Google: Our self-driving cars would be tip-top if you meatheads didn’t crash into them

Google's self-driving car engineers have released their first progress report, and they said they will publish monthly updates from now on. "We’ve made a lot of progress with our self-driving technology over the past six years, and we’re still learning," the report [PDF] stated. "Every day we head out onto public streets so we …

[Jeff Lewis]

"You're driving it wrong."

Seriously though, the real question is: would these accidents have been less likely with a human behind the wheel?

From the sounds of it, not really. About the only iffy one is the rear enders, mainly because people drive by pattern and if the person behind the gCar was thrown off of unexpected behaviour by the gCar, they may have reacted incorrectly - which might not have happened if they were behind a regular car.

[Charles 9]

But then you read the part at the bottom about the bicycle swerving in front of the G-car, a textbook example of unexpected behavior, yet the G-car reacted correctly and AVOIDED it.

[Blank Reg]

Sure, but can the self driving car anticipate the "unexpected" behaviour? Quite often a real driver can tell what another driver might or might not do based on how they are driving,where they are looking/not looking and just their general driving style. I've avoided plenty of accidents because I could see another driver was an inattentive idiot so I was prepared even before they pulled their bonehead move for the day.

[Anonymous Coward]

Let's give it their first law: Always behave as if the other vehicles are out to kill me. Worked for me as a cyclist, motor or bicycle. In one case I was run off the road in a rocky part of Southern California by this idiot lady. Her son had died riding a motorcycle. The motorcycle California Highway Patrol officer was not amused. Neither was I but that's beside the point as I was never called for the trial.

Get programming people!

[Charles 9]

That's a thought. A computer-driven car can be programmed to assume the worst: that a car might suddenly stop in front of them, swerve into them, cut into the narrow gap you normally leave for the first instance, assume the end of a blind curve can be roadblocked, and so on. And make all the car's driving actions work under those assumptions. That way you don't need cues to be prepared for trouble: you're prepared in any event.

[Neil Barnes]

@Charles9 - the problem is that taken to extremes, the car would sensibly decide never to exit the garage.

[Lee D]

Interesting game. The only winning move is not to play.

[Pookietoo]

Re: you're prepared in any event.

So the car leaves a safe gap behind the car in front, which is promptly filled by another car, so it has to slow to establish a safe distance again, and another car cuts in ... all the while it's decreasing the gap and increasing the closing speed of vehicles behind it.

[fruitoftheloon]

@ Pookietoo: Re: you're prepared in any event.

Pookietoo,

Yay, but that is focussed on vehicles that have five rings on the bonnet/hood...

Cheers,

Jay

[Truckle The Uncivil]

Inevitable consequences? Manual drivers will take advantage of computer-driven by ignoring them to a certain extent knowing that they will get out of their way. Computer-driven cars will then take longer for the same trip...

[Christoph]

"Let's give it their first law: Always behave as if the other vehicles are out to kill me."

But can it obey the Second Law?

When all else fails, and whatever you do you're going to hit something. try to hit something cheap.

[J Bourne]

When all else fails, and whatever you do you're going to hit something. try to hit something soft.

There you go fixed that.

[petur]

exactly, I encounter idiots not obeying stop signs every day. The reason I've not had an accident (yet) is because I can guess pretty well if such an idiot is coming up. Is called experience....

[Anonymous Coward]

Yup...

As a biker, you develop a "sixth" sense as to the behaviour of other road users very quickly - either that or you become road-kill.

I've had my fair share of accidents over the years (and yes, some were my fault), but I've avoided far more simply by being able to anticipate when that driver in the 4x4 with the "Child on board" sticker is about to do something stupid.

Many of these things will be down to almost imperceptible behaviour, something that will be impossible to encode into the self-driving car simply because as humans, we would find it very hard to describe quite why we knew the muppet was about pull that bonehead move of the day.

[Schlimnitz]

Re: Yup...

Shouldn't be too hard:

if(CarInFront.Is(Make.Audi) || CarInFront.Is(Make.BWW))

[graeme leggett]

Re: Yup...

"if(CarInFront.Is(Make.Audi) || CarInFront.Is(Make.BWW))"

But this is google so advertising revenue dictates that CarInFront=Toyota

[Dr. Mouse]

Re: Yup...

As a biker, you develop a "sixth" sense as to the behaviour of other road users very quickly - either that or you become road-kill.

I agree. It is amazing how much the realisation of imminent pain/death can improve your perception.

My bike instructor put it this way: If you have an accident in a car, you'll dent the bodywork. If you have one on a bike, it's going to f*****g hurt! This is a great motivator to be aware of your surroundings, notice the guy on the roundabout who has not spotted you, back off approaching a blind bend, and beware of sheep who think that the best place to be on a foggy day is sat in the middle of the road.

[IglooDude]

Still there may be an advantage to the GoogleAI: it constantly checks for texting and call behavior from all nearby cars and via comparison with registration information identifies vehicles with smartphone-distracted drivers so as to maintain increased separation.

Or at any rate, that's why Google says they want all that data, right?

[Thought About IT]

Cyclists

"One cyclist veered in front of the car at night, and the software was clever enough to stop immediately to avoid a crash."

I should bloody well hope so! That's the very least capability a self-driving car must have.

[Christoph]

Re: Cyclists

If it's using radar it can even detect the stealth cyclists who have done everything they can to make their bike invisible - no lights, no reflectors, dull-coloured clothing, etc. I came within inches of hitting one of those because I only spotted him at the last moment.

[Anonymous Coward]

Re: Cyclists

"If it's using radar it can even detect the stealth cyclists who have done everything they can to make their bike invisible"

Maybe. But I wonder how well the radar works with the carbon fibre bikes ridden by true cycling fundamentalists?

[Charles 9]

Re: Cyclists

Decently well, I would think. Radar doesn't rely on metal, and to stealth a vehicle requires a combination of radar-deflecting design and radar-absorbing paint, and they'd still be of limited use in multistatic (they'd spot the dead zone against the background) or mobile (the case here, it can hit differing angles) detectors.

[Anonymous Coward]

Re: stealth cyclists

I wouldn't have thought they'd bother anyone for long - being only milliseconds from death, their numbers must be decreasing pretty rapidly. I mean, they're no more visible than not-particularly-visible pedestrians, and if they dared venture out on the roads there'd be a bloodbath.

[Anonymous Coward]

Re: stealth cyclists

I wouldn't have thought they'd bother anyone for long - being only milliseconds from death, their numbers must be decreasing pretty rapidly. I mean, they're no more visible than not-particularly-visible pedestrians, and if they dared venture out on the roads there'd be a bloodbath.

I like your pure Darwinistic view of the world. Of course, the insurance may look at this differently (and so would the stealth cyclist) but as a long term problem solving strategy it's flawless.

[JLV]

Re: stealth cyclists

maybe he was a fan of Stephenson's protagonist in Zodiac - "if they can't see they can't hit you" school of thought.

I'd like to know how the google cars compare, statistically, to human drivers for same mileage & same conditions. And let's not forget statistical confidence, mostly wrt sample size.

[TeeCee]

Re: Cyclists

Actually it's a capability that only self-driving cars have.

By the time a normal driver reacted to that, it's likely that the idiot cyclist would already be under the wheels.

[Anonymous Coward]

New driver

Wonder how this compares to new drivers?

As far as the stats are concerned are new drivers more or less likely to be involved in an incident? Is it because new drivers drive to rule or make poor decisions?

[Voland's right hand]

Re: New driver

Pretty bloody good.

If memory serves me right, the probability of a new driver to have an accident in the first year in a metropolitan area is > 50%. This is for London which is of course infinitely worse than mountain view.

As far as Two more were down to other cars not obeying stop signs, my late dad used to have a couple of sayings: "There is a special alley in the graveyard for those that thought they have priority" and "I have priority, but does the moron on the other side of the intersection know that".

[Anonymous Coward]

Defensive driving

The arrogant writer has his blame game on. Google's approach seem to be getting the rest of the world to conform to its ideas about roads and how cars should behave, so that their ideas will "work" - but that won't work because, unlike the search space, they have no monopoly. What they didn't report here was that on their first cross-country trip, they "discovered" that the road surface markings were different in each state, read; the guys developing this haven't a clue about the larger world of driving outside of their home town, which is typical of CA parochialism, and they want you to trust your life to them. They can barely manage well paved orderly thoroughfares. They don't have a monopoly on the roadways, although they're trying desperately to buy influence in Sacramento in order to get permission to further menace the already menacing freeways of CA with yet another home-brew custom car.

[Richard Jones 1]

Re: Defensive driving

There are several types of contributors to this section, those who feel they have additional insights to a situation, those who have alternative experiences and those who seek additional information. I was unable to decide which you tried to be. I was left with the impression that you simply wanted to expose your prejudice and preconceived idea of how the world should be.

Put simply if you are starting from any location, it does not matter which you will initially profile that location not say what New York or for that matter New Delhi has if you are in California, Surely it would be basic common sense to start of with the right profile? Have done that and got the thing working, expansion to areas where all sorts of variables add complexity would be the next stage. Should new profiles be kept as a monster data base or paged in and out according to how many specific data points are noted. Road marking may well vary from State to State, that is a variable that will need to be covered, the style of trees the dampness from bone dry to water logged, the closeness of buildings indicating congest town or open country all have to be mapped in some way to build a risk profile. Hopefully new drivers live long enough to acquire the skills that will allow them to take such issues in their stride. Returning for a moment to road markings, they may well be part of a set standard per local variation. However, wear and road works can also affect their presence and appearance as can the presence of high sided vehicles and other variables. All such variables do have to be considered when controlling a vehicle. Knowledge of wind speed should also alert drivers to the risks of different situations, how many do take account of such issues - not many if they are of the wetwear variety that I see weaving about in cross winds. Some, (many?) cannot even detect changes in the road layout ahead causing precipitous manoeuvres when roads narrow. One ended up facing fully the wrong way after braking too hard while doing two or three other changes of direction. They did not hit anyone.

A while later close by similar stupidity resulted in five deaths. No autonomous driver-less control required, that was down to wetwear meathead driving.

This is a research project, for such someone to complain that research finds things out is does not show a high level of understanding of why research is carried out. I may be the only one who expects research to discover aspects that are not obviously apparent to desk jockeys, but someone will probably tell me I am wrong on that point.

I did once avoid being rear ended on a country road, I stopped due to stationary farm traffic, the meathead behind was talking, not looking I could see they would not stop so I drove onto the grass verge, my movement drew the other driver's attention to the road, he braked and steered out towards oncoming traffic stopping two cars in front of me along side the stationary traffic. Would his error have happened with a Google car - I doubt it would.

[Anonymous Coward]

Re: Defensive driving

Put simply if you are starting from any location, it does not matter which you will initially profile that location not say what New York or for that matter New Delhi has if you are in California, Surely it would be basic common sense to start of with the right profile?

Interesting that you're fully prepared to overlook the fact that Google happens to have world's largest database of road layouts in Streetview.

[fruitoftheloon]

@ Idiot ac: Re: Defensive driving

Dear Ac,

Having considered your narrative, may I make a few observations:

- if googles plan to develop autonomous cars is so shit, why don't you show them where they went wrong? I am sure they would appreciate your sage thoughts

- were it not for homo sapiens pushing the boundaries (and occasionally tripping over them) WE WOULD BE STILL LIVING IN CAVES watching TV by candlelight...

It is called 'evolution'

Btw how did your last Nobel committee nomination go??

Regards,

Jay

[fruitoftheloon]

@1980s coder: Re: @ Idiot ac: Defensive driving

I was being ironic (perhaps too ironic...)

Also I will lookup a nipkov disc tomorrow, it sounds jolly interesting.

Have one on me.

Cheers,

Jay

[Charles 9]

Re: @1980s coder: @ Idiot ac: Defensive driving

"Also I will lookup a nipkov disc tomorrow, it sounds jolly interesting."

While you're at it, look up "Magic Lantern", which actually did use candles at first.

[fruitoftheloon]

Charles the ninth: Re: @1980s coder: @ Idiot ac: Defensive driving

Charlie,

Will do, thanks for the pointer!

Cheers.

Jay

[Phuq Witt]

Re: @ Idiot ac: Defensive driving

"...Eh? So we would have developed television and yet still be living in caves?...

Whoooooooshhh!!!

[fruitoftheloon]

Phick with: Re: @ Idiot ac: Defensive driving

Pw,

Yay, with you there bro, one takes pleasure from the fact that someone else f'ing got it...

Have one one me, btw would you believe it, my favepub in da village, £2.50 per pint..

Cheers,

Jay

[Kirk Northrop]

Re: @ Idiot ac: Defensive driving

Or the TV that uses a light bulb. Oh, hang on. TVs don't use light bulbs. Silly you.

[Anonymous Coward]

Re: Defensive driving

The arrogant writer has his blame game on. Google's approach seem to be getting the rest of the world to conform to its ideas about roads and how cars should behave, so that their ideas will "work" - but that won't work because, unlike the search space, they have no monopoly

I noticed the downvotes here, but in support of your statement it's worth observing that other, non-Google efforts don't appear to have Google's problems of realising roads may differ from those at home. This may be because those companies already build that rather important part, the car, and thus have a far more developed insight in what a car may encounter on a global basis because they already sell globally and as a consequence had to deal with roadside issues.

Not that that is forgivable for Google: as it created world's largest database of actual street footage in Streetview I would like to hear how it is possible that it appears they entirely ignored that in their development.

[Meerkatjie]

Re: Defensive driving

To me it sounds like they are starting simple and adding complexity as they need too - the benefit of this is if they only add more complexity once the previous stuff is working then they know in which area they need to look to find the problem.

There is a possibility of painting yourself in a corner with this approach so it will be interesting to see how they (or anyone else building self-drive cars) deal with this.

[PassiveSmoking]

Re: Defensive driving

Wow, just... wow.

How did you find time out of your busy schedule smashing up spinning Jennys to write this?

[T. F. M. Reader]

Are humans more cautious around them?

Are Google's self-driving cars clearly marked? I don't know, but I assume they are. Even if not, I suspect the folks in Mountain View (where they test the cars) can recognize them by now.

How does that change the behaviour of other (human) drivers nearby. If you see a self-driving car a bit ahead in the next lane, would you instinctively steer away to reduce the chance that something unexpected happens? An experienced driver can usually guess the intentions of other drivers by observing them - we all do that instinctively. If that Toyota in the next lane looks suspicious, let me give it some space, eh? If you don't know how the robot will behave, will that make you behave more carefully? And if you get used to it?

In other words, is there a bias that makes accidents involving Google self-driving cars less likely?

I have never driven next to a self-driving car, so I have no idea how I would change my behaviour.

[Anonymous Coward]

Re: Are humans more cautious around them?

I would have thought most people would take the opposite approach, driverless cars will be fair game for cutting up, tailgating, pulling out in front of, etc, even fun games seeing how much you can force them to take evasive action, swerve dramatically and so on.

[mathew42]

Re: Are humans more cautious around them?

While I can see the fun in playing chicken with a google car I suspect google would have very good documentary evidence of your behaviour.

In local town, the school bus drivers had a simple arrangement with the local police. If a car was seen driving carelessly around the bus, the local constabulary would pay a visit to the driver and scrutinise the car for defects. For most kids their first car was barely roadworthy so finding defects was trivial.

[Anonymous Coward]

Re: Are humans more cautious around them?

"Are Google's self-driving cars clearly marked?"

Well, if the article photo is accurate, the only thing you might mistake a G-car for is this

[Destroy All Monsters]

Re: Are humans more cautious around them?

the only thing you might mistake a G-car for is this

Wait until it deploys katana-armed robotic arms.

[DropBear]

Re: Are humans more cautious around them?

"I have never driven next to a self-driving car, so I have no idea how I would change my behaviour."

For a crude approximation, s/self-driving car/police vehicle/ and you have your answer.

[Zog_but_not_the_first]

This is just typical of what I've come to expect from El Reg

Refusal to take a press release at face value, snarky comments and good questions.

Keep it up.

[Len Goddard]

Assumptions

If all the cars on the road were automated in the same way and could thus make reasonable assumptions about the behaviour of other vehicles, then automated cars might well be safer. Unfortunately that is not the case and is unlikely to be so for some time (if ever) so for the forseeable future the only safe assumption is that every other road user is a homicidal lunatic (can you tell I used to ride a motorcycle?) and so any half-way reasonable AI would decide the safest thing to do would be to stay at home.

Seriously, a good driver relies not just on his eyes and ears but on a whole host of subliminal clues, some of which are probably below the level of concious thought. I once avoided a very nasty accident because I could smell the diesel fuel spill on the junction in front of me even though I could not see it. I doubt the googlemobiles have asomatic sensors. (For those who have never encountered it, diesel fuel on tarmac is at least as slippery as wet ice).

[TeeCee]

Re: Assumptions

The diesel smell thing only works on a motorcycle anyway. Impossible to smell through a charcoal-impregnated pollen filter and damned difficult even in a car without one, as you're just nowhere near as up-close-and-personal with the air around the vehicle.

The Googlebots are no worse off than the rest of us (when on four wheels rather than two) here.

Personally I reckon that, when on two wheels, I'd be far happier if cars were autonomous than not. The autobots are never going to do the not-seeing-me-even-when-looking-straight-at-me thing that meatsack drivers are so fond of and if I do fuck up and miss spotting one, it's far more likely to react immediately to cater for my stupidity than the meatsack is. I also suspect it's infinitely less likely to suddenly turn right without indicating.....

[DropBear]

Re: Assumptions

"The autobots are never going to do the not-seeing-me-even-when-looking-straight-at-me thing that meatsack drivers are so fond of"

If our experience so far with machines thought us anything at all, it must be that machines are exceedingly good at not making many of the errors humans make; they unavoidably do, however, come up with brand new and unexpected ways to fail.

[fruitoftheloon]

@Len Goddard: Re: Assumptions

Len,

I am with you on that one matey.

A while ago I had a gratis motorbike lesson, basically to see if I had the basic coordination needed.

The instructor asked if I had ever riden a motorbike at all, my answer was "no, but I have regularly cycled through central London", his reply was "you'll be fine then..."

Cheers,

Jay

[Doctor Syntax]

Re: Assumptions

"If all the cars on the road were automated in the same way"

But would they be? Some would be on v1.0, some on 1.1, some on 2.0, some on SP1, some on SP2, some wouldn't have had the update applied for some reason...

[IanDs]

A good driver does do all sorts of things that an autonomous car would have difficulty doing as well.

But most drivers on the road are not good drivers, they're inattentive rule-ignoring phone-texting showing-off unskilled meatbags who crash with monotonous regularity. Replace these drivers by even reasonably competent autonomous cars and the accident rates will plummet -- yes there will be situation they can't deal with, but these are few and far between in comparison.

It's easy to conjure up such "autonomous cars can't cope but I can" situations, but if these are massively outnumbered by the "autonomous cars don't get tired or drive like idiots" situations -- which they surely will be -- there will be far fewer accidents and deaths on the roads.

Then the insurance premiums for robot cars will drop, and those for the few remaining meatbag drivers will go up massively as they have far more accidents -- and this cost difference will go up and up as the robot cars get better. Eventually only the rich will be able to afford to drive their own cars -- maybe this will become as unacceptable as smoking is nowadays when it becomes obvious to everyone how big the risk to other (innocent) people is...

(biggest non-disease cause of death nowadays : car accidents)

[Anonymous Coward]

I would imagine the insurance premiums somehow reflect the underlying risk. Just by having a significant proportion of faultless robo drivers won't make the meatbags drive more recklessly thus raising premiums. There would just be fewer accidents and fewer payouts and reduced insurance 'take'.

Unless of course the robocars drive like a short sighted 80 year old and piss off all the meatbag drivers causing them to drive recklessly to get past them.

[Destroy All Monsters]

Just by having a significant proportion of faultless robo drivers won't make the meatbags drive more recklessly thus raising premiums.

You seem to assume that the insurance market is a monopoly. Not so. "Do you own a robo-car? Our premiums are 30% less than the competition, and we will thrown in a free lawyer in case of accident against meatbags. Call 1-800-DRIVERATOR now!"

[Charles 9]

"A good driver does do all sorts of things that an autonomous car would have difficulty doing as well."

Can you cite some specific examples of things that humans could do easily that no reasonable amount of machine sensors and training could do as well? Because I strongly suspect there's actually very little true intuition (probably the one thing machines can't replicate) in driving and that it's mostly a matter of subtle cues we're trained to recognize: cues that a well-sensored machine could be trained to notice as well.

[Grumpy Fellow]

The Snow Belt?

One aspect that they might be missing out in California is driving in a snowstorm, especially at night. It really changes everything when most of your vision is filled with an almost opaque wall of snowflakes flowing toward you. I'm not sure exactly how I see where I'm going in the snow. I follow the tire tracks or tail lights ahead of me if there are any. Otherwise I try to distinguish between the drifts and the bar ditch as best as I can. If there are phone poles I try to remember how far away from the road they are. I once had to turn around leaving a small northern plains town in a snow storm because once the phone poles ran out, there was nothing at all to drive by. Another thing that an autonomous car might have trouble with is getting around a break down or accident where someone is directing traffic by hand. I hope that they program the cars to not close off side driveways when approaching a traffic light. The car should stop short and flash the lights to let a car out, not close up the gap and block them in. For some reason, when I read about autonomous cars, I always flash back to the movie Airplane, with Otto the Autopilot.

[Charles 9]

Re: The Snow Belt?

"One aspect that they might be missing out in California is driving in a snowstorm, especially at night. It really changes everything when most of your vision is filled with an almost opaque wall of snowflakes flowing toward you."

Well, in regards to inclement and especially extreme weather (like the blizzard you describe), there's a fair chance of the car outseeing the human because with the imaging technology available to day, to say nothing of down the road, the car can likely "see better than the average person". Now, I'll grant you GPS at its current state probably isn't accurate enough to distinguish between one side of the road and the other, but perhaps with infrared and radar imagery, it'll probably have a better shot of seeing through the snow and getting an idea of where the road's supposed to be.

What you did at the end also indicates what the ultimate failsafe should be for a car that is unable to navigate. If, in spite of all its abilities, it cannot find the way forward (such as not being able to discern the road due to heavy snow, a washout, or an unexpected road change, it should find someplace safe, alert the driver that it cannot continue, and perhaps suggest requesting assistance. As a last resort (if going forward is a must), it could turn the matter back over to the driver while advising proceeding with utmost caution. That way, at the least, if something happens, it's not nearly as likely to be a big something.

As for your breakdown issue, a smart car should be able to discern it as an obstruction. But I'll grant you maneuvering will be difficult if traffic is reduced to one lane for both directions as a result. The best solution would be for a policeman to direct traffic, and their dress codes can be updated to make their signals easier to distinguish.

[IanDs]

You're correct that it's all a matter of observation and anticipation and intelligence -- currently a robocar is probably better at the first, not so good at the second and certainly worse at the third (assuming the mythical "good driver").

Even as this improves with time and programming there will always be a small number of unexpected cases not covered by the program where a good human driver will do better, but the number of these will fall continuously as the robocars get better.

However the vast majority of "accidents" (the police don't call them that any more) aren't caused by these rare cases and/or good drivers, they're purely predictable and caused by bad drivers inattention/stupidity/texting, and in all these cases robocars will do way better than humans very soon -- or even now, looking at the roads every day.

[Medixstiff]

"But most drivers on the road are not good drivers, they're inattentive rule-ignoring phone-texting showing-off unskilled meatbags who crash with monotonous regularity."

My friend used to be a smart *ss and ask if we had hit 100KM/h when I used to take off from the lights and get into the left lane, then hit cruise and watch everyone behind me.

However once he went through the Police academy and started doing the same thing, he didn't like it when I started asking "are we up to 100KM/h yet?" and defensively went on about being told that most accidents happen at traffic lights and being taught to get a head of the idiots and pull into the left lane"

Funny how he does something and it's right, I do the same thing and I'm a hoon or worse.

[Intractable Potsherd]

I'm not sure about the insurance premium thing myself. I have a feeling that auto-driving cars will be adopted by a large percentage of the people who cause accidents - those who drive only because they have to. The ones who actually hate driving, but for whom the alternatives are poor. The ones who really aren't paying attention, or don't have sufficient confidence (how many readers here know at least one person who will not drive off their "usual" routes? I know two - neither of them will drive on motorways because they don't like merging with other traffic, nor will they drive in the nearest city (despite a perfectly good ring-road and well-managed traffic)). The ones that would rather be texting/talking/working than driving. The ones that our motorcycling colleagues complain about.

Good drivers will be good drivers regardless, though they (we) will die out because people won't have chance to develop the skills.

[Zack Mollusc]

Google Cars will NEVER cause an accident

Google Car hurtles into orphanage, many orphans die.

Google Lawyer: 'We have many billions of dollars and also our telemetry shows that the orphanage was entirely at fault'.

Judge : 'Not Guilty. Orphanage will pay Google's laughably inflated legal bill. Next!'

Usher : 'Next case is Google v Tour de France Peleton.'

Google Lawyer: 'We have many billions of dollars and also our telemetry shows that the cyclists were entirely at fault'.

Judge : 'Not Guilty. Widows of the Tour de France Peleton will pay Google's laughably inflated legal bill. Next!'

[Charles 9]

Re: Google Cars will NEVER cause an accident

"Google Car hurtles into orphanage, many orphans die."

Show just how something like that would happen more often than with a human driver (and I've PERSONALLY witnessed a drunk driver jump a parking bump, ricochet off a support pole, and drive INTO the front door of a C-store) and without outside help (such as being forced off by a human driver).

"Next case is Google v Tour de France Peleton."

How would a Google Car be permitted on a Tour de France course? And what about the spectators that are between an outside car and the course itself? And like I said, what's to stop a human from doing the same, only more frequently due to inattention, inebriation, or both?

[Anonymous Coward]

Re: Google Cars will NEVER cause an accident

You know this is theregister.co.UK don't you? ;-)

Parking bumps.. Support poles.. C-stores? What the heck are those? :-D

[Charles 9]

Re: Google Cars will NEVER cause an accident

The bump (or stop) is the thing about 3 feet wide and 6 inches high that's supposed to stop a parking car from going too far and going over the pedestrian-only sidewalk.

http://i.imgur.com/wiOdIGY.jpg

A support pole is one that may be holding up the front overhang of a store, keeping it from falling on you.

And the "C" is short for "convenience". As in SPAR or best-one on your side of the water. Imagine a crazed driver jumping the kerb and crashing into a SPAR, and you'll get an idea of what I saw at a 7-Eleven.

(Not the incident I saw, but still a good example)

http://chicago.barstoolsports.com/files/2013/03/711.jpg

[Anonymous Coward]

Re: Google Cars will NEVER cause an accident

"The bump (or stop) is the thing about 3 feet wide and 6 inches high that's supposed to stop a parking car from going too far and going over the pedestrian-only sidewalk."

Bollards.

...as in that's what we use in the UK to stop such things. Between that and most on-street parking being parallel rather than perpendicular to the kerb.

[Anonymous Coward]

When the cars get autonomous enough

Car to the owner: "- I won't enter the streets today. Too much traffic. Order a taxi or take a bus."

[Anonymous Coward]

Re: When the cars get autonomous enough

Basically a fuchikoma.

[Anonymous Coward]

To those here hailing the arrival of Google cars

It's OK to be afraid to drive or to acknowledge your insufficient driving skills, you may go and use one of these. I'm not buying your argument that you are an excellent driver and all others are idiots so all of us should be transported/controlled by mighty Goo(gle).

Me, I'm confident I can drive myself a car and I'm planning to do it for a long time.

[Anonymous Coward]

Re: To those here hailing the arrival of Google cars

And then AC got involved in a big pileup on the highway.

[hplasm]

Re: To those here hailing the arrival of Google cars

"I'm confident I can drive myself a car ..."

Nothing personal, but that right there is the heart of the problem.

[IanDs]

Re: To those here hailing the arrival of Google cars

There are a few people who drive well and use observation/anticipation and haven't had an "accident" (see below) for hundreds of thousands of miles, the problem is that *everybody* thinks they're a good driver -- but at least 95% of them are wrong.

The reason the emergency services don't call them "accidents" any more is they very rarely are, they're almost always caused by inattention/distraction/stupidity. If robocars can get rid of these "accidents" the road casualty/death rate for them will probably drop by at least a factor of 10.

When -- not if -- this happens, being allowed to drive on the road may end up more like being a pilot -- a privilege that you don't get to do it without a lot of training, and where any idiotic driving or infractions get your license taken away. After all, cars kill far more people than planes, a two-ton lump of metal being driven at high speed is a lethal weapon.

Don't get me wrong, I don't want to see the "right" to drive taken away, but just because we all take the huge number of road deaths and injuries for granted doesn't mean it'll stay that way once technology has a workable fix for this -- the same happened with trains and then planes, cars are next.

[Intractable Potsherd]

Re: To those here hailing the arrival of Google cars

Actually, I sort of agree with you. Once being able to drive is no longer a near-necessity, it will be easier to increase the standards of the driving test so that it is more rigorous. People who are currently forced to do something they don't want, or don't have the innate ability, to do can still have the freedom that comes with car ownership without needing to pilot a couple of tonnes of metal. That is all to the good.

However, regarding your comment about people who have driven hundreds of thousands of miles without an accident - I'm one, and I know many others. I don't class myself as an excellent driver, and I am not as good as many of my friends,but I have the intuition, the skills and the ability to anticipate the road. I also know when I am not well enough to drive, and so make other arrangements - something I'm very lucky to be able to have (I won't be forced to lose a day's wage, or my job, by not turning in).

[IanDs]

Re: To those here hailing the arrival of Google cars

Anyone who had driven hundreds of thousands of miles without an accident (me too) has to be pretty good at things like observation and anticipation and not driving around with a phone glued to their ear or texting or reading a paper, which by most standards makes them an excellent driver even if they couldn't keep up with a racing driver on a track -- or wanted to...:-)

[Breen Whitman]

"The other example given was Google cars dealing with cyclists who didn’t obey the rules of the road."

A quick search of YouTube reveals that cyclists like to wear cameras so that they can berate all around them and then place the footage online.

The issue is that it reveals what bad road users cyclists are. Worse still they seem in a hightened aggressive state, and are not averse to beating a happless motorists, even mums with children in the car.

So yea, the software has to deal with that too.

[JeffyPoooh]

AI is hard

Call me when a self-driving car can win a rally.

[toxicdragon]

Re: AI is hard

Im calling!

https://www.youtube.com/watch?v=M2AcMnfzpNg

[IanDs]

Re: AI is hard

He said win, not come last...

[Intractable Potsherd]

Re: AI is hard @ toxicdragon

Now compare those speeds with, say, Dakar!

However, the use-case for self-driving cars is never going to be competition - it would be more boring than F1 ...

[toxicdragon]

Re: AI is hard @ toxicdragon

You know for a tech site I have wondered this a lot. What is wrong with you? It is cool. An autonomous thing did something. thats two marks on the cool scale. All it needs to do is have jet boosters and glow in the dark for the full set.

Show some wonder in the world.

[Medixstiff]

Good luck bringing automated vehicles to Western Australia, I saw one idiot flip his bike on Saturday trying to do a wheelie to impress another motorcyclist, I can honestly say, that over the last 5 years of cycling to work, I would have died at least once a month in that time period if I had trusted Perth drivers to do the right thing.

"The report also highlighted some of the smarter aspects of the cars' software. Google cars can identify emergency vehicles, for example, and automatically give way in a fashion many fleshy drivers are irritatingly unwilling to do."

When my best friend was working at WAPOL, he said that you do not legally have to give way to emergency vehicles, because if you have an accident, say by going through a red light to allow an emergency vehicle to get through the intersection, it doesn't cover you against blame by insurance etc.

[Intractable Potsherd]

Re: givig way to emergency vehicles

"Google cars can identify emergency vehicles, for example, and automatically give way in a fashion many fleshy drivers are irritatingly unwilling to do."

If they can do this properly, then it will indeed be a good example of programming. Pulling over is only part of the job. Doing it in the right place (e.g. not just before a blind bend or bridge), and at the right time (so that other road users don't start overtaking you) is just as important.

[graeme leggett]

Re: givig way to emergency vehicles

Now I'm wondering how they programme them to recognise the hand signals of a horse rider.

And there was that time I was on a coach that had to follow a stray cow down the Glasgow-Ardrossan road.

[Lionel Baden]

I love the idea

But the day I cant drive my 2cv on the open road, it will make me very sad .....

[IanDs]

Re: I love the idea

Me too :-(

Just expect that insurance -- if you can get it -- will cost you a *lot* more when 10% of the cars (those driven by humans) are responsible for 90% of the road accidents/deaths. And if you run someone over, prepare to be treated as a criminal pariah who killed because they wanted their "freedom" to drive.

Very much like the public attitude to smoking and making children breathe it nowadays -- this used to be one of the "freedoms" enjoyed by smokers, now it's seen as illegal and child abuse.

[fruitoftheloon]

IanDs:freedomRe: I love the idea

Ian,

Yay, smokers and their freedom, there is an interesting concept.

Wifey lost ~30% of her lung capacity due to a relatively common condition.

If she (involuntarily) gets a lung full of someone else' smoke, a trip to casualty may beckon.

How would you reconcile that with 'smokers rights'?

Just wondered....

Regards,

Jay

[IanDs]

Re: IanDs:freedomI love the idea

I'm not in any way in favour of smoking, I hate it -- I was simply pointing out that what society finds acceptable and normal can change drastically over a relatively short time. As well as smoking I could equally have picked not wearing a seatbelt or drunk driving, all are things that used to be seen as normal "freedoms" but are now seen as dangerous and unacceptable by most people.

[Anonymous Coward]

Curve Projection

Surely it cant be to hard if you can track an object over time (obviously more than 3 data points required, the more the better). You can do simply speed and curve projection (like i said in basic form 3 would work, drop the oldest on next read and re project). This also gives you the speed and projected speed.

Driving patterns can be fairly easy with maths to replicate, you just have a routine that works on distance and the greater the distance available for breaking feeds into the rate of breaking. In emergency situations where an extended breaking period was deemed safely possible (read in other input and calculate based on distance to target). Graceful breaking isn't that hard.

What hard like someone else said is to the write in all the finer subroutines that govern projections based over time based upon target being tracked, ie if target has previous increased rate previously to that of expected target it would simply read in the targets previous offset level and average it back across time (previous points) to give an expected result.

The more you open a can of worms on reading in the finer subroutines that allow you to fine tune projects on other targets the more things you discover you have to compensate for i guess.

But this is realistically only a part of what the smarts are about, its about doing what drivers take for granted and do automatically (well most of us), humans take all the complicated sensory from the car, feeling traction, weigh distribution and putting that to our advantage to go faster and/or go safer, some to a varying degree sadly.

Its really a golden time as things like cruise control, 4 wheel stability and breaking control has come a long way from the more crude systems of the past, making things possible.

Cars that are not only smart enough to drive you to the shops and avoid all the other idiots on the way, but could also put in a good time round a track in your name if you wanted, without sliding you into a tree at the first attempt at 50 on wet corner because its not programmed to react or compensate or expect for reduced friction on the road surface.

Everyone is scared of the day the privilege of driving will start slipping away and will be taken from us the moment its generally considered safer than manual control.

So what's going to happen when those of us out there with the know how start to back engineer the safety limits (raising the bar maybe, rate of acceleration, breaking, etc, etc) in the system and to also add in our own subroutines :) everyone loves to tweak right, every one wants theirs to be that special one, the ones that's just that bit faster than the rest.

You can take the petrol head out of the drives seat, but you cant take them out of the car... :-p

[Charles 9]

Re: Curve Projection

"So what's going to happen when those of us out there with the know how start to back engineer the safety limits (raising the bar maybe, rate of acceleration, breaking, etc, etc) in the system and to also add in our own subroutines :) everyone loves to tweak right, every one wants theirs to be that special one, the ones that's just that bit faster than the rest."

Signed firmwares and trusted paths will probably nip that in the bud. Just look at where Android is going now with Marshmallow's dm-verity and Google's move to centralize the core OS.