Nearly half of Americans don't want a self-driving car

I Robot. You passenger.
The future is here.
I saw an autonomous car being tested this morning, in Kalifonia. I'm not sure if it was autonomous mode, but it seemed to have problems deciding when pedestrians were going to enter the crosswalk. It was way more conservative than live drivers I've seen. The problem is that I don't think autonomous cars can tell which way people are facing, so they don't know which street the pedestrians are going to cross; in front of the car or parallel to it.
a lot of decisions during driving is through eye contact with other drivers and peds. I don’t think the AI people have that figured out yet.
 
My opinion is I'll wait for the technology to be out in the wild for a few years so they can get the major bugs worked out, then I'll see where there are and consider buying one. I'm also fairly dubious about how well they will be able to handle all the rural backroads that I frequently travel on.

Then too it will all depend on the implementation. I've tried out some of the driver assist stuff on some models and found it to be nice, others I've found it to be from useless to annoying.

Actual rural roads are much easier. They are significantly more predictable, same for highways. City/urban areas are where the real problems await.

Tim
 
I thoroughly understand people that like driving and don't want to give it up to a machine.

What I don't understand are people that think it won't/can't work in our lifetime. How long was it between Kitty Hawk and the first landing on a moving ship? (14 years).

And I don't understand people that think AVs, once generally available, won't be safer that human beings. I was witness to a four car pile up three blocks from my house on Wednesday, involving a truck full of muriatic acid. The major road was shut down for hours for miles because some idiot ran a red light.


For the Kitty Hawk analogy: That has already occurred. Limited use, experimental AVs are out there. I do not doubt that technology. It took another 30-40 years or so to get to modern carrier with jets. This is where we are trying to get to with AV and i don't think it will happen soon. Various reasons.

There is no doubt that AV will be safer. Even the current batch is safer. But it's not a relevant to people. There is a whole new level of "who's fault it is" that freaks people out. That will take a generation or two to get over.
 
.

I often have trouble figuring out what pedestrians are going to do. Many times I need to turn right at an intersection and there will be a pedestrian facing the cross street that I will be turning on to, but they are face down into their smart phone and acting like they are about to cross... or not. Yesterday a guy was standing there and after the light turned I hesitantly started to go, then he stepped forward and I stopped and he stopped and I went again and he went again, then when we got close he looked up from his phone and gave me the finger. I would have yelled at him but he had his ear buds in and probably wouldn't have heard me anyway.

A coworker installed two directional air horns aimed at roughly 45 degrees from the front. To avoid noise limit issues, he has a dedicated buttons for each horn.
They are awesome, I have seen kids buried in phones not look up at a crosswalk, do the start stop dance once. And he hits the button, and they freak out.
I would love to install them on my car....

Tim
 
a lot of decisions during driving is through eye contact with other drivers and peds. I don’t think the AI people have that figured out yet.
Neither have most people.

Tim
 
.

I often have trouble figuring out what pedestrians are going to do. Many times I need to turn right at an intersection and there will be a pedestrian facing the cross street that I will be turning on to, but they are face down into their smart phone and acting like they are about to cross... or not. Yesterday a guy was standing there and after the light turned I hesitantly started to go, then he stepped forward and I stopped and he stopped and I went again and he went again, then when we got close he looked up from his phone and gave me the finger. I would have yelled at him but he had his ear buds in and probably wouldn't have heard me anyway.
Nice straw man memes. Interestingly, one is proving to be reversing itself. The telephone vs Telegraph thing. Communications is moving more and more back to small text notes. Not unlike a telegraph.
 
If autonomous can reduce losses by 10% and increase efficiency by 10% its a huge win. Can we get rid pf hail storm claims too? Can these cars literally find their own shelters instead of being parked on the damn street everywhere getting hailed on, trees falling on them etc? Can autonomous cars eliminate the need for a 4-way stop sign? Can they finally make a 4-way stop sign work where humans simply cannot comprehend?

Can autonomous cars take my elderly grandparents to the doctor? Can a blind person get into an autonomous car and go somewhere?

Can autonomous cars shuttle my kids around without driver's licenses?

When they do park on the streets can they automatically park somewhere avoiding narrowing the damn street down to one lane? Humans can't do it, a nice long stretch to my apartment and only one home owner on that stretch parks on both sides blocking the **** out street right near a corner so its a freaking PITA anytime cars converge there.

I can't wait till most of these questions are answered "yes"

Screw manual driving
 
Last edited:
Nice straw man memes. Interestingly, one is proving to be reversing itself. The telephone vs Telegraph thing. Communications is moving more and more back to small text notes. Not unlike a telegraph.
Just an interesting set of quotes. I am sure there were quotes saying the opposite which turned out to be true. Some depend on the time - no one WOULD want a computer in their house back then due to size capability and cost. The bank President's quote to Ford's lawyer was against the recommendation of his staff.
 
Aircraft emergency landings will be easy, integrate ADS-B into the cars, problem solved. All of the cars on the freeway automatically pull off to the shoulders and blink their lights in time like a runway as soon as you declare an emergency! Technology.
 
I refuse to even buy a car that automatically locks my doors when I put it in drive.
Pretty much every car tracks you and phones home these days. You can tell the manufacturer to not track you. But they will still get the data.
 
And there’s the rub. Neither manufacturers nor federal/state/ local governments should be in the business of mandating self-driving/autonomous vehicle as the only approved vehicle allowed to operate on public roadways.
That ship has already sailed. Fed/state/local already determine what equipment a vehicle must have when traveling on public roads. And many states have vehicle inspections (and/or emissions inspections) on a regular basis (whe I live its annual inspections and every 2 years on emissions)
 
Nice straw man memes. Interestingly, one is proving to be reversing itself. The telephone vs Telegraph thing. Communications is moving more and more back to small text notes. Not unlike a telegraph.
Did you by chance reply to the wrong thread? I'm not sure how the above applies to my post.

I'm middle-of-the-road when it comes to AVs. I will happily point out their flaws and benefits. Still, personally, I would love to see them come to fruition.
 
I will happily point out their flaws and benefits. Still, personally, I would love to see them come to fruition.
They will only come to fruition when all the nation's highways, roads, streets, etc. all have uniform lanes, striping, embedded sensors etc. And that's only for good weather. Now throw in hard rain, snow, blizzards, dust, etc. and all bets are off. The budgeting requirements to set up the proper infrastructure for the autonomy would be astronomical and would require trillions upon trillions of $$$. Autonomous cars/trucks are all nice in theory, but in reality... they're a NOGO.
 
.

I often have trouble figuring out what pedestrians are going to do.
So do I. Pedestrians are unpredictable. How do the sensors figure out whether a figure standing at a corner is waiting to cross in one direction or perpendicular to that direction, looking at their, phone, or waiting for a bus or Uber? Then there are bicyclists, lane-splitting motorcycles, vehicles that stop in a travel lane to unload or parallel park, construction zones, emergency vehicles, city busses, etc.
 
That ship has already sailed. Fed/state/local already determine what equipment a vehicle must have when traveling on public roads. And many states have vehicle inspections (and/or emissions inspections) on a regular basis (whe I live its annual inspections and every 2 years on emissions)

Doesn’t mean I have to like it or approve of it.

More importantly, if a government were to only allow fully self-driving vehicles on a public roadway, what happens to all the non self-driving vehicles? Motorcycles? Antiques/collectibles?

That’s a pretty hefty economic and environmental impact. Several hundred million non self-driving vehicles aren’t going to just disappear into thin air. And, like eminent domain, the owners are going to have to be made whole.
 
Well, you have to ban the human landing the 172 also... :)

Actually most of the static stuff is not to hard. (I realize there are cases where current cars haven't done to well.) Deer, etc. humans don't do to well with either.

All that said, I don't think it's as easy as most of the techies make out. (And I am one, albeit one with gray hair.)

At some point it will be about economics and that's game over. As was pointed out in the verbal discussion, over the road trucks would be an excellent case. Trucks could move 24 hours a day and the computer won't fall asleep at the wheel. They might need extra autonomous only lanes for a while, but with trucks it's only cargo. I suspect similar things will happen with commercial aircraft: cargo will go autonomous first.

John
Agreed that cargo will be the proving ground, because insurance on the goods will be pretty easy. That will provide the proving grounds to further develop the technology for passenger cars. I suspect that the cargo -> passenger transition will go very, very quickly.
 
Pretty much every car tracks you and phones home these days. You can tell the manufacturer to not track you. But they will still get the data.

Luckily with OnStar at least it's easy to fix. Open OnStar box, remove Cellular modem, close and reinstall box.

4 problems:

OnStar GPS powers the GPS antenna and the in-dash nav uses it, rewiring the in-dash nav direct to the antenna makes it work. The Cellular modem was also the GPS receiver.

No compass, apparently rather than $5 on a mag compass module they use GPS track and wheel speed information. Oh well.

Red Light on the mirror, remove the wire if you care.

Sets a DTC in the computer, but as the MIL doesn't come on I don't much care. Maybe the dealer will whine when I take it in for one of these recalls.

Admittedly I'm not paranoid enough to use a spectrum analyzer to see if they hid any other transmitters, but due to them being cheap, er, cost optimized, I'm pretty sure this is the only one.
 
Doesn’t mean I have to like it or approve of it.

More importantly, if a government were to only allow fully self-driving vehicles on a public roadway, what happens to all the non self-driving vehicles? Motorcycles? Antiques/collectibles?

That’s a pretty hefty economic and environmental impact. Several hundred million non self-driving vehicles aren’t going to just disappear into thin air. And, like eminent domain, the owners are going to have to be made whole.
How many horses and buggies do you see on public roads? Yes there are exceptions for tourist vehicles, and you might still see them in rural areas. I don't think we are talking about 5 years in the future; maybe 30 or 40 years.
 
They will only come to fruition when all the nation's highways, roads, streets, etc. all have uniform lanes, striping, embedded sensors etc. And that's only for good weather. Now throw in hard rain, snow, blizzards, dust, etc. and all bets are off. The budgeting requirements to set up the proper infrastructure for the autonomy would be astronomical and would require trillions upon trillions of $$$. Autonomous cars/trucks are all nice in theory, but in reality... they're a NOGO.

Actually weather is one of the easier problems to solve based on some posting/blogs and discussions with an engineer who works on the problem. Between radar, laser range finding, GPS, street maps.... the "car" can actually see better than humans and react faster to changing weather/road conditions.
The major problem with the systems, is dealing with people, animals... the random events and weather is not considered "random".

Doesn’t mean I have to like it or approve of it.

More importantly, if a government were to only allow fully self-driving vehicles on a public roadway, what happens to all the non self-driving vehicles? Motorcycles? Antiques/collectibles?

That’s a pretty hefty economic and environmental impact. Several hundred million non self-driving vehicles aren’t going to just disappear into thin air. And, like eminent domain, the owners are going to have to be made whole.

Not at all, they are not taking your car away. And driving is well documented in law and court as a privilege, not as a right. So you can keep your car and drive on private land only....

Tim
 
Actually weather is one of the easier problems to solve based on some posting/blogs and discussions with an engineer who works on the problem. Between radar, laser range finding, GPS, street maps.... the "car" can actually see better than humans and react faster to changing weather/road conditions. The major problem with the systems, is dealing with people, animals... the random events and weather is not considered "random".

Unless you imbed sensors in every single road in America, there is no way in hell radar/lidar, and GPS based systems are going to be able to take into account all road and weather conditions. Regarding animals, humans, etc., they're actually becoming more adept at programming those scenarios into the AI databases. Will they be able to program every scenario? Probably not in our lifetimes.

Besides, all it will take is a few nefarious hackers to cause massive pileups and deaths and the entire autonomous vehicle program will be relegated to the dust bins of history.
 
As cars become more efficient or move to electric, the government will need to collect money somehow, thus, they will want all cars to have transmitters on them. Thus, I doubt they will be much of a road block. But, it will happen after most of the baby boomers are gone
 
Unless you imbed sensors in every single road in America, there is no way in hell radar/lidar, and GPS based systems are going to be able to take into account all road and weather conditions. Regarding animals, humans, etc., they're actually becoming more adept at programming those scenarios into the AI databases. Will they be able to program every scenario? Probably not in our lifetimes.

On this we will disagree. There is nothing needed in the roads. The sensors on the vehicles already outstrip what our eyes can see.
Even my parents Subaru Eyesight (Subaru brand of driver warning system) can see around the truck in front of you, at night and in the rain better than our eyes can.

Tim
 
On this we will disagree. There is nothing needed in the roads. The sensors on the vehicles already outstrip what our eyes can see. Even my parents Subaru Eyesight (Subaru brand of driver warning system) can see around the truck in front of you, at night and in the rain better than our eyes can.
Ya might want to rethink that statement ;)

EyeSight Limitations and Stereo Cameras Precautions:

EyeSight Limitations*:


EyeSight needs time to recognize an object or potential obstacle. When the speed differential with an obstacle in front is greater than approximately 20 mph, it may not be possible to avoid a collision.

Sun glare, inclement weather or a dirty windshield will hinder EyeSight operation. A cracked windshield, oil film on the glass or reflections may impede EyeSight performance. In these conditions, EyeSight may not operate properly and EyeSight may temporarily stop operating.

EyeSight may not recognize low-contrast objects, those with regularly spaced patterns, no pattern, horizontal lines, backlit objects, very small objects (less than 3 feet tall) or non-standard shaped vehicles, such as cement mixers or car-carrier trucks.

Non-approved aftermarket vehicle accessories, such as hood protectors, and water-repellent glass coatings should not be used with EyeSight-equipped vehicles. EyeSight may not operate properly and EyeSight may temporarily stop operating.

Stereo Cameras Precautions

The stereo camera lenses are precision components. When the system detects that the stereo camera lenses are dirty, no EyeSight functions will be activated.

Always observe the following precautions when handling them:

1.) Do not touch or try to clean the EyeSight camera lenses inside the vehicle. Extra precaution should be used when cleaning the inside of the windshield. Over-spray from cleaners may impair or even damage the cameras' lenses.

2.) Do not remove or disassemble the stereo cameras.

3.) Do not change the positions where the stereo cameras are installed or modify any of the surrounding structures.

4.) Do not install any interior rearview mirror other than a Genuine Subaru rearview mirror (such as a wide-type mirror). Also, use the rearview mirror so it does not obstruct the stereo cameras. Failure to do so may affect the stereo cameras' field of vision and could prevent the EyeSight system from functioning properly.

For more detailed information, please consult your EyeSight Owner's Manual.

*EyeSight is a driver assist system that may not operate optimally under all driving conditions. The driver is ALWAYS responsible for safe and attentive driving. System effectiveness depends on many factors such as vehicle maintenance, tire condition, and weather and road conditions. See the Owner's Manual for complete details on system operation and limitations.
 
On this we will disagree. There is nothing needed in the roads. The sensors on the vehicles already outstrip what our eyes can see.
Even my parents Subaru Eyesight (Subaru brand of driver warning system) can see around the truck in front of you, at night and in the rain better than our eyes can.

My Outback has Eyesight, and while it works well in many cases, it does questionable things in others. Worst case is moving along the interstate and 70 or so, and traffic is stopped ahead. It will stop, but it waits so late that the stop is almost a full effort stop. And the lane keeping is like playing ping pong, it continually hunts left and right. Trust, but verify is a very good motto when driving with these systems.

The supposed testimonials where people say the eyesight saved them when people pull out, etc., scare me. These people must be lousy drivers with poor situational awareness. I've been driving the car for a year and a half and not once has the system intervened to save me. Yes, people do boneheaded things in front of me, but I'm always watching. Maybe 700k+ miles on motorcycles has honed my sense of what is going to happen, but people who react so late that something like Eyesight intervenes to "save" them scare me.
 
@azblackbird

And your point is? How does limitations from one of the most technology backwards car companies mean that we need to spend trillions on road sensors and make all roads identical?

Tim
 
@Bill Jennings

No question on the limitations, and that it is not perfect, or have the best judgement.
My point, sensors are already better than humans, and that is on average car. There is no need to install sensors on every road for every mile.
Now dealing with the data, and being able to process it will be the hard part.

Tim
 
And your point is? How does limitations from one of the most technology backwards car companies mean that we need to spend trillions on road sensors and make all roads identical?
Kinda like the way our air traffic control and the associated equipment and technologies used are somewhat uniform these days. The same type technologies (more or less) will have to apply to autonomous vehicles unless you want massive mayhem on your hands.
 
Last edited:
There is no need to install sensors on every road for every mile.
I suggest you go take a tour of a modern day factory that uses robots for their parts and inventory delivery. For true autonomy regarding vehicles, that's the way the system will have to be set up or it will never work.
 
Kinda like the way our air traffic control and the associated equipment and technologies used are somewhat uniform these days. The same type technologies (more or less) will have to apply to autonomous vehicles unless you want massive mayhem on your hands.

We already have mayhem on our roads. Have you driven anywhere in the USA recently where there is more than one car per mile?
ATC has no bearing on cars. ATC came into existence because pilots were unable to see other aircraft. Therefore a combination of procedures and technology allowed for the creation of an external and centrally managed control structure.
Cars never had this history, they work in two dimensions following predictable paths. Therefore control was effectively decentralized to each vehicle with traffic lights and other tools only providing flow control, on directional control for any one car.

The difference between central and fragmented/decentralized control between the industries has created widely different solutions and requirements.

Tim
 
I suggest you go take a tour of a modern day factory that uses robots for their parts and inventory delivery. For true autonomy regarding vehicles, that's the way the system will have to be set up or it will never work.

Why? I can tell you the requirements are not the same. You still have not said why the Subaru Eyesight system, limitations and all shows that you need sensors on every street.

Tim
 
While autonomous vehicles may soon make 90-95% of paved road driving automated they're still going to have serious issues with exceptions like a wreck on the side of the road and people being flagged around, road construction, knowing you can safely pass that tractor on a 2 lane road(or for that matter ever passing on a 2 lane road) and all the other stuff that humans are 'good' at.
 
You still have not said why the Subaru Eyesight system, limitations and all shows that you need sensors on every street.
As Bill stated, his car (as do most other level 2 and 3 cars) "ping pong" all over the road as it's trying to read and interpret the lane striping and other surrounding data. Embedded sensors much like you'd find in a modern day factory that uses robots would remedy this problem and would keep the vehicles in a more stable and uniform status.
 
As Bill stated, his car (as do most other level 2 and 3 cars) "ping pong" all over the road as it's trying to read and interpret the lane striping and other surrounding data. Embedded sensors much like you'd find in a modern day factory that uses robots would remedy this problem and would keep the vehicles in a more stable and uniform status.

LMAO, by your very statement you show you do not know anything about the technology or the systems.
The ping pong is has NOTHING to do with the sensors or even the basic data processing of the sensor data, it has everything to do with the programming of the response to the sensor data. The ping pong is caused by the system correctly determining the boundary condition (e.g. side of the road; the line, dashed or otherwise...) and having a very simple algorithm to move you away from the boundary; which naturally sends you toward the other boundary....
Here is the basic test you need to verify the sensors are affected by weather: drive on a road and let the car driver assist show a boundary condition. Then drive on the same road and see how some change in weather condition (or some other variable) prevents the sensors from finding the boundary.

For grins and giggles, I have done some programming in this space, so I think I have a pretty solid basic understanding of how it works (or more accurately does not).

Tim
 
The ping pong is has NOTHING to do with the sensors or even the basic data processing of the sensor data, it has everything to do with the programming of the response to the sensor data. The ping pong is caused by the system correctly determining the boundary condition (e.g. side of the road; the line, dashed or otherwise...) and having a very simple algorithm to move you away from the boundary; which naturally sends you toward the other boundary....
That's what I as inferring. Glad you're up to date on this stuff. Obviously you know the limitations of autonomous vehicles and know that we're still light years away from them being fully functional with out massive infrastructure upgrades.
 
I have a 2018 Honda Odyssey. Among other things it has 1) Lane Keeping Assist (LCAS) and 2) Collision Mitigation System.

The LCAS does not work at all on wet roads in the dark. Dry roads, OK. Wet in daylight, OK. Wet at night (even with artificial lighting) nope. And I don't mean occasional lapses, it behaves as if the sensors no longer see the lane markings.

The collision mitigation system (CMS) is an interesting case. It will slow the car (very abruptly) to 22 mph, then it releases the brake and you're on your own. This is by design according to the manual. It has engaged when gaining on a car in the left lane slowing to turn left on a four lane road (we were in the right lane) but the road curved right. Fortunately the car behind us was not too close. It has distracted me and almost caused an accident because I was approaching traffic in my lane at a stop light and I planned to move into the left turn lane (where no cars were queued) it flashed the "BRAKE" signal and beeped so I looked down to see what was wrong-right at the point where I would have been moving over into the left turn lane. When I looked up I was uncomfortably close to the cars in front and swerved sharply into the left turn lane. The CMS is good at detecting things directly in front of the car. But not judging whether they are actually in the path of the car. And I have no idea why it releases at 22 mph. I guess they think I'm safe below that speed.
 
That's what I as inferring. Glad you're up to date on this stuff. Obviously you know the limitations of autonomous vehicles and know that we're still light years away from them being fully functional with out massive infrastructure upgrades.

Why the infrastructure upgrades again? They are needed for many reasons. But AV is NOT one of them.

Tim
 
Why the infrastructure upgrades again? They are needed for many reasons. But AV is NOT one of them.

Here ya go Tim... these articles explain the "smart road" technologies that I'm referring to more than I ever could. As I stated previously, to have true autonomous vehicles the nation's roadways will need to be upgraded to a more standardized format.

https://atelier.bnpparibas/en/smart-city/article/autonomous-vehicles-require-smart-roads

https://www.fastcompany.com/90140902/smart-roads-are-coming-do-we-need-them

https://www.forconstructionpros.com...ars-away-are-you-ready-to-upgrade-smart-roads

https://multimedia.3m.com/mws/media/1373651O/3m-connected-roads-brochure.pdf
 
Doesn’t mean I have to like it or approve of it.

Don't have to like it, but we've ceded that power to government. And given that they are public roads, their roads=their rules.

More importantly, if a government were to only allow fully self-driving vehicles on a public roadway, what happens to all the non self-driving vehicles? Motorcycles? Antiques/collectibles?

That’s a pretty hefty economic and environmental impact. Several hundred million non self-driving vehicles aren’t going to just disappear into thin air. And, like eminent domain, the owners are going to have to be made whole.
Well, maybe.

I think if we watch the rest of the Positive Train Control saga, we'll have an idea of what might stand. PTC will be required of all trains. It's giving fits to those that own/operate old steam engines on excursions over mainline track.

Not sure that the "like eminent domain" argument will stand. You still own the vehicle, you just can't drive it on a public road. Firearms haven't been required to be compensated where a state adopts new restrictions - you have to turn them in or take them out of state... and that's closer to a "taking" than denying the right to drive on public roads. As it is now, you have to have toll tags to drive on the newer toll roads - and denying access to non-payers is not considered a "taking". And in the media business, there was no compensation from the government for forcing TV stations to go digital.

I don't like the idea of a government mandate like this either, but there's enough case law that it could be required. Of course, that might also bring about Trump II, so it might not happen...
 
Here ya go Tim... these articles explain the "smart road" technologies that I'm referring to more than I ever could. As I stated previously, to have true autonomous vehicles the nation's roadways will need to be upgraded to a more standardized format.

https://atelier.bnpparibas/en/smart-city/article/autonomous-vehicles-require-smart-roads

https://www.fastcompany.com/90140902/smart-roads-are-coming-do-we-need-them

https://www.forconstructionpros.com...ars-away-are-you-ready-to-upgrade-smart-roads

https://multimedia.3m.com/mws/media/1373651O/3m-connected-roads-brochure.pdf

Everyone of those articles is by a company trying to sell something, of course they are going to say it is required.
Look at https://waymo.com/ over 8 million miles on existing roads. And that is just one company.
Sensor tech is way ahead of the decision technology to drive safely, especially in highly random environments like an urban area.
Look at Bosch, the mobility group actually makes a lot of the lane assist, speed monitoring and other safety technology. And this tech has been in high end cars for over a decade.

Tim
 
Back
Top