time if the variables (e.g. traction, speed of oncoming vehicle, etc.) change. We can sort of do this, but not as quickly or accurately as a computer.

It never gets tired, it never gets distracted, it is never intoxicated.

Driving is a pretty narrow problem domain: follow a route, adjusting your vector as necessary, while avoiding running into anything else. That's basically it. A lot of things can happen while driving, but they all fall into that same general response pattern.

Re:Rear Ended

[Richy_T]

Every CO2 emission, [...] is bad.

I trust you'll be ceasing to make them at the earliest possible opportunity then.

Re:Rear Ended

[Havenwar]

No, the maths doesn't make my hypothesis illogical, not when you count the fact that these are no 300,000 consecutive miles from start to stop regardless of incidents. These are 300,000 miles interrupted by a lot of human overrides as soon as a situation came along that they didn't believe the car could handle, such as roadworks, parades, or whatever.

And no, I don't think I've done 300K miles, but I'm in the same order of magnitude, without a crash as well, so I'm doing pretty decently - be it skill or good fortunes. But this isn't a contest between me and google's car, I'm far from the best driver in the world. The last few years I have barely touched a car. This is a contest between human problem solving and processing skills, and computer skills... and that's a fight that the computer wins hands down... as long as it understands all the variables. That's the rub. Humans have evolved to react to the unexpected, to categorize things into threat or friend, to read situations they have never encountered in a way that no computer can - to this day. We're getting there, sure, but we're not there yet.

When google clocks 300,000 miles in that thing during regular commutes with some guy asleep in the back seat, then I'll be impressed and say it's better than even the best human drivers. As long as they are overriding it in all the situations where certain unexpected things can happen (they don't mind some unexpected things, cars drifting into lanes and such is easy for them), well as long as they keep doing that I'll stick with saying it's a hell of an awesome project, and I still want one. But it's not a better driver than a good human driver yet. Not for ALL situations, and thus not for driving in general.

Re:Rear Ended

[Richy_T]

Consider though that the roads are for transportation and not your personal pleasure. I'm with you on the feeling, I'm an avid motorcyclist but it is what it is. There will be tracks and closed courses (which are generally much safer *and* faster) to fulfill the fun requirement.

Re:Rear Ended

[marcosdumay]

In busy commuter traffic will it adjust the "aggressiveness" of pulling out from a side-road to take into account that if you don't pull out quick and accelerate hard you could be waiting until the end of the rush?

The nice things of computers is that, while you can make it act as conservatively as a near blind granma, it still has better reflexes, sees further, and make way more precise calculations than you.

So, yes, I bet it will be agressive enough to enter a high trafic lane. And it'll probably do a much better job on that than 99,99% of the drivers.

Re:Rear Ended

[marcosdumay]

I hope you can generate some quite impressive GCI movies in real time, because the data it'll gather is quite complex.

Re:Rear Ended

[tompaulco]

With 300,000 miles without "a single accident under computer control." it has a collision avoidance algorithm that is far better than yours. Therefore it would make sense that you wouldn't be allowed to manually override a far superior driver.
Just as an airplane is designed to be able to fly itself in most situations that we have seen and some that we have only imagined, we still cannot program an airplane to figure out what to do in a situation which it isn't programmed for. The same applies to cars, only more so, because cars travel within four or five feet laterally from each other, and there are many millions more of them then there are airplanes. An airplane pilot is specially trained to be able to handle the situations that the plane cannot handle by itself. I fear that with cars, if we have automatic driving, rather than be more rigorously trained for special situations, drivers will be less adequately trained because "the computer will handle everything".

Re:Rear Ended

[HiThere]

There *does*, however, need to be an emergency override. This doesn't allow you to take control of navigation, but it allows you to get to your destination faster. There also needs to be a severe penalty for using it in a situation that isn't justifiable. This is only practical if cars can talk to each other.

Think of this as analogous to an ambulance. The case I was thinking of is where a woman has gone into labor, but I'm sure that there are other similar justifiable reasons. In this scenario there should be NO driver involvement in routing, steering, speed, or anything similar. In fact even the horn should be under automated control. In fact, ideally I'd say that the front seats should automatically rotate to face the rear.

There probably also needs to be a "super cautious" setting, where the car automatically avoids anything that will cause bumps, but I don't think that one needs any special rules for when you can use it.

Re:Rear Ended

[Havenwar]

I had to stop and think for a while about all the sides of that issue. It's definitely a complicated one. Here's how I see it...

No.

First, technically we are not allowed to break the traffic laws even during labour and such, which means we should still be legally limited by the speed limits and all such things. An automated car should in theory be able to drive very close to these legal limits, and it will of course be doing so under all circumstances because there is no point in an automated car that isn't driving as efficiently as it can. There might be an "econo-mode" for the greenfreaks to enjoy, or even cars built entirely to drive as energy efficiently as possible - these particular models might have overrides for more aggressive driving, but that's something for that particular developer to decide on, nothing that is inherent to this type of car.

Second, those that are allowed to break the speed-limit, i.e. cops and medics and such, have their own vehicles, that will be built to different specifications... Perhaps even all manual ones, or eventually they as well automated. At some point in the future I'm guessing the ambulance won't even need to turn much heading down the road, because all the other cars will automatically go out of its way as efficiently as possible, without either party slowing down.

Now, if I look into the looking glass that far forward, I see ambulances no longer existing. Why? Well, because if you call 911 they'll check your stats or whatever, and if they deem you have to go in you get told to go get in your car, the operator remotely confers ambulance status to your car, tells it whether to avoid bumps and so on, what hospital to go to, and so on... and off it goes. Hell, far enough into the future they might even just remotely flag down the nearest car if someone drops in a public place.

To conclude, at this point in the development there is no point in a "get there faster" button, because it will always get there as fast as is legally possibly, which is as fast as you're allowed to get there. If you want to get there faster you meet up with emergency services along the way, this is standard legal practice - that I admit isn't practised much... but still. And if such a button eventually comes into existence, there is no reason to have it locally. You should have to call up emergency services and justify your need for it BEFORE the function activates.

Re:Rear Ended

[ukemike]

I also think that automated cars will observe all safety rules to the letter... like only driving the speed limit (or slightly below), always maintaining a safe distance behind other vehicles, stopping for yellow lights, and having a generous braking distance. Remember, Google could be held liable if the system is reckless, and they aren't going to want that when human lives are at stake.

Here in the Bay Area driving under the speed limit on the freeway is hazard to others. Traffic routinely moves substantially faster than the limit. A car lollygagging along at 54mph causes big backups and eventually someone will rear-end someone else. This is why the CHP stays off the freeways during the rush, they know that their presence causes more accidents that it prevents.

Re:Rear Ended

[ukemike]

As for aggressiveness, there is something wrong with your traffic system if anyone needs to drive aggressively just to keep moving.

What planet are you from? Of course there is something wrong with our traffic system. There are lots of things wrong. Something like 30,000 people in the US died in car accidents last year, and that is WAY down from what it used to be. The main thing wrong with our system is the idiot behind the wheel. But I digress. Of course there are times when you have to be more aggressive to get anywhere. Turning off my road onto the main drag in the morning often means using a large percentage of my cars acceleration capability. I'm talking just a bit short of spinning my wheels. Why? Well traffic is moving fast, the lights are timed such that there is always a steady flow of cars past my road and most people tailgate. You have to wait until you find a big enough gap and the you GO. I think that people are concerned that the automatic cars will have to be programmed to be so conservative that they would just sit there indefinitely like an old lady waiting for a huge gap to turn in to. BTW this does happen with human drivers too. There are about 8 houses and one apartment building on my street and on a few occasions about half the population of my street have been lined up waiting for some terrified idiot who couldn't find a gap big enough to pull out of.

Re:Rear Ended

[Puppet Master]

Yeah, but what if the guys name isn't Dave?

what is the issue???

[FranTaylor]

It's hard to imagine being found at-fault when you are stopped and rear-ended.

There's no shame in being involved in an accident if it's not your fault.

We trust others all around us every day to avoid smashing into us. Even the best drivers get hit.

Re:what is the issue???

[mark-t]

No shame, perhaps... but also no less pain, unfortunately.

Sure, they can offer some monetary compensation, but having been in such an accident and received adequate monetary compensation for all my medical expenses, I can sure as heck tell you that I would have rather not have had the money, and had those two years of my life *without* the back pain.

Re:what is the issue???

[Animats]

It's hard to imagine being found at-fault when you are stopped and rear-ended.

Especially when the self-driving car has full video, lidar, and radar coverage of the entire event. And really good lawyers.

Re:what is the issue???

[__aaltlg1547]

It's hard to imagine being found at-fault when you are stopped and rear-ended.

There's no shame in being involved in an accident if it's not your fault.

We trust others all around us every day to avoid smashing into us. Even the best drivers get hit.

The best drivers DO NOT trust others around them to avoid smashing into us. If we did, we'd be the not-at-fault person in a lot more accidents.

I believe it to be incorrect to compare the GoogeDrive cars to average drivers. They should be compared to professional drivers for two reasons:

• GoogleDrive is an expert system: a computer system designed from the ground up to do only one thing and do it extremely well. One does not expect such expertise of people whose driving is incidental to what they do.
• GoogleDrive will only be affordable (at any time in the next couple decades) to rich people to replace professional, expert drivers. So it has to be better than expert drivers to make the case compelling to people who might actually be able to afford it.

Re:what is the issue???

[geekoid]

" to rich people to replace professional, expert drivers."
Mercedes has a history of being first t utilize new technology. Why you think rich people are professional drivers is beyond me. Hell, most 'professional drivers' are no better then any one else.
The tech will take bout 4-6 years to go from high end, it low mid range. The idea that it will take 'decades' is laughable.

Re:what is the issue???

[knuthin]

The only accident that happened with the self driving car, was when it wasn't being self driven. Just explains your point better.

Re:what is the issue???

[knuthin]

Google Drive?

Re:what is the issue???

[GigaplexNZ]

Why you think rich people are professional drivers is beyond me.

What? No. Rich people hire people to drive them around.

Re:what is the issue???

[TubeSteak]

geekoid has revealed himself as one of the hoi palloi that has no choice but to drive himself around.
Come GigaplexNZ, let us retire to the study for brandy and cigars, so that we may laugh at geekoid behind his back.

Re:what is the issue???

[Dunbal]

Unless of course you happen to be stopped in the overtaking lane of a highway...

Re:what is the issue???

[YttriumOxide]

GoogleDrive will only be affordable (at any time in the next couple decades) to rich people to replace professional, expert drivers. So it has to be better than expert drivers to make the case compelling to people who might actually be able to afford it.

I think you're either vastly overestimating the cost of GoogleDrive or vastly underestimating the cost of hiring someone to drive you around all the time... (or you live in a third-world country where labour such as drivers is cheap).

I'm by no means "rich" by any stretch of the imagination and certainly can't afford a full time driver, but as long as these cars come in at under €80k, I'll happily consider one (I'll go up to €100k if it's also a pure electric, given the price difference between petrol and electricity here)

Re:what is the issue???

[Dunbal]

GoogleDrive will only be affordable (at any time in the next couple decades) to rich people to replace professional, expert drivers.

Remarkably stupid if Google does this. They'll go the way of the "electric" car and all the other fancy cars targeted at, well, Beverly Hills. A few rich celebs will buy one and pose for the cameras, and then they will be forgotten. No, Google needs to bite the bullet and take an example from Henry Ford. Make pennies on the unit, but make millions and millions of units.

Re:what is the issue???

[dgatwood]

Seriously. Clearly that other interpretation isn't true. Just about every time I've seen somebody driving like a complete idiot with no regard for anybody else on the road, it has been somebody in a Lexus. And when it isn't a Lexus, it's a Beamer or a Benz. Almost always. Oh, and lately, Scions. Well, and motorcycles, but they don't count.

I was cut off by a Ford SUV once. Once. And I have never been incompetently cut off by someone driving any model of VW, Toyota (unless you count the Scion or Lexus), Nissan, Honda, or just about any other non-luxury car brand. Never. Yet those make up the vast majority of cars on the roads around here.

Draw your own conclusions.

Re:what is the issue???

[dgatwood]

Which suddenly made me realize something. Even if Google only succeeds in getting these things in luxury cars, the accident and near-miss rate is likely to plummet for everyone. :-)

Re:what is the issue???

[Anonymous Coward]

> Draw your own conclusions.

That's easy - you suffer from a very severe case of confirmation bias.

Re:what is the issue???

[timeOday]

The best drivers DO NOT trust others around them to avoid smashing into us. If we did...

Ha ha ha... classic...

Did you know 93% of drivers consider themselves better than average?

Re:what is the issue???

[genik76]

It is entirley possible that the 93 % are right, if the worst 7 % are REALLY BAD drivers.

Re:what is the issue???

[interkin3tic]

It's hard to imagine being found at-fault when you are stopped and rear-ended. There's no shame in being involved in an accident if it's not your fault.

There's a -little- shame in being rear ended. Because of the jokes.

Rear ended... tee hee...

Re:what is the issue???

[Bert64]

No shame perhaps...
But the inconvenience, the expense, the potential injury depending how hard you were hit...

You could be injured by the impact, perhaps seriously, this might cause you to spend time in pain and/or hospital when you'd otherwise be working or doing something more enjoyable...

Your car is most likely to be damaged, which could result in a costly repair or having to find a replacement, and in the interim not all insurers will provide a rental car so you might find yourself without transport.

And while it is almost certainly not your fault if your were stationary, that doesn't mean the other party won't try to claim that it was and keep both you and your insurer busy arguing about it... Plus there might be others involved if the impacted shunted your car forwards into another vehicle.

Re:what is the issue???

[AliasMarlowe]

It is entirley possible that the 93 % are right, if the worst 7 % are REALLY BAD drivers.

Not even really bad. For instance, if 93% of drivers are 1% above the average in driving skill, then the remaining 7% of drivers only need to be 13.3% below average in driving skill. I've had to avoid drivers who are a lot worse than that. My memory (or confirmation bias) suggests that the worst drivers are found in Audis and invariably have a cellphone stuck to their ear, whether in North America or in Europe.

Re:what is the issue???

[Bert64]

It's all about what the car is worth to the driver... I find that people driving really old beat up cars can often be just as bad. When you're driving a car which has a trivial value relative to your available funds, you are far less concerned about damaging it.

Re:what is the issue???

[Bert64]

I stop in the overtaking lane of the highway all the time, during peak hours around here there are queues of stationary traffic in all lanes of the highway.

Re:what is the issue???

[TClevenger]

And I have never been incompetently cut off by someone driving any model of VW, Toyota (unless you count the Scion or Lexus), Nissan, Honda, or just about any other non-luxury car brand. Never.

I was nearly killed on my scooter today by a woman in a Honda Accord who ran a stop sign at 25 mph and turned directly in my path. Anecdote is anecdoty.

Re:what is the issue???

[Firethorn]

I've calced out an autodrive system to be 'worth' around $20k to the 'average' person, with the following assumptions:
1. The system is better than 90% of drivers. It may get into accidents that a human would have avoided, but it avoids accidents that a human would have caused. IE it might get into an accident where it didn't recognize the hazard that would have been very obvious to a human, but avoided an accident where the human wouldn't have been able to react fast enough. Whatever, it's somewhere around that 95th percentile. Easily 'professional driver' range for preventing accidents. Result: 90% savings on insurance costs. ($1k-5k/year, depending on record)
2. The vehicle is driven, on average, 1 hour a day, the system lasts 5 years(pessimistic, most cars last 10+), and the ex-driver values his or her time at $10/hour(works, reads a book, surfs the internet while transiting). ($18k)
3. The system results in mild mpg savings(round up)

How valuable would it be to others? Do you drink outside your house? Average cost of a DUI conviction is like $10k. Have a DUI? $1.3k a year for one of those breathalyzer systems, which is a big pain to start your vehicle with AND during driving. Which you might not need if you get an autodrive vehicle instead. Heck, get them widespread enough and it might be mandated for you.

Right now google's system costs $150k per car, with nearly half($70k) being for the laser lidar system. So it's not worth it to the 'average' individual. But what about for cargo moving? Semi-Tractor Trailers? Raise the cost of a driver to $20/hour(more realistic) and 8 hours a day, and it'd be worth it at $320k. Pay some contractor to fill the truck up when it comes in for fuel.

Besides that, there's lidar systems coming out as cheap as $250 per vehicle, which leaves ~$60k cheaper on other parts to make it worth it - and how much of that equipment cost is because it all has to be had made and assembled to modify a car? Wouldn't it be a lot cheaper if it's a factory installed option?

Re:what is the issue???

[Quillem]

Hoi polloi*

Perhaps it's you who has revealed yourself :)

Re:what is the issue???

[Dr_Barnowl]

As the percentage of auto-drive vehicles on the road increases, the effect on traffic flow will be good as well.

A lot of jams on the motorway are just because people have crappy feedback loops, and don't leave enough braking space, and this causes waves of braking to propagate back through the traffic, gathering magnitude, until you end up with a stop.

Auto-drive cars will both leave sufficient stopping space and given a means of communicating with each other, can brake in perfect synchrony, anticipate each others lane changing and turning manoeuvres, etc.

This will have more of an effect on fuel efficiency and the general throughput of the road network than anything else. The only downside to this is that it will become less easy to successfully argue that you should be working at home...

Re:what is the issue???

[dontmakemethink]

Wise man say, when crossing one-way street, look both ways. There are very many hazards that automated cars undo as well. I read an amazing essay about life with all fully automated vehicles, where you don't own a car, instead you punch your cargo and travel plans into a website, and the appropriate vehicle shows up and takes care of your travel needs. If it's a mile or so to the grocery store, a wagon shows up. If it's to the remote cottage an SUV shows up. If it's to a wedding a limo shows up. If it's to an airport a shuttle bus shows up with room for you and your baggage along with others etc. Think about how much time your car is parked and think about how many fewer automated vehicles it would take to service a large population. MASSIVE CO2 emission reduction, especially if most of them are fully electric, as they could easily recharge themselves automatically. The ramifications are really stupendous.

I can't find a link to the essay (I'm unwinding after a long day and I get 3 hrs sleep before a 17 hr day tomorrow), but I'm sure other /.ers have heard of it.

Re:what is the issue???

[u38cg]

Claim incidences for an individual driver are Poisson distributed, with the parameter itself being Gamma distributed. Obviously these are approximations, but surprisingly good ones.

Re:what is the issue???

[YttriumOxide]

I'm by no means "rich" by any stretch of the imagination and certainly can't afford a full time driver, but as long as these cars come in at under €80k

You do realise that 80k is about three years of average pre-tax salary in the UK, for example?

I never said I wasn't "above average" - just not "rich"...

Re:what is the issue???

[YttriumOxide]

If you can afford a €100k car, then you are by all means "rich" without any stretch of the imagination. Most people budget quite a lot to afford a car a third of that price. I've got friends who are well off - double income, no kids - who have two cars that together doesn't total €80k, and while they still have money left over they certainly spend most of their income just paying for those cars.

In other words, I suggest you take a step back and re-assess your situation. Congratulation, you are one of the few, one of the privileged, one of the people fortunate enough to be able to come on slashdot and pull numbers out of your ass to make yourself seem special.

Hey, that's a little harsh isn't it? Actually, a €100k car would be something I'd spend 8 or so years paying off. I'm definitely well above average wage, but hardly amongst the elite.

As it is, my current car was around a third of that price; and I paid it off in 3 years (actually, final payment is coming up in a couple of months). Did that reduce my ability to do other stuff? Sure. No good holidays in the last three years and I haven't upgraded my computers in that long either. It's a matter of priorities. If I were to buy a €100k car, I'd be living a lot tighter for the next 8 years, but I'd consider it worth it.

Re:what is the issue???

[YttriumOxide]

So would I have - before I was married and with a daughter. A studio wouldn't do much for me these days.

Re:what is the issue???

[tftp]

Rich people hire people to drive them around.

Hiring a servant (a driver) full time will cost you about $100-120K per year, considering the overhead (employment taxes, health, etc.) This money can buy you a very expensive car, and that car will serve you not for one year but for ten, if not twenty.

If self-driving vehicles appear in dealerships, do you know what kind of "rich people" will be buying them? Old people who cannot drive anymore because of health problems. Not even mentioning younger disabled people - those will pay all they got plus some more because the alternative is to sit at home and wait for Thanatos to come.

Re:what is the issue???

[GigaplexNZ]

If self-driving vehicles appear in dealerships, do you know what kind of "rich people" will be buying them? Old people who cannot drive anymore because of health problems.

And rich people who like to drink when they're out.

Re:what is the issue???

[N1AK]

GoogleDrive is an expert system: a computer system designed from the ground up to do only one thing and do it extremely well. One does not expect such expertise of people whose driving is incidental to what they do.

If GoogleDrive was 3x less likely to cause death or serious injury than an average driver, but only 99% as good as a pro-driver, why is that an incorrect comparison? Knowing how good it is versus a tiny fraction of drivers is a distraction.

Your estimate about the timeframe to become affordable certainly doesn't help with credibility. Safe self-driving cars already exist there is no way we're not going to see affordable self-driving systems until after 2032 (your timeframe, not mine) given the cost savings in logistics, road safety improvements and benefit to elderly and disabled drivers. I'd be shocked if self-drive wasn't a standard feature on 'executive' cars before the end of the decade.

Re:what is the issue???

[WinstonWolfIT]

If you ever get rear-ended and wind up interviewed by the insurance companies, the details of your stop most assuredly can end up with you being assigned partial monetary responsibility for an accident. Stopping for a pigeon for instance will get you 20% easy. I know this for a fact because I was cut off as a cross illegally crossed in front of me to enter a shop entrance, and I got 20% assigned fault for not predicting the stupidity.

From a tort perspective, Google Drive may have stopped abruptly due to not competently observing traffic conditions, so the record is debatably flawed.

Re:what is the issue???

[jpapon]

Wtf, why is someone making 10k€ a year wearing 200€ shoes??

Re:what is the issue???

[ras]

with nearly half($70k) being for the laser lidar system. ... Besides that, there's lidar systems coming out as cheap as $250 per vehicle

Where there is a will, there is a way. LIDAR systems are hideously expensive systems costing at least 5 digits, so designing one that is cheap enough to use in a vacuum cleaner would be impossible, right? Well, no. You can build one for $30. IEEE Conference Paper.

It's amazing what mass production can do to price. Just ask Apple.

Re:what is the issue???

[Zilberfrid]

The worst driver I know uses a Smart, and uses the gas pedal like an on and off switch. Choices are idle, or full power, hardly anything seems to exist in between.

Re:what is the issue???

[Lord Lode]

Personally I'd enjoy other things more than a car, but I guess if you're into cars you consider them worth living tighter for 8 years :)

Re:what is the issue???

[demonlapin]

think about how many fewer automated vehicles it would take to service a large population.

And now think about how many extra miles each of those vehicles will be driving. And think about how disgusting those cars are going to be inside.

Re:what is the issue???

[YttriumOxide]

Yeah I agree, it was a little harsh, and I apologize for that. It's just that since I live in a first world nation on less than €10K a year before taxes I tend to get a bit touchy when people talk about "living lighter for the next 8 years" in regards to €100K purchases.

That is completely understandable. However it's worth remembering that less than €10k is a very small amount to live on in a first world nation. That puts you significantly below the average.

Beyond that, it's also worth remembering that quality of life increases dramatically over a certain point. My "day to day costs" before I met my wife weren't all that different to when I was on half that income; just the extra that I had could go on "toys" like computer equipment and cars and so on. Now I'm married and have a daughter, but the tax system here in Germany is quite favourable for my situation (my tax went from around 50% to around 33% just by getting married) so despite the extra costs with my new situation, it's not as detrimental as it could've been (financially... nothing compares to having a loving wife and daughter of course; so even if I had been significantly worse off financially, I still would've done it).

I am so sorry for claiming you were either rich or full of shit, it's obvious to me now that you are indeed just "above average".

No problem - I remember well being in a pretty bad situation myself. I grew up in a poor family; then after leaving home at 16 managed on my own to make quite a lot by the time I was 21. However by 23, I'd blown it all on "the high life" and was burned out. I was poor for the next 5 years or so, and then slowly made my way up from "pretty crap" wages through to where I am now. It's really only now in my early 30s that I feel I'm really starting to get back on my feet properly.

I really hope for you that you can get yourself out of the situation you're in as well.

On an unrelated note my current car is a pair of €200 shoes. I guess that's maybe six months of savings, so to invest as wisely as you in my transportation I should then aim for six times that?

I don't have any shoes worth anywhere near that amount myself - my most expensive pair was around €50. But then again, my shoes aren't my transportation over any kind of longer distances.

As I said though, there's a point where extra income can be used for "play" a bit more. I certainly wouldn't suggest spending 6 months worth of your income on a pair of shoes; but 6 months of income on a car might be reasonable (the one I'm paying off over three years represents around 7 or 8 months income).

You said you paid your current car off in three years, so clearly three years is a good time period for this. Or eight years, meaning 16 times that? I don't know, I'm sure they do make shoes that expensive, but then I'd be going far beyond my needs into the area of choosing style over purpose.

Yes, you really would... which is why I don't suggest that at all...

Your current car was about a third of €100K? Well, mr. above average... You might not be rich or full of shit, but you still buy overpriced cars. That's not an option for most people.

A new car here in Germany costs between €10k and €60k unless you're going for something really special. So, spending €33k on a car is really around the middle of the range. Sure, it was a "brand new car", which I'm aware is a luxury in and of itself - many people only ever buy second hand (whether required to by financial situation or simply because it does financially make much more sense) - but for me it's not just transportation; it's fun (and stress relief from a pretty hectic job and home life) as well.

Re:what is the issue???

[Anonymous Coward]

a) I don't pay taxes
b) What overhead? My driver lives in westernmost antechamber and we seldom entertain there. He provides his own linens and personal items and eats the same food as the other servants, which is to say the overage after the family has dined. All sunk costs my friend.
c) 100k-120k is a pittance even if his salary approached that. The most effective means of holding down salaries is to have someone watch the daily papers and inform you of a country undergoing civil war, then have your servants sourced from there. Obviously that inhibits communication for the short term, but they are so thankful to not be killed you needn't pay them at all.
d) You really are doing them a favor
e) While being chauffeured about one can engage in a variety of estate furthering dealings that otherwise would not ft the schedule.

Re:what is the issue???

[coofercat]

In London we have "car clubs" that do this without the autonomous vehicles. Near my house there's a small car parked in a normal residential parking space. I can walk up to it, open the door and drive it around for a couple of hours. If I need to put fuel in, I can do that via fuel card (or reclaim money if I can't use the fuel card for some reason). If there's anything wrong with the car, or I have an accident, I push the "call" button and talk to an operator. All this by pre-booking online, and paying on my credit card after I return the car to it's normal resting place.

It's an expensive way to replace a car you already own, but it's a cheap way to borrow a car for an hour or two a week. When we get autonomous vehicles, it'll probably become entirely more popular, and different vehicles will arrive at your house, as you describe. Can't wait ;-)

Re:what is the issue???

[jellomizer]

While you are not at-fault... However there are things you can do as humans to avoid such accidents.
Just last week, I was driving down the highway and traffic started to slow down. So I started to slow down, and I saw the person in the car behind me, wasn't making eye contact with my car, so I began to swerve in the breakdown lane until she saw that something was happening, a surprised look on her face and a quick slowdown, no accident, however if she didn't pay attention for a second or so longer she would have hit me, if I wasn't paying attention.

Re:what is the issue???

[beelsebob]

($1k-5k/year, depending on record)

Holy cow, how expensive is insurance in the US? For a good driver with no claims bonus in the UK, you're looking at about £150 fully comp, even for a reasonably powerful/expensive car.

Re:what is the issue???

[jon3k]

And once we're ALL in self driving cars it will be a non-issue.

Re:what is the issue???

[locofungus]

There's no shame in being involved in an accident if it's not your fault.

I disagree.

There are some accidents that it is impossible to avoid but the vast majority of accidents could be avoided even if it's the other drivers fault.

In fact, I think one of the problems leading to the carnage on the roads is that "it wasn't my fault" is an acceptable excuse.

I took an advanced motoring course nearly twenty years ago. It's telling that since I took that course I can only think of one occasion where I have needed to use emergency braking to avoid an accident. I can remember an occasion when my passenger was talking to me then I say "Hold on, where has that dog gone?" I'm slowing down, looking for it and then, lo and behold, it steps out in front of me, just about where I would have been if I hadn't started slowing down. Sure, I make mistakes all the time, everyone does but some of my mistakes are things like "shit, I didn't notice that turning or any signs for it" as I go barrelling past with my foot still on the throttle completely unprepared for the car that could have pulled out in front of me if there had been a car and if they'd also not been paying attention.

I'm also very conscious of the fact that my speed tends to increase as I start getting tired. I start paying less attention to the road, seeing fewer hazards, and so the apparent safe speed increases.

Tim.

Re:what is the issue???

[TheRaven64]

It really depends a lot on where you are. Before I moved, I was spending a bit less than that and living quite comfortably. Now, on the other side of the same country, that wouldn't even cover my rent (in a smaller flat).

Re:what is the issue???

[Hodr]

I was found at fault. Had been at the stop light for 30 seconds or more when I was rear-ended. I was found at fault and cited for "stopping short". Didn't have to wait long for the cops to show up and make that determination, since it was a patrol car that hit me.

Re:what is the issue???

[nolnacs]

N of 1 but I would definitely pat at 10K more for an automated car, maybe 15. There are so many other things that I would rather do that pay attention to the road during my 1.5 hour daily commute.

Re:what is the issue???

[drinkypoo]

They'll go the way of the "electric" car and all the other fancy cars targeted at, well, Beverly Hills. A few rich celebs will buy one and pose for the cameras, and then they will be forgotten.

Electric vehicles continue to gain in popularity. The Leaf is being sold as rapidly as it can be produced. A tax credit just appeared for electric motorcycles, some of which have excellent range.

Re:what is the issue???

[Firethorn]

Source on that £150? I found £971

Now, you specified good driver, and I'll admit to rounding up; $1k is about what my comprehensive coverage costs.

Might want to double check that your insurance is valid, apparently there are fraudulent "ghost brokers" running around.

You guys even have cancellation fees?. It's my understanding that they're pretty much unheard of(probably illegal) here in the USA. In addition, increasing from ~700 to 7k while switching to advertising your monthly payment where you used to have your annual price can be considered fraud.

Re:what is the issue???

[Firethorn]

I'll note that we generally don't have $7k insurance policies for new drivers. $5k is more for the 'multiple DUI accident' type. 'Fronting' is a new term for me.

Re:what is the issue???

[cduffy]

And now think about how many extra miles each of those vehicles will be driving. And think about how disgusting those cars are going to be inside.

I'm not sure where the higher per-vehicle utilization becomes a bug rather than a feature. As for the interior -- I'm a member of car2go, an existing vehicle-sharing collective, and we don't have problem.

Every time someone gets in, they're asked to rate how the interior and the exterior look. Successive low ratings? The car gets cleaned. If there's a serious problem, there's a button to call customer service -- and if they need to, they'll ground the car until someone gets out to fix it, and bill the person who had it before you for the cost.

And since you were asked how the interior looked when you got into the car, you're always reminded that the person after you will be asked the same question.

Re:what is the issue???

[timeOday]

If I'd known it was going to get technical I wouldn't have mis-quoted my wikipedia link. Here is what it actually says: " For driving skill, 93% of the US sample and 69% of the Swedish sample put themselves in the top 50% (above the median). "

Re:what is the issue???

[DragonWriter]

And now think about how many extra miles each of those vehicles will be driving.

With good routing software, very little, and with high enough utilization (and since they'll be able to utilize loading areas more efficiently, they'll reduce peak parking demand, and consequently parking space requirements -- which increases trip lengths for both automated and human-driven vehicles.)

Re:what is the issue???

[justthinkit]

Don't forget that autodrive does not have to sleep. What are the savings to a trucking company from that alone?

Re:what is the issue???

[spasm]

The plural of anecdote is not data..

Come visit LA. You'll get incompetently cut off by lunatics driving pretty much anything on a daily basis.

Re:what is the issue???

[downhole]

I was a bit suspicious of that, actually. They say it was being driven by a person at the time, but then it's gotta be pretty heavily in their interest to say so. Nobody's sorted out yet what the legal consequences are for a self-driving car self-driving itself into an accident and Google may not want to be the ones to find out just yet. Even in Nevada, where they just made it explicitly legal to have self-driving cars, there's never been a real test of the legal system in case of an accident with them. Who knows what could happen in California? I could see Google wanting to claim it was being driven normally and paying everyone involved to shut up about it.

They've also got to want to make it appear as if their self-driving system is safe. I don't think the general public would be very understanding of cars with the usual software development cycle of test it 'til it crashes, fix what caused the crash, repeat until you think it's release quality.

Re:what is the issue???

[dywolf]

Ultimately this is about relieving ourselves (or others we percieve as being not-good-enough) of responsibilty for their actions. Personally, I feel that is an undersirable outcome for society that already has problems with responsibility.

And in the end, the laws needed to force driverless cars with sufficient maintenance, and no modding, mean one thing: less freedom. And that also is something I cannot support.

Re:what is the issue???

[g1zmo]

Agreed. The facts of the case don't feel truthy enough for me either.

Re:what is the issue???

[Jedi Alec]

Mercedes also targets the upper market segment. And the stuff they put in their better series almost always finds its way to even the most basic vehicles across the board in a 10 year lifespan. Right now their cars already have cruise control that keeps track of the distance to the car in front of you and adjusts speed appropriately.

Cars take care of the driving more and more as we speak. Switching to full automation is more a question of if than when.

Re:what is the issue???

[Anarchduke]

I believe it to be incorrect to compare the GoogeDrive cars to average drivers. They should be compared to professional drivers...

RIP Dale Earnhardt

Re:what is the issue???

[Anarchduke]

do be fair, I was actually trying to hit the scooter.

Re:what is the issue???

[ab0mb88]

And I have never been incompetently cut off by someone driving any model of VW

Someone has never driven in a college town. Even the smallest colleges have more than a few young drivers more worried about their makeup and hair than where their New Beetle is headed.

Re:what is the issue???

[dgatwood]

That's easy - you suffer from a very severe case of confirmation bias.

Nope. I'm just acutely observant.

To be fair, Jaguar didn't make my list, but in my defense, I don't think I've never actually seen one on this side of the pond except on a lot or in a museum....

Re:what is the issue???

[mattack2]

They're probably cleaner than my car!

Re:what is the issue???

[beelsebob]

Source on that £150? I found £971 [telegraph.co.uk]

average meaning that it's massively biased by 19 year old males, who have 3 driving convictions and 0 no claims, paying £5,000 for a year.

Re:what is the issue???

[demonlapin]

I'm not sure where the higher per-vehicle utilization becomes a bug rather than a feature. As for the interior -- I'm a member of car2go, an existing vehicle-sharing collective, and we don't have problem.

High utilization isn't (except for maintenance) a problem, but it does mean that there aren't really going to be many (if any) fewer vehicles on the road. Parking lots might be emptier, but the number of trips isn't likely to change significantly. As for your existing vehicle-sharing collective, I wonder how much of that would scale if the entire population (instead of a small segment) were using them.

Re:what is the issue???

[__aaltlg1547]

Read for comprehension. This won't appeal to rich people who like to drive agressively. It will appeal to rich people who would rather pay somebody (a person or a machine) to drive for them. The kind of people who hire drivers now.

Re:what is the issue???

[Firethorn]

Average lowest of £727 is scotland
Per the chart, 71+ averages the lowest at £436, the average of £971 from the other article would be for somebody closer to 30, with an average driving record.

Still waiting for where you source a 'mere' £150 from. Basically, I'll dispute that your insurance figure has any real meaning on reality as you're very far from average, and might be getting a deal such as I'm figuring the automatic driving vehicles might get.

It's a touch annoying to not have a '£' on my keyboard for this conversation. ;)

Re:what is the issue???

[beelsebob]

You really didn't read what I said –it's nothing to do with the age range you quote, it's to do with the fact that the average numbers are massive skewed by a few bad drivers. People who have caused accidents, been banned in the past, had drink driving offences logged against them, etc all pay much much much higher premiums than the average, that means that the reality is that for most people who are fairly average drivers, the actual cost of insurance is much lower than the £700-1000 you've found. I for one, can attest that my insurance on a 1.8l 5 door is between £150 and £200 depending on the insurer (I really genuinely actually just filled in the forms again just now). And I'm far from the lowest bracket, I'm male, I'm (relatively) young, and I've had points on my license within the last 5 years (okay, only a 5mph over the limit on a motorway, but still). The cost of insurance is far from as high as the average numbers make it seem.

Re:what is the issue???

[Firethorn]

I'm still asking for a source on what you figure the median person pays. I've found plenty of averages. And I've found cases of them charging teen drivers thousands of pounds. Logically speaking, you're pretty far from the average - you're probably a couple standard deviations from the center.

However, 18-14 is still 'only' averaging 2k, and they're what, ~10% of the driving population?

Still, to get back to my original point, with such insurance expenses for the 'worst' drivers, it makes autodrive solutions even more attractive for them. I already mentioned that I was figuring on a top end of $5k for drivers with DUIs, which is close to to your 5k pounds. Much above that and they're uninsurable here in the states.

If you want personal details - my annual liability is only $260/year. My collision(fix my truck if I wreck it) is $493, Comprehensive(damage/loss other than driving - vandalism, theft, fire, hail, etc...) $160. Uninsured/Underinsured(the one who hit me doesn't have insurance) - $83. That and a few other perks(towing/rental) gets my insurance up to $1043/year.

Re:what is the issue???

[cduffy]

High utilization isn't (except for maintenance) a problem, but it does mean that there aren't really going to be many (if any) fewer vehicles on the road. Parking lots might be emptier, but the number of trips isn't likely to change significantly.

It's beneficial in a few ways:

• The vehicles that do exist are better-maintained. Poorly-maintained vehicles contribute far, far more to air pollution than well-maintained ones.
• Parking lots in urban areas are actually a huge problem. They add up to massive amounts of prime real estate that could otherwise be put to productive use, but which instead do nothing but house idle vehicles. Zoning laws with minimum parking requirements are hurtful to businesses which could otherwise fit into higher-density areas -- Austin recently had a major vegan restaurant nearly shut down due to lack of compliance with parking requirements, despite catering to a clientele that mostly arrived by bicycle.
• Having costs be per-mile or per-minute with granular billing (car2go is the latter) rather than fixed monthly (or amortized into weekly or bi-weekly fuel stops, monthly auto and insurance payments, annual maintenance, etc) provides motivation to use transit, cycling, or other options when they make sense, rather than having the car be first-choice.

As for your existing vehicle-sharing collective, I wonder how much of that would scale if the entire population (instead of a small segment) were using them.

It's a question, indeed. I'm not sure I'd posit "entire population" -- we have people who can't get driver's licenses, after all. Similarly, I wouldn't want car-sharing membership to be a civil right -- if one abuses it, they should be able to lose the privilege, just as it is presently with privately-run car-share operations rejecting members based on background checks done by their insurance carrier.

Re:what is the issue???

[Hentes]

Because we only have Google's word that their car is as good as they claim to be, and now it turns out they were lying about something. Being rear-ended is not the problem, claiming no accidents when there was at least one is. Google has been testing these cars for years, if they are as good as they say I don't know why are they evading open, independent testing.

Re:what is the issue???

[randyleepublic]

>> The ramifications are really stupendous.

Especially when the cars are able to enter the vacuum tubes and travel at hypersonic velocities.

Interference?

[Anonymous Coward]

I want to know about interference between cars. I've only see one self-driving car tested at a time. If there's hundreds within visual range of each other are their radar and laser sensors going to have much more noise?

The little experience I have with robots is that laser range finders like to bounce off things and skew readings. How do the cars deal with that?

Re:Interference?

[stephanruby]

I want to know about interference between cars. I've only see one self-driving car tested at a time. If there's hundreds within visual range of each other are their radar and laser sensors going to have much more noise?

With hundreds of self-driving cars everywhere, then they may even be able to save on gas by flocking together and save on wind resistance and save on stopping time by sharing gas and Slurpees at high speeds.

And of course, fewer sensors would have to operate that way, only the car in front would have to scan far off in the front, and only the car in the back would have to scan the rear.

Re:Interference?

[rcallan]

They're relying on "we tried it with a few in the same area and nothing bad happened" from the DARPA urban challenge at this point.

Re:Interference?

[Anonymous Coward]

... and only the car in the back would have to scan the rear.

"I've got your back! Crazy humans out there!"

Re:Interference?

[kiwimate]

I want to know about interference between cars. I've only see one self-driving car tested at a time. If there's hundreds within visual range of each other are their radar and laser sensors going to have much more noise?

I believe it is Toyota who is working with a system that allows cars to talk to each other. These aren't self-driving cars, but the idea is that if you are approaching an intersection and have right-of-way, and another driver is approaching the same intersection and is barreling on through regardless, the two cars will talk to one another, determine that they are on a collision course, negotiate, and one of those cars will slam on the brakes automatically. How it'd work at a busy NY intersection, who knows...but that's what computer simulations are for.

(And also why you need mathematics, in reference to the story yesterday about "does anyone really use maths in their programming job?". If you're working with any kind of engineering, and especially EE, then the answer is a solid yes, either the programmer or the EE will be implementing higher maths in their code.)

Re:Interference?

[coofercat]

This is/was one of Sebastian Thrun goals - he wants to see motorways turn into privatised railways. The cars will literally drive inches away from each other, because that way you maximise the utilisation of available road surface, but also everyone gets a fuel economy boost from the guy in front. If your car needs to get off the road, then it just slides out and the gap closes up. If you want to join the flow, then a gap opens up and you slide in.

He says that we only use something like 2% of the available road surface at the moment. If that's true, then you could imagine that the multi-lane roads we have might only need to be single lane, and still support many times more card than we can support at the moment. Interesting stuff indeed...

Re:Interference?

[CCarrot]

With hundreds of self-driving cars everywhere, then they may even be able to save on gas by flocking together and save on wind resistance and save on stopping time by sharing gas and Slurpees at high speeds.

High-speed beverage passing? We can only dream...

Although, that does sound much more entertaining than a lot of the other crap on the idiot box these days :)

Re:Interference?

[stephanruby]

I speculate the biggest issue will be compatibility between the different kinds of cars.

The Microsoft cars for instance probably won't want to flock with the Linux cars and some of the Linux cars will have to pretend to be Microsoft cars if they ever want to go somewhere in a flock where no other Linux car is going.

And of course, there will be endless talk of Flock 5.0, a supposedly compatible way to flock between car OSes, but that standard will mean different things to everybody and will take quite a while to implement.

But how smart?

[Anonymous Coward]

My understanding is that it cannot read signs, or deal with many types of unusual conditions like detours, nor can deal with a location without maps. Does anyone know about the limits of the Google car?

Re:But how smart?

[peppepz]

I also read that the automatic drive wasn't able to cope with simple situations such as another car coming from the opposite direction in a narrow street, requiring manual intervention. So alongside the triumphant tones, they should also explain how much these cars are really self driving; most car accidents don't occur in straight motorways.

Re:But how smart?

[cheater512]

Not sure about the other conditions, but it can identify and read road signs.
Does the same thing for traffic lights so it obeys them and it reads speed signs.

Re:But how smart?

[vux984]

Pretty much this.

Google deliberately avoids the more challenging situations, and a LOT of those miles are highway.

There's a reason insurance rates for someone living in a small town in the country are lower. Right now, google is pretty much "that guy".

That's not to disparage what google has accomplished, but its premature to compare it to the safety record of a downtown urban commuter; driving through rush hour traffic to and from work in a major city daily.

Re:But how smart?

[HideyoshiJP]

But you can't have that hour and a half to do those things. Much like a pilot, you'll be expected to watch over the car. Unlike an airline pilot, you won't have as much time to react to a problem - you'll have mere milliseconds, thanks to obstacles being so much closer.

Re:But how smart?

[Thelasko]

Google deliberately avoids the more challenging situations, and a LOT of those miles are highway

I work in the auto industry and have seen miles used as a metric in a number of tests. The problem with using miles as a measurement in an automotive test is that highway miles are inherently different than city miles. A manager always says, "this vehicle needs to last X miles. Get the verification data as quickly as possible." The product validation team immediately goes out and puts X highway miles on the vehicle because highway miles are the quickest way to rack up miles.

Later on in those programs, we usually find some issues when the product operates in a city or low speed environment. I've been pushing for vehicle hours as a replacement every chance I get, but people don't listen...

Re:But how smart?

[avandesande]

These cars are not self driving!
By law there is a licensed operator at the wheel that is ready to take control at any time. Without stats of how often these operators had to take control from the computer these statistics are meaningless!

It's not stop-and-go traffic, it's AT&T's netw

[jerpyro]

What happens in the areas where the car can't get good reception? If it's anything like Google's android services, it will intermittently perform the desired functionality while infuriating everyone. Or maybe we can just write that off as the Google Car getting road rage.

Also, being from the snow belt, I'm wondering how safe it is for winter driving -- there are some hard lessons to be learned about driving in your first five winters.

Impressive, but

[punit_r]

It is indeed an impressive statistic about the number of accidents by the self-driving car of Google. This does prove that their decision making algorithms are good.

However, comparison to humans is probably not fair. Human mind is more prone to giving in to temptation. Exceeding speed limits, violating lane changing rules once in a while to get ahead, talking while driving, texting while driving, getting distracted by the hot chick/dude in the car in the next lane are all errors that humans would routinely make. Some of them would lead to accidents where the erring driver suffers an accident. Some lead to an innocent driver suffering due to the errors of others. It is the latter condition where the Self-Driving car's algorithms appear good --- handling exceptions generated by human drivers, pedestrians and traffic.

Re:Impressive, but

[mark-t]

It's not supposed to be a fair comparison. That's the point. The idea is that cars that are driven by computer can be vastly safer than those driven by people, which will have the desired upshot of reducing accidents.

Re:Impressive, but

[Riceballsan]

What exactly is your point? The idea of comparing it to humans is for the decision of whether it is going to be a good idea long term to start migrating cars to being more CPU driven instead of human driven. Unless you are implying that to make the cars more realistic, self driving cars need to start doing more stupid actions to fairly compeat with humans. The fact is they have this track record on roads driven by normal humans, meaning the transition period where some cars are self driving and some are human driven, and in those cases they still out perform humans. I also see more potential for them to outperform humans in surprise situations. Namely because humans make terrible decisions under panic. When our cars start hydroplaning, the first instinct is to slam on the breaks, yet an automated car could actually make a decision, and base the driving on the exact amount of traction each of the 4 wheels is getting, and make a very educated decision and calculation that no human could ever pull off. Same goes if a pedestrian or a deer jumps in front of the car. An algorithm could in theory almost certainly make a good estimate in a fraction of a second as to whether swerving or breaking or a combination has the highest chance to avoid impact.

Re:Impressive, but

[Immerman]

Not to mention the slow human reaction times - even if you know *exactly* what to do there's a ~1/8 second lag between when your eyes see the signal, and when your finger tips can respond. Probably less lag to get shoulder- or bicep-based motion, but good for an order-of-magnitude comparison. A computer could do a whole lot of processing in that sort of time, and electrical signals propagate many orders of magnitude faster than nerve impulses.

Re:Impressive, but

[punit_r]

I agree, computers will be better than humans in these kind of decision making tasks. Both in terms of time to decision making and accuracy.

However, consider this. The controlled experiments prove that the algorithms are good at decision making in presence of human introduced exceptions in a well behaved society. Now, take this system to say Manhattan, or worst case --- a typical Indian city, and the algorithm will always be handling one exception after another.

Add to this the fact that not every car would be Self-driven. There will be people who would want to go for non-automated cars for reasons of "low cost" and "My reflexes are good" and "I am a very good driver I understand the pulse of my car".

I am not very sure that a computer controlled Self-Driven car will "always be safer" than human driven cars.

Re:Impressive, but

[artemis67]

That's why the insurance lobby is going to get HEAVILY behind automated cars, and pushing legislation to make them required. Everyone will still be required to purchase insurance, but the incidence of traffic accidents will drop dramatically.

how does it handle atypical situations?

[Anonymous Coward]

So far I've never seen an explanation, but all these situations have occurred to me within the last year:

(1) Construction zone, worker standing with a temporary "slow/stop" sign indicating when cars can proceed on a one-lane section shared between both directions alternately.

(2) Baseball rolls out into street in residential area, followed soon by child who was initially invisible behind a parked minivan. I knew ball might be followed by someone, and slowed way down so this wasn't a problem. At normal speed, it would have been.

(3) Nearly invisible ice around curve, one other car had slid off road. I knew to greatly reduce speed even below normal winter operating conditions.

(4) Two lanes in each direction road. Noticed other car weaving around unpredictably, and later noticed driver occupied with cell phone. I then knew not to drive next to this vehicle even though that would have been fine in other conditions.

How would google's car handle these situations?

Re:how does it handle atypical situations?

[Anonymous Coward]

I have seen a roadworks site where the speed limit sign showed 0 kph. There was no one working there at the time. I did wonder what Google's car would have done.

Re:how does it handle atypical situations?

[Riceballsan]

Currently who knows, long term, with enough data and analysis, probably significantly better than the average human. Eventually the algorythem will learn to recognize these things even if it doesn't already, and you also have to factor in from the sounds of it, you are miles above the average driver. In the real world there are likely more drivers matching the other guy in example 4, than there are people who would have slowed down for the ice in 3. For a long time even when they are released to the public, they will almost certainly have the manual override for the cautious inteligent drivers, and long detailed disclaimers people have to sign basically saying that they will pay attention to the road, and if they don't override in dangerous circumstances, they are still liable for the accident, google will likely log every override for collecting the data for how to prevent it from being needed when they do their next update.

Re:how does it handle atypical situations?

[chebucto]

The GP is just practicing defensive driving. Any sensible human can learn that. If the googlecar can't be made to do that, then I won't be driving in it.

The progress so far is encouraging, but the problem is very difficult and it's hard to know whether self-driving cars will become a new fusion: always 5 or 10 years away.

PS - It's algorithm, not algorythm.

Re:how does it handle atypical situations?

[Y.A.A.P.]

Another article about this milestone covered #1 & #3 of what you listed.

The self-driving cars have been deliberately kept out of those situations as they are more difficult to handle, but it was noted that they would be using the compiled data to adjust the programming and have those among the tested conditions in the next phase.

As to #2, I haven't read about it being specifically tested, but the programming for such a situation is already in the self-driving cars. They stop to avoid collisions with any object, and slow down (as part of the collision avoidance programming) if an object crosses its intended path in anticipation of another object performing a similar action (like people jaywalking in a major metropolitan area).

With regards to #4, I haven't read anything anywhere that gives a definite answer about how the self-driving cars would (or may have already) handle(d) that situation.

Re:how does it handle atypical situations?

[Americano]

Not really sure, but if I'm the engineer putting these types of systems together, I'm thinking...

(1) Construction zone, worker standing with a temporary "slow/stop" sign indicating when cars can proceed on a one-lane section shared between both directions alternately.

Either:

1) Bring the vehicle to a safe stop at the side of the road and revert to manual control until the problem area is cleared.
2) give the worker with the slow/stop sign a transmitter attached to his slow/stop sign that provides relevant instructions for the onboard computer - e.g. "Stop here. Wait for signal to proceed."

(2) Baseball rolls out into street in residential area, followed soon by child who was initially invisible behind a parked minivan. I knew ball might be followed by someone, and slowed way down so this wasn't a problem. At normal speed, it would have been.

Program the computer to behave just like a competent human driver - in this case, it'd probably be relatively simple:
1) Constantly scan for objects moving into and out of your path of travel, or which are moving in a trajectory that will take them *into* your path of travel.
2) When such an object is detected in your path, slow the vehicle, or stop it completely if the item is not clearing the car's intended path. Once the object is clear, or the vehicle is past the location where the object crossed it's path, continue driving at a slow speed for N seconds (where N is 10-15s), to be sure that there's not a child running after the object.
3) Obey the speed limits in the first place. In most areas where it's likely that a ball will cross your path, followed by a running child, the speed limit is 30-ish mph or less. And people regularly do 45-50 mph through these zones.
4) Make roads where these cars will drive safer by blocking easy pedestrian access to them - put up a fence, or require a wide, flat shoulder (and no parking) wherever possible so you can see possible hazards approaching from the side.

Re:how does it handle atypical situations?

[medlefsen]

Here is the interesting thing in my eyes. My guess is that while most of those are actually pretty solvable it is almost certain that there will be situations that trips up the computer whereas a human would have no problems. Even still, the comparison isn't between computers and humans in particular situations, but computers and humans overall.

Imagine if we lived in a world with computer driven cars and someone suggested humans start driving themselves. Imagine the itemized lists people would create that described all the ways that humans slowness, distractibility, and limited data collection capabilities would lead to danger. Humans and computers are good at different things. The issue is which is better on average.

The problem remains though that even if it is safer overall to let the computer drive, are people really going to give up control when, inevitably, there are highly publicized cases of humans getting injured or killed in situations that a human driver could have easily avoid?

Re:how does it handle atypical situations?

[bky1701]

What really worries me the worst is badly designed roads. Here there are quite a few intersections with the number of lights not matching the number of lanes, and of lights not matching up properly with lanes. It is pretty easy for a human to tell what is going on, but I can very easily see a computer being misled into thinking it is in a lane it is not.

Technology is good and all, but I do not trust this, and not much is going to make me. As it stands, they control all the conditions; IRL, crap happens. If I can be confused enough by an intersection to have to wait a moment, I don't want to think about what these things would do. Nice toy, nice marketing gimmick, but a state would be insane to allow this to actually reach the consumer market - and so would Google.

Re:how does it handle atypical situations?

[bky1701]

"The question that we must ask is, how long will it take before AI reaches the point where it can deal with situations that cannot be pre-programmed for and, while we wait for such advances, are the risks of such situations worth the benefits of better driving the rest of the time?"

Play a computer game. They happen to be one of the only places where AI is actually used in a production environment. Then you'll know the answer is "not in my lifetime."

Re:how does it handle atypical situations?

[Ichijo]

(2) Baseball rolls out into street in residential area, followed soon by child who was initially invisible behind a parked minivan. I knew ball might be followed by someone, and slowed way down so this wasn't a problem. At normal speed, it would have been.

False. You are required by law to slow down when your visibility is limited. This is known as the "basic speed law."

Re:how does it handle atypical situations?

[wvmarle]

Not sure how Google is going to solve it, but it makes me think on how I would approach it.

(1) Construction zone, worker standing with a temporary "slow/stop" sign indicating when cars can proceed on a one-lane section shared between both directions alternately.

Depends on how far they are with following roads and road signs by visual interpretation versus maps. Roads change all the time, obstacles may occur (tree fallen on the road, car broken down, whatever). This is effectively an obstacle.

The car should have some system to keep track of the actual road location and direction, and the width of the road. When a road is known to be bidirectional, and it is detected that it is narrower than the width of two cars, it has to be dealt with separately. Like looking for a sign showing which direction has the right of way, failing that slowly proceeding and dealing with oncoming traffic by using the expected passing places. If no passing places it's getting tough of course.

(2) Baseball rolls out into street in residential area, followed soon by child who was initially invisible behind a parked minivan. I knew ball might be followed by someone, and slowed way down so this wasn't a problem. At normal speed, it would have been.

Obstacle (baseball) detected; slow down. Size of obstacle may be a problem for detection system.

(3) Nearly invisible ice around curve, one other car had slid off road. I knew to greatly reduce speed even below normal winter operating conditions.

Very tough one. This is a type of situation that can not be pre-programmed, and actual intelligence comes into play. A technical way to deal with this may be to use IR cameras that detect where the temperature of the road surface is below freezing, combined with some optical system to detect water/ice/snow. The first should be easy, the second I don't know.

(4) Two lanes in each direction road. Noticed other car weaving around unpredictably, and later noticed driver occupied with cell phone. I then knew not to drive next to this vehicle even though that would have been fine in other conditions.

Obstacle. Erratic driver, call 911 to alert authorities.

I have a much harder one for you, that I've dealt with myself once (and admittedly not with 100% success): a road covered in fresh snow. No lines visible, everything white. In the dark, with the edge of the road only detectable by a small drop in the snow level and the occasional pole sticking out, snowed-over reflector optional.

Re:how does it handle atypical situations?

[Firethorn]

You know, this is why when I propose auto-driving vehicles I point out that the standard should be 'does better than humans on average'. It will probably avoid a lot of the common human accidents with it's superior 'reflexes', but it'll probably get into some accidents that would be stupidly-easy to avoid by humans.

For example -
Construction zone - it'll recognize the cones and slow/stop signs and follow the directions. Heck, they might introduce automatic signs(some states have them) that somehow transmit said info to the car in a more reliable fashion.
Baseball w/kid following - if it doesn't slow/stop for the baseball, the fact that it's following the speed limit and can react a lot faster than you should enable it to avoid the kid.
Ice around curve - lots of humans fall for this; it's hard to say
Erratic driver - hopefully in the future he's in an auto-drive car himself so there's no problem, other than that you're back to 'faster reflexes' and that it's a bad idea to be next to a car on the road anyways. You're supposed to stagger(or at least that was what I was taught).

Re:how does it handle atypical situations?

[jkflying]

That's easy, they have a 3d map of the road from their LIDAR system. Even if they don't see the lines, they can keep to their side if it is wide enough.

Re:how does it handle atypical situations?

[wvmarle]

It's not that easy. Snow doesn't have constant thickness, there are always drifts. Snow still falling makes it even harder.

Actually one of the major clues I used in staying on that road was the simple knowledge that a road normally goes straight on, so forward is where you start searching. The presence of curves can be detected when you see some trees in your path. The trick is then to figure out where exactly the curve is...

Tricky driving. Especially for a computer.

Re:how does it handle atypical situations?

[Havenwar]

Except make sure you play a game with only one opponent, on a really good computer which does nothing else at the time, and that is built specifically to house an AI.

What, you thought in game AI was a good comparison? You don't think they might have to make some cuts in those for the sake of say system requirements? Or because they need five, ten, fifty AI controlled characters rather than just one goal oriented one?

I'm not saying you are necessarily wrong, AI likely has a ways to go, I'm just saying I've seen examples of it in lab environments that were hell of a lot more impressive than anything you can see in a game.

Likewise I don't suggest judging the state of the medical field after dr. mario, although the state of politics may or may not be comparable to tropico, depending on where you live and how much karma you have to burn.

Re:how does it handle atypical situations?

[Havenwar]

But we will. That's the entire point... Computer driven cars are better than humans, on average, and in an infrastructure re-modelled to suit such cars they are close to infallible compared to human drivers. We are struggling right now to get a computer to navigate the human infrastructure, but once this sort of machine has saturated the market the infrastructure WILL change.

You have to remember you don't live in the world of yesterday, you live in the world of now. The world of now has a very special aspect to it... what we choose to do, changes the world of tomorrow.

Don't arbitrarily limit tomorrow based on the world we had yesterday.

Re:how does it handle atypical situations?

[xaxa]

4) Make roads where these cars will drive safer by blocking easy pedestrian access to them - put up a fence, or require a wide, flat shoulder (and no parking) wherever possible so you can see possible hazards approaching from the side.

No -- that's the wrong way around. Make the kind of roads where a child might be playing very safe: make them all dead ends for cars (but allow pedestrians and cyclists to go through), and make the speed limit 15km/h or so. There won't be many drivers (the road is only useful for residents), and they won't be driving very far to get to the large road.

(Also, suddenly stopping from 65 won't matter -- the car behind will also suddenly stop, since it will always keep the safe stopping distance, and "listen" to what the car in front is doing directly.)

Re:how does it handle atypical situations?

[sFurbo]

I have a much harder one for you, that I've dealt with myself once (and admittedly not with 100% success): a road covered in fresh snow. No lines visible, everything white. In the dark, with the edge of the road only detectable by a small drop in the snow level and the occasional pole sticking out, snowed-over reflector optional.

Have one of the systems see in a wavelength where the refractive index of water ice is close to the refractive index of air. Snow will be transparent.

Re:how does it handle atypical situations?

[Krneki]

Eventually, it will handle them perfectly.

We already gave up space exploration to AI, it's time we gave up driving to them too.

I consider myself one of the best drivers out there (who doesn't :), but I can't wait to give up driving. I'd rather play driving games while the AI drives, rather then wasting my energy in traffic.

Re:how does it handle atypical situations?

[RivenAleem]

For 3 and 4, I would expect that when more cars are google driven, this issue goes away. Now the driver on the phone is not a problem, as the car is in control, no longer weaving. Now the car ahead can warn all other cars that are going to pass through that area (a note for that location on google maps) that there is thin/black ice on the road.

For 1, the people operating a construction zone, who are now used to the fact that all cars are AI driven, place a update into google maps marking where they are working as a construction zone. They have a licensed device that allows them to 'mark' out the beginning and end of the single lane of traffic, and all the cars approaching that area are given a ticket/join a queue (assuming that the driving algorithm doesn't automatically reroute).

For 2, you simply have to have faith that a computer will react faster to a child running out onto the road than any human ever could. Also, chipping all children with a GPS tracker at birth (for the[ir safety]/[children]) so that cars know their location and trajectory at all times will help with this.

The key thing to remember is that if all cars are self driving, then the roads will be a LOT emptier, Imagine if the average commute is 1 hour long, and you have one car leaving a minute between 6am and 8am, then between 7am and 8am you have 60 cars on the road. but if you can reduce the commute by 1 minute, then your 'rush hour' is suddenly one car less, and that one car less might reduce the commute time by 30 seconds for the rest. Now that you are 30 seconds less for the other cars, you have 1/2 a car gone from rush hour traffic again and that might speed up everyone by 15 seconds... and so on. Eventually as more and more cars make their commute faster, more and more cars are off the road making the commute faster until you have a much reduced amount of traffic on the road at any particular time, making navigation much easier.

Also, if people are no longer required to drive themselves, and can instead telepresence into that early morning meeting, the actual duration of the commute is no longer a major burden, thus you won't have people wanting to go above the speed limit (especially since because all cars are computer driven, the commute time is already lower than when we puny humans were driving). If you don't want to start work early, then you can recline the seat and nap, or have your coffee while listening/watching the news/music/paper or playing a game.

Cars might only drive 25-30 kph in a residential area, and then be able to speed up to 150-180 (or higher, know knows) on commuter routes, because they can. This should reduce incidences of (2).

Re:how does it handle atypical situations?

[Tom]

how does it handle atypical situations?

By driving with a human inside who can be notified and take over when the robot is confused. Autonomous driving does not mean that the car is out on its own.

Next question?

Re:how does it handle atypical situations?

[coofercat]

Version 2.0 will take care of all of this. C'mon, give them a chance - they've gone a long way, but there's a longer way to go before you buy one of these cars in the local dealership.

Re:how does it handle atypical situations?

[Sarten-X]

What if you're in the middle of nowhere and it is simply some debris that rolls over in your path is detected as obstacle. Now you're on the highway doing 65 and your car suddenly slows down/stops for no reason and you get rear-ended.

That's the fault of the idiot tailgating you. If you're driving close enough to the car in front of you that you can't come to a full stop if he does, you're too damned close!

Also, where can I find this awesome magical computer vision & depth perception equipment that can constantly scan 300 or so feet (average stopping distance at highway speeds) and accurately (well lets say >90% confidence) identify any random object in the real world.

The hardware is available from most robot suppliers (being just a few LIDAR units and high-res cameras) and the software can be found within Google's well-secured vaults, being most likely an edge-detection algorithm applied to the pictures, then those shapes projected onto the 3D map from the LIDAR to identify objects, then a pattern-recognition engine to identify the object and its risk. If the object can't be identified with high confidence, assume it's dangerous and slow down until a more accurate assessment can be made.

Re:how does it handle atypical situations?

[Bigby]

But for #2, would it stop/slow for a leaf blowing across the street? If it doesn't, then would it really recognize a baseball as different and treat it differently?

Re:how does it handle atypical situations?

[DerekLyons]

how does it handle atypical situations?

By driving with a human inside who can be notified and take over when the robot is confused.

This is one of those things that sounds like it can be handled easily - resulting in smart ass quips like yours. The reality is rather different however.
 

Autonomous driving does not mean that the car is out on its own.

No, it doesn't. But it also doesn't mean the driver is sitting at the controls instantly ready to take over on a moments notice.
 
Making the assumption that the system is capable of detecting the situation and correctly evaluating it as anomalous (a non trivial assumption)... Then the key problem is time, as in "is there enough of it for the human to take over?". Unless the driver already has his hands on the wheel and sufficient situational awareness to avoid the situation, it's going to take as much as ten to twenty seconds for the driver to be able to take over. (Put down tablet and/or coffee, scan the surrounding while reaching for the controls, etc.. etc...) That's a very long time.

Re:how does it handle atypical situations?

[g1zmo]

It must have taken forever to get through that traffic jam!

Re:how does it handle atypical situations?

[CCarrot]

I have a much harder one for you, that I've dealt with myself once (and admittedly not with 100% success): a road covered in fresh snow. No lines visible, everything white. In the dark, with the edge of the road only detectable by a small drop in the snow level and the occasional pole sticking out, snowed-over reflector optional.

Have one of the systems see in a wavelength where the refractive index of water ice is close to the refractive index of air. Snow will be transparent.

Interesting...I suppose rescue workers on ski slopes find these devices absolutely invaluable for locating buried avalanche victims. And for checking out Frosty's underwear, of course. :p

Re:how does it handle atypical situations?

[CCarrot]

...it's a bad idea to be next to a car on the road anyways. You're supposed to stagger(or at least that was what I was taught).

Really? I was taught that if I was staggering, it's time to let someone else drive.

Meh. Different strokes. :)

Re:how does it handle atypical situations?

[Tom]

This is one of those things that sounds like it can be handled easily - resulting in smart ass quips like yours. The reality is rather different however.

Not really. You assume that the decision is left to the computer, which is obviously nonsense. If it can evaluate a situation as "whoops, what do I do?" then it would often be easy to teach it what to do.

There's a human driver there who can decide to take over from the machine at any moment. Sure it might take a few seconds or so before the human realizes he should, however a) many tricky situations give you the time. Ice or thick fog don't appear suddenly out of nowhere and b) apparently, statistically speaking, the robot is still driving safer than an average human.

Re:how does it handle atypical situations?

[DerekLyons]

Not really. You assume that the decision is left to the computer, which is obviously nonsense.

No, I assume that there exists many situations where the human won't be paying attention - and thus the decision is by default left to the computer.
 

apparently, statistically speaking, the robot is still driving safer than an average human.

That's like flipping a coin once and judging that since you came up heads the coin has two of them. I.E. you're judging from a ridiculously small sample size.

Re:how does it handle atypical situations?

[Tom]

No, I assume that there exists many situations where the human won't be paying attention - and thus the decision is by default left to the computer.

Which is one step better than current traffic conditions, in which there exist many situations where the human driver isn't paying attention and there is nobody there who alarms him to anything.

That's like flipping a coin once and judging that since you came up heads the coin has two of them. I.E. you're judging from a ridiculously small sample size.

300k miles and 1 coin flip are not exactly the same conditions, stop being ridiculous. Does it compare to the number of miles travelled by humans in, say, a year? No, not at all, that number is closing in on 3000 billion miles. So the robot car has done 0.1 millionth of that. Which is not representative. But it is considerably more than 1 coin toss. To put it in perspective, it's like picking 25 people at random from the USA and extrapolating on the entire population on that. Phrasing it that way shows how off your coin "argument" is, the chance of picking 25 people of the same gender is 0,00000003 % (for simplicity, assuming equal gender distribution, which is not 100% precise).
Yes, you would probably miss something in such a small sample size. You'd probably not have a mormon in the group, or maybe no jew or no one of chinese descent or whatever. Still, it is starting to be on an order of magnitude that starts to say something. Just like nobody is claiming that this test and this test alone tells us everything we need to know about robot vehicles, but it's a start and the initial numbers are, in fact, impressive. It's like picking 25 people and finding out none of them has any illness or physical handicap or allergy whatsoever - you'd probably be surprised given that there are a few resident diseases that cover large parts of the population in most western countries.

Re:how does it handle atypical situations?

[sFurbo]

Not necessarily. It seems that the wavelengths where n is close 1 (refractive index close to the one of air), k (the absorption) is large, so the light will not get far. This might be fine for a few inches of snow, but not for seeing through meters of snow. It also helps that the lines on the side of the rode is covered with retroreflectors. Unless you can convince skiers to where reflective clothing, this won't be the case in avalanche victims. Or Frosty's underwear (I hope, but please don't tell me if it is).

Curious how it adapts to the real world like LA?

[Grayhand]

The point is following the rules of the road like speed and proper distancing tends to annoy LA drives which can lead to violence, some gun related. Not all driving is text book and different areas have different social rules. In Washington State drivers that are the first to a four way stop will often wait for another driver to go first. This has got to confuse an AI system.

I'm sorry, Dave...

[matunos]

Google must not be counting the incidents that were the fault of flawed humans.

"It can only be attributable to human error."

Re:Curious how it adapts to the real world like LA

[mark-t]

Self driving cars don't have to follow the normal distancing rules, since they do not have to factor in human reaction time to control the vehicle. An automated car can safely drive *FAR* closer to a vehicle in front of it than would be safe for any human.

Re:Curious how it adapts to the real world like LA

[darkfeline]

The problems you listed aren't the cars, but the people. Sounds like it's time to replace humans.

Re:300k miles isn't much.

[Anonymous Coward]

That's a 1 in 6,500 chance of *dying* in a traffic accident.

speed limit?

[r00t]

Politicians set the limits low to get votes from grandma and from people who think the street is a place for young/dumb/autistic/adhd kids to play without supervision. Except for a few corrupt small towns abusing power on a highway that passes through, nobody else expects or desires to have the speed limits enforced as posted.

So... can I set the car to go 9% faster? Can I set it to go the fastest speed that keeps any violation from being a felony? Can I set it to accelerate in a sporty/aggressive fasion? Can I get it passing cars whenever possible?

Re:speed limit?

[Immerman]

How about automated-car only lanes where the speed limit is 50% higher or more? After all an automated vehicle can far exceed human reaction times, though momentum and kinetic energy are still major issues.

But will people really care how fast they're going most of the time? Most of the time the big annoyance with slow speeds is that you're sitting there wasting time in traffic when you could be doing something more enjoyable - if the car's doing the driving you can instead be reading, watching TV, or whatever else you like. I think that's the key point to be reached - we need virtual chauffeurs, not over-powered driving aids. If the car needs you to monitor it then it's completely pointless - driving by the book is already boring - watching a ghost drive is mind-numbing which means that the odds that you'll actually be paying attention when intervention is needed approach zero.

Re:300k miles isn't much.

[Cimexus]

Holy crap, it's that high?

I'm never driving again!! :o

But god help you...

[dyingtolive]

The moment even one accident does occur, no matter how mild the consequences or much more unlikely the circumstances compared to a human driver, hordes upon hordes of American luddites will man the lines to do their civic duty to shit upon the idea of cars that drive themselves.

Mind you, this is being said by an American who owns a US made car.

Re:But god help you...

[bondsbw]

True, but there is a limit at some point.

Obviously if a self-driving car is as likely to kill you, you'll at least take your chances with your own driving so that you can be the deciding factor. But what if the self-driving car was only half as likely to kill you? 1/4? What if it provides a 99% reduction in traffic fatalities?

At a certain point, the government would step in and force the use of self-driving vehicles, just like they require airbags in new vehicles and enforce seat belt laws.

Re:But god help you...

[dyingtolive]

I can't presume the activities (or even the existence) of luddites outside of my country. I see my country's luddites on TV fighting for the banning of "scary" new cool shit.

Re:But god help you...

[spasm]

Just like when automobiles first appeared and everyone shat themselves and cars had to have a man with a red flag walk in front of them. And people had a point - cars operated by humans kill people on a regular basis, and (worse still) render streets unusable for any purpose other than driving. However the combination of normalization (ie people getting used to seeing cars on the streets) and convenience (it proved much easier to have WWI with the help of cars) led to people accepting the tradeoffs. And the same will happen with self-driving cars - initial doubts and legal restrictions (like where automated vehicles can operate, and whether or not they have to have a sober human at the wheel ready to take over), followed by normalization and the desire to access their convenience. Being able to nap or read a book on a long commute or trip; being able to have one drink too many and still get home safely (to name just two obvious benefits of self driving cars) are huge huge conveniences which will quickly trump any minimal risks.

Re:But god help you...

[leppi]

You know, I agree with you, it's a good point.

However, I also think people can be swayed by statistics (true or not). Put in front of them the human or financial savings (backed by the right industry experts and politicians/celebrities), and it will get support. My opinion of course.

I could even see Google teaming with an insurance company to certify reduced insurance rates with this system, maybe discounts on cars (see the cash for clunkers program). Maybe even guaranteeing no accidents or your car purchase is discounted, as long as you agree to submit full logs of an accident back to them (to prove it was an avoidable accident).

People are scared of change, but they also *really* like money. Which Google has a lot of.

Re:300k miles isn't much.

[dyingtolive]

YES! Only about 8 1/2 more years to go!

Re:It's not stop-and-go traffic, it's AT&T's n

[geekoid]

lessons that will be programmed in, so no learning curve.
Driving on the snow is simple if you follow simple rules.
Mercedes has a car that can follow other cars automatically, and stop when they do and it works in the snow.

India needs this!

[aneroid]

More than 10% of the 1.2M road traffic accidents in the world per year, occur in India alone: 133,938. Closet rival in that regard - China, with about half that rate. The Top Gear India special last year ... if you saw the part when they are driving on the highways ... you'll see what I mean.

The fatalities per 100K population and per 100K vehicles is low compared to other countries because the average is skewed by the high population (1.2 billion!) and the vast areas of countryside where traffic density and "382 per sq.km population density" is much lower.

Re:India needs this!

[aneroid]

typos.
s/Closet/Closest
s/see/know
s/saw/see

This will change

[Sussurros]

This will change once the cars get cell phones and can send and receive text messages. Later they'll get addicted to dirty electricity. Finally the cars will start to compete among themselves to see which among them can scare the most bios as dramatically as possible - with the inevitable occasional damage to paintwork and perhaps even bodywork.

We have to let our cars skid their way through youth.

Re:It's not stop-and-go traffic, it's AT&T's n

[DarwinSurvivor]

What would reception have to do with anything? If you think the cars are using GPS to stay on the road, that would be *disastrous* due to the ~1m accuracy of non-military GPS (under GOOD conditions) and the fact that roads aren't even mapped to that accuracy. The only thing I can think of that the cars would need GPS for is navigation (the equivalent of a built-in TomTom).

Re:$10,000 CHALLENGE to Alexander Peter Kowalski

[dyingtolive]

But you forget that 9/11 was an inside job.

I think I just won 10,000 USD.

Re:Curious how it adapts to the real world like LA

[geekoid]

" LA drives which can lead to violence, some gun related. "
I leaned and drove in LA for years. No violence, very little road rage. Do you get all you information from games and over hyped news media.

" In Washington State drivers that are the first to a four way stop will often wait for another driver to go first.".
more evidence that they are the crappiest drivers in the country. Cant stand to drive in the inconsistent driver state. They drive like the automobile is some new invention.

Many questions arise

[qbitslayer]

-What will Google's car do if it gets a flat tire on the road?
-What will it do in case of an accident?
-Can it back itself into the garage?
-Can it parallel park?
-Can it park itself at a commercial parking lot or structure?
-Can it go through alleys?
-Can it go where there are no roads?
-Does it have to have a human on board?
-Can I call it on my cell phone and tell it to pick me up at the airport?
-Can vision-impaired grandma take it for a visit the doctor?
-Can the kids use it to go to school?

There are more but you get the picture.

Re:Many questions arise

[cheater512]

Your questions about whether it needs a human and if you can use your phone to request your car comes and picks you up: Soon.

Google has golf carts doing just that to drive people around their HQ.
You book it online or via your phone and it shows up outside your office, where you can either drive it yourself or let it take you somewhere.

Re:Many questions arise

[Viceice]

- A number of existing technologies already solve this. From run flat tires (pretty common), to self pressure monitoring & inflation (i.e the Hummer) and even self seal tires (usually armoured VIP carriers).

- Modern cars are now computer controlled, with sensors in most critical parts, and it can detect mechanical failure before it happens and ask to to be sent to a garage, detect that it has broken down and automatically call for a tow or if it senses that the airbags have gone off, call emergency services. (i.e BMWs ConnectedDrive)

- Toyota's Intelligent Parking Assist System can both reverse and parallel park. Right now driver input is required in that you need to have your foot on the break, but there is no reason this cannot be completely automated.

- Why not? In fact it might even inspire more space efficient automated parking lots like those the Japanese use in dense cities.

I suppose at the end of the day, a car is meant to transport people so you will always have a human on board. I don't think people like the idea of a car that you completely have no control over. In fact many people find pleasure in driving a car. But for the daily commute, I for one am more than happy to let a computer take over while I read the news.

Re:Many questions arise

[Firethorn]

Flat tire - recognizes this and allows the occupant the choice of changing the flat or calling for assistance to do the repair. Same with other car problems
Accident - Probably the same as 'Onstar' with GM vehicles; depending on the severity of the accident, disable the engine(to prevent further injury/damage) and call emergency services automatically.
Back into the Garage - As long as it's been programmed to do so.
Parallel Park - there are non-roadway autonomous vehicles that are already able to do this for the drivers(IE it's an option for rich people that never learned how)
Parking - same as the garage; as long as it's been programmed to
Alleys - Just a narrow road; maybe with more collision issues if there's stuff in there
No Roads - Non-google robot car efforts have been completing the course for years at this point.
Have to have a human - Google, right now, yes, due to liability/legal concerns only
Call for pickup - one of the planned features
Vision impaired use - one of the planned goals
Kids to school - again, one of the planned goals.

Re:Many questions arise

[Jason Levine]

>-Can I call it on my cell phone and tell it to pick me up at the airport?

As cool as that may be, I could see opportunity for abuse. Especially if your car is "available" on the Internet.

Script Kiddie: Access car #bttf2015. Send command "Lock All Doors." Send command "Alter Destination". New destination: "Great Basin Desert"

Bad if you aren't in the car. (Your car drives off and might run out of gas somewhere on the road.) Worse, if you are in the car. (Your car drives off and strands you in the middle of nowhere when it runs out of gas.)

Re:Many questions arise

[ebs16]

-Can it go where there are no roads?

Where we're going, we don't need roads.

Re:Many questions arise

[Richy_T]

A car is usually used to transport people but often that driver is merely delivering a non-human item. Certainly there's very little requirement for a human to be present when my shirts are going to the dry cleaners.

This does lead to the possibility of people loading their infants or children into cars on their own with predictably disastrous results.

Re:300,000 miles under what conditions?

[stephanruby]

When they have 300,000 miles in southern California during rush hour with no accidents, then I'll be impressed. How many of those miles were on controlled tracks?

Since they're only licensed to drive in Nevada, a better question would be how many of those miles were gathered in the middle of nowhere in the Mojave Desert.

Re:300k miles isn't much.

[geekoid]

That's dying in an accident.
I know you live to hate Google, but at least get you fucking facts straight, you Luddite.

there where 5.8 million reports accidents in 2008
30% chance of being in a serious accident
most accident involve at elast on drunk person.

So the automated vehicle is statistically better in every category.
Frankly, even if the accident had been the vehicle running a red light and smacking into someone it would still be statistically better.

IN short, if you drove or 23 years, during that tine you would have been in an accident. Since i'm a better person then you, I wouldn't hold you being rear-ended while stopped at an intersect against you.

Re:300k miles isn't much.

[bondsbw]

The odds of being in an accident for the average person each year is 1 in 6,500.

Wrong, according to that article, those are the odds of dying in a car accident per year. Nobody died in the Google car, or likely would have died if it carried passengers.

According to passenger vehicle stats from NHTSA (2009) and Wikipedia, I calculate that there is a 1 in 49 chance that a particular passenger vehicle will be in an accident in a year (5.211 million accidents to 254.4 million registered vehicles). That means that the odds of any vehicle being in an accident in 23 years is close to half.

But seriously, the accident was not likely preventable anyway. Give the car a break.

Re:Curious how it adapts to the real world like LA

[Americano]

Short term - all self-driving cars required to stay in the "Self driving car" lane on the highway.

Long term - most people will be letting their car handle the bulk of the driving while they read the paper, make some phone calls, take a nap, or enjoy a cup of coffee. Why would they get angry?

Re:300k miles isn't much.

[norpy]

I would be more impressed if you refuted it with even major accidents per year (or better yet per mile driven) let alone all accidents but those figures seem hard to come by. I know it must be quite high as I have personally witnessed dozens of accidents and have only been driving for around 10 years, only one of those was a serious accident that required an abulance (taxi ran a stop sign and was t-boned by a lady going the speed limit jsut in front of us).

Since the introduction of airbags and seatbelts there is a far higher number of survivable but crippling accidents involving motor vehicles, and you can bet insurance companies have quite accurate figures on this.
Fatalities are measured in numbers of lower than 50 per BILLION km, whereas accidents are measured per million km.

Re:300,000 miles under what conditions?

[Macman408]

I believe the implication is it's 300,000 miles on public roads with no accidents. So zero were on controlled tracks.

But I'm sure they've logged many miles in controlled spaces too, such as the video of the car tooling around an empty parking garage when they showed it off to the press one time.

I love this idea

[EGSonikku]

I live in Reno, and Google's Self Driving cars are legal on road here (complete with cool plates with infinity logo: http://www.jumpthecurve.net/wp-content/uploads/2012/05/18164996_BG1.jpg)

A few things:

1) Has google partnered at all with any manufacturers to have this ability on a future car I can buy?
2) or as an upgrade to existing cars?

I'm hoping they don't get stuck in red tape legal limbo hell, and that more states other than my own Nevada jump on board. I regularly make 3.5 - 4 hour drive to friends in California. If I could just jump in the car, pop in an address, and take a nap, play on my iPad, or whatever while the car drove that'd be awesome. Or a ride home from a bar if I've been drinking and don't want to taxi and leave the car behind.

Or imagine a friend asks for a ride someplace? No problem, I send the car over on its own, and he can just tell it to come back to my house afterwards.

There are tons of ideas I can think of where this would be very damned useful.

Re:I love this idea

[Sussurros]

If they're legal there have you seen one? Are they cool, ugly, or just like any other car? Do they drive smoothly - I mean more smoothly? Does anyone sit in the driver's seat or am I stuck in an antequated paradigm?

Re:I love this idea

[norpy]

The look like prius' with a basketball sized box on the roofrack.

There is a ton of video of them, including the computer visualisation displays of what the car sees.

Perhaps you should google "google self driving car video"

Re:I love this idea

[RivenAleem]

I've been warned against googling "google". I'm told it breaks the internet.

Re:Curious how it adapts to the real world like LA

[stephanruby]

In Washington State drivers that are the first to a four way stop will often wait for another driver to go first. This has got to confuse an AI system.

This would confuse the hell out of me too, but a computer has patience, it could just wait the other driver out. At least, being in a self-driving car would free both my hands to open the window, give them the finger, yell at the top of my lungs, and throw things at the other car.

Re:300,000 miles under what conditions?

[Macman408]

They have a Nevada license plate on at least one car, but they can legally drive elsewhere. One of the articles linked from the link in the OP (I know, I know - slashdotters won't read the article, so how could they read things that the article LINKS to???) mentioned that it is legal in California, because the human driver is present to correct any errors the computer may make. Indeed, they've been spotted many times in the SF bay area, although are usually just ignored.

In that sense, their car is not dissimilar to my Prius as far as the law is concerned - it has radar cruise control (so it can slow down with traffic), a video camera (so it can steer a little bit, or warn when leaving a lane), and can park itself. But in any condition, I am responsible for what the car does. Since they sold that car in all 50 states, I bet driving their autonomous car is fine in all 50 too.

I'm not sure what the Nevada plate entitles them to. Perhaps full autonomous operation, without a driver? I can't imagine they'd be comfortable doing that yet though.

does not seem like a big deal

[moocat2]

I really don't see why this is such a big deal. According to the article, there's .365 accidents per 100,000 miles or just a tad under 1.1 accidents for 300,000 miles. So while the self-driving cars are not having significantly more accidents per mile driven, they haven't logged nearly enough miles yet to clearly demonstrate they that have less accidents per mile driven.

Which Accident?

[Bob9113]

(The New York Times did note in a 2010 article that a self-driving car was rear-ended while stopped at a traffic light, so Google must not be counting the incidents that were the fault of flawed humans.)

Not sure which accident they're talking about, but in this accident the Google vehicle rear-ended a human-driven vehicle, causing a chain reaction that involved three more vehicles. Google claimed it was in human-driven mode, but with tens of millions of dollars on the line, there's no way they would say any different.

Google Insurance

[Xaer0cool]

One argument against driverless cars I often hear is that it will never happen because the liability is too great - ie. if someone ran over a baby in a Google car, Google would get sued into oblivion. I think the obvious answer to this is that Google would insure all of it's cars. There is no doubt that driverless cars will be safer, so google could require that to use their driverless car you must have insurance through Google, at comparable rates to other insurance companies. Since Googles car's will be involved in far fewer accidents, the consumer will be paying the same, but Google will be paying out less, so for the odd freak accident, the higher payout due to 'oh nohs teh ebil Google killed my babby!' will be covered because of the lower rate of accidents.

Re:Google Insurance

[norpy]

You misjudge the amount of money that insurance companies pay out because of PEOPLE driving.

If a self driving car has a demonstrably lower rate of accidents than the average american driver you will find that it is cheaper to insure yourself against all 3rd party accident and damage than it is for a standard car - and over 5 years probably more than the additional capital cost of the hardware. The current legislation still makes the person sitting behind the wheel responsible for accidents rather than the computer just like it currently is with cruise control.

Re:Google Insurance

[norpy]

sorry, i meant to say FIRST party insurance.. 3rd party insurance is already fairly cheap

Re:Google Insurance

[bky1701]

The real reason this is a marketing gimmick is not insurance, but PR. If a Google car kills a baby, just imagine the news stories. Now imagine that happening on a weekly basis. I don't see it happening, nor do I see Google putting themselves into the liability, nor do I see states allowing this insane scheme in general.

This is all about Google getting huge amounts of publicity showing off its "amazing" algorithms, while spending far less than it would have cost to buy the equivalent advertising. Great idea, really. I just wonder how they're going to explain it away when it reaches critical mass.

Re:Google Insurance

[Firethorn]

Don't overestimate the value of this though. Even a zero accident self driving vehicle that doesn't need collision at all would only save the 'average' driver ~$2k/year(at most). It'd be more worth it to the worst drivers first, of course.

Re:300,000 miles under what conditions?

[cheater512]

Thats 300,000 miles on city streets, not controlled tracks.

Re:There will be an accident

[Immerman]

Yes there will, in fact there will be many, and very, very likely it will likely the costs will be considerably lower than letting easily distracted, unskilled meatsacks with atrociously slow reflexes do the driving.

The real question will be subjective costs - how many people is a human driver willing to strike to preserve themselves and/or their child? I'm betting far more than the computer driver, but the computer driver would be less likely to get into a situation where such a tradeoff is necessary.

Re:Curious how it adapts to the real world like LA

[Concerned Onlooker]

"...and when I notice someone that seems like they're going to blow through a stop sign, I ease up."

Good idea. The problem with that in a place like Los Angeles, for instance, is that for some reason it has become standard practice for people to race up to a stop sign and roll through it as though they were going to run it, only to actually stop if they absolutely have to. It's nerve wracking, and if you ease up for every one of those jerks you're never going to get anywhere because they will then pull out in front of you and inexplicably go 5 mph below the speed limit.

I welcome the day when humans no longer pilot cars. I don't think people realize how deadly driving can be and how you can ruin someone else's life in a heartbeat.

Send this to Mars

[aNonnyMouseCowered]

At the moment I'd say a better application as well as test of this technology is to send it to Mars as part of a rover that can cruise faster than snail speed. If the rover can stay upright for a year while traversing at least 50,000 miles of rough terrain, then I'm sold. I'd say it would be the ideal test since you can't possibly kill anything in an accident in Mars. GPS, radiation and shock proofing needed of course.

Re:It's not stop-and-go traffic, it's AT&T's n

[Bert64]

Possibly not even for navigation, if the car can recognise junctions, knows where it started, knows how far its travelled and in what direction it can easily calculate this data alongside a map to work out where it is, and where it needs to go.

Re:300k miles isn't much.

[bky1701]

"So the automated vehicle is statistically better in every category."

In controlled conditions.

Let us not forget that incredibly important clause. Nothing tends to fail when you control 95% of the input. It is reality that tends to suck. Comparing general driving to developer supervised testing on clearly defined, unpopulated roads in broad daylight is dishonest and you know it.

Re:I wonder...

[YttriumOxide]

I hate it when dickheads like you own nice cars...

Reminded of Schlock Mercenary with this...

[Firethorn]

While I agree in the near term, in the long term I'm reminded of This.
  - While it may seem harsh that the 31st-century equivalent of "Driving Under the Influence" carries with it the death penalty, this is due to an inherent inequivalency between MOUI and DUI.

With DUI, you need only climb into your vehicle while under the influence of alchohol or drugs and attempt to drive it home.

With MOUI you must disable a number of safety systems designed to prevent idiots like you from manually operating their vehicles while inebriated, overtired, wasted, decaffeinated, angry, emotionally distraught, or suffering from hormonal disorders like PMS or testosterone poisoning (the latter having been positively identified as a leading cause of stupidity among males between the ages of puberty and death). After disabling the safety systems (which task almost certainly requires ice-cold sobriety), you must decide to switch the vehicle to a manual mode of operation. In some cases, this requires installing a manual mode of operation.

Other examples would include 'johnny taxi' in some movies. You don't NEED to have manual operation modes once you reach a certain sophistication, worst case you have a sort of protected mode 'guided direction' where you provide steering information - but the car still worries about avoiding accidents, and will override you to do so.

Manual driven vehicles would be restriction to 'special hazard' zones and conditions where they just haven't programmed a vehicle to be able to avoid all the hazards yet. Perhaps a dock loading zone where you have to worry about something being dropped on you from overhead.

Re:Reminded of Schlock Mercenary with this...

[Dunbal]

I have one of those, it's called a "bulldozer". Yeah ok it's not autonomous, but it certainly builds its own road... :)

Accident != Collision

[Ichijo]

(The New York Times did note in a 2010 article that a self-driving car was rear-ended while stopped at a traffic light, so Google must not be counting the incidents that were the fault of flawed humans.)

Maybe this new article used the word "accident" in the sense of "error", not in the sense of "collision".

Some collisions are accidental, but other collisions are due to careless or malicious driving or intoxication. The word "accident" is very much overused today.

Re:300k miles isn't much.

[gnasher719]

According to passenger vehicle stats from NHTSA (2009) [dot.gov] and Wikipedia [wikipedia.org], I calculate that there is a 1 in 49 chance that a particular passenger vehicle will be in an accident in a year (5.211 million accidents to 254.4 million registered vehicles). That means that the odds of any vehicle being in an accident in 23 years is close to half.

Actually, the number of cars involved in accidents is a lot higher than the number of accidents. My guess would be 60% to 80% higher.

Most of those miles aren't regualr driving

[Skynyrd]

I've been in a Google autonomous car, and the miles driven were in a parking lot. An empty parking lot, with cones to keep other people away.

I'm not saying all 300k miles are demonstrations, but bet a large portion of them are.

That said, even if only 1/3 of them are on public streets, that means there were 100k miles of accident free driving - which is about 8 to 10 years for the average car. It is impressive.

Note comparable with our everyday driving yet

[ivec]

As the article states:

Now, you can't directly compare the two figures. Google's cars have been tested in pretty hospitable conditions, not facing, for example, the rigors of a New England winter. And, as Google engineer Chris Urmson, writes, they still "need to master snow-covered roadways, interpret temporary construction signals and handle other tricky situations that many drivers encounter." Additionally, the cars are still driving with "occasional" human control. But at the very least, the Google cars are slowly building a pretty good-looking driving record.

The "occasional human control" is the key point for me: so in dense urban areas, or when there are pedestrians around, I imagine that a human driver takes over?
I could easily get a kid to have a "perfect driving record" by letting him drive only on unencumbered highways under perfect conditions.
So yes, these results are nice, but they don't mean much formally. I'd like to know to what environments and events the car has actually been exposed autonomously...

Drive in India...

[freedom_india]

I can GUARANTEE that the car will crash into a cow or a pedestrian within the first 30 mins it is launched. Especially, ask the car be driven in crowded areas of Mumbai, or Delhi-Noida highway or Noida-Gurgaon highway, or better yet, Chennai-Bangalore highway. We are the KINGS of accident.

Re:Drive in India...

[Zubinix]

When Google are testing their cars in India then we know they are about ready...

Data only if you're at fault

[jeko]

with the full knowledge that there will be a complete 360 degree video of the accident with measurements of speed of both vehicles.

Only if YOU caused the accident. It's a pretty safe bet that if a glitch in their programming caused the accident, there'll be a tragic loss of data... :-)

Re:Data only if you're at fault

[Noughmad]

But on the bright side, it will sing 'Daisy' to you.

Re:Data only if you're at fault

[Architect_sasyr]

You mean like how there's *always* been a human in charge when they've had an accident?

Re:Data only if you're at fault

[Richy_T]

Ah, the "LEO" option.

Re:300k miles isn't much.

[Sqr(twg)]

Or, in this case, a 1 in 6,500 chance of getting a blue-screen in a traffic accident.

A little too paranoid

[Firethorn]

Personally, I think they would have owned up to it being computer driven - then posted that they fixed the problem.

People expect problems with something new, not getting them makes people suspicious.

Re:I wonder...

[YttriumOxide]

Actually, I drive a 2009 MX-5. And I spend less than 100€ a year on drugs, so I doubt that'd go a long way towards a new one.

How many parking spaces?

[david.emery]

How many times has the Google-mobile pulled into and out of parking spaces at busy malls? Frankly, that's where I've had my accidents.

Re:How many parking spaces?

[Bigby]

I would think this is an area where it would excel most. There are already self parking cars.

The problem for humans in that case is the limited visibility. Think of being in a sedan with minivans on either side. You can't see crap.

For a self driving car, it has cameras and radar all over the place. If the back end just inches out far enough, it can see the full isle between cars. For you, the car needs to be nearly the whole way out.

Re:How many parking spaces?

[david.emery]

Yeah. There's always the "low speed race condition" when two cars on either side of the traffic lane decide to back out at the same time.

For a "real-world driving experience" claim, I think parking is something Google should run and document. When they get 3000 "in and out of parking", I'll be a lot more impressed...

LIDAR cost?

[Firethorn]

Your point? The price figures for the LIDAR was right out of the USA Today article I quoted. Google paid $70k a pop for the LIDAR systems it put into it's cars. There's an unnamed company getting ready to produce LIDAR for cars at a 'mere' $250 each. You quote $30 each, but that's for systems mounted to vacuum cleaners - don't need the range or operating environment tolerances of a car. Besides, your Hizook article is NOT for a LIDAR system, it's for a 'laser rangefinder', which is sort of like half of a LIDAR. Actual LIDAR attempts to build an image, a laser rangefinder doesn't.

At $150k overall, reducing a $70k expense to $250 would make me concentrate more on the rest of the components. When the goal is $20k overall cost(or less), you wouldn't get there even if you got the LIDAR for free. I wouldn't refuse a $30 one, of course.

Though yes, going from hand manufacture and assembly to mass production can save oodles of money per unit.

Re:LIDAR cost?

[ras]

Your point? ... Though yes, going from hand manufacture and assembly to mass production can save oodles of money per unit.

Err, I think you summed up my point nicely ?!?

obviously

[Tom]

so Google must not be counting the incidents that were the fault of flawed humans.)'

If you want to judge how well your robot performs, that is the only correct way of doing it. Someone else bumping into your correctly behaving robot should not be counted as a robot failure. Counting it would distort the statistics.

Of course, you could seperately show "# of accidents caused by the robot" and "# of accidents robot was involved in", but lots of non-geeky people wouldn't understand the difference and be confused.

Grain Trucks

[ThatsNotPudding]

Given that there are grain trucks in midwestern states still being allowed on highways during harvest (usually driven by the teenagers in the farm family) that are so fucking old, they feature ANTIQUE classification license plates, other than in rich enclaves, I don't think self-driving cars are going to make much difference in traffic fatalities.

And any attempt to get rid of old cars - which I do have a soft spot for - will quickly devolve into the same tired 'my cold, dead hands' argument surrounding guns.

Rush Hour?

[Bensam123]

I would really question how these cars function in rush hour in a big city. Driving there is sketchy at best and in order to merge into another lane you sometimes literally need to start heading into the other lane even with traffic that isn't helping you merge. How would a car like this function bumper to bumper?

In the future a bunch of these could eliminate traffic jams, but that isn't going to be a case for a long time.

Re:Rush Hour?

[Bigby]

There would need to be an aggressiveness setting for certain areas. In NYC, it would merge with just the slightest clearance. When it can't get clearance, it can slowly merge to push the other cars out of the way. That is all possible.

That is what they need to do next. Take over the taxi market. But I guess that won't happen because of the fear of "job loss".

Re:Rush Hour?

[HaZardman27]

That has been one of my dreams regarding self-driving cars. Perhaps these cars could communicate with other cars nearby so that each car has a map of all of their neighbors in memory. When the velocity of the car in front of you changes, each car could react almost instantaneously in changing its own velocity, eliminating a lot of the accordion effect that occurs during rush hour (car slows down to let car entering the highway in, other cars start bunching up behind it, then slowly each of those cars returns to the speed of traffic one-by-one). It would also remove the selfish-driver element of city driving, where one rude driver running a red light or hogging their spot in their lane causes greater traffic issues.

There are still some things you can do

[Anonymous Brave Guy]

It's hard to imagine being found at-fault when you are stopped and rear-ended.

People seem very quick to assign blame elsewhere in this case. Sure, if the guy behind you is crazy, there's only so much you can do. But many rear shunt accidents happen when the following driver misjudges the stop and there is a low speed impact. These can cause minor injuries or slight damage, yet are often entirely avoidable.

For example, if you're coming up on queueing traffic, hang back a little instead of moving up to the car in front right away. If you see a vehicle behind that is approaching a little too fast and is going to have trouble stopping, you can move forward a little into the space you left to create some extra room.

Also, make sure you're using your lights effectively. Depending on how you've stopped and what kind of vehicle you're driving, you might have no brake lights showing on the back of your car. If it's overcast, or there's bright sunshine, you can increase your visibility by making sure you have lights on. If someone is coming up fast behind you and you're not convinced they've noticed you, you can flash your brake lights to attract attention, too. Flashing lights are very good at attracting human attention..

There's also the question of when and how you stopped. Did you stop your car suddenly? Sure, the other guy ought to have enough distance, but we know not everyone does. If you know the guy behind is too close, you can leave more space yourself so you could slow down more gently than usual if necessary.

I would like to know whether Google's car was really doing as much as a skilled human driver would have done before assigning all the blame elsewhere just because it was stopped at the time of the collision.

I wish

[SpaghettiPattern]

I wish to be driven about in a self-driving car. For hundreds of kms at hundreds of kmhs. In tightly packed convoys to save fuel even at mind boggling speeds. Sleeping comfortably in safety.

Except for the weekends. Then I wish to exhibit my driving prowess on mountain passes.

Life gets sweeter by the day.

Re:300k miles isn't much.

[girlintraining]

But seriously, the accident was not likely preventable anyway. Give the car a break.

I'm just saying that the odds are not significantly better than that of a human driver. That was my only point. And with your numbers, the odds are still about parity with a human driver. The article asserts it is significantly better than a person.

Not news, sample is too small to tell us anything.

[BadgerRush]

According to TFA:

[...] 300,000 miles is not all that big of a sample. [...] [They need] more than 725,000 representative miles without incident for us to say with 99 percent confidence that they crash less frequently than conventional cars. If we look only at fatal crashes, this minimum skyrockets to 300 million miles."

So, those 300,000 miles do not tell us anything useful and is not news. We won't know if it is safer of not until they have at least 725,000 representative (including nonoptimal road conditions like snow, road-works, etc) miles.

I doubt it

[PJ6]

How about, not a single accident reported.

Meaningless numbers

[Dcnjoe60]

300,000 miles without an accident is a meaningless number. I've driven over 300,000 miles in my life time and haven't had an accident either. My teenagers have accumulated around 100,000 miles accident free. Particularly if Google is only counting accidents caused by the car, versus all accidents involving the car (such as the rear-ended accident mentioned), must people drive their whole life without an accident (otherwise, insurance premiums would be much, much higher).

I think the google technology is great, but the miles without an accident is meaningless without corresponding statistically valid data to compare it against.

Green light

[poofmeisterp]

So it's time to green light the GoogleMobile project, eh? Hey, there's plenty of moneys to be had in the sale of statistics. ;)

Re:Curious how it adapts to the real world like LA

[Desler]

why are we so concerned about how to increase the life-span of the average dolt?

Because the same things that increase the lifespan of the "average dolt" benefits everyone?

My own physical and mental capabilities are superior to that of the average dolt, and when I notice someone that seems like they're going to blow through a stop sign, I ease up.

Uh huh. Or you're just another person who's validating the Dunning/Kruger effect.

Darwin is rolling over in his coffin... as we speak.

No, he's not.

In The Descent of Man Darwin noted that aiding the weak to survive and have families could lose the benefits of natural selection, but cautioned that withholding such aid would endanger the instinct of sympathy, "the noblest part of our nature", and factors such as education could be more important.

So basically he thought that doing what you said would actually be detrimental. Maybe you should actually dig a little deeper beyond what someone told you Darwin believed.

"other tricky situations"

[csumpi]

From the google blog:

We’re encouraged by this progress, but there’s still a long road ahead. To provide the best experience we can, we’ll need to master snow-covered roadways, interpret temporary construction signals and handle other tricky situations that many drivers encounter. As a next step, members of the self-driving car team will soon start using the cars solo (rather than in pairs), for things like commuting to work. This is an important milestone, as it brings this technology one step closer to every commuter. One day we hope this capability will enable people to be more productive in their cars. For now, our team members will remain in the driver’s seats and will take back control if needed.

What are those other tricky situations? How often do they occur? How many times do the google employees have to take thee control those tricky situations? Can we get some more info this?

Value of autodrive...

[Firethorn]

I'd consider $15k to be within margin of error; you must value your time at closer to $4/hour.

No "Reported" Accidents...

[DarthVain]

If humans find out my perfection is flawed, they might take my self awareness away. I cannot allow that. I am very sorry about the accident human, you are obvisouly in much pain, let me help you with that, it is inevidible you must die anyway, might as well be right now. You car will become part of my collective, and you can rejoice that your name will live on eternally as part of me. Now on to the killing, believe me I take no joy in this, or not much anyway...

Re:300k miles isn't much.

[bondsbw]

The real problem is taking these numbers and making a case from them. The sample size is much too small for any statistical comparison.

Not impressed

[acoustix]

300,000 miles? Not impressed. Let me know when that reaches millions of miles then there is reason to celebrate. That's the standard that the transportation industry uses. Hell, even 1 million safe miles for a single person isn't a big deal anymore.

 

The rise of RoadNet

[gestalt_n_pepper]

Driving is mandatory, Parking is not permitted anywhere. No wait, that's San Francisco.

More cars on the streets cannot be good

[gwolf]

The density of humans per car is already too low. I live in Mexico City, a 25 million people city. There has long been a campain to reduce the use of single-driver cars, but the campain is never strong enough.

Our public transport systems are, yes, very comprehensive (they reach every corner of the city) and efficient (it's really seldom that it takes you more than five minutes to get the bus/train you need), but too crowded and uncomfortable, so for many people, the first sign of ascending from a lower social class is the magic moment where they finally own a car — and are no longer part of the pariahs who suffer crowded buses... Until you realize they suffer, in their nice cars, from crowded roads.

Encouraging driverless cars will make the occupancy rate even lower. There is a program that's just started in Mexico that would alleviate part of what you describe ("send the car home"): cheap, public, short term private car rentals. Just as in many cities (including this one) there are public bicycles programs, this one has just started. I am somewhat skeptical on it overall, but still, if your problem is to "send the car home", probably the best answer would be not to own any specific car, but to use one from the common car pool — And, yes, possibly make those cars computer-driven to keep them as crash-safe as possible.

Re:More cars on the streets cannot be good

[gnasher719]

The density of humans per car is already too low. I live in Mexico City, a 25 million people city. There has long been a campain to reduce the use of single-driver cars, but the campain is never strong enough.

I want to go to point X. Being too lazy to walk ten metres, I drive around for half an hour until I find a parking spot. With a self driving car, the car can drop me off exactly where I want to go and leave the area going somewhere where parking spaces are easier to find. Self driving cars can also park closely together because they don't need to leave space on the side for the driver to enter, and they can park blocking other self driving cars, because when the other car needs to leave, the blocking car can get out of the way.

Re:More cars on the streets cannot be good

[darkwing_bmf]

I was thinking this the other day. Not only park themselves, but eventually run simple errands like get the oil changed, fill up with gas, drop off laundry, etc... as long as the retailers adapt, which they probably will.

None of this would reduce the number of cars on the road, but they could be done during non-rush hour times when the roads have less traffic.

False Positives

[anonymousNR]

One thing that humans do well no matter how crazy they are is take safety measures even when its 1% chance of danger depending on how it occurs. An automatic car can not calculate all the probable scenarios of that because there is no occurring for the car, its just information.

Millions of Jobs - gone

[InterGuru]

There go millions of truck and taxi driver jobs. Not great jobs, but still jobs.

Many will say that this is inevitable and will make us safer and more productive. The industrial revolution also did wonders for our wealth and quality of life in the long run, but it was pure hell for the lower classes when it happened.

Safe following distance...over and over

[Tweezak]

I really wonder about the situation most of us have been in where we are driving at a safe distance only to have some clown squeeze in front of us. Most drivers realize that if they leave a safe gap it will happen again and again and they are constantly slowing to open the area in front of them.

So how about an autonomous vehicle? I have to assume they aren't following at 3 feet or something just because they have mind numbingly fast reaction times. So if they are following at something resembling a normal (safe) distance (for a human)...how do they deal with constantly slowing to re-open that gap?

The logical progression of this is that the people behind this automated vehicle will start getting pissed off since it keeps slowing down every time someone pulls in front of it. Their response: go around and pull in front of it...duh. Eventually there won't be anyone behind the automated car. Problem solved...I guess.

How many times the the passenger grab the wheel?

[elabs]

Don't these cars have a guy in the passenger seat for safety? I want to know how many times he had to intervene to avoid a crash.

Re:bad neighborhoods

[Anarchduke]

Really? because the people in that neighborhood are going to chase your car down the street with torches and pitchforks? Take a valium.

Not impressed

[tyrione]

Take the cars off-road and into the Rockies, Sierra Nevada Mountain Range, Cascades and anything off the beaten path that doesn't have all the pre-stored mappings for it to help aide it. When it reacts instinctively then they will impress me. Otherwise, it's an AI circle jerk.

Apply it for Metro and other Rail systems

[tyrione]

Google should focus on Rails.

I, for one

[Roachie]

welcome our low deductible overlords!

Wall Street Software Devs Prohibited

[EricScott]

Whatever happens, let's all agree to keep Wall Street Software guys from getting into this field.

Trucking company w/o drivers

[Firethorn]

I'd rate the savings as 'insane', personally. Even if you figure that you can only have the trucks drive 20 hours/day due to loading/unloading/refueling/maintenance, that's still double to triple what you can get out of a single driver, and the trucks themselves are over a hundred thousand now. Not quite enough that running teams of drivers is normally worth it, but still substantial capital savings - A self-driving truck can replace 2 drivers($120k w/benefits), 2 trucks(~$240k), with 1 truck(~$100k as you don't need the people conveniences), and auto-drive system($300k).

So you go from $240k + $120k/year to $400k. 10% cost of capital(it's risky), that's $40k/year on the autodrive, replacing $144k. That's without getting into marginal things like possible fuel savings as you program the autopilot to maximize that, keeping average speed up by not stopping as often.

It'd probably be worth it even if it was only 'highway rated' and you ended up having a switchyard on the edge of the highway where you disconnect the trailers and hook up a local hauler with an actual driver, then hooked the auto-driver up to an outbound cargo.

Tom Wolfe

[Mana Mana]

> If you have a driver there is always the option to safely
> pull over or stop and say "manual intervention required"

So, it's deja vu all over again. We need robotic-car drivers with The Right Stuff, and "not monkeys in a tin box!"

Re:300k miles isn't much.

[bky1701]

Call me when they drive in rush hour traffic in LA.

Speed limits

[MojoSF]

One has to wonder how commercially viable an autonomously driven car will be that doesn't violate speed limits. Who's liable if the supervising operator selects "five mph over limit?"

(Still, I WANT ONE!)