There's no way to manually take over just in case something goes wrong? Foolish design, at best.

@1 One of the other articles about it cited data that having drivers take over has proven worse in the real-world tests of the existing fleet.

The belief that people are better drivers than robots is apparently not supported by facts.

@5 overreact much? We need better Public transportation but not going to replace cars completely

If you thought taxi drivers were upset about Uber, just wait until they get a load of this.

I really do not at all in any way comprehend all the excitement and fawning over google's self driving cars. Why so much time, money, attention, news coverage of a transportation system that can only move 2 - 4 people at a time while relying on roads, traffic corridors, highways, etc? Why aren't we trying to fix the real transportation problems in America with meaningful transit options and infrastructure?

I would love nothing more than to see cyclists complain about how dangerous self driving cars are, even though accidents still drop off sharply. Aggressive cyclists will continue to be aggressive and accidents will continue to happen, none of which will be the fault of the car. Oh and cars cant hit and run. If it detects a collision, its not going anywhere until a police officer releases it.

I have a feeling they'll use very small versions of these cars to deliver things in addition to people. Amazon will have drones, and Google will offer a fleet of delivery vehicles. They pull up outside your place, you enter a code on your smartphone and a window opens with your stuff. Like a movable Amazon locker.

#10: In the USA, cars kill more people under 45 than any other reason. Cars kill more people under 18 than ALL DISEASES COMBINED. Worldwide, cars kill almost twice as many people as all violence combined, including murder, assault, and all war.

If self-driving cars turn out to be safer, which is expected, and popular/accessible, they'll be one of the most important public health inventions ever.

@4, when all cars are self-driving, they'll probably ALL be tailgating, so they won't stand out... essential for more efficient use of roadways. But I'm pretty sure it'll be a lot safer anyway.

Assuming we're not going to redesign our infrastructure to be more human-centric, I'll take electric robots over manually operated combustible engines any day. Life will at least be quieter then. Until they start adding artificial horn honks and beeping noises to them in order to appeal to neanderthal sensibilities, that is.

They only go 25 mph -- which is great. They can be the last mile to standard transit (buses, LINK and Sounder).

If you need to carry stuff from the mall, just have it shipped in one with no seats.

In fact, you could just build a self-drive shopping cart that rolls itself down the sidewalk, just like the ones my mom used to steal from the grocery store.

You'll pry the steering wheel of my pickup from my cold dead hands.

I don't trust automotive electronics enough even to have power windows. If, instead of a stuck window it was a stuck/hacked/defective computer? No. No. Hell no.

Anyway, who asked Google to R&D this ridiculous thing? It sure as hell wasn't anybody with a real car.

All cars aren't brand-new all the time. Things age.

Manufacturers upgrade parts in existing models, sometimes in the same year. There's already a lot of software in cars, but this takes things to a much higher level and requires much tighter integration.

And, when they improve a part, they usually stop making the earlier one. Not so much of a problem, usually, but who is going to recalibrate the software if you need a replacement? Most mechanics aren't very good, and they're particularly hopeless with control systems.

Look at all the automotive recalls of recent cars. Over the last 3 years I think it's in the millions. And look at how the average car ages. A self-driving car will be orders of magnitude more difficult to mechanically maintain because of all of the added controls and actuators.

Not to mention the software upgrades, which will be nearly continuous. How much time does your Windows laptop spend now being upgraded by Microsoft? Going to be a pain if you want to go for a drive and your car is offline, updating.

In practice, in real life, over the service life of a vehicle? I'm a pragmatist. I don't see this as workable. We just don't have the infrastructure. Auto mechanics would have to be replaced with highly-qualified engineers, and even then there are too many vulnerable failure points. This is the sticking Toyota throttle, to the googleth power.

Gotta agree with SeattleBlowsMe on this one. Imagine what happens when the Chinese army hacks the system and makes all the cars stop. Or all execute a simultaneous 90 degree left turn.

You want mass adoption of these? Insurance discounts. Game, set, match.

@13: Gonna cuntpaste my response from the LAST time you made that fuckstupid claim!
Not to spoil your little world with facts, but the number one cause of death for people between 1 and 44 is "unintentional injury". Below 1, mortality is most commonly from "congenital abnormalities", between 45 and 64 people are most likely to die of "malignant neoplasms" a.k.a. cancer, and above 65 the leading killer is "heart disease". (Source.) Data from the same year suggest that roughly 1/4th of those accidental deaths are car-related. In that case, cars would take third or fourth place in most age categories.
Don't get me wrong; I have a certain distaste for cars, and they represent a serious public safety issue, but you only weaken your cause when you say shit without making sure it's correct.

@17: I'd argue this point with you, but if you don't believe in biology, neurology, psychology, geology, or computer modeling, you're unlikely to be receptive to the idea that computer scientists and engineers know what they're doing maybe. Fun fact! There are computers in your car already and a malfunction of one of them could put you in danger of life and limb. These things are designed and tested carefully and when defective ones are released to the market it is a Big Fucking Deal.

#23

A few years ago I drove a 1991 Pontiac Gran Prix V-6, bought used. It has so much power the car would twist with torque when it revved.

Also, as I found out through numerous repair bills and searches through repair manuals, it was rife with computers. I believe it was the first year GM -- who pioneered the idea -- started using chips to control ignition and other functions. There were four separate computers spread around the engine. At one point I had to replace an electronic barometer (that tuned the gasoline mix for the elevation). There was even a computer built into the standard equipment radio! (Which I imagine is a good place for it...unless you want to replace the radio.)

But to echo your point. Yes, over two decades of computer control.

Sure hope it runs down all the Google Glassholes.

@22

Nope kiddo.

I believe tools are only as good as their operator.

And I believe that when an academic says something at odds with real world experience I can and should ignore the academic.

At any rate I was like you once. Argue for arguments sake, not because it advances anyone's understanding? Yeah, we all do that when we're young. As I get middle aged proving myself to be right when I already know I am isn't as interesting.

@2 I'm not sure how having a human take over in the case of computer failure is worse than letting the (now broken) computer continue on.

Look, automation is great, but even airplanes have the ability to let pilots take over. I don't see why cars are any different.

None of the poll options match, because the reality is that automated vehicles are coming and the impact will be both positive and negative.

They won't replace public transit, because for urban environments buses and trains simply move more people than individual cars, and are if anything easier to automate because they can be assigned fixed routes.

They won't replace personal cars for a long time, because they'll be more expensive per mile for the foreseeable future. They won't reduce the number of cars purchased because while you can use an automated car for more hours out of the day, that means they'll wear out more quickly and need to be replaced more quickly (cabs last about 5 years max from what I've heard).

They will allow people to live further out, promoting sprawl, but not that much further out (because we have already developed within the areas that are a manageable driveable distance), and will at the same time allow some people to drive less often because they won't have a car in their own driveway. They won't necessarily be more convenient because it takes a couple minutes to drive one to your house, instead of just getting in your car and driving away.

What they will do is almost entirely eliminate traffic fatalities, vastly reduce the need for parking, kill most driving jobs, and free up time that people now use driving to be wasted instead on our electronic devices. The best effects will likely be in improved architecture and street design. That's assuming that the automated car companies don't pass laws that allow for higher speed limits and other changes that facilitate cars over people, which is a possibility. But I would still bet that our cities and neighborhoods get a lot more pleasant to live in.

It looks compact enough to double as a coffin after the inevitable car smash.

It looks compact enough to double as a coffin after the inevitable car smash. Dual purpose.

Do the Air Bags come filled with disclaimers when the car runs over preschoolers in a sudden rainstorm and gust of wind ... Or are those optional?

@25 We already know your willfully ignorant no need to remind us.

*you are* ugh

Can't wait to see how fast these things get hacked.

The possibilities are endless... send people places, send *many* people to the same place at the same time, alter speeds and other parameters...

And cyclists will just do what they already do and give the thing a good hard kick or smash it with a bike lock, which these will hopefully interpret as a collision and stop until the police release it :D

@25: ...what? Yes, tools are only as good as their operators. No, that doesn't have any bearing on this discussion here.
Also, there's a difference between discounting an expert opinion when it is prima facie incorrect (e.g., ignoring the weather report when it is in direct contradiction with what you see out your window) and discounting an expert opinion just because you're too ignorant to understand it (e.g., your opinions on global warming, sexual orientation, evolution by natural selection, et cetera).

@22 "There are computers in your car already and a malfunction of one of them could put you in danger of life and limb. These things are designed and tested carefully and when defective ones are released to the market it is a Big Fucking Deal."

Unless it's GM. Then it's a "Nothing to see here!" deal, even after dozens of deaths. Lets face it, computers break down often enough, and they're programmed badly quite often. While I wouldn't worry too much about a catastrophic failure in these systems, I am expecting some unintentional bugs, some mass usage issues that didn't come up in testing, and some shenanigans as soon as the folks who like to do such things start dissecting the hardware and software.

They certainly won't lack for early adopters, but like most people I think I'll wait until gen 2 or 3 to actually get in one. And stay off my lawn! :D

@33 they hacked other cars in less than 24 hours, I wouldn't considering buying one unless google had a hack-a-thon where they let people try and take over the car without having physical access to the hardware. I eagerly await Google's response to my comment on the slog here.

From Wikipedia - "Google's robotic cars have about $150,000 in equipment including a $70,000 LIDAR (laser radar) system."

More than I thought. Even if they can bring the price down 50% for the market system, and get some automakers to sell them, I won't be able to afford one on top of a car's base price.

@38 What I want to know is how well that LIDAR system deals with heavy rain or snow. A lot of noisy, irrelevant data thrown at the system isn't going to end well, I'm afraid.

Your PC is run by computer...notice how many times it has errors when you're doing something?
Your Smart Phone is run by a computer...ever see it hiccup a bit and take a while to load something?

So you have a car run by a computer and you're going 65...want to risk it having a glitch and killing you?

And people may be flawed, but I have yet to meet a driver who could potentially be hacked by a kid in a dorm room in Singapore. Wait until the boys at 4Chan figure out they can hack automated cars.

And this will never catch on. People like to feel the road beneath them and it's just unsafe having a machine that is easy to fuck with in control of a moving heavy object.

22- What are you arguing? I didn't say cars were the #1 killer of newborns or old people. I said they're the #1 killer in the US of people, as a group, up to age 45.

Look at this table on page 3. Add up the numbers for all age groups through 45. #1 killer. http://www-nrd.nhtsa.dot.gov/Pubs/811620…

@42: Your numbers are from 2008/9. I'm talking 2010 numbers. And malignant neoplasms are on the rise, while motor vehicle deaths are declining.

Hey guys,

within the scope of my master thesis, I want to submit a questionnaire to assess what do you think about Google's self-driving car.

I would greatly appreciate your participation in the survey!

If you have any questions,please feel free to contact me!

https://docs.google.com/forms/d/11rSv2oZ…

Since 2009, Google has pursued the idea of a fully-autonomous, self-driving, driverless or robotic car.The idea of a fully-autonomous vehicle implies that the driving functions are all done by an inboard computer which is able to master the speed, headways, braking and manoeuvres of the vehicle. The National Highway Traffic Safety Administration (NHTSA) distinguishes between five levels of automation (0-4).
The automated driving systems by traditional car makers equal NHTSA's level 2 where the driver is to be able to take over control at all times. Google's self-driving car corresponds with level 3 or limited self-driving automation where drivers are expected to be available for occasional control. Its long-term vision is level 4 or full-self driving automation where the role of the driver essentially shifts from being an active driver to the one of a mere passenger that is chauffeured around from A to B. Would you be willing to adopt Google's self-driving car? Please follow the link to participate in an online survey that studies potential consumer willingness to adopt Google's self-driving car.

https://docs.google.com/forms/d/11rSv2oZ…