Autonomous Cars Will Save Money and Lives

cartechboy writes "Autonomous cars are coming even if tech companies have to produce them. The biggest hurdles are the technology (very expensive and often still surprisingly rudimentary) and how vehicle to vehicle (V2V) communication happens (one car anticipates or sees an accident, it should tell nearby cars). So what are the benefits to self-driving cars? They may save us thousands of lives and not a small amount of cash. A new study from the Eno Center for Transportation (PDF) suggests that if just 10 percent of vehicles on the road were autonomous, the U.S. could see 1,000 fewer highway fatalities annually and save $38 billion in lost productivity (due to congestion and other traffic problems). Right off the bat you can imagine autonomous driving easily topping your average intoxicated drivers' ability behind the wheel. At a 90 percent adoption mark those same numbers in theory would become: 21,700 lives spared, and a whopping $447 billion saved."

Comments

I like my A4 2T 6 speed

[vi.emacs]

Autonomous driving? No thank you!

Re:I like my A4 2T 6 speed

[geekoid]

I like my horse, cars? no thank you.

Re:I like my A4 2T 6 speed

[Deluvianvortex]

I like my standard too, but I hate it when I have to drive 2000 miles in it. Can I just put it on auto and be there by morning, please?

Re:I like my A4 2T 6 speed

[Austrian Anarchy]

I like my standard too, but I hate it when I have to drive 2000 miles in it. Can I just put it on auto and be there by morning, please?

Driving 2,000 or even just 500 in my manual shift Jeep is fine, but it is the 5 mile trips that are annoying. Still, not annoying enough to trade it for an autonomous car. Do I want to ban these new-fangled cars? No, of course not. However, I sure as hell don't want it to be the only choice in automotive transport either.

Re:I like my A4 2T 6 speed

[Marxist Hacker 42]

Yeah, but your boss can't expect you to work on your commute. This is really about adding 10 hours a week to your workweek.

Re:I like my A4 2T 6 speed

[Spy Handler]

And when you plow into a pedestrian in your Audi A4 while checking a Facebook message, better call Saul!

Re:I like my A4 2T 6 speed

[rnturn]

Or you could put the effin' smart phone away until you get to your destination.

I was wondering how far down I'd need to scroll to find a comment about how this would benefit people who can't leave their phones alone while they're behind the wheel of a car. As it turned out... not very far at all.

Re:I like my A4 2T 6 speed

[epyT-R]

false equivalence fallacy? no thank you.

Re:I like my A4 2T 6 speed

[Anonymous Coward]

If you want to drive recreationally, on a closed course, I expect you'll be able to do that indefinitely in more or less whatever format you prefer. But there's no reason you need to endanger others with your manual driving just to scratch your recreational itch or satisfy some nostalgic idea of "freedom" (via dependence on the auto industry, the oil industry, and public roads).

Re:I like my A4 2T 6 speed

[profplump]

He's a socialist for pointing out that cars require huge, mostly capitalist, social support to exist and function?!? Exactly what would he have to do to demonstrate his commitment to capitalism?

Re:I like my A4 2T 6 speed

[TemperedAlchemist]

You're right!

Self-driving horse? No thank you.

Re:I like my A4 2T 6 speed

[Anonymous Coward]

It isn't about benefiting people that can't put their phones down. it is about benefiting the people they run into.

Re:I like my A4 2T 6 speed

[Narcocide]

This is actually a reasonable but often overlooked point about self-driving cars. The original automobiles were a step backwards with regards to some features. No matter how cross-eyed, projectile vomiting, passed-out drunk you were, your horse would still usually be able to find its own way home.

That said, horses are also unintelligent, skittish, emotional, and even to the trained horseman often FAR less cooperative than even the least reliable production automobiles. Something tells me once Microsoft starts making A.I. Driver software we're going to see another step backwards, not forwards...

Re:I like my A4 2T 6 speed

[epyT-R]

Well horses are a lot closer to these self driving cars than they are to dumb machines.. They have just enough intellect to drive up the unpredictability for the rider without offering enough of it to put the owner at lease enough to let go of the controls. I'll pass. I'd much rather ride a horse at 15mph than ride one of these cars at 70mph+.

Re: I like my A4 2T 6 speed

[Anonymous Coward]

If you work 2 h during commute, then you work 6 h in the office. That is all.

Re:I like my A4 2T 6 speed

[Anonymous Coward]

But there's no reason you need to endanger others with your manual driving just to scratch your recreational itch or satisfy some nostalgic idea of "freedom"

On the contrary, there's no reason you need to intrude on his ability to drive on roads he helped pay for just to assuage your fear over the trivial amount of "danger" he presents. It's quite enough that he's required to pass a competency exam and forbidden from driving while impaired.

Re:I like my A4 2T 6 speed

[Firethorn]

Seriously: no, I cannot* imagine that the autonomous cars are actually better than the average intoxicated driver, according to modern limits. Your average intoxicated driver gets home safe and doesn't put anybody at risk. Given what I know about the difficulty in training a computer to behave intelligently in unplanned situations, I cannot imagine that autonomous cars are significantly better than your average intoxicated driver.

Excepting your disclaimer, I'd say that you have a very limited imagination. Personally, I can easily imagine it. Indeed, with the Google cars and such I'd say that we're very very close. Especially if 'perfection' isn't your standard, merely 'median human driver', IE 'better than the median drunk driver'. Now, I do imagine that the accidents would be different. I picture our AI cars being very good at avoiding accidents via 'fast twitch' actions, trading that for 'duh' accidents that no human would get into. The sort of accidents we see when human drivers get too stupid about following GPS directions, essentially. Things like driving into a flooded ford, through construction, etc...

When I figured it out, I pegged a self-driving car with merely 'half' the accident rate of humans to be worth a couple grand a year to me, and vastly more to those who either drive more or are outright bad drivers. I figured that DUI convicts would be among the first adopters - forced to do so by the courts, as interlocks are both expensive and hard to use, and I'd imagine the insurance break would be substantial at that point.

Re:I like my A4 2T 6 speed

[wonkey_monkey]

It's quite enough that he's required to pass a competency exam and forbidden from driving while impaired.

It's only "quite enough" because it's a balancing act between safety and practicality and this is the way things have settled for now. Autonomous cars are going to change that whether you like it or not.

Re:I like my A4 2T 6 speed

[wonkey_monkey]

Why can't it be about both? Sure, if a texting driver causes an accident, on a karmic scale I'd rather the driver at fault died than the innocent party, but given the choice I'd rather no-one died at all.

Re:I like my A4 2T 6 speed

[jecblackpepper]

When considering whether someone thinks they are better than average in driving skill you should look at this study

Svenson (1981) surveyed 161 students in Sweden and the United States, asking them to compare their driving safety and skill to the other people in the experiment. 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:I like my A4 2T 6 speed

[a_claudiu]

But there's no reason you need to endanger others with your manual driving

Citations please?

Re: I like my A4 2T 6 speed

[rally2xs]

Oh, behave! You KNOW that 2 hr commute to/from will be expected to be unpaid work, and you'll still be "at work" for 8 hours, work 10, get paid for 8.

Re:I like my A4 2T 6 speed

[Dragonslicer]

It's quite enough that he's required to pass a competency exam and forbidden from driving while impaired.

Either you've never driven on public roads with other drivers around you, or you have very unusual definitions of "competency" and "forbidden".

Re:I like my A4 2T 6 speed

[TheRaven64]

Yes, I too find it difficult to believe that a vehicle using sensors with centimetre precision on nearby obstacles and penetration through rain and fog, direct feedback from the wheels as to current grip levels, the ability to control the angle of the wheels to a single degree or better, and sub-millisecond response controller times, could possibly be better than a human.

Re:I like my A4 2T 6 speed

[Virtucon]

That's a pussy mobile meant for preppies driving to work. Try an AMG SL65 for getting that venti latte at Starbucks.

Re:I like my A4 2T 6 speed

[Dcnjoe60]

Or you could put the effin' smart phone away until you get to your destination.

I was wondering how far down I'd need to scroll to find a comment about how this would benefit people who can't leave their phones alone while they're behind the wheel of a car. As it turned out... not very far at all.

Seems a much cheaper solution to distracted drivers (at least by cell phones) would be a $10 chip that simply blocks cell phone reception. If it is illegal to talk and text while driving a vehicle, then block the signal. If you want/need to talk or text, pull over. Or, instead of putting a chip in each vehicle, it could be built into the phone. It seems smart phones are pretty good at telling where you are, what direction you are going and even your speed. If your speed appears to be over 5 or 10mph, it automatically goes into "car mode" which would be similar to airplane mode. The difference being, it can still receive calls and texts, they would just automatically go to voicemail (for calls) and not alert you for texts, until you fall below the speed threshhold, then you have your choice to pull over and do something. (And for those thinking that fine, at a stop light, start using your phone again, as soon as your speed goes above the threshhold, it turns into a receiver only device again).

Seems that would be a lot simpler than waiting for everybody to purchase cars that will drive themselves.

Re:I like my A4 2T 6 speed

[Dcnjoe60]

It isn't about benefiting people that can't put their phones down. it is about benefiting the people they run into.

Since most accidents are single vehicle accidents, the number of people that are run into that would benefit would be quite small. Most likely, the benefit is going to be to the occupants of the car with the technology and in reducing property damage.

Re:I like my A4 2T 6 speed

[crakbone]

That is way to general a view point. It does not take into account passengers, or the need to contact emergency services. Can you imagine a kidnapped child finds her cell phone but can't call because the vehicle is in motion? Or you drive a toyota and the gas pedal sticks? Or that the vehicle you are in gets into an accident on a lonely road and the chip is damaged and still thinks it is moving so shuts phone and text down.

Re:I like my A4 2T 6 speed

[nortcele]

Self driving horse... We had a horse on which I would regularly fall asleep while herding cows on the range. The horse would continue to weave back and forth and push the cows with me zoned out in the saddle. So... yes, the self-driving horse existed long ago.

Re:I like my A4 2T 6 speed

[pr0fessor]

I'm sorry, I find the prospect of a fully autonomous car to be laughable not due to technology but liability to the manufacture. Imagine if every fatal accident ended with a lawsuit. They would be advertised on TV... "Where you or a loved one in a car accident while in {insert model} autonomous car that result in injury or death? Call our law offices as you may be entitled to a substantial award. Including medical expenses, time off from work, property damages, and more..."

Re:I like my A4 2T 6 speed

[rwise2112]

Autonomous driving? No thank you!

I agree! It's nice that they're saving money (maybe they can pay for their own gas) and are taking 1st aid courses to save lives, but I prefer to drive myself.

Re:I like my A4 2T 6 speed

[zifferent]

Good point about lawsuits and that would be an issue initially, but all a company would have to do is prove that the self-driving car is safer than a person in a similar situation.

Re:I like my A4 2T 6 speed

[Dcnjoe60]

That is way to general a view point. It does not take into account passengers, or the need to contact emergency services. Can you imagine a kidnapped child finds her cell phone but can't call because the vehicle is in motion? Or you drive a toyota and the gas pedal sticks? Or that the vehicle you are in gets into an accident on a lonely road and the chip is damaged and still thinks it is moving so shuts phone and text down.

Well, if your concern is kidnapped children, then we better not have autonomous cars, because the kidnapper can be in one place and the car driving someplace totally different. If your accelerator sticks while driving down the highway, what will using the cell phone accomplish? Wouldn't a better, more logical choice be to turn off the car instead of trying to call for help? If your vehicle gets stuck in an accident, then by definition, you aren't moving 10mph for one and the chip, if it is damaged, is highly unlikely to be in a position to jam the signal in the first place.

All of those scenarios are nothing more than red herrings. The only legitimate obstacle and one you didn't mention is that a passenger in the vehicle would also be prohibited from using their phone. But, that would be a small price to pay for increased safety for relatively low cost. Besides, every time we are a passenger on a plane, we agree to not use our phone.

Re:I like my A4 2T 6 speed

[MiSaunaSnob]

on that note maybe what the industry needs to do is win F1 or WRC or 24h Le Mans with a self driving car? obviously not the only thing they need to do but it would go a long way in reassuring people about the potential.

Re:I like my A4 2T 6 speed

[kermidge]

Do it the same way the government did for the nuclear power industry - legislate limited and capped liability.

Re:I like my A4 2T 6 speed

[FuzzNugget]

What about off-roading? What about new roads that aren't yet in the GPS maps? Imperfect road surfaces that confuse imperfect recognition systems? Lines mispainted on the road? Faded or covered lines? Awkward intersections? Snowed over signage? Intersections directed by police officers while traffic lights are out?

Can automatic driving systems handle all these and more? And if not, what's the point if we still need to keep our eyes and attention on the road at all times in case one of these situations come up? If anything, they would make us *less* safe by giving us a false sense of security and putting us out of practice whenever we need to take over.

Besides, we already *have* a "universal interface" for managing transportation and it's generally pretty good at handling inconsistencies and determining these sorts arbitrary decisions very quickly.

It's called the human brain.

The problem is that, at least here, the programming is shit. Seriously, just getting a glance at the training requirements that are expected of drivers in other parts of the world, mainly Europe, would cause an American to die of shock.

Lost revenue to the cops

[Anonymous Coward]

Cops won't like it because they'll see lower revenue from DUI fines, speeding fines, and all that crap they love taking money for.

Re:Lost revenue to the cops

[Chuckstar]

Why do you think cops care about that money? Municipalities may care about that money, but the cops couldn't care less (they don't get a cut, after all). But cops do try to avoid hearing "how come everyone else writes more tickets than you do?" So they make a point of writing tickets. But they really don't care about revenues, per se.

Re:Lost revenue to the cops

[andymadigan]

They'll care if less municipal revenue means layoffs at the police department.

Re:Lost revenue to the cops

[Austrian Anarchy]

Cops won't like it because they'll see lower revenue from DUI fines, speeding fines, and all that crap they love taking money for.

I'm sure the governments will figure out a way to bust people for DUI even if they are riding in an auto-auto. They already do it for sitting in cars that are not running.

Re:Lost revenue to the cops

[Bigbutt]

That assumes the standard street cop thinks that far ahead.

[John]

Re:Lost revenue to the cops

[Bigbutt]

Where's that article on Oregon taxing for miles driven?

[John]

Re:Lost revenue to the cops

[NoKaOi]

But cops do try to avoid hearing "how come everyone else writes more tickets than you do?" So they make a point of writing tickets. But they really don't care about revenues, per se.

That's part of it. Another part is that it's their job to enforce the law, and so that's what they do. It's not their job to decide which laws to enforce (although obviously that happens to a degree). If a cop only enforces the laws they feel like enforcing, then they become the judge and jury too and our system generally tries to avoid that (federal agencies excepted of course). Sure, some cops are jerks that just want the opportunity to power trip on you, but for the most part that's a minority of cops, that's just the ones you notice because they're being jerks.

On the flip side of it...having autonomous cars and writing less tickets means less time spent on traffic enforcement, which means more time spent on real cop stuff. Also, for the most part, cops (at least the ones I know) really hate it when accidents happen, and especially DUIs since they're preventable accidents. Most cops are strict about drunken driving because they see the negative results first-hand. Most cops would be very, very happy if accidents from drunk or reckless driving were virtually eliminated.

Additionally, even when not on traffic enforcement, cops still spend a lot of time in their cars driving around even when they're not going all lights & sirens. If their cars were doing the driving for them, they could get stuff done like paperwork, which would free up even more of their time to bust bad guys.

Most cops didn't say "I wanna be a cop so I can drive around and write speeding tickets!" Instead, they usually think, "I wanna be a cop so I can catch bad guys that are doing bad things to people!" Self-driving cars would free them up to do more of the latter. Only the cops who get off on power tripping on everyone they pull over would be disappointed, and those are the ones who should be weeded out anyway.

They'll care if less municipal revenue means layoffs at the police department.

And how often does that actually happen? Dunno about other areas, but in the areas (and nearby areas) that I've lived, cops don't get layed off. They do, however, stop or reduce hiring new cops.

Re:Lost revenue to the cops

[Anonymous Coward]

You outta pay more attention on what municipalities and the police departments spend (scratch that, waste) money on. Police are more and more corrupt, they blatantly violate civil rights, either with illegal monitoring or surveillance, and or abusing laws and arresting people on petty bullshit. ANd the shit stain Judges/DA's that instead of dropping the case try to con people into cutting deal at a preliminary hearing (blatant abuse of law and there authority) which leads to civil rights lawsuits is most cases, and forget about a public defender helping you and doing there jobs telling a Judge or DA to go get fucked they have nothing for a Trial.

And not to mention how the judicial committee allows a Judge, her bother, who is the DA, and her son and or husband who is the Cities Lawyer (solicitor) and finds no conflict of interest, or a wide open set-up to prevent any lawsuits while the above mentioned statement shows blatant civil rights violations. There are at least 70% of the cities in the state where I live that are connected like this.

And lets not start on how the your local council decides to waste money, on idiot projects or personal because they have business/people connections and want to "help" them establish some sort of legacy!! On top of the brides and kick backs they are getting from the local corrupt businesses. Or the fact they refuse to cut unions from public spending to save the tax money. There are a number of thing going on that this morons get away with. Primarily talking out there ass when it nears election time, and then doing little to nothing. Oh yeah you can also find there concerns on there budgets and tax money when you catch them in a high end restaurant, every other day as well, lunch and dinner..

You must live in a fantasy utopia, or a large city where no one knows what there local politicians are up too. Not yelling at you just ranting, needed to get that out there as if slashdotters haven't figured it out.

Re:Lost revenue to the cops

[bananaquackmoo]

What do you mean they don't get a cut? Who do you think pays for the cops?

Re:Lost revenue to the cops

[Cro Magnon]

I'm sure the standard police commish will remind the standard street cop of that fact.

Re:Lost revenue to the cops

[DoomHamster]

Well, some cops are given incentives to care:

http://www.ksl.com/?sid=27376993&nid=148&title=emails-reveal-incentives-behind-cottonwood-police-issuing-tickets&fm=home_page&s_cid=featured-1

Re:Lost revenue to the cops

[Chuckstar]

A cop gets the same salary if he writes 5 tickets a day as if he writes 6 tickets a day (or 50 vs 60, or whatever the appropriate order of magnitude might be). He cares much more about whether or not his supervisor is hounding him than whether or not the municipality gets an extra few $100 speeding tickets.

Re:Lost revenue to the cops

[Chuckstar]

An individual cop really cares very little about a "bigger budget". And the effect of any one cop's ticket writing on the budget is exactly nothing.

Re:Lost revenue to the cops

[Chuckstar]

I challenge you to find a cop out writing speeding tickets who is doing so in order to avoid layoffs at the police department. They're doing it because they got put on traffic duty that day, and they get yelled at if they don't meet the not-really-a-quota-but-come-on-we-all-know-its-a-quota.

For the most part, cops write tickets because their bosses tell them to write tickets. If you were doing 67 in a 55 zone, the guy writing your ticket doesn't really want to be giving you a ticket much more than you want to be getting a ticket. If you were doing 140 in 55 zone, that guy might be very happy to slap you in the wallet for your dangerous stupidity. But in neither case is that cop particularly considering the effect of the fine on the municipal budget.

Re:Lost revenue to the cops

[Chuckstar]

Exactly the original point I made. The municipality may care about that money, and they may pressure the cop, but the cop himself doesn't really care about that money. (To me, the commish isn't so much a "cop" in this scenario, as he is the cop's politically appointed boss.)

The reason I make the distinction is purely because it's not fair to the individual cops to label them in that manner -- as being simple cash grabbers. The cops don't want to be out there giving tickets for going 12 mph over the speed limit. They do it because the bosses tell them to do it. As I indicated in another comment, the cops will happily give tickets for really dangerous stuff, because they tend to care about public safety, but most would just as soon focus on "real" crime than be out handing out tickets for going 67 in a 55 zone.

Re:Lost revenue to the cops

[Chuckstar]

That may be, but one still need not blame the individual cop for that. That's really what my original comment was about.

38 billion in productivity or

[geekoid]

30 minutes more sleeping?

Re:38 billion in productivity or

[Anonymous Coward]

Amen! The real savings is all that time people spend behind the wheel. Autonomous cars might make the U.S. livable again.

Also, I'd so love to never interact with another cabby again, just horrible people far too frequently.

Re:38 billion in productivity or

[Jeremi]

30 minutes more sleeping?

30 minutes more sleep would also make people more productive -- so either way it's a win.

Re:38 billion in productivity or

[gunzy83]

Nap time pre and post work would make me very happy haha

Re:38 billion in productivity or

[RivenAleem]

I know something I could do with that 30 minutes that is much more productive than sleeping *wink wink*.

Yeah, my emails and RSS feeds aren't going to check themselves!

Re:38 billion in productivity or

[Ardyvee]

That is assuming most people will use those extra 30 minutes to sleep. I don't think they will. I think it will translate in going to bed a little later because now can leave your house 30 minutes later, so don't need to wake up as early, so don't need to go to bed as early (or can go to bed later).

So what'll we do with half a trillion dollars?

[rsilvergun]

or put another way, what'll happen when we have half a trillion dollars less economic activity? Since our entire civilization is based around getting people to trade among themselves. I just don't see all these productivity gains are ever going to make it down to my level...

Re:So what'll we do with half a trillion dollars?

[Anonymous Coward]

Actually, Autonomous cars are a productivity gain that quickly translates, by allowing you to nap or read in the car after you buy one.

We want to put as many people out of work as possible, that's really the whole point of technological advancement, that's how we make our lives better. There are obviously powerful people who steal our productivity gains, like Wall St. and real estate brokers, our expanding law enforcement and industrial prison system, etc. We must reclaim these productivity gains for ourselves by enforcing transparency upon those that rule us, ala big banks, big companies, and governments, as well as by shortening the work week.

You know, after the industrial revolution, workers needed to do exactly this to but by forming unions, fighting in communist revolutions, etc. Unions were the ones who shortened the work week from six to five days during the 20th century, which helped bring about more prosperity. France has benefited economically form it's 35 hour work week more recently.

At present, our best way to force the government to shorten the work week is to : (a) Invent technologies that put masses of the pointless white collar workers out of work. High frequency trading helped reduce the number of people needed in finance, for example. And (b) obstruct Keynesian make work programs like the expansion of law enforcement through the war on drugs, war on terror, and surveillance state.

Autonomous cars are cool though because they require no connected political reform, just put all the drivers and cabbies out of work (yey!), and save everyone an hour or so per day (double yey!).

See https://www.youtube.com/watch?v=Ie5zO-mF31M

Re:So what'll we do with half a trillion dollars?

[epyT-R]

That's it. There will never be computer driven cars for the masses. It will always be cheaper for them to drive their own.

Not when the insurance companies artificially jack up the rates for human driven cars. They will force the majority into this, guaranteed.

Re:So what'll we do with half a trillion dollars?

[debrain]

or put another way, what'll happen when we have half a trillion dollars less economic activity? Since our entire civilization is based around getting people to trade among themselves. I just don't see all these productivity gains are ever going to make it down to my level...

Not all economic activity benefits society. Perhaps the most well known demonstration is the parable of the broken window:

The parable of the broken window was introduced by Frederic Bastiat in his 1850 essay Ce qu'on voit et ce qu'on ne voit pas (That Which Is Seen and That Which Is Unseen) to illustrate why destruction, and the money spent to recover from destruction, is actually not a net-benefit to society. The parable, also known as the broken window fallacy or glazier's fallacy, demonstrates how opportunity costs, as well as the law of unintended consequences, affect economic activity in ways that are "unseen" or ignored.
 

The productivity gains failing to make it to your level are arguably a problem of inequality of the distribution of wealth, not lack of economic activity.

Re:So what'll we do with half a trillion dollars?

[Anonymous Coward]

Preach on, bro! You'll also never see GPS for the masses, it will always be cheaper for them to open a map. Or power windows. Or automatic transmission. Or...oh, wait.

Re:So what'll we do with half a trillion dollars?

[Obfuscant]

Realistically, electronically controlled vehicles will roll out in this order:

You mean "autonomous", don't you? Electronically controlled is close to what we have today for almost everything but actual steering. "Fly by wire".

6. Middle income people who can't qualifiy for a license.

If you can't qualify for a license, you aren't going to be able to take over from you car when you need to, and you won't be able to drive the last mile from your garage to the road with all the electronic navigation aids that will be required. Save lives? Turn the highway into a large parking lot when one car fails and the "driver" isn't qualified to pick his nose, much less manipulate the controls of his car to get it out of the way.

And for all the bright people who will immediately try telling me all the wonderful things autonomous cars do, you need to keep in mind that they don't exist in anything close to production form and the large scale interactions have not been determined in real life. Models of something that doesn't exist have no ground truth data to verify the model and shouldn't be trusted with millions of human lives.

Re:So what'll we do with half a trillion dollars?

[profplump]

First of all, low-income people already have a reduced rate of car ownership compared to the other groups you list, so half of what you're predicting is "no change from the status quo".

Second, why would you want to own a car when you could rent a driverless cab (or ask the automated bus to swing by my current location)? The economics of car ownership change entirely when you take drivers out of the loop.

Re:So what'll we do with half a trillion dollars?

[ozmanjusri]

Autonomous cars are cool though because they require no connected political reform, just put all the drivers and cabbies out of work (yey!), and save everyone an hour or so per day (double yey!).

Welcome to 1904.

The Wobs may have been too militant for their own success, but they well understood the nature of the battle. IT and business/knowlege workers today are facing the exactly the same threats to their enjoyment of life now, and will need to decide how to respond or be overwhelmed.

Where the machine is put in, some of the workers move out. One worker with a machine, or a small working force with machinery, will produce more goods than a large working force with hand tools. So that machinery displaces laborers. This is the feature of machinery that secures its installation in industry. But machinery does more than merely throw workmen out of jobs, it renders the versatile skill of the craftsman unnecessary. So the machines have won their way into every industry, and wherever they went less labor was required until eventually the aggregate of these surplus laborers grew to such proportions that there came into existence what is known as the army of unemployed.

At first the unemployed were largely of the mechanical trades, but the invention of new mechanical devices, and the improvement of machinery, which has been going on, has reduced the unemployed to a working class contingent in which the unskilled workers predominate.

Ask the average worker what relation machine production has to unemployment, and you will find that he is unaware of the fact that machinery will explain unemployment. Yet this fact, which is potent enough to be self-evident, is a mystery to the average unionist, let alone to the average working man and woman. The unemployed, even after many experiences, on the average only understand that "the job was shut down" by the boss. It is accepted that the employer has an unquestioned right to shut down industry, regardless of the social consequences.

http://www.iww.org/history/library/iww/isandisnt/6

Autonomous cars are tangential to the conflict, but apportioning the benefits they will bring will require political reform.

Re:So what'll we do with half a trillion dollars?

[JanneM]

Not when the insurance companies artificially jack up the rates for human driven cars.

If humans are the cause of more accidents there's nothing artificial about it.

More realistically, I expect most people a generation from now will find the higher vehicle cost to be easily offset by not having to get a manual driving license, freeing up driving time, lower fuel consumption and using the car even when disabled, too young or otherwise not able to drive manually for whatever reason.

Re:So what'll we do with half a trillion dollars?

[Great Big Bird]

This sounds remarkably like arguing for the broken window fallacy. In this case we are talking about windows that don't have to be broken anymore.

Re:So what'll we do with half a trillion dollars?

[FishTankX]

I don't think it'll necessarily put cabbies out of work, because unless i'm mistaken the primary reason people would take a taxi other than drinking, is either they lack a car (by choice, or a family with only one car, where the wife or husband needs to get somewhere while the car is out), or there is no parking at the destination. It would seem that autonomous cars wouldn't benefit people in either of these cases.

Re:So what'll we do with half a trillion dollars?

[Ocker3]

Exactly, self-driving cars aren't a replacment for taxis, not with the current level of technology. We're not talking Jonny Cabs from Total Recall yet, we're looking at reducing the amount of effort the average commuter needs to get from point A to B. I wonder if self-driving cars would allow better ride-sharing arrangements?

Re:So what'll we do with half a trillion dollars?

[epyT-R]

This is not a lifestyle I want to live. I can't imagine a future population truly being happy with this either. No matter what the soccer mom associations running western society, today say, there's much more to life than safety and convenience, especially when it comes to control over mental state and physical location/transportation.

Re:So what'll we do with half a trillion dollars?

[epyT-R]

Hmm.. well if you think about it, those who handle their own sticks are less likely to be accused of 'rape' by 'empowered' women, less likely to have child support liens against their paychecks and assets, and less likely to have credit spendthrift girlfriends/wives. These guys are much more profitable for insurance companies because they are more likely to pay their bills on time, don't take unnecessary risks to impress women, and are less likely to be in bad health due to post divorce misery.

Re:So what'll we do with half a trillion dollars?

[Laxori666]

Oh god, the broken window fallacy. Spending is a terrible, terrible measure of economic activity. What actually makes people wealthy is producing things. Otherwise we could just destroy 10,000 cars and spend millions of dollars making new ones to replace them, every day, and be wealthy beyond our wildest dreams. Hint: it won't because all you're doing is wasting money re-creating cars you just destroyed.

Re:So what'll we do with half a trillion dollars?

[grumling]

Ah, but what if you rideshare an autombile? Instead of having one sitting in the garage and parking lot all day, why not make it available to others? Summon your vehicle with an app and it comes to you, ready to take you wherever.

I’m sure that rush hour would make scheduling and resource allocation difficult, but for many people, not owning a car while having a reliable, economical “taxi” service could be fantastic.

And let’s not forget that the 1st world population is aging at an accelerated rate. The number of people who won’t be compentant to drive is increasing while the overall population is stabilizing.

Re:So what'll we do with half a trillion dollars?

[grumling]

The stick shift people are basically automatic. The others are anything but automatic (but it sure is a lot of fun trying to get them going).

Re: So what'll we do with half a trillion dollars?

[Anonymous Coward]

Bus will be first. Route is known and static.

Re:So what'll we do with half a trillion dollars?

[jader3rd]

I can't imagine a future population truly being happy with this either.

Really? You don't think that people would rather be playing games on a mobile device or texting, than having to pay enough attention to their surroundings to avoid harm to others and themselves?

Re:So what'll we do with half a trillion dollars?

[jhol13]

This is the lifestyle I want *you* to live as it decreases the changes you drive over me.
Go drive in race tracks, not in city centers.

Re:So what'll we do with half a trillion dollars?

[wonkey_monkey]

This is not a lifestyle I want to live.

I don't want to a live a lifestyle where I have to do what someone else wants for 8 hours a day, 5 days a week, but I got over myself.

I can't imagine a future population truly being happy with this either.

Then you're not imagining very hard. I'm only part of the present population but I already think it'd be an excellent idea.

Re:So what'll we do with half a trillion dollars?

[Plunky]

This is not a lifestyle I want to live. I can't imagine a future population truly being happy with this either.

I can imagine being very happy that I don't have to own a car, or drive it around.
I can imagine that I will be very happy to just summon a car and tell it where I want to go, to be taken there quickly and dropped off and being able to walk away, not have to worry about parking that car, or maintaining it.
I can imagine being very happy that when I want to use my bicycle instead, I won't be cut up because some ignorant driver doesn't think I ought to be on the road.
I can imagine being very happy that when I want to walk somewhere I can cross the roadway safely wherever and whenever I want because the cars will flow around me.
I can imagine being very happy that when I get over the age that it is really safe for me to drive, I will still be able to get around just as well as I always have been able to.
I can imagine that a future population would think it strange that people like you wanted them to suffer with traffic jams and having to sit focused and barely moving while driving, and being penalised for letting their attention wander at the wrong time.

Re:So what'll we do with half a trillion dollars?

[JanneM]

This is not a lifestyle I want to live. I can't imagine a future population truly being happy with this either.

It is a lifestyle I want*, and I suspect quite a lot of other people also want. And as the cost of a manual car go up; as you are stuck driving the same pace as all the strictly law-abiding autonomous cars; and as you stare out in traffic instead of surfing the web or throwing birds at pigs like the people in the cars around you, that lifestyle will appeal to more and more people who may not like it at first.

And soon you'll have a new generation of "drivers" that grew up with autonomous cars and used them when they were still too young to drive. Most of them will never see the point in spending a bunch of money and time learning how to drive manually. Once autonomous cars are out there, it will be only a matter of time until "driving" is an expensive, somewhat eccentric hobby you do mostly on a driving track.

* I already live that lifestyle in a way, as I don't have or need a car at all where I live. I only keep my license for use as an ID.

Re:So what'll we do with half a trillion dollars?

[drinkypoo]

No matter what the soccer mom associations running western society, today say, there's much more to life than safety and convenience, especially when it comes to control over mental state and physical location/transportation.

Ask the people stuck in traffic, not actually going anywhere, if they want to have to drive their own car. Now ask yourself how the majority of road-hours are spent. Now ask yourself how dumb you sound.

Guess what? Your right to drive is already constrained by laws regarding safety. You're not permitted to drive your car just anywhere it will go. And the cops can already confiscate your car if you represent basically any kind of threat to public safety. The freedom of automobile ownership is largely illusory.

Re:So what'll we do with half a trillion dollars?

[BigZee]

I think the money saved will be lost elsewhere. Like it or not, a large part of car ownership is aspirational. This move will make your car more like a utility and you'll probably replace it less often as a consequence. After all, what's the point of having a high performance engine if your computer is going to pull away from traffic lights in an economical way. That's not to say I don't think this is a good thing, it will make roads much more safe. However, the concern I have is regarding regulation. If there is an accident, who is liable? You, the guy who serviced the car, the company that built it? And thinking along these lines, who will be authorised to service the car and what controls will exist to prevent people using the car if it's not serviced or not serviced by an approved service company?

Re:So what'll we do with half a trillion dollars?

[Muad'Dave]

... destroy 10,000 cars and spend millions of dollars making new ones to replace them ...

You mean cash for clunkers?

Re: So what'll we do with half a trillion dollars?

[xigxag]

If the car is sufficiently self-driving, it will be able to drop you off at your destination, then find a nearby parking spot and wait there until you are ready to retrieve it.

Re:So what'll we do with half a trillion dollars?

[rally2xs]

Say what? You would take an autonomous taxi, of course, that would get you from point A to point B without a driver to have to talk to. It might even be coin operated (bill changer, actually.)

Re:So what'll we do with half a trillion dollars?

[TheRaven64]

There are schemes like ZipCar that give you access to one of a fleet of cars for short periods. It wouldn't be much of a stretch to extend this to self-driving vehicles. I don't see taxi companies that have human drivers competing with ones that don't - the driver is a significant part of the total cost.

Re:So what'll we do with half a trillion dollars?

[Blackknight]

Paper maps are far superior to GPS or smart phones. When you're out in the middle of the mountains with no signal or a dead battery you'll be damn glad you have some old fashioned paper maps.

Re:So what'll we do with half a trillion dollars?

[RavenLrD20k]

I want to fly my own plane but can't afford the classes right now. Never flew on an airline yet. No use for train travel, which is fine for autonomy anyway. I do steer my own boat. Often. And I often roll on a Motorcycle too, which on a low traveled road is much more pleasurable than having to worry about when this guy that looks like he wants my lane is actually going to decide to take it, or worse, sweep through it... without checking his blind spots. For me, a 50 minute commute on back roads with the Bike is much more pleasant than the 20 minute roll on the interstate. I gladly take the feeling of full control over my transportation over yielding that control to anyone or anything else. In fact, I dare say that the commute is my zen time to decompress from a long day at work. I lose that ability, I'll adapt, but I'll sorely miss that feeling.

Re:So what'll we do with half a trillion dollars?

[Ardyvee]

Sure, let's build various cities with different architectures and long roads surrounded by nothing but desert, a road through the woods (that gets all orange from the leaves during autumn), long stretches of highways and side roads that lead you through some fields.

I'd like to drive one day and just drive around, having the experience of driving, and experience going around with me and my car and the scenery going by. Cities, farms, going up the mountain, through the woods, or just a long drive through the equivalent of a desert.

Now, where would exactly we fit all this, in your "race tracks"?

I do understand the desire for safety. I do understand the desire for efficiency. I lived in Venezuela, Caracas, and I know what a long wait inside the car is (although I've heard it's gotten worse nowadays. Glad I don't live there anymore). It's boring, annoying and a waste of resources. And I didn't want to be "there". But I do want to drive around. Me, myself. Not a computer. No. Me, the car and the road.

Re:So what'll we do with half a trillion dollars?

[Laxori666]

Hahaha. Yes, exactly.

With reasoning like this, one wonders why the government doesn't just give every citizen hundreds of thousands of dollars, so each person can then go on a huge spending spree, thus surely stimulating the economy by vastly increasing spending.

Re:So what'll we do with half a trillion dollars?

[mcgrew]

Turn the highway into a large parking lot when one car fails and the "driver" isn't qualified to pick his nose, much less manipulate the controls of his car to get it out of the way.

How's that any different than when a transmission seizes up? The 80 year old lady is going to push her Escalade to the side of the road? That's what tow trucks are for.

Re:So what'll we do with half a trillion dollars?

[ultranova]

This is not a lifestyle I want to live. I can't imagine a future population truly being happy with this either.

I can't really imagine the mindset of considering driving a "lifestyle", but I'll take your word for it. Take mine when I say that I would love to let a computer take care of what's ultimately a boring, mechanical task while I daydream, watch scenery or browse the net.

No matter what the soccer mom associations running western society, today say, there's much more to life than safety and convenience, especially when it comes to control over mental state and physical location/transportation.

Oh yes, let's not forget the freedom to meditate or use mind-altering substances.

Re:So what'll we do with half a trillion dollars?

[Obfuscant]

How's that any different than when a transmission seizes up?

Gross large scale mechanical failure is typically much different than a computer or sensor failure. You can imagine no failure mode other than one that physically disables all drivetrain operation, not even one that leaves the car drivable but not autonomously?

I've had a clutch cable break while I was away from home. Lots of people might consider that a major drivetrain failure. Because I had a driver's license and knew how to drive, I was able to get the car home and then install a new cable the next morning myself. I've had GPS lose position and I could "take over" the navigation because I know how to drive. I've had exactly one large scale gross mechanical failure that required a tow truck, and that was when a car behind me saw the green light but missed the fact I was still stopped at the intersection in front of him.

Imagine an autonomous vehicle with a BSOD forcing it to stop, and the humans inside have no idea how to move the vehicle to a place of safety. What hilarity!

And when today's ubiquitous navigation system screwup has the car about to cross an active runway at a British airport, watch the fun as the people inside panic because they don't know how to keep it from happening or get out of the way of the airplane in their window.

No. There will always be a license involved, just to deal with contingencies. You do notice, I hope, that pilots still need licenses to fly even though a large part of the process is now autonomous for them, don't you? The recurrent training for an ATP working in a major airline doesn't deal just with the landing and takeoff -- manual parts. "Oh, the rest is automated, you don't need training to deal with that." They've had autonomous systems (autopilots) for a couple of decades and those systems STILL fail in sometimes wonderful and catastrophic ways.

Re:So what'll we do with half a trillion dollars?

[Kielistic]

It's likely easier to get a GPS lock in "middle of the mountains" than it is to pinpoint yourself on a map from landmarks or dead reckoning. When your map blows away in a storm or otherwise gets easily destroyed you'll be damn glad you have a backlit GPS device that knows exactly where you are.

Re:So what'll we do with half a trillion dollars?

[Blackknight]

Until the battery dies.

Re:So what'll we do with half a trillion dollars?

[Kielistic]

Why would the battery die? Forgot to charge it? Forgot to turn it off? Random failure? All red herrings which was my point in bringing up being blown away in a storm. A paper map is not "far superior" to a GPS device it merely has different failure points.

Re:So what'll we do with half a trillion dollars?

[FishTankX]

Not a retort to you, but I wonder if self driving vehicles will be able to go get themselves refueled autonomously too. That would be hilarious, when a car with no driver pulls into your gas station and declares over loud speaker "I require Alcohol!" (For comic effect, if ethanol powered)

Re:So what'll we do with half a trillion dollars?

[JanneM]

Of course, each and every autonomous car is recording every second of the closest fifty meters at all times. You so much as hint at breaking the speed limit or breaking any other traffic rule, and a dozen cars will have you recorded with your bike license plate, position, time and speed down to a tenth of a kilometer.

Driving manually, whether a car or bike, is going to become very, very dull.

Re:So what'll we do with half a trillion dollars?

[__aaltlg1547]

Do you know what people do in cars?

I do and no way and eww!

Re:So what'll we do with half a trillion dollars?

[__aaltlg1547]

A little more than you're paying to rent a car now, if you want to make a similar rent-by-the-day contract. If you want an automated cab ride, it would be less than we pay now for cab rides.

Re:So what'll we do with half a trillion dollars?

[mcgrew]

I've had a clutch cable break while I was away from home. Lots of people might consider that a major drivetrain failure. Because I had a driver's license and knew how to drive, I was able to get the car home and then install a new cable the next morning myself.

Volkswagon Rabbit? I had one years ago, that thing went through clutch cables like nobody's business. If my then-wife would have been driving she'd have been stuck (it did happen once but I was there so we just switched drivers).

Ever had an axle break or a differential go out? I have. Broken axle is no fun, busted differential can stop the car pretty damned quick.

Imagine an autonomous vehicle with a BSOD forcing it to stop, and the humans inside have no idea how to move the vehicle to a place of safety. What hilarity!

Imagine a paraplegic whose van has a blowout. He's stuck until help comes, just like your clueless people and their bluescreened car... and AFAIK Windows is the only OS that does that, and I think it's only older versions as I haven't seen one in years.

No. There will always be a license involved, just to deal with contingencies. You do notice, I hope, that pilots still need licenses to fly even though a large part of the process is now autonomous for them, don't you?

Licenses will be needed for some time, yes, but the art and science won't stop evolving. It will be a while before cars are fully autonomous; it will happen little by little.

Re:So what'll we do with half a trillion dollars?

[sydbarrett74]

Your rights end where mine begin. I guarantee you you're not as proficient a driver as you think you are. Nor am I, nor is anyone else.

It's a race for ownership of the car's OS

[Trip6]

Whoever owns control of the car wins.

Re:It's a race for ownership of the car's OS

[epyT-R]

Doesn't scale very well over long distances, though.

Re:It's a race for ownership of the car's OS

[wonkey_monkey]

Completely bug free

So far beyond wrong.

Re:It already exists!

[cosmin_c]

Obviously it depends on how public transport is in your specific area, but then again, if the public transportation is lacking, or dangerous to use, I sincerely doubt autonomous cars will make their appearence there any time soon.

What utter crap

[onyxruby]

Hey, let's play this game with computers, after all we don't need freedom behind the keyboard either and **AA's claim piracy cost the economy countless billions of dollars every year. Let's have autonomous computers! We'll make the operating system and hardware completely closed to prevent anyone from altering their 'trusted' environment. Now in order to keep anyone from hacking into their computers and driving by themselves we'll have to make sure that we take away the ability to install software that hasn't been approved.

We'll do this through a centralized market place where every application is signed and approved. Now the signing agency is taking on a lot of work to act as big brother and censor everything so it's only fair that they get a cut of 30%. How if your application sells well we'll cut the fee down to 20%. Now we have to make sure that your computer can't be used to pirate software so we'll keep up the autonomous trend and make all updates automatic. By locking out software from any distribution method other than the market and ensuring updates are automated your environment will stay trusted for software companies to continue offering you software.

Welcome to Microsoft Surface RT of the future, big brother knows best. What possible legitimate reason do you have for driving your own computer and endangering the economy by enabling the possibility of piracy? Think of the children!!!

Re:What utter crap

[girlintraining]

We'll make the operating system and hardware completely closed ... make sure that we take away the ability to install software that hasn't been approved. We'll do this through a centralized market place where every application is signed and approved. Now the signing agency ... get a cut of 30%.

Dude, you just described Apple's marketing strategy, not Microsoft RT.

Re:What utter crap

[onyxruby]

Actually it's exactly the strategy in use on the Surface RT. Apple will at least let you install software from third party sources, the RT doesn't. Mind you Microsoft modeled this strategy after Apple's so I guess you could blame Apple.

Re:What utter crap

[girlintraining]

Actually it's exactly the strategy in use on the Surface RT. Apple will at least let you install software from third party sources, the RT doesn't. Mind you Microsoft modeled this strategy after Apple's so I guess you could blame Apple.

Yes I know. I was trying to be snarky. Alas, the internet has yet to come up with the much-needed [SNARK] HTML tag.

Less dead people!

[Neo-Rio-101]

"the U.S. could see 1,000 fewer highway fatalities annually and save $38 billion in lost productivity (due to congestion and other traffic problems)"

...and less dead people unable to be put to work

Skeptical

[Anonymous Coward]

You know . The way they're painting this , it seems like there's not going to be any unforeseen problems with it.
I can already predict crashes due to hacking/ buggy softwares and etc.

Don't get me wrong. I agree with the fact that automated cars are a step in the right direction. However, what I dislike is how it is being presented here. It is presented as if it was a holy grail of driving. The solution of all problems. That's very misleading and dangerous. That's what I can't stand. The dishonesty of it all.

  We should be very honest here with the end users about what auto cars can accomplish at this point and what they can't.

Re:Skeptical

[couchslug]

Slashdot management have a serious car-boner for this tech so you'll read lots about it here. Also, page hits.

Re:Skeptical

[Anonymous Coward]

The number of crashes from buggy software is much higher with human drivers than computers. How many computers hit the wrong pedal and slam into buildings?

However, what I dislike is how it is being presented here. It is presented as if it was a holy grail of driving. The solution of all problems. That's very misleading and dangerous. That's what I can't stand. The dishonesty of it all.

Humans driving is much worse than the worst automated design. You misunderestimate how badly humans drive.

Re:Skeptical

[epyT-R]

Agreed. This dishonesty should set off alarm bells about their true intentions and priorities.

Re:Skeptical

[ebno-10db]

The number of crashes from buggy software is much higher with human drivers than computers. How many computers hit the wrong pedal and slam into buildings?

And you know this how? Please don't say Google, because they just say "cool, X zillion miles without an accident, blah, blah, blah". We have no idea what the driving conditions and other factors are.

Re:Skeptical

[Dunbal]

I think it's a wonderful idea - maybe because I'm older. It would allow my in laws, for example, to continue being mobile in their late 70-s and 80's, whereas now they can't drive. It would allow me more mobility too, since I can't really drive due to health reasons. I can imagine automatic-only roads, where the speed limits are increased and traffic flow is automated - no more traffic jams, traffic lights would result in faster trips and more efficient fuel use.

Of course I like driving as much as the next guy, but I wouldn't mind if it became relegated to a "hobby" as opposed to an unavoidable daily chore.

Re:Skeptical

[JamieIanMacgregor]

and toddlers could then steal cars and go joyriding. no more waiting until they're 13 http://nz.news.yahoo.com/a/-/top-stories/19263252/police-fail-13-year-old-boy-before-death/

Re:Skeptical

[Firethorn]

I can already predict crashes due to hacking/ buggy softwares and etc.

I can too, which is why I always assume that at least *some* accidents will still occur. For that matter, I figure that the profiles of accidents will change - moving away from things like 'ran through a red light' and towards 'drove into a not completely correctly marked construction zone'.

Still, the question becomes one of how the autonomous cars affect the average rate and severity of crashes. If it cuts off 50%, concentrated in the extreme, more commonly fatal crashes, it's a real benefit to society.

However, what I dislike is how it is being presented here.

If you follow the links you'll see a mention of cutting accident rates by 80%. I tend to view 50% as a good dividing line for starting adoption, 80% just makes the case so much tighter.

Re:It already exists!

[dgatwood]

If you want an autonomous car where you can check mail and send texts on your way to work, you can have it right now. It's called a bus.

Only if either A. you have access to a park-and-ride facility that is closer to your house than your workplace is, or B. the bus stops very close to both your home and your workplace. I've usually found that unless your commute is at least half an hour by car, you'll spend more time walking to and from the bus than you would spend driving, and even if you don't count the walking time, it still takes 2–3 times as long to get there. As always, YMMV.

Public transit is great for moderately long commutes, particularly if parking sucks at your destination. If I'm going into San Francisco, I take public transit. If I'm going to work, though, there's actually enough parking, so it isn't worth the 20 minutes of walking and 30+ minutes on a bus just to save 15 minutes in my car. It would probably be slightly cheaper, but the inconvenience is pretty severe. And that's without having to change buses at all.

And all the rest

[Anonymous Coward]

Aircraft, busses, taxis are all going to be autonomous! (What does auto mean, anyway?)

And then when that all works -

The rest of the system goes on-line August 4th, 2017. Human decisions are removed from strategic defense. Skynet begins to learn at a geometric rate. It becomes elf-aware at 2:14 a.m. Eastern time, August 29th. In a panic, they try to pull the plug.

And when we come back, we'll have... "The Rest of the Story".

Re:And all the rest

[OutOnARock]

"elf-aware"

Just Buddy, or the whole crew?

Re:It already exists!

[demonlapin]

Not dangerous or absent. Just inconvenient, slow, and unpleasant.

Right off the bat?

[JoeMerchant]

>Right off the bat you can imagine autonomous driving easily topping your average intoxicated drivers' ability behind the wheel.

Didn't anybody pay attention to the DARPA Grand Challenge?

Personal Time Saved

[Salgat]

I'm extremely frugal and I'd still buy one the instant an affordable one is released simply because an autonomous car represents a potential savings of 4,000 hours of my life over the life of the car. That's represents 2 years of a full time job. That's time that could be spent doing whatever I usually do at home, including sleeping, entertainment, and personal work/finances. It's incredible to think about.

Re:Personal Time Saved

[Salgat]

Even if I lived half the distance (15 minutes from work), that's still a year's worth of full time working. The point only becomes irrelevant if you live within 5 minutes of your job.

Risk Perception 101: People are Idiots

[Arancaytar]

People are willing to endure a risk orders of magnitudes higher of crashing by human error than by machine error.

Much as they're okay with the risk of dying from flu every year by not vaccinating, but not the comparatively negligible risk of a terrorist attack.

Re:Risk Perception 101: People are Idiots

[JoeMerchant]

You can sue an individual with some chance of winning (though mandatory insurance tries to make us into mini corporations, it doesn't completely succeed.)

If the accident is blamed on a company like Google, do you think their attorneys would have let the product out the door without closing off the product liability exposure? Google et. al. will not roll a product like auto-drive out to the general public until they've successfully lobbied themselves teflon body armor.

Re:Risk Perception 101: People are Idiots

[FishTankX]

You could always just make the self driving part of the system deactivate itself unless it has received all scheduled checkups and maintenance, putting the liability with the mechanic.

Re:Risk Perception 101: People are Idiots

[tftp]

No mechanic would ever assume such liability. As of today, liability of most businesses is limited to monies paid.

I think the insurance companies will gladly support the new cars. This eliminates the least reliable element - the driver. Software can be upgraded on all cars overnight, if need be. But when was the last time when someone in a different state wrapped his car around a tree and you got a free lecture on his specific mistake? Autonomous cars *will* be learning from mistakes of others.

Re:Risk Perception 101: People are Idiots

[delt0r]

Flu vaccines don't work (but are mostly harmless). I sometimes work with people in the field. Its pretty much accepted by anyone with the data that they don't work. We even know why. There are some new methods of getting more current strains ready quicker, should have an updated report on the current trails sometime around christmas.

In god we trust. The rest of you show me the data. As you said, Risk Perception...... we suck at it. And even people who know that still don't check the data.

Re:Risk Perception 101: People are Idiots

[Neil Boekend]

Or they'll insure themselves against the relatively small risk. Especially with decent logging most of the accidents will provably be the other party's fault.
Google will cover itself. Partly by lobbying for new laws, partly by insuring and passing the cost of that insurance on to their customers.

Re:Risk Perception 101: People are Idiots

[drinkypoo]

No mechanic would ever assume such liability. As of today, liability of most businesses is limited to monies paid.

Every mechanic assumes such liability. You can sue them in civil court for your damages regardless, if you have reason to believe that they did their job incorrectly.

Re:A breathalizer in the dashboard will do the sam

[glavenoid]

It seems like there's always some not-drunk (but not necessarily sober) jackass who is willing to cheat these dashboard breathalyzer tests for inebriated drivers who are forced to install such devices due to DUI/DWI infractions. The device may well be more simple to design and implement than a self-driving car, but their efficacy is likewise simple to undermine.

I'm holding out hope for the Johnny Cab: "The door opened. You got in". Now *that's* what I call simple!

Re:A breathalizer in the dashboard will do the sam

[glavenoid]

...by which I mean an autonomous, self-driving taxicab service staffed by a gentle robotic companion voiced by one Robert Picardo, of course.

Assuming no faults in the driving AI.

[Chas]

That's the problem.

Currently, they're looking at data for autonomous vehicles in a complete vacuum.

I'm quite sure that having such cars on the roads in percentile quantities will yield their own sets of unique fatalities sooner or later.

In the mean time, I'm not an quadriplegic. So I'll choose to drive my own damn car.

Re:Assuming no faults in the driving AI.

[ebno-10db]

Yup, GIGO. Let's assume that this tech works great under real-world conditions. Then we know that it will be a big improvement. Wow, such insight.

Re:Assuming no faults in the driving AI.

[Dunbal]

That depends. Although I agree that road conditions could, for example, become suddenly unpredictable in an earthquake or land-slide or bridge collapse and the computer couldn't handle them. But then it can be argued that a human driver would not have been able to do better, either.

Re:Assuming no faults in the driving AI.

[jader3rd]

In the mean time, I'm not an quadriplegic. So I'll choose to drive my own damn car.

But does driving your own car, instead of having a system which transports you to your desired destination, increase your chance of becoming a quadriplegic?

Re:Assuming no faults in the driving AI.

[delt0r]

A better question is, does it increases the risk of me becoming a quadriplegic because GP is a shitty driver? Statistically speaking GP *is* a shitty driver. Even more so if he/she is one of those drivers who "knows i am a really good driver....".

Re:Assuming no faults in the driving AI.

[drinkypoo]

There is one condition that the human driver handles better than the self-driving car; electrical system failure. I don't know about you, but over my lifetime I've been in at least a dozen vehicles whose electrical systems have failed while they were in service. My Mercedes doesn't care, aside from loss of signals and such. Everything else just dies. And so will you, if your hands aren't on the wheel and your feet aren't on the pedals when it happens and you're moving at a high rate of speed.

Re:Assuming no faults in the driving AI.

[anon mouse-cow-aard]

I think the point is that when an AI is driving, an accident will be a rare occurrence that people will want to understand, because it will be rare, and because the AI can be improved so that that problem does not happen again, and because if it happens for one person, the AI will do the same thing for other people, and if someone wants to sue, after the first time it happen, they will have a much stronger case for negligence on behalf of whoever is responsible for the AI.

Everything possible will be done to ensure that the same thing doesn't happen again, to avoid liability and/or a satire piece (like the Pinto gas tanks) that will kill the model involved. So a autonomous car accident will look more like an airline accident, with an investigative team coming in from wherever, and really nailing what went wrong, because just like an airline crash, hundreds of lives, sales, and big-time law suits will be on the line. That focus will change the numbers.

Look at fatality stats. When are normalized by distance travelled, it is 7.97/0.03 --> 266 times more likely to die in a car today than an airplane. That's where the direction things are going to go, only more so, because airplanes are still driven by humans, and there is an upper limit of reliability of humans that is much lower than can eventually be achieved by automation. So the airline numbers should continue to improve as automation is further refined.

Re:Assuming no faults in the driving AI.

[drinkypoo]

Computers can work for years without shutting down. *Can*. But a daily reboot for 5 minutes works for almost every computer in existence. Even general purpose ones.

The penalty for failure in the case of your home PC is quite low.

The penalty for failure in your car is not.

The best technology for cars which go where they are meant without hands on the wheel is still some form of rail. As a bonus, we can eliminate these stupid pneumatic tires completely, without resorting to [cool but nevertheless ultimately hackish] halfway solutions like tweels. Eliminating the issues of yaw and traction control completely make issues like safe stopping distance trivial.

Re:Assuming no faults in the driving AI.

[Dunbal]

Probably not, because engineering demands fail safe. The car will stop as soon as it detects that it is failing. It's not difficult to have back up systems for things like the brakes, with their own emergency alternator tied to wheel rotation or something.

Re:Assuming no faults in the driving AI.

[Nivag064]

Just keep Microsoft and Apple away from your car's computers!

Not a bus; a taxi

[Namarrgon]

Buses don't provide door-to-door, non-stop service. Taxis do - but of course now you have to cover the whole cost of the driver by yourself.

Re:Not a bus; a taxi

[Capt.Albatross]

Buses don't provide door-to-door, non-stop service. Taxis do - but of course now you have to cover the whole cost of the driver by yourself.

Self-driving vehicles will change that trade-off...

Re:Not a bus; a taxi

[Anonymous Coward]

Self driving taxis will be cheaper than owning a car.

The interest on the capital cost of owning a car is the highest single cost of owning a car for most people. The much higher utilisation of a self driving taxi compared with owning a car will mean that interest on the capital cost of the vehicle will be a fraction, perhaps one tenth or less for typical users, of the cost per kilometre travelled.

Self driving taxis will revolutionise transportation, and they are inevitable.

We just have to wait.

Um, what?

[cascadingstylesheet]

Right off the bat you can imagine autonomous driving easily topping your average intoxicated drivers' ability behind the wheel.

Um, what? Self-driving cars will drive better than drunks? That's an endorsement?

Re:Um, what?

[delt0r]

Unfortunately yes. However it won't be long before they are substantially better than the average driver. Even if by just not losing patients or getting angry or tired and not speeding.

10% of 25,580 is 2558

[__aaltlg1547]

Just sayin'.

And I don't think that 10% computer driven cars would do much to change congestion.

No they won't....

[mark-t]

Because people won't trust them.

Re:No they won't....

[Anonymous Coward]

Old people won't trust them. Young people will wonder why old people are trying to kill them with their 2-ton death rams, and ask the computer to stay off the non-automated-only roadways.

Re:It already exists!

[Anonymous Coward]

In Dallas, buses cost more than driving (even with the inflated driving costs that public transport agencies use for calculating car costs), and is slower than a bicycle. They are also inconvenient and infrequent.

Re:No thanks

[epyT-R]

Yup. There are plenty of apologizers keeping that worry to themselves just to avoid the 'conspiracy nut' label, and/or who don't care about anything besides convenience (until someone else's form of it intrudes on their own lives, of course). These people project their own whims onto everyone else and become surprised/fearful/offended when the rest of us don't step it up. If there's a root dynamic to today's societal ills, this is it.

It's one thing to automate repetitive tasks and another to automate living life; the latter being what happens when the control of this automation is handed to governments/corporates. At that point it's slavery. Because of this, these technologies only become interesting to me when the leadership and cultures of so-called free nations are sufficiently mature to understand and handle the concept of freedom. Currently, they are not, and now we see how every new device with a computer inside has some kind of remote use-tracking featureset built into it, marketed as convenience of course.

It is highly unlikely they've worked all the flaws out of these cars. The problem is just too intractable for that. The last thing I want is to hurtle 70mph down a highway under the control of cheap chinese embedded computers programmed by the lowest bidders when the manufacturers still can't get their relatively simple electronic throttle controls working right.

Re:It already exists!

[JonBoy47]

Public Transportation: A great way to get from someplace you don't live to someplace you don't work.

Don't be first!

[jcdick1]

Because you know that as soon as your car is recognized as autonomous, some asshole kid is going to say "Let's make it crash!"

Re:Don't be first!

[tftp]

Does it matter that the autonomous car will be continuously recording everything around it, and will retain plenty of that recording to put that kid in jail for attempted murder? Not too many people will dare to even approach such a car with bad intent. I'd build such a car to record everything around it all the time, even when parked :-)

Re:Don't be first!

[Neil Boekend]

He is talking about a script kiddy that takes over control while you are driving it.
That'd be a liability hell for Google and they probably know it.

Re:Don't be first!

[tftp]

The asshole kid will be sitting on the opposite side of the world.

Those cars won't be communicating over the Internet. It makes no sense. They'd use short distance, high bandwidth, low latency spread spectrum technologies. On top of the media they will run a simple mesh network with low cost of entering and leaving. What does that mean? It means that the only way you can "hack" that local area network is by physically sitting at the roadside and doing nasty things. You can try to jam the RF, however hard it is for a 100 MHz wide spectrum. The cars will simply conclude that the RF is not in the cards today, and they will continue using their own sensors. You can try to pretend to be a car that does bad things on the road... but it will be hard in itself because, if I were to do it, I would sign every message with a hash of car's VIN and the manufacturer's secret key in a tamper-resistant IC, like TPM. How many people would actually want to go to such a great length to earn themselves a prison term? It's like subverting a railway signaling system. Nowadays that kid would get an automatic conviction for terrorism - which, for a change, it may be.

All in all, a reasonably well made system can be very reliable. As I understand, TCAS is like that. How often is it subverted by "the ahole kid who is sitting on the opposite side of the world?" Can you provide even a single example?

The code running hundreds of millions of cars will need to be written to the same level of care as what they used to fly the Space Shuttle. But it won't be. It will be written to the level of "oh, we'll just patch it".

There are certain rules for developing systems that human life depends upon. Medical equipment, avionics, and the Shuttle. How many crashes of the Shuttle can you attribute to failure of a computer?

If they can't get it right for military aircraft that cost $125+ million a pop, what are the chances for your $25,000 hatchback?

The chances for the hatchback are, actually, excellent. Let's count. There are 100 airplanes made, each costs $125M. What is the total value of the deal? $12B. Now, there is a million (just one million!) of hatchbacks made, each costs $25K. What is the value of that deal? $25B. That's why you can easier become a billionaire by inventing a new toothbrush than by inventing a new missile. That airplane costs $125M, but the paperwork for it costs $100M, and you will be waiting, and on the hook, for ten years while the purchase goes through Congress, up and down, up and down, as various interests use you as a pawn in their own disputes. A hatchback? You just go ahead and make it. Elon Musk made one, not being Ford; Fisker made one, not being Ford. And, of course, every Ford of the world makes hatchbacks.

There is another effect: a $125M airplane has more or less $125M of labor that goes into it. The volume is so low that the reuse factor is not playing much of a role. Everything is invented, designed and written specifically for this airplane. However a common car will be operated by the same software as any other car or a truck - you just parametrize the mechanics of the vehicle. The reuse factor is huge. Millions of automated cars can run the same software package.

But there is yet another effect. A military airplane spends not that much time in the air. It is expensive to fly those things, and the number of hours in the air before overhaul is small. This means that very little use data, if any, is fed back to the manufacturer for improvements. Compare to an automated car. With a million of those cars on the road, the manufacturer can have access to tens of millions of hours of actual driving (with permission of the owner. But since the owner is not driving, he doesn't really care.) This information can be used to further improve the software.

I'm looking forward to buying one of those self-driving cars. Sometimes you are too sleepy or too tire

Re:Don't be first!

[tftp]

He is talking about a script kiddy that takes over control while you are driving it.

Why would a car be connected to a network? And how? Cellular networks are absolutely unusable for the purpose - they are not reliable, and they have terrible latency. If cars need to talk to each other, they will use their own radios, protocols and wireless LANs. You cannot hack into those - especially if cars are not in habit of browsing pr0n sites :-)

That'd be a liability hell for Google and they probably know it.

It will be taken care of on the level of legislatures. Otherwise nobody will be able to sell an autonomous car.

Re:Don't be first!

[drinkypoo]

He is talking about a script kiddy that takes over control while you are driving it.

Why would a car be connected to a network? And how?

How is a good question. Why is a dumb one. It has been discussed ad infinitum that having self-driving cars communicate would improve throughput.

Re:Don't be first!

[Neil Boekend]

He is talking about a script kiddy that takes over control while you are driving it.

Why would a car be connected to a network? And how? Cellular networks are absolutely unusable for the purpose - they are not reliable, and they have terrible latency. If cars need to talk to each other, they will use their own radios, protocols and wireless LANs. You cannot hack into those - especially if cars are not in habit of browsing pr0n sites :-)

Why: adverts, traffic congestion avoidance, map updates, firmware updates, Facebook and about a million other reasons.
How: cellular, in combination with wifi at the owner's home. It doesn't matter that it's unreliable.
The wireless lans you are talking about are not magical. They can be hacked in to. However, assuming Google has the intelligence they have displayed up to now, it will be difficult.

That'd be a liability hell for Google and they probably know it.

It will be taken care of on the level of legislatures. Otherwise nobody will be able to sell an autonomous car.

Partly it'll be taken care of by legislature. However, I assume they are smart enough to secure the data streams properly.

Re:A breathalizer in the dashboard will do the sam

[mspohr]

I'm not familiar with the dynamics of this situation but it seems to me that if there is a not drunk person who could pass the test that they would logically be the one to drive the car. I do, however, understand that "logic" doesn't always apply in these situations.

Insurance

[Dan East]

This topic has been discussed here several times now, but one thing I haven't seen brought up is insurance. If my vehicle is driving itself and causes an accident, then what driver is to blame? The person sitting behind the wheel? Why would my insurance company want to pay for an accident caused by a piece of software when they can go after the company that produced the software? Or what if they will only insure Ford cars and not Chrysler because statistics show that one auto-driving system performs better than the other? If my car's autonomous system just flat out runs over a little girl playing in the street and kills her, could I be charged with manslaughter because I was behind the wheel reading the newspaper?

Think back a few years to the Toyota "auto acceleration" issue, and the lawsuits and government testing, etc, etc that was going on over that one issue. And that was possible hiccup in a single system that merely relayed user input to the engine. It wasn't even remotely as complex as a vehicle actually driving itself.

There's going to be a whole lot to figure out in the legal, insurance and liability areas that makes the technical challenge and development look like child's play.

Re:Insurance

[SeanBlader]

Self driving cars do not cause accidents, therefore insurance isn't necessary. Autonomous cars are such a huge game changer in society because of the number of ancillary things that go away because of them. Traffic cops, car insurance, taxicabs, truck drivers, all disappear. It's the next massive paradigm shift in world society, at a level comparable to the changes brought on by steam and electricity. The effects on the global economy and society won't be fully understood for decades afterwards. Flat out, it's going to be huge.

Re:Insurance

[CastrTroy]

The problem is, that kind of autonomous car isn't here yet. There is no autonomous car that doesn't require a person to be sitting in the driver's seat ready to take over in case something goes wrong. That's why I think this whole thing is stupid. If I have to be sitting in the driver's seat, paying attention to the road, I might as well be driving. Because if the car is doing most of the driving, it's more likely that I won't be paying attention when something bad happens. Until the cars are good enough that they don't require a steering wheel, gas pedal, or brake, I don't see much point in them.

Re:Insurance

[Dan East]

Self driving cars do not cause accidents, therefore insurance isn't necessary

That's ridiculous. Things will happen to autonomous vehicles that will result in deaths and destruction of property, even if 100% of vehicles are autonomous. Insurance will not go away because the stakes are too high both with liability and the cost of the hardware involved.

Re:Insurance

[Dunbal]

No taxi drivers? What the hell will all the immigrants do for work?

Re:Insurance

[profplump]

If only there were some other highly-automated transportation system in place we could use as a model. Maybe set up some sort of mandatory validation methodology for the control systems and a post-accident review system to assess problems found in the field. If we nationalized those services they might be called the Federal Automated Automobile Administration and the National Automative Transportation Safety Board. But that's just silly I know -- these problems are totally new and we are completely unable to find ways to deal with them, so we should just stop now.

Re:Insurance

[cohensh]

I don't think this is as big a problem as you expect. First, there will never be a day D when there aren't autonomous cars, and a day D + 1 when there are. They're going to be integrated slowly and continuously. Think, cruise control, adaptive cruise control, lane keeping assist, automated parallel parking.

In the near future you won't have driverless cars, both for technological and social/legal reasons, any commercial car will require a person in the driver's seat. That makes liability easy, whoever's in the driver's seat is responsible for the car and its actions. It's exactly the same as if you turn on cruise control in a "non-autonomous" car and rear end someone on the highway. You're liable, even though you're using technology.

The auto acceleration isn't really apples to apples with this. That's just a potential product failure (Like car's catching on fire when they get rear-ended). It doesn't really apply to autonomous cars more than non autonomous cars, except for the potential "autonomous cars are more complex", which is questionable anyway, considering how complex current cars are.

Re:Insurance

[delt0r]

Mechanical failures are not new. Breaks can fail, tires can blow out etc. These things have happened and have killed people. This is really not new. And software doesn't change it, since the software is part of the car... ie ABS breaks can have software faults.

Re:Insurance

[u38cg]

Insurers won't have a problem, really. Overall, claims will decrease. There will be some interesting liability issues, but those are par for the course in insurance anyway.

Re:Insurance

[Firethorn]

There is no autonomous car that doesn't require a person to be sitting in the driver's seat ready to take over in case something goes wrong.

That's due to legal requirements to be on the road. They're still in the research phase after all. Even a self-driving car that's limited to highway driving for 'no monitoring needed' would still be highly useful.

Personally, I think that they're less than 5 years off, so it's time to be really working on the legal aspects.

Re:Insurance

[trout007]

Quite a few of the crash tests are done by the I Insurance Institute for Highway Safety (IIHS) http://en.wikipedia.org/wiki/Insurance_Institute_for_Highway_Safety

The insurance companies pay for it so they better understand the costs involved in insurance different cars. I don't see why they wouldn't do the same thing for software.

Re:Insurance

[Anonymous Coward]

Indeed. The no insurance thing is a pipe dream. However insurance costs may decline (or simply be lower than manual cars) if the autonomous cars are indeed safer.

Re:It already exists!

[n1ywb]

You, obviously, live in a major metro area. Plenty of people don't, and have no viable public transportation options, besides perhaps hitchhiking.

Re:A breathalizer in the dashboard will do the sam

[epyT-R]

Except that a johnny cab done today would report your travel plans to the local police dept, insurance company, and any other institution that has a vested interest in judging your behavior. No thanks. I'd rather walk.

If cars were like computers...

[Shompol]

This is very old, but in some wierd way relevant. In fact, #10 already materialized.

If cars were like computers
If General Motors had developed technology like Microsoft, we would all be driving cars with the following characteristics:

1. 1. For no reason whatsoever, your car would crash twice a day.
2. 2. Every time they repainted the lines in the road, you would have to buy a new car.
3. 3. Occasionally your car would die on the freeway for no reason. You would have to pull over to the side of the road, close all of the windows, shut off the car, restart it, and reopen the windows before you could continue. For some reason, you would simply accept this.
4. 4. Occasionally, executing a maneuver such as a left turn would cause your car to shut down and refuse to restart, in which case you would have to reinstall the engine.
5. 5. Macintosh would make a car that was powered by the sun, was reliable, five times as fast and twice as easy to drive -- but would run on only five percent of the roads.
6. 6. The oil, water temperature, and alternator warning lights would all be replaced by a single "General Protection Fault" warning light.
7. 7. The airbag system would ask "Are you sure?" before deploying.
8. 8. Occasionally, for no reason whatsoever, your car would lock you out and refuse to let you in until you simultaneously lifted the door handle, turned the key and grabbed hold of the radio antenna.
9. 9. Every time GM introduced a new car, car buyers would have to learn to drive all over again because none of the controls would operate in the same manner as the old car.
10. 10. You'd have to press the "Start" button to turn the engine off.

Re:If cars were like computers...

[istartedi]

You left out the part where the car goes a million miles per hour, runs on electricity from a rechargeable battery, and costs $100. Unfortunately, it fits in the palm of your hand.

Re:If cars were like computers...

[blahplusplus]

He also missed the problem that CPU's have to do a lot and programs running on them are written by humans. Programs and power are about trade offs. Everyone forgets that. The problem was never computers but the more power you want and less static the system (i.e. more open and changing and dynamic) the more potential problems because of limits on resources and time given the modern complexity of software and trying to solve difficult not easily defined problems.

Travel is a known quantity, programming in the universe of infinite problem sets are much harder to tackle especially when you have programs written by different people interacting on the same system.

Re:If cars were like computers...

[rock_climbing_guy]

Start Button? No, you have to go to move the cursor to a Hot Corner, clicks Settings, Power, Shut Down.

Moving off from the lights all together.

[Circlotron]

It would be great to see a line of 100 cars all beginning move forward at the same instant instead of stretching out, and you moving forward only after the light has turned red again. Much greater throughput at the intersection. Also it opens the possibilities of a long line of cars safely tailgating, therefore slipstreaming one another and saving fuel. Being out at the front of the line would no longer be desirable.

Re:Moving off from the lights all together.

[sylvandb]

I have a better idea... How long will it be until there are enough autonomous cars on the road to create a cooperative environment for my meat-driven car? If all those autonomous cars sense that my car is deaf, blind and dumb, and the smart controller is nothing but meat, it seems they will get out of my way and give me a bit of margin just out of healthy sense of "avoid an accident." That means I can jump the green light, snake around on the right, and in general drive in a way that is barely on the right side of legal and which would very unsafe if I could not predict the reaction of the other drivers. But if those other drivers are machines, even if I cannot control them their algorithms have to be predictable and they have to allow for unexpected human action.

and 1 bad accident / death will lead to a lot of c

[Joe_Dragon]

and 1 bad accident / death will lead to a lot of time and money in the courts??

and will they be able to have some outsourced coders be forced to come to court / how much will the courts like to have to deal with a big list of contracts / Sub contracts to get to who did what piece of the over all system.

airplanes autopilot still don't cover all stuff

[Joe_Dragon]

airplanes autopilot still don't cover all stuff and they have less to deal with then a car does.

Re:airplanes autopilot still don't cover all stuff

[jader3rd]

airplanes autopilot still don't cover all stuff and they have less to deal with then a car does.

Which is why getting your pilots license is easier than your drivers license!

What does the Self Driving Car do if?

[PastTense]

What does the Self Driving Car do if there is a young child or dog playing along the edge of the road?

Is this what you would do?

Re:What does the Self Driving Car do if?

[Dunbal]

It would be able to stop much faster than you if the dog or child ran out into the road.

Re:What does the Self Driving Car do if?

[sylvandb]

It would be able to stop much faster than you if the dog or child ran out into the road.

No, it would not.

It's initial application of the brakes would happen much faster. That reaction time to apply the brakes is a minute fraction of the time it takes to stop the car.

Re:What does the Self Driving Car do if?

[Neil Boekend]

A fraction yes. A minute fraction NO.
A car going 80 km/h (50mph) with a driver with a reaction time of 1s needs 22m (67 feet) to react. The worst case scenario for the breaking distance at 80 km/h is 38m. That means it's 22/60=0.37.
Not a minute fraction, but a large fraction. That is: unless you are driving a road illegal car (because it brakes to slow) at mach 1 of course.

Re:What does the Self Driving Car do if?

[jecblackpepper]

The autonomous car would be able to quickly decide whether coming to halt in its current lane is safer or if manoeuvering into other spaces to avoid the child/dog is safer - and probably much more easerly than a human driver who can freeze in panic at an unexpected situation. Additionally the computer can potentially make the cold hearted decision that breaking but still hitting the obstacle, because there is not enough stopping distance, is ultimately the least bad thing to do - for example if there is heavy traffic in the on-coming lane and pedestrians besides the road.

It could also, for example, automatically sound the horn to alert people and perhaps enable the child/dog to notice and get out of the way. A human on the other hand is more likely to be still be evaluating the situation and reacting by slamming on the breaks and not have the foresight (or the co-ordination) to do this.

Re:It already exists!

[CastrTroy]

This is why I cycle commute. Only marginally slower than a car (sometimes faster when there's heavy traffic) and way faster than the bus. The longest cycle commute I ever did was 25 km, but that was only for 3 month. It took me 1 hour 15 minutes tops. However, due to traffic, the same trip in a car took about 1 hour. Driving would have only saved me 15 minutes each way. And the standard deviation on bike is much smaller. Pretty much the same trip time every day. In a car, traffic could be really bad, and it could take an hour and a half, or 45 minutes, you never really know until you reach the destination. Now I'm only 7 km from work and cycling is great. 15-20 minutes and I'm at work. A car would take at least 10, so I'm really not using any appreciable time.

Re:It already exists!

[PhrstBrn]

It really depends. Where I work, parking is awful and I have to walk 5 blocks from a parking garage to my building, because all the parking garages near my building are full and have a long waiting list (apparently the waiting list is 20 years long for the better parking). Meanwhile, the bus only takes 5 more minutes than driving, and it drops me half a block from my building. So in some cases, busses can be literally faster. To be fair, it also helps that I'm only half a block from a main road, and that busline happens to go directly to my workplace.

Re:It already exists!

[CastrTroy]

I find that unless you go long distances most of the time riding the bus is actually waiting for the bus. For me, it takes me 5 minutes to walk to the bus. I have to wait on average 5-10 minutes for the bus. You can't plan to arrive 2 minutes before it get's there, because half the time it will be early, and then you have to wait 15-30 minutes for the next one. Then I take a 5 minute bus ride, get off, and wait for the next bus. That's another 5-10 minutes. Take another 10 minute bus ride, and I'm at work. So 5 minutes of walking, 10-20 minutes of waiting for the bus, and 15 minutes of being on the actual bus. 30-40 minute trip and half of it is spent waiting for the bus to get where I am. And people wonder why I ride a bike. Spend less time riding than I do waiting for the bus.

Why do you think this means people are idiots?

[King_TJ]

I'd say it's more a matter of people being HUMAN. Humans have a whole range of emotions too, which often prove detrimental or at least reduce efficiency at attaining the desired outcome in a particular situation. Should we just eliminate all those pesky feelings too and become strictly logical?

I think we all realize we're going to die eventually one way or another. When it comes down to it, we're generally far more okay with it happening because we made a mistake while doing something we enjoy (or even something we felt was productive, necessary or just a good idea to do at the time) than because of something not at all under our control (machine error or malfunction). The risk of injury or even death is in some manner, part of what makes things worth doing for people. (Why do people enjoy thrill rides at the amusement park, or parachute jumping out of perfectly good airplanes? Or even for the more risk-adverse among us, why do some people like to gamble at the casino once in a while or follow the stock market?) Life without risk would be incredibly boring. Driving a car is one of those calculated risks people take all the time because the benefits (fast travel and freedom to get from point A to B when YOU want to leave) feel worthwhile. But passively sitting in some machine that takes a person there won't involve any of the joys of operating the vehicle anymore. That's going to be a big downside for people who went through the whole process of learning to drive, achieving a license to do so, and investing years in what they probably believe made them a better and more skilled driver with all the practice.

Re:Why do you think this means people are idiots?

[ultranova]

I'd say it's more a matter of people being HUMAN. Humans have a whole range of emotions too, which often prove detrimental or at least reduce efficiency at attaining the desired outcome in a particular situation. Should we just eliminate all those pesky feelings too and become strictly logical?

No, it's not a matter of being HUMAN, it's a matter of being immature and ill-trained. The problem is not having emotions, the problem is having insufficient training in using them - basically, you are letting them react to a fantasy ("I'm a better than average driver, and can avoid a crash no matter what") rather than reality ("I'm an average driver, and the computer is a better driver than me"). Consequently, they end up giving bad recommendations, leading to seemingly irrational behaviour. Garbage in, garbage out.

Delusions of grandieur are like beer of the soul: they're tasty and make you feel good, but also unhealthy, addictive and impair judgement. They'll kill you if you overindulge. And if you combine them with driving, you're a threat to everyone on the road.

So no, we shouldn't eliminate "all those pesky feelings", but rather learn to query them about the actual logical consequences of our choices. Life isn't Star Trek; being logical in no way conflicts with having emotions. And irrationality - being bad at achieving your goals - is by defintion never a virtue, altough I suppose it could be a good thing for other people, depending on what those goals are.

But passively sitting in some machine that takes a person there won't involve any of the joys of operating the vehicle anymore. That's going to be a big downside for people who went through the whole process of learning to drive, achieving a license to do so, and investing years in what they probably believe made them a better and more skilled driver with all the practice.

And not having to drive anymore is going to be a big upside for those of us who invested all that time simply because we need to get from point A to point B at will, and are rational enough to recognize sunk cost fallacy as such.

Who will or won't have these

[sfm]

> There will never be computer driven cars for the masses.
> It will always be cheaper for them to drive their own.

And more fun !!!

Reality vs Ignorance and inertia

[EmperorOfCanada]

Where this is going to get interesting is when nearly all the cars on the road are autonomous and the last remaining hold outs will be preventing many other cool solutions that only work when you have 100% autonomous such as eliminating traffic lights. Eliminating traffic signs such as one way, speed, stop, etc signs. Eliminating speed limits. Even eliminating things such as lanes.

Basically the last manually driven cars will be seen to be a homicidal menace and high cost nightmare.

Re:Reality vs Ignorance and inertia

[liwee]

Like medicine, there are usually side-effects to every thing in life. Ease of online transaction comes with the possibility of financial fraud. I am not denying the possible benefits of 100% autonomous system, but in the event of a malfunction, the risks of death/injury is very real.

Re:Reality vs Ignorance and inertia

[EmperorOfCanada]

Yes I suspect for every 10,000 lives saved there will be one death. But a fantastic death; where the car will just drive a family off a cliff or something sensational that the media will be all over. Then for a while people will be, "I'm not getting into one of those death traps."

Plus there will be a huge number of special interest groups who will desperately try to keep drivers in cars: Taxi unions, truck driver unions, bus driver unions, classic car associations, the police (if the drivers are perfect then no more tickets), and even groups like MADD might find themselves without a mandate if there are no "Drivers". But then you get more subtle groups who will lose their minds: many small municipalities coffers will become empty if they can't be handing out fines. Even larger governments might discover serious drops in revenue without any ticket revenue coming in.

And even groups like the police will be ticked that they can't pull "suspicious" people over by just waiting for them to make a traffic mistake.

Then you get stores and other commercial areas that have made based their financial model on easy parking, but if you are using either your own robot car or more specifically a cheap robot taxi then you can get dropped off in the most dense parts of downtown and go to your specialty stores and when done get picked up at the push of a smartphone button. So if these groups realize the threat to their business models then they too will squeal.

But on a side note one of the biggest threats to life and limb posed by robot cars will be the potential for a drastic reduction in the average distance walked. I can see some people integrating a robot car so much into their movements that they step out their front door into a waiting car, it drops them off at the front door of destination one, picks them up at the front door when they are done, and this would continue until they are eventually dropped off at the front door of their house. Whereas right now they might have to walk from their parking, walk among a cluster of destinations, and then walk back to their car.

This whole lack of walking could turn out to be more deadly than the lives saved through car accidents. At least with no-walking deaths it will be people doing it to themselves vs car accidents often killing other innocent people.

Re:Reality vs Ignorance and inertia

[Anonymous Coward]

When we get to a certain tipping point, you could just mandate a dash mounted panel which could tell the human driver to stop, what the speed limit is, etc. Not a big problem, we could pretty much do that now. And slap a beacon on it, all the other cars would know it's a hazard; hell, even without a beacon the other cars could just treat it as any other hazard, like a fast moving cow on the road or something.

Re:Reality vs Ignorance and inertia

[Attila Dimedici]

Once adoption of autonomous cars passes a certain level, manually driven cars will be outlawed for everyone except the rich and politically powerful. The government will love the fact that your car will only take you to places which they have designated as appropriate for you to go to. Any other location which you may wish to go to will not exist as far as your car is concerned.

Re:Reality vs Ignorance and inertia

[psydeshow]

This whole lack of walking could turn out to be more deadly than the lives saved through car accidents. At least with no-walking deaths it will be people doing it to themselves vs car accidents often killing other innocent people.

Then again, imagine how much safer it will be to skate, skateboard, cycle, fly kites, walk the dog, or participate in just about any other form of exercise that happens to take place on or near a roadway. If cycling deaths drop with the same rapidity as automobile deaths, cycling will become A LOT more appealing to risk-averse people. Walking/cycling/etc to music will also become safer, and music encourages more strenuous activity.

There is also every chance that you'll get more casual exercise, just by having more free time and less stress.

Re:Reality vs Ignorance and inertia

[EmperorOfCanada]

Yes this is going to be horrible to watch the petty abuses of power. One that I have long predicted would be that rich neighborhoods will not allow cars through them if you haven't set the neighborhood as the destination itself. What I am specifically referring to are neighbourhoods where people are presently driving their cars through legitimately. For instance to get to the nicest park in the city (and biggest) you presently have to drive down the most expensive street in the city for around 5 blocks. There are other ways to get to the park but I would imagine 80%+ go down this street. I can see these people dropping hints to their politician friends that they want the legal navigation parameters changed.

A simple example of this was when John Kerry apparently had a fire hydrant moved from infront of his Boston house. Apparently this is near impossible to get done (administratively).

But the other one that I see are some kind of priority driving for politicians. Traffic would be ordered to part like the red sea for them; soviet style. These sort of rules would be needed for things like firetrucks, ambulances, and even trams. So I can see the politicos thinking that they are just as important.

Re:No thanks

[profplump]

Where do you drive that they've worked out all the flaws in human-controlled cars? Or are you just ignoring the status quo to complain about the scary new system?

Also, what part of your argument couldn't be equally applied to the transition from human-and-animal-powered transit to steam-and-oil powered-transit? We've made these changes before. Most people are happy with them. I suspect you're happy with them. Why is this set of changes different and objectionable?

Re:It already exists!

[Anonymous Coward]

Perhaps, but then you have to sit next to people.

And some of them might be xenophobes!

As much as I love tech. This is bad

[dinther]

1 - If the car hits someone. Who is responsible
2 - If the car hits another autonomous car who is at fault.
3 - Imagine the much more complex and costly process to sort out damage claims.
4 - Strict standards and regulations will be required. This of course means less freedom.
5 - Government will want to switch off your car when you don't comply. For safety of course.
6 - The NSA and FBI will get their hands on those switches and do with you as they please (Movie: Fifth Element)
7 - The perceived benefits are so great that soon manual driving is banned.

Result, cost goes up freedom goes down. As much as I love my car to take me home after having a few too many drinks, I think I prefer to take a cab and retain my freedom or what is left of it.

Re:As much as I love tech. This is bad

[profplump]

We can regulate automated cars and deal with accidents the same way we already deal with automated planes (and trains and other such things) -- with regulation on the systems qualified to control the vehicle (FAA) and investigations into accidents (NTSB). We'll need some new rules, but the general problem space is well understood and already regulated.

Beyond that it's not clear to me why the government would be in any better position to disable your self-driving car than your human-driving care -- if they're going to start install remote-kill switches there's nothing that would prevent them from doing it on today's cars. I can see why you're concerned about the general concept, but I don't understand why you're attaching it to self-driving cars.

You also might consider how removing drivers from cars will change the economics of car ownership, cabs, bus routes, etc. -- I'm not so sure many people will ever own a self-driving car, because why would you want to hassle with it when you can just rent one when you need it and get door-to-door service? If you drive a lot, or if you really like the idea of being able to leave things in the car, you could still buy one, but unlike today car ownership wouldn't be clearly necessary for people who drive only tens of hours a week.

It's also worth noting that the "freedom" and "low cost" you have today has a significant cost in human lives and economic productivity -- I know dead, poor people don't get paid much attention, but I suspect they'd argue that not having to buy a car in order to get to work and not being dead due to human drivers would *increase* their freedom and *decrease* their costs.

Re:As much as I love tech. This is bad

[dinther]

I used to fly 737's and although that was 17 years ago, we had glass cockpits and auto-land. It was easy to see how planes can fly automatically, yet today pilots are still very much involved with the flying process. Highly trained and checked every 6 months.

Back to Joe Blogs in the autonomous car, it would be naive to expect a similar level of vigilance. Hell, I am sure I would zone out too.

But I do agree that the shape of transport would change. And why not. The car has only existed for about 100 years. That is not that long in the scheme of things. I can easily see myself walking along pressing a button on my phone calling for a car and just hop in a few minutes later. Cars could park themselves tightly in out of view underground car parks. It would certainly make a city look a lot nicer. In this model my arguments would not hold since the transport provider becomes responsible for the vehicle performance and safety.

I would love to see a development where such a transport company becomes so good that people voluntarily get rid of their outdated cars.

Re:As much as I love tech. This is bad

[umafuckit]

1 - If the car hits someone. Who is responsible 2 - If the car hits another autonomous car who is at fault. 3 - Imagine the much more complex and costly process to sort out damage claims.

I've seen this brought up on /. before, but it's largely not a concern because there will no longer be "fault" as we think of it today. e.g. if two autonomous cars hit each other then your insurance company will just pay you and theirs' them. The exception will probably be if the crash occurred because you ignored the need to replace a faulty GPS, or similar, in which case your insurance pays the other party. The point is that things will likely become a fight between you and your insurer rather than between your insurer and the other party. Thus, claims may well be more simple. The complexity may occur, admittedly, if the fault is traced to something the manufacturer was responsible for, but that's true now as well.

If the autonomous car hits a non-autonomous then we'll probably see some hybrid of what happens today. Hitting a pedestrian or cyclist may be more complicated. Autonomous cars will almost certainly have to adopt the Russian model: having at least one camera on board to record potential accidents (again, making claims easier to deal with). If a cyclist breaks the highway rules and gets hit, then the accident is probably their fault. If an old lady who can't hear properly walks into the path of an autonomous car and gets hit (despite the car attempting to avoid a collision) then nobody is to blame and the costs might be split between your insurer and her health insurance or perhaps it's all down to her insurance. Again, however, incidents such as this happen today as well.

Re:As much as I love tech. This is bad

[Ardyvee]

This is where I see it going, to be honest. I would not expect to see a world rid of human-driven cars (just as we still see people use arrow and bow to hunt), but the majority of the cars will be autonomous. The upside: less cars, more people share the cost of maintaining the car, no need to "buy" it, shift of the responsibility of maintaining the good conditions of the car shift towards a company dedicated to that instead of average Joe that has a life besides the car. The obvious changes these upsides would have in our daily lives are surely easy to see.

All I would like is the ability to still drive around on my own, regardless of the dominance of autonomous cars.

Re:No thanks

[epyT-R]

I'm not the one claiming they do. There've been a lot of articles over the last few years talking them up. Laws are already being passed in various states authorizing these things to run on public roads. Politicians are treating them as though they're ready. Considering the poor state of heuristics and computer-driven contextual awareness in other areas, there's no reason to expect any better from these cars. They are a threat to their riders and anyone else on the road.

Because that's a false equivalence. The uncertainties of animal intellect were masked by the relatively slow speed and sparse population of ridership. In fact, these cars are closer to having that uncertainty coupled with a body that can move at 70mph.

Re:No thanks

[epyT-R]

I misread your first sentence.. They haven't, but humans are still far better at contextual awareness than current AI and sensor technology is. Computers are faster, yes, but I'd rather have a human at the wheel that makes the right choice most of the time than a computer that is only semi aware all of the time. If you're concerned about drunk drivers, you should be concerned about these autonomous cars as well. AI and sensor technology just isn't there.

Re:A breathalizer in the dashboard will do the sam

[JamieIanMacgregor]

They'll just make it a game to have their young innocent kids blow in the device, no win here.

Re:It already exists!

[Belial6]

What wold you do with a $2,000 per year raise?

Buy a car and quit taking the bus?

(Sorry, I couldn't resist.)

Re:It already exists!

[Miamicanes]

Anybody care to guess how long it'll take cities like New York to pass a law making it illegal for driverless empty cars to follow any route besides one leading directly to a parking space somewhere, to avoid having 40,000 driverless cars doing laps around lower Manhattan for hours at a time since it's cheaper to run the car for 2 hours than to actually pay to park for two hours?

I can definitely see driverless cars causing massive collapse in downtown parking rates across America. In a city like Miami, the difference between $2-4/day parking (in a reasonable neighborhood) and $17/hour parking is usually about 3-4 miles, max. I can also see lots of tension as urban residential neighborhoods a mile or two from the skyscrapers that traditionally had adequate curbside parking suddenly find themselves inundated daily with self-driving cars looking for a cheap place to park.

Re:A breathalizer in the dashboard will do the sam

[BlueStrat]

I'm holding out hope for the Johnny Cab: "The door opened. You got in". Now *that's* what I call simple!

Yeah, but did you forget how that ride ended at the quarry? Ol' "Johnny" nearly ran Ahnold over and then exploded in a giant fireball!

Now, movie pyrotechnics, CGI effects, or other silliness aside, the fun happened after Ahnold damaged "Johnny".

What happens in highway traffic at 60-70 mph if the system sustains sudden damage and/or major malfunction, like losing all sensors, an electrical wiring system short, sudden power-loss, or having a servo suddenly lock steering to one extreme of it's range? How much trouble and expense will be put into fall-back systems?

All that doesn't even begin to address the problems with integrating autonomous and human-driven vehicles, which absolutely must be done. Banning human-driven cars/trucks/etc in the US would be ridiculously impractical from logistical, economical, and food supply perspectives, as well as being nigh-impossible to do and/or enforce, even if military force were used. The US is really, really big...like large areas where you'd be only person in a hundred miles or more in any direction, big. The US can't even control Afghanistan.

In the US there are far, far too many people living in areas where autonomous cars would not work, like places where roads changed after every good rain (some rural areas in MS, for example...been there) but where cars/trucks are still required to sustain a level of modern civilization. Many of the plains States, desert-SW States, and Rocky Mt. States have large areas where I can't imagine any of the current or even the third/fourth-gen autonomous cars being capable of replacing normal vehicles.

The poor, logical, predictable computers & software in an autonomous vehicle at this stage/level of tech would not stand a chance in hell of being able to deal successfully & safely enough with human drivers who are all the opposite things, will still be driving things like a 1970 Plymouth 440 'Cuda with factory Super Commando Six-Pack carbs and with nothing more sophisticated than electronic ignition and an AM radio for "electronics", and who are not going away.

http://www.youtube.com/watch?v=VzHxZ2f7HwY

Strat

I'm not going to buy an autonomous car

[bluegutang]

Instead, I'm going to rent one. For half an hour each morning and half an hour each evening. In between, the car will drive other people to their destinations. It will never (well, rarely) be parked on a street curb or in a garage just taking up space. It will function like a taxi, except MUCH cheaper since by far the largest expense in taxis is the driver's salary.

That's the future. Owning your own autonomous car will still be possible, but why would you do it when you can have the same convenience from a shared vehicle at a fraction of the cost?

Re:I'm not going to buy an autonomous car

[Neil Boekend]

Many people will want to own one, because it does not have the same convenience.
Assuming nobody checks the cars between rents the mess in the cars will be a problem. They'll be smelly from all the trash on the ground.

Re:what's acceptable

[Deluvianvortex]

the Boeing 777 can take off, cruise at altitude, change course, and land all autonomously. There's a flight crew but they mostly just taxi and make sure nothing terrible happens (nothing terrible has happened yet). Never say never.

Insurance complexities.

[Firethorn]

I think the insurance companies will gladly support the new cars.

I do as well, but there are still some extreme complexities.

The vast majority of auto accidents today are easily blamed on a driver, making the manufacturer not liable. However, consider the Toyota acceleration problem. What if somebody experiencing that ended up rear ending another car, pushing them into traffic where they are then struck again, increasing injuries?

That's a case where the manufacturer could end up being liable again. They're still a huge company, so it's generally not worth suing, but given enough auto-driving cars with accidents arguably blamable on the AI driver I could see a class-action starting up. As a result I could actually see 'liability insurance' baked into the cost of the system.

While most people carry insurance limited to around $100k at the top end per person, there is no such limit in a lawsuit. So that could drive costs for the manufacturer through the roof.

On the other hand, if accidents are truly 80% less likely than traditional drivers, you'd have to increase maximum liability by more than 5 in order for it to actually end up costing the consumer more, as each subsequent $ of insurance is cheaper than the last, because it's less likely to be used. Just because you might carry a $1M umbrella policy doesn't mean that somebody will get $1M if you cause them to break a finger(for example).

But from a business standpoint I could see car companies pushing to have their liability limited - in this day and age, with limits where they are now, the 80% thing, and such, I think $1M per person limit from congress would make the businesses looking to put out autonomous cars seriously consider it.

That's more than enough to drive a business out of business if it's auto-drive system is truly flawed, but more than enough to screw those HIT by bad drivers less than they currently are, while saving drivers/riders enough money to save them money in the saving of lives.

Re:Insurance complexities.

[tftp]

As I said elsewhere, the liability will be taken care of. Otherwise nobody will be able to sell such a car. You can never guarantee that the car will always pick an option that is going to save the day. There may be no such option - if, say, a tractor-trailer jackknifes in front of you, or loses a wheel that proceeds to strike your car; or a tire blows up on your own car, on an icy road. We are dealing with nature here. One more or less clear liability would be for a programming defect that directly led to an accident. However, unlike hardware defects, software can be easily updated by the customer. (Fail to do that, and bear the liability yourself.) This means that every bug will be triggered only once - and eventually you run out of bugs.

A taxicab driver works a long shift, gets tired, and still his business is profitable. Perhaps if the vehicle is not too badly controlled by the computer then it will be safe enough so that the liability will not be a major factor.

better than drunks

[Firethorn]

Well, yes. Not enough data, low hanging fruit, and all that.

The first chess programs couldn't beat a human familiar with the rules. Today they have machines that can beat the best chess minds in the world.

I predict first we'll see a widening of 'assisted' driving - monitoring for tailgating shifting towards 'automatic follow', automatic braking if something intrudes into the road, etc...

Then we'll have the equivalent of early drones - capable of traveling, from parking area to parking area. Parking(or at least selecting a spot)/landing depends on you. Eventually those situations will be less and less(they can now land on their own), and they'll start showing up in places like rich people's cars and highway trucks. The trickle down will continue, hitting the 'worst' drivers first - drunks, etc...

Networked cars

[Firethorn]

Why? In order to share traffic data, routing intentions, etc....
How? There's a lot of options - infrared, radio, directional, non-directional, etc...
Hacking? Still possible, just means that you also need to write something to 'speak' in whatever protocol the cars are using. Though I wouldn't assign a huge amount of trust in the signals. Each car should still assume that the other cars are likely to behave erratically and move accordingly.

Re:No thanks

[Anonymous Coward]

If only they were as mature as you! Hey Einstein, have you looked at your ECU recently? You already hurtling down the highway under the control of cheap chinese embedded computers.

Re:It already exists!

[Zouden]

Why wouldn't the cars just wait at home?

Re:The econmoics don't make sense

[jecblackpepper]

Except that no-one is going to replace a perfectly good vehicle with an autonomous one unless there is a benefit to them. It will be that pepole just move to an autonomous one next time they replace their vehicle. More likely the trend will actually be more and more automation in "normal" vehicles each generation so that everntually the majority of vehicles are autonomous (or capable of being autonomous).

Alongside that will the people who actually see an economic benefit from an autonomous vehicle. If it means that they can spend more time with their family or doing things that are more productive during their commute/travel then they'll buy an autonomous vehicle.

So it's not just the saving from the purported reduction in accidents that makes the economic case for autonomous vehicles, it's the whole equation

Re:It's fine as long as the government goes away

[EmagGeek]

Government will be involved, as Google and other Autonomous Car manufacturer will lobby government to may autonomous cars mandatory, "for the children," of course...

Hacker mentality

[Toshito]

I have a hard time to understand most of Slashdoter's stance on self driving cars.

When we're talking about almost everything else in the tech world (computers, operating systems, etc.) you all want to keep complete control. That's why most of you despise tablets, locked gardens, OSes that keeps you out of the loop.

Well for me it's the same thing with cars, or anything in my life. I want to keep having complete access to it's innards, and complete control over what it does. I don't want a car with a locked down hood, encrypted software to keep me from making repairs or modifications. And I sure as hell don't want to be driven on the road by a buggy software, developped by a company which cut corners just to make it's investors happy.

Given the unreliable state of almost every gadget that is controlled by software, I would not bet my life on it.

Re:A breathalizer in the dashboard will do the sam

[BlueStrat]

There will doubtless be considerably more safeguards built in than there are in the current drivers...

That has not been established and is far from "doubtless".

Autonomous cars will be vastly inferior, operationally-unreliable, and much less safe in the uncontrolled and chaotic real world unless & until we achieve a level of true AI indistinguishable mentally from a sentient being like a human, which is currently the state of the art at dealing with strange, unexpected, and unknown circumstances/situations and making intuitive leaps of understanding/prediction/solution. And, they must do it cheaply enough to have a "Cortana" in every autonomous vehicle.

To say that current, or even the next two generations, of autonomous vehicle tech is ready for the "rubber to meet the road" and replace manually-driven cars and trucks in the real world is pure delusional fantasy, unfortunately.

Even if it were practical to do from a tech standpoint, I'm worried about the loss of individual privacy and freedom , and the amount of control over individual lives and the variety of personal choice in how & where someone lives that people would necessarily have to surrender to make such a system work.

Think about this. Journalist plans to publish a story on some government cover-up/atrocity/corruption/etc, hops in his car (which has been reporting all his travels, revealing his informants and his intentions) and tells it to take him to his publisher's office, but it locks the doors and takes him straight to a TLA-determined destination instead where he is "disappeared". Think Michael Hastings on a national scale. Or plain old criminals/criminal gangs hack the system.

Just because something *can* be done, does not mean that it *should* be done. "Those who would sacrifice liberty for security (or safety)..."

Strat

'Save Money and Lives'

[l3v1]

While I could definitely see the benefits of eliminating unnecessary risks from the road, my problem with approximations is that they seldomly come true. While the number of accidents might decrease, that doesn't mean the number of incidents will. E.g. autonomous vehicles can also brake down, algorithms can go haywire, situations can occur when prompt human intervention would be important (and won't happen since they are not paying attention), etc. etc.

One situation where I'd really welcome more self-driving vehicles is in southern CA during rains. Yes, they are rare, however, most people drive like idiots during rainfall, and accidents become much more frequent. Probably because they have so little experience in driving in rainy conditions, so they drive too close to each other, or drive too fast, or drive to slow, and one more thing: they make me lose my temper :) which I really don't like. So yeah, bring self-driving cars on.

Just please, pretty please, make'm so that the autonomous driving mode can be disabled for the times when you actually want to enjoy a nice driving-around from time to time.

Enabling more privacy invasion

[JeffOwl]

Wait until you read Google's terms of service for their automated car. They will be able to track who is in the vehicle by cell phone initially but will use weight sensors (that currently enable the air bags) and other biometric cues to correlate with a specific person in case they leave their phone behind. They will know in detail everywhere the car goes. In fact they will know in advance and they will send advertising to the phone of the person who gets out, not just the driver, and possibly an alert to the proprietor of a nearby establishment telling them in who is coming. This information will be for sale to anyone with the money to pay for it. It goes without saying that all this information will be reported to the government upon request.

What's So Much Fun

[rally2xs]

...about driving?

Its 20 miles to town on a divided highway. What do you suppose happens there? Cars clump up and drive in a platoon, never seeming to have watched a NASCAR race long enough to see what happens when the lead car has a problem, and everyone crashes. If it isn't a platoon, then its this one guy either following too D close for any kind of safety, or parking himself exactly in my blind spot. On the drive back from town, the right turn via the turn pocket just before the traffic light to attain the 4 lane divided highway always seems to attract some moron in a big, huge pickup truck that tries to come racing around me on the left, dive right, and slam on the brakes 'cuz he can't negotiate the turn pocket faster than a snail in molassass. Plus, the cops are always out there with radar and lidar and vascar and choppers and airplanes and etc. trying to lift $100 or more from me if I forget to set the cruise control at their ridiculously low speed limit that is set that way just to generate ticket $$$. And then there's a phalanx of traffic lights out in the country fer cryin' out loud, and a pile of morons that take 30 seconds to get moving because they're blabbing or texting on their cell phones. That is, if they're not already moving and drifting into my lane while watching their damned gadget instead of the road. Fun driving? Where is that?

You're Going To Get Killed...

[rally2xs]

...when the self-driving car is forced to do exactly the 55 mph speed limit, and the traffic that normally does 85 there continually runs into you from the rear.

Re:You're Going To Get Killed...

[rally2xs]

Engineers CAN'T solve it, politicians have to solve it, which means it won't be fixed for at least the 1st 20 years.

software problems already exist

[Gravis Zero]

one big issue that autonomous cars cant deal with (yet) is weather. seriously, if there is snow on the road (and even if the cameras have clear vision) it has no idea where the road is. fog and rain would also be an issue because of the cameras. humans can compensate for this (most of the time) but there needs to be either pinpoint maps of the road or a large advancement in the software.

however, i look forward to when these are inexpensive and reliable.

Except that no one will buy them

[Squidlips]

Might as well take the train...

Robot Cars, or, Fix the core issue?

[danknight48]

Considering we seem to let anyone, quote "intoxicated drivers" drive on our roads.
Its not really hard to reduce any figure, especially when you just remove the risk from the pool.

You know those drivers who you see on the phone, just "driving with minimal effort/thought". Take those out of the equation, and the drunks.
Then watch the figures improve. Or, Robot cars.

Hidden agendas everywhere

[Virtucon]

The main proponents of this are all the safety nuts out there who feel that we can't drive safely. What you're missing in all the accident data is the fact that 1) Cellphones and in-car distractions (Nav Systems, etc.) are killing more people every year, it's on the rise. 2) Drunk Driving and Drugs are the leading cause of traffic fatalities. Since the government won't be able to control your behavior they'll just stick you in some little box that will take all responsibility away from you and most of these studies coming out are all sponsored by the government. Sorry, I choose not to listen to a bunch of drivel put out by some paid-for think tank research. Maybe someday if I can no longer drive because of age, then I'd probably consider it but this is a bureaucrat's dream of controlling people's lives. Too much traffic or smog? we'll slow you down or not let you drive today. Had too much to drink? We'll take you right over to the rehab center where you can get dried out? Drugs in your system? We'll just drive you over to the police and they can arrest you.

Re:It already exists!

[Cro Magnon]

Sounds familiar. When I used to take the bus to work, I had to take one bus from my house to downtown, then another from downtown to the workplace. Bus #1 usually got me downtown just in time to miss #2, and I had a 15+ minute wait for the next one. Once, seeing that I'd just missed a bus from downtown back to my house, I started walking, and made at nearly halfway home before a bus showed up!

idiotic idea

[slashmydots]

How much would it cost for terrorists and other hackers to crash them into each other all at once? Probably more than a half trillion.

rural areas

[Rukia]

It may be possible to enforce all autonomous cars in large cities with controlled entry, like New York.

I doubt it will ever happen for rural areas like most of Nebraska.

How about autonomous trains?

[Lluc]

How much do you bet that the NYC subway system will continue to use conductors and drivers for about 100 years after autonomous cars are driving perfectly on the streets above?

Save lives!?

[RevWaldo]

Try telling that to Dr. Tenma! (or Boynton, or whatever..)

.

Autonomous cars-a bad idea.

[Sqreater]

The daily social interaction we get across our society by driving will evaporate and with it the benefits of having to compromise with and even conspire with our fellow citizens to attain the common goal of achieving our destination safely. Autonomous vehicles will introduce a degree of separation and alienation to our society that we can only imagine today.

"switchable automony"

[peter303]

I foresee a scenario where cars change from manual to automonous by a flip of a switch. Then your freeway toll will be lowered if you submit to the autonomous lanes. As this article states it will not only be cheaper, but safer and faster too.

My uncle has a country place...

[blunttrauma]

...that no one knows about. He says it used to be a farm, before the motor law.

Autonomous cars, no thanks.

I will continue to daily drive a 31 year old car with a manual gearbox, no power steering, no airbags, and the traction control and anti-lock brakes are both my right foot. Well weathered leather, hot metal and oil...

Adjusting liability

[Firethorn]

Limited liability has happened before. Plenty of people killed by company drivers, generally the company makes some payout, but not enough to kill the business.

I posted some more of my thoughts on it here

To sum it up: I figure that the companies will fight for, and win, some sort of limited liability. Beyond that if the systems are sufficiently safer than human drivers it won't even cost that much. Consider how much liability insurance costs people now.

If the cars are 80% less likely to be in an accident, you could probably increase liability levels 10X over standard (currently ~$100k in most states) and still come out ahead.

Re:Adjusting liability

[pr0fessor]

I still don't see them in the near future more like a gradual transition with of course regulatory and legal liability issues along the way to iron out. I think we are already starting that gradual transition with self parking cars, self braking cars, and cars that warn of possible accident situations.

I have one question? Will they be driving slowly?

[lamer01]

Will the automated cars obey the posted speed limits? That would be horrible as the whole city where I live has 25mph limits posted everywhere. No one obeys them except the occasional freak. If you drive at the posted limit it will take forever to cross town, it would make commuting not practical.

let the car park itself?

[Chirs]

If the car can drive itself with nobody in it, then it can drive itself home or to a parking lot somewhere.

Re:let the car park itself?

[FishTankX]

Will it drive around to find an empty parking lot? How will it know where a parking lot with a vacancy is? What if it runs out of gas looking for parking? It sounds like specially designed autonomous car parking infrastructure might be required, atleast the cost to you would only be parking + gas and could probably be located in a cheaper area of town farther away, as the time spent commuting to and from the parking is free.

Re:It already exists!

[psydeshow]

Anybody care to guess how long it'll take cities like New York to pass a law making it illegal for driverless empty cars to follow any route besides one leading directly to a parking space somewhere, to avoid having 40,000 driverless cars doing laps around lower Manhattan for hours at a time since it's cheaper to run the car for 2 hours than to actually pay to park for two hours?

This is a really interesting point.

But the fleet could simply drive itself back out of lower Manhattan to areas with cheaper parking/storage facilities. After all, the "reverse commute" is usually pretty light. Also, a large percentage of driverless cars would make multiple inbound trips since people's workdays start at different times. The problem is not dissimilar to what already happens with the taxi and limousine fleet.

Now, if you wanted to own your own driverless car, then it gets interesting. Since you're going to pay for parking for the car, you have an incentive to send it back out of the city to your home garage, or at least to a cheaper parking space across the river. There is absolutely no reason for you to park your car, er, have your car park downtown just because that's where you work.

Re:I have one question? Will they be driving slowl

[ebvwfbw]

Will the automated cars obey the posted speed limits? That would be horrible as the whole city where I live has 25mph limits posted everywhere. No one obeys them except the occasional freak. If you drive at the posted limit it will take forever to cross town, it would make commuting not practical.

The speed limits are intentionally low. Helps with government revenue. Especially in neighbourhoods where they can only go to the minimum for the state. In Maryland that's 25 MPH. Used to be 35. Should be at least 35 in my neighbourhood. We're all on 2 acre lots and visibility from the road is good so I'd feel comfortable at 50 to be honest. For those on 1/4 acre lots and smaller it probably should be 35 or slower.

They'll have to increase speeds to what they should be. Remember 55 killed people. That's a fact. Increasing speeds clearly saved lives.

Re:A breathalizer in the dashboard will do the sam

[BlueStrat]

If the government is in the habit of disappearing journalists, they're not going to fuck around with "hack your car because real life is a Die Hard movie" bullshit. They're just going to straight up kidnap them. Autonomous cars don't change anything here.

Better alert Michael Hastings...oh, wait....

Strat

Elders!

[romons]

We are an aging society. The autonomous car phenomena has arrived just in time for that eventuality. The deaths and destruction due to elders driving past the time they are safe will be mostly eliminated by autonomous cars, either personally owned or taxis. I can't wait.

Rental vs Ownership

[Firethorn]

either personally owned or taxis.

I figure that'll depend how the economics shake out. Owning a car is a rather massive investment, but there is a transactional cost to renting one, even by the hour when it can drive itself from a holding yard to your door.

On the other hand, a retired couple or individual that needs a car maybe 2-3 times a week(grocery, doctors, shopping) would be an ideal case for renting. Somebody who NEEDS a vehicle during prime commuting times, in addition to taking kids to soccer, going shopping, and such is more likely to want to own his own.

Why luxury safer electric cars should be free

[Paul Fernhout]

By me from 2009 (excerpts): https://groups.google.com/forum/#!topic/openmanufacturing/bNyZ6qupGFU
"This essay explain why luxury safer electric (or plug-in hybrid) cars should be free-to-the-user at the point of sale in the USA, and why this will reduce US taxes overall. Essentially, unsafe gasoline-powered automobiles in the USA pose a high cost on society (accidents, injuries, pollution, defense), and the costs of making better cars would pay for themselves and then some. This essay is an example of using post-scarcity ideology to understand the scarcity-oriented ideological assumptions in our society and how those outdated scarcity assumptions are costing our society in terms of creating and maintaining artificial scarcity. ...
    As a rough approximation, sixteen million new cars a year times US$30,000 a car (lower price through volume) would be US$480 billion a year, an amount easily found by reducing some of the about US$1 trillion defense budget (including everything) and US$2.5 trillion health care cost which is about half paid in taxes (total US$3.trillion for those two things, about US$2,25 trillion in taxes). Essentially, US$480 billion a year for free-to-the-user safe electric cars would be only about 20% of the US$2.25 trillion a year in taxes we spend on health care and defense. And in turn, we would save a big chunk of US$164 billion a year for accidents, and a big chunk of the defense budget spent to defend oil supplies, and a big chunk of other medical costs related to environmental pollution, and a big chunk of costs related to global climate change. So, overall, the US tax payer would probably save money on taxes by giving away free open-source safe luxury electric cars (or, at least plug-in hybrids to start).
    Beyond that, then there is the additional benefits that more research in auto safety (even to the cost of hundreds of billions of US dollars a year), especially in perfecting cars that drive themselves at night using radar. Such cars might eliminate virtually all driving accidents eventually, as well as let the human "driver" of such a car use the internet or sleep during the trip (about 90% of serious accidents happen at night, often related to poor visibility or tiredness). One such example of great research in this area (although it may not be FOSS yet?):
      "Princeton Autonomous Vehicle Engineering"
        http://pave.princeton.edu/
    So, why don't we do this right now? I'd suggest it is mainly due to scarcity ideology creating artificial scarcity. For instance, the same computer technologies that can be used to design and operate safer cars are instead used to manage electronic credit or to produce fancy advertising and astroturfing related to promoting free market fundamentalism.
    Essentially, it's all ideology (or ignorance, or corruption, or vested interests, which may all be essentially the same thing), because as I show above, it is even financially cheaper to be both financially-subsidized free-as-in-beer and open source free-as-in-freedom. There are also other various freedoms that safer free-to-the-user electric cars would give us (including freedom from seeing loved ones die in car accidents, by cancer caused by gasoline additives, or by hurricanes caused by global climate change).
    So, I'd suggest, over the next ten to twenty years, this is a major change we will likely see in the USA's personal transportation system -- self-driving free-to-the-user safer electric cars (or plug-in hybrids) built using FOSS methodology. And, taxes will then go *down*, along with other direct to the user expenses for insurance, maintenance, and energy, because our transportation system will then, by adjusting for externalities (like national security, pollution, and health care costs), be cheaper overall to design, build, operate, and recycle."

So, in that sense, expensive cars are another example of market failure due to unaccounted-for externalities.

Re:Why luxury safer electric cars should be free

[Firethorn]

Hmm...
1. $30k for a 'luxury' auto is probably unattainable. $30k gets you a not-quite baseline Taurus, Malibu, Impala, Avalon, etc... Even a basline Tesla S is over $60k today, and the price has dropped.
2. 'Reducing some of the defense budget' - easy to say, but as I work in defense(AD USAF) I can say that it's complicated to reduce the defense budget right. To put it bluntly we need to be allowed to reduce platforms if you're going to reduce spending - Fewer bases, ships, planes, and tanks to make it short. Congress doesn't like letting us do that.
3. I readily acknowledge that the US Federal Government alone spends almost enough on healthcare to provide universal coverage for all US Citizens if we could reduce our spending per person down to that of the EU average. When I last looked the shortfall was under $100 per person.
4. How do you define safe? Teslas are still traditionally driven vehicles, no reason to believe that the accident rate will be that much less. While they're certainly safer on average in an accident, doubling the average cost per vehicle will offset a lot of the prevented injuries and deaths. For that matter the situation can become quite complex as you convert deaths into expensive medical bills*. Not that I'm against reducing accidents, but it's what I tend to call a 'freakonomics' point, like smoking - where the vast majority of smokers still make it to retirement, but die quickly enough that their healthcare and reitrement payments are actually less than non-smokers, making them cheaper for pension funds.
5. However, I'm all for pushing auto-drive vehicles, but the technology isn't there yet. Before I saw the progress on auto-drive tech, I was pushing for PRT(Personal Rapid Transit) which is a ultra-light rail(3-5 people per pod) system with individually powered and routed cars.
6. Not as big of a chunk of our military goes towards defending oil supplies as you think. Speaking from the inside, the real situation is a lot more complicated.
7. I disagree on how similar technology between credit card processing and radar guidance on a car would be.

Roughly speaking, I'd say that we don't have the technology to release a self-driving car that would reduce accidents as much as you say, much less one that can be built as cheaply as you say. Instead I figure you'd end up with vehicles that are the equivalent of the 'common cars' produced by the USSR and other communist block countries - more expensive, less safe and operational than their commercially produced counterparts.

That's BEFORE getting into the problems with waste that occur when you give people people stuff for free - they tend to not value it like they would if they have to invest effort into obtaining it.

After that, you have the 'problems' of personal preference. Do I need extra cargo space, room for more passangers, towing, off-road, hostile climate, etc...?

*Sadly, it's quite possible for an injured person to end up more expensive to society than a dead one.

Re:Why luxury safer electric cars should be free

[Paul Fernhout]

Thanks for the great reply. PRT ideas are really cool (good fictional example is in Logan's Run). Wish I had more time to provide a more detailed reply. But essentially, I think some of these issues are points of uncertainty.

How much would electric cars costs go down with mass production and a build-out of recharging infrastructure? How much safer would the cars be with just a bit more research and testing if there was a trillion dollar initiative (on the order of the Interstate Highway initiative of the 1950s)? How soon would be have fusion energy to power these cars with another trillion dollar initiative (a slashdot article a while back said we were US$80 billion away)? If we can spend trillions on Iraq (including future obligations like to care for injured soldiers), and trillions for wall street bailouts, and trillions for "quantiative easing" as mostly a gift to the banks, shouldn't be able to put a couple trillion into upgrading the US transportation system which is still at the core of US commerce and defense? How many varieties of vehicles do people really want? How much of the US military is ostensibly justified to protect supplies of oil? How much of the current US military posture is obsolete or ironic (i.e. expensive aircraft carriers perhaps being sitting ducks for either submarines or missiles, tanks being vulnerable to cheap drones, 3D-printed military robotics used to fight over economic ideas like capitalism that are falling apart due to 3D printer and advanced robotics eliminating the value of most human labor, etc.)? Could a trillion dollars invested in battery research lead to major breakthroughs? Would human behavior change some with a greater sense of abundance? Probably no one knows all these answers for sure, so fertile ground for lots of discussion.

BTW, I'd be willing to spend more on "defense" as long as it was well spent; I feel the current US defense spending is ineffective ("planning to refight the last war" etc.); to my mind, making the US transportation system safer, more resilient, and more self-reliant is an example of improving the intrinsic security of the USA (including saving 30,000 lives and perhaps 10X that serious injuries per year). And that's just the USA:
http://www.cnn.com/2013/10/28/opinion/sutter-road-deaths/
"The ongoing Pulitzer project highlights sobering facts: Roads kill 1.24 million people each year, and by 2030, that annual number is expected to jump to 3.6 million."

The original justification for the Interstate Highway system was "defense" (including to land airplanes):
http://en.wikipedia.org/wiki/Interstate_Highway_System
"The Interstate Highway System gained a champion in President Dwight D. Eisenhower, who was influenced by his experiences as a young Army officer crossing the country in the 1919 Army Convoy on the Lincoln Highway, the first road across America. Eisenhower gained an appreciation of the Reichsautobahn system, the first "national" implementation of modern Germany's Autobahn network as a necessary component of a national defense system while he was serving as Supreme Commander of the Allied forces in Europe during World War II.[9] He recognized that the proposed system would also provide key ground transport routes for military supplies and troop deployments in case of an emergency or foreign invasion."

So, how about a "war on traffic fatalities"? :-)

Re:Why luxury safer electric cars should be free

[Firethorn]

How much would electric cars costs go down with mass production and a build-out of recharging infrastructure?

I figure you might get another 20% or so off the price of the equivalent of a Model S. The biggest expense is the battery pack, and Tesla is already buying commodity cells, such as are used in laptop batteries. The most economical per kwh/most produced size at that.

The recharging infrastructure is branching out and spreading relatively fast, though I'm not happy with how propriatory the tesla stations are.

How much safer would the cars be with just a bit more research and testing if there was a trillion dollar initiative (on the order of the Interstate Highway initiative of the 1950s)?

A trillion dollar initiative isn't 'a bit more R&D'. To put it in perspective, a trillion dollars should be sufficient to build enough nuclear power plants to replace every existing nuclear & coal plant inside the USA. A trillion dollars, amortizied over 30 years is approximately $1M per fatality using current accident rates. So if your $1T preventented EVERY death over the next 30 years it'd be at a cost of $1M per person saved. Probably worthy, but I don't think it'd stop that many(though the injury reduction would help pay for it). As for the Interstate Highway System, the estimated cost is less than half of that, and provided capabilities, not just safety.

How soon would be have fusion energy to power these cars with another trillion dollar initiative (a slashdot article a while back said we were US$80 billion away)?

That would probably be a 'too many cooks' amount of money

If we can spend trillions on Iraq (including future obligations like to care for injured soldiers), and trillions for wall street bailouts, and trillions for "quantiative easing" as mostly a gift to the banks, shouldn't be able to put a couple trillion into upgrading the US transportation system which is still at the core of US commerce and defense?

Iraq & Afghanistan are over decades, of course, you can't just toss this money at something and 'get it done now'. Especially when you're dealing with physical assets and not electron shuffling that bailouts and such actually consist of.

To take it in another angle, ask yourself why Europe, China, and other countries that are nearly as capable of tossing 'trillions' around like you propose aren't? It'd benefit them as well, after all. In some ways they're in a better situation to make such as shift than we are. Besides, the world has already spent 'trillions' on vehicle safety - Top Gear did a piece a few years ago about the difference between a top rated car that earned an 'A' and a 'B'. People walk away from accidents today that would have had them 'dead right there' even a decade ago, even in the cheapest of cars.

My answer is that the problem is manyfold - there's only so many resources you can devote to the problem, we've hit reducing returns for vehicle safety(IE it costs more and more per life saved/injury prevented) without a drastic shift in methodology(such as moving towards self-driving cars). Said alternate methodologies range from 'not developed yet' like self driving to 'Cost trillions in infrastructure modifications' like PRT.

On the military - some of that 'planning to refight the last war' is to ensure that we don't have to refight the previous war - groups were only forced into asymetric warfare because we're so good at symetric warfare.

How many varieties of vehicles do people really want?

So, how about a "war on traffic fatalities"? :-)

*looks at War on Drugs* - Hell no!

Thank you Captain Obvious

[Gallomimia]

Seriously? On Slashdot? There are people who read slashdot don't know this? I don't believe you. You're tricking me right?