Don't we normally experiment on human subjects in "controlled" environments?

I mean, Tempe or Scottsdale - these are upscale places where people with real bing walk around.

Universities and prisons are where most psychological and behavioral testing is done.

How many prisoners can you maim in an afternoon? How many liberal arts students are too dumb to not get hit by a robot car? Sounds like several PhD projects right there.

I'll assume uber and these other companies are using their own product to ferry their children around, right? Watch while little Billy runs after a ball while UB#8381735 is backing up the driveway.....

Sure, sure. We didn't know that we needed pop-tarts or diet cherry pepsi, and people spend billions on these products alone (at least I assume it's in the billions - you get the idea).

But, really, has anyone here, for example, woken up one day and said, "hey, I sure wish I had a self-driving vehicle"? Real people, not corporations, for which the goal is to eliminate all human employees. Is it that awful to actually pay attention to, you know, driving, during your commute? Is it that much of a hardship? Is it that onerous?

I don't want to depend on the reliability of alternative transportation, unless I live in a very good mass-transit city.

I should be able to get in the car and tell it to take me to the doctor, the grocery, the bingo parlor. Any place my old heart desires.

I can't wait to get my hands on one.

and yes spend a day in LA traffic and get back to me on how onerous a commute can be.

Assuming, among other things, that the three responses are representative of the country overall.

The "among other things" include the realistic ability to buy such a vehicle. I'd like to buy a top of the line Volvo, but that isn't going to happen. People may want to buy electric vehicles, but without a nationwide network of recharging stations, use may very well be limited.

But points are taken, even if I'm not going to include myself among those interested. Thanks.

the end goal is to summon one on demand. Much like taxis now but far more efficient.

Was just reading an article that stated currently there are huge portions of our land that are dedicated to cars that could be eliminated. Parking garages no longer needed if the cars remain on the road at all times. Garages at houses no longer needed. Energy consumption will drastically be reduced. you no longer drive a pickup to run to the grocery store but instead take a small energy efficient for one person vehicle to the store and back and then wave good bye as it drives off for the next person.

By simply switching to autonomous driver-less vehicles we could eliminate a huge portion of our fossil fuel needs.

There are just a ton of benefits over and beyond simply improving road safety. This page has a video that goes into many of the benefits that we could expect even though it is a bit dated at this point.

http://bigthink.com/ideafeed/googles-self-driving-car-is-ridiculously-safe

Thanks.

During my federal career, I managed spare parts, but mostly mechanical (they're more fun), but the managers who had the electronics like radar always had a heck of time keeping up with supplying the right parts to keep the system operational. Just based on my many years of talking to and watching my electronics manager friends, I'd never use a car that was guided by radar. It's not just that it can break down, but things as simple as weather or dust can disrupt it.

To me, much more logical would be to supplement the sensors used for accident avoidance and parallel parking with AI learning from experienced drivers. I don't think we'll ever get to a true self-driving vehicle, but I think there can be plenty of driver assistance if needed.

and they don't just rely on radar.

They use radar and lasers and sight in combination with continuous monitoring that stops the car at any point of failure. They have already logged millions of miles on the road with up until this recent accident not a single fatality and only one at fault accident which happened at 2mph.

They are already moving cargo on the roads. This video is from two years ago.

on the road. What happens when the number increases? That's multiplying the chances for the technology to make the wrong decision. These units will fail. Do you know how many hours after use that will happen? What happens in snow or sleet or rainy or cloudy? Are there going to be double and triple redundant systems? There are a lot of practicle questions to be answered.

Better than human is and already they are proving much better. the more of them on the road the safer it gets as they communicate with each other and advise each other of conditions both in front and behind themselves.

Did you watch that video? The driver is in the back seat reading a news paper while that truck cruises merrily down the highway with no one at the wheel.

That was two years ago and the amount of them on the road has only increased since then and the tech continues to just get better and better.

Of course there are lots of practical questions. Do you think they are not being considered? This is not new. This has been in the works for over a decade now.

At this point they have been operating in many cities for over a year fully autonomous and so far we have a single fatality that appears to be the pedestrians fault not the vehicles. And of millions of miles traveled now by these vehicles I believe there is exactly one accident that was the fault of the vehicle and it was a 2mph bump into a bus.

In the case of the waymo cars (googles version) redundancy is built in with backup systems constantly running checks waiting to take over in the case of a failure.

I'm worked on new systems before, including new military aircraft and new avionics. They always look great in testing or in limited use.

The fact is the system failed and killed a person.

Possibly but it is also entirely possible it could not have been avoided by anyone human or not. Breaking distance is a thing and if you throw something in front of a car that is inside the breaking distance it wont matter what was driving it the car can not stop in time.

Maybe it was a glitch that caused it.

One thing I know for sure in the end we will know exactly what happened because all of that data is logged and retrievable.

Again perfection is not the metric. Significant improvement is and so far they demonstrate significant improvement even in these early stages.

Does that mean there are other areas where they ARE still testing?
and what about other driverless cars and trucks, I wonder.

I can predict with 99.9% certainty that her family will have no problems finding a lawyer.

As far as the liability for this pedestrian fatality, there was extensive discussion about it before the testing was even allowed. What I don't know is whether the liability is different for different states.
I'm waiting to hear the ruling on this.

If they could build an autonomous robot that just pickpocketed your wallet, they'd skip the whole transportation part.