Do you feel safer with self-driving cars on the road?

[Spinnor]

If driving is such a terrible experience for you

I'm not in fear of driving. I drive slow, try to be careful, avoid driving at night if I can, avoid city driving if I can. Knock on wood, have not had an accident since I was a teenager. I have said that I worry while blinded by oncoming traffic I will not see someone dressed in dark and walking too close to the road. My work requires me to drive a work truck around and since I work for myself, yes I have to drive. I am only fearful for my childrens future.

 

[OCR]

When traditional cars are outlawed, only outlaws will have traditional cars.

 

[DS2C]

Honestly I would feel safer. Not that I trust them, but I would trust them more than people. I live on a highway and when it snows, I shovel my sidewalk. I've almost been killed multiple times by people veering off onto the side of the road because they're looking into their lap to their cell phones. Whenever I'm driving, I'll sometime glance at the person next to me, and probably 8 out of 10 times they're messing with their cellphones. It's a real dirt bag thing to do because it puts others at risk.

 

[Spinnor]

Knock on wood, have not had an accident since I was a teenager.

Did though have some close calls in city driving. My luck may be running out.

 

[Dr. Courtney]

Brace yourselves, its here.

"Uber to buy 24,000 specially-adapted Volvos in bid to develop fleet of driverless cars"

24,000 cars at over $48,000 each for paid transportation services may not be the first step toward widespread adoption. In any case, for me, it's there, not "here" because I am unlikely to drive/ride in any of the cities served.

Uber is interested in this technology, because it is cheaper for a company not to pay human drivers, not because it is safer. We're still a long. long way from this uncertain and expensive technology being widely adopted by the vast majority of vehicle owners who do not pay a driver and who prefer not to spend (of cannot spend) $48,000 on a car.

 

[Spinnor]

We're still a long. long way

$50 bucks says that in 7 years self driving cars will have an accident per thousand miles traveled ratio better than legal American drivers taken as an average.

Will you take me up on the bet assuming I have made it precise enough for your liking? If you want, take drunks out of the equation, they should not be driving.

 

[Dr. Courtney]

$50 bucks says that in 7 years self driving cars will have an accident per thousand miles traveled ratio better than legal American drivers taken as an average.

Will you take me up on the bet assuming I have made it precise enough for your liking? If you want, take drunks out of the equation, they should not be driving.

Comparing with the American average (in anything) is a pathetically low bar.

$100 the accident rate of self-driving cars is higher than the rate for my wife and I.

$100 that the average price for self-driving cars is at least $5000 higher than for traditional, driver-controlled cars.

$100 that the installed base is less than 10% of private passenger cars in the US.

 

[Spinnor]

Comparing with the American average (in anything) is a pathetically low bar.

I don't think that is true, if self driving cars have a better overall averaged safety record then all American drivers then with self driving cars on the road there will be fewer accidents?

$100 the accident rate of self-driving cars is higher than the rate for my wife and I.

The above bet won't work. Most people go many years or even their entire lives without an accident. We need the statistics of many people to make meaningful comparisons. And women are I'm guessing statistically safer drivers.

$100 that the average price for self-driving cars is at least $5000 higher than for traditional, driver-controlled cars.

A quick Google search shows the above bet would be a bad one to make depending on our timeframe. On 11/21/2024, 7 years from now, I bet you $100 that the extra cost to make a self driving car will drop below $5000.

$100 that the installed base is less than 10% of private passenger cars in the US.

That is a hard one and I think I will agree with your assessment above, less than 10 percent adoption of self driving cars in 7 years.

 

[jack action]

I don't think that is true, if self driving cars have a better overall averaged safety record then all American drivers then with self driving cars on the road there will be fewer accidents?

The above bet won't work. Most people go many years or even their entire lives without an accident. We need the statistics of many people to make meaningful comparisons. And women are I'm guessing statistically safer drivers.

You are missing his point. If the self driving cars are above average but below the best of human drivers, why would the best drivers want to pay extra to be driven by a self driving car?

Then, once everybody realizes that, guess how many people will think they are among the best drivers ...

A quick Google search shows the above bet would be a bad one to make depending on our timeframe. On 11/21/2024, 7 years from now, I bet you $100 that the extra cost to make a self driving car will drop below $5000.

Maybe I can help you save some money. Here what the quick Google search reveals:

IHS Automotive forecasts that the price for the self-driving technology will add between $7,000 and $10,000 to a car’s sticker price in 2025, a figure that will drop to around $5,000 in 2030 and about $3,000 in 2035

That is a hard one and I think I will agree with your assessment above, less than 10 percent adoption of self driving cars in 7 years.

Not to burst your bubble about the future popularity of autonomous cars, but:

IHS predicts that annual sales between 2025 and 2035 will jump from 230,000 to 11.8 million. That’s about 9% of all the world’s auto sales in 2035. Seven million of those 11.8 million vehicles will rely on a mix of driver input and autonomous control, with the remaining 4.8 million vehicles relying entirely on computers to get around. Combined with vehicles from previous model years, IHS also forecasts that there will be 54 million autonomous vehicles on the road by 2035. When will sales of autonomous cars outnumber those of conventional cars? IHS expects this tipping point to occur by 2050. By then, IHS says the majority of vehicles sold and those in use are likely to be autonomous, with conventional vehicles becoming increasingly rare.

So we are talking 20-30 years instead of 7 years (and they are not all fully autonomous, most still requiring driver input).

 

[Spinnor]

You are missing his point. If the self driving cars are above average but below the best of human drivers, why would the best drivers want to pay extra to be driven by a self driving car?

Then, once everybody realizes that, guess how many people will think they are among the best drivers ...

Yes, everyone thinks they are above average except me. It is her point by the way. Once they are better then the average driver there should be no problem with them being in the road. I will offer you the same bet I made to the Dr., In 7 years self driving cars will have a better driving record as a whole then humans. Take the bet?

 

[kyphysics]

Yes and no, but I voted no.

If the cars are in bracketed lanes (like a train on a track), where they have to follow a certain path, then I'd feel safer with that.

But if the car has to make decisions in an "open" driving environment, then I wouldn't feel safe. The cars would need artificial vision (an area my friend is doing his Ph.D. work on) and there are so many kinks that have to be worked out for that to be safe. I don't believe A.I. vision could adequately identify all threats and non-threats properly and make correct decisions.

 

[Spinnor]

Not to burst your bubble about the future popularity of autonomous cars, but:

I agreed that the adoption will likely be slow for regular passenger cars. I wrote,

"That is a hard one and I think I will agree with your assessment above, less than 10 percent adoption of self driving cars in 7 years."

 

[Spinnor]

there are so many kinks that have to be worked out for that to be safe

It helps that world wide thousands or is it 10's of thousands of engineers are working on the problem of making self driving cars safe.

A google search says I am probably way off,

"Moore, who previously spent eight years at Google and ran the company's Pittsburgh office, estimates that there are 1,000 to 2,000 people in the city working on autonomous driving." That is just Pittsburgh.

From, https://www.cnbc.com/2017/09/16/pit...s-200000-pay-packages-for-robotics-grads.html

From, https://www.google.com/search?q=how...ome..69i57.13855j0j8&sourceid=chrome&ie=UTF-8

 

[Buzz Bloom]

At the present time, since there are still very few self-driving cars (SDCs), not enough experience has yet happened to allow for currently hidden dangerous bugs to be found and fixed. Therefore, I feel very slightly less safe. As the number of SDCs grows, I think it will become increasingly more dangerous for a while until many of the hidden dangerous bugs are found and fixed. Then it will begin to become gradually safer, and eventually it will become safer than it is today.

 

[Qamerash]

I got my driver's license, but I didn't dare to drive to the highway, especially when the traffic was heavy

 

[Spinnor]

I got my driver's license, but I didn't dare to drive to the highway, especially when the traffic was heavy

Off topic, you are driving a car and you come to a stop at a T-intersection in the U.S.A. where we drive on the righthand side of the road. You want to make a left turn. You look left and right to make sure no traffic is coming. Your last look for oncoming traffic before entering the intersection should be to the,

1, left
2, right?

Why?

 

[Stavros Kiri]

Off topic, you are driving a car and you come to a stop at a T-intersection in the U.S.A. where we drive on the righthand side of the road. You want to make a left turn. You look left and right to make sure no traffic is coming. Your last look for oncoming traffic before entering the intersection should be to the,

1, left
2, right?

Why?

Left [again], to minimize time ... (and avoid possible accident)
(Sometimes, I first look right, then left. Or 3 times [left (to make initial decision), right, left (again, for final decision)].)
+ after have entered the intersection and have passed safely left side, check right side again for last time, to avoid high speed incomers ...

A machine would have sensors for both directions at the same time, I assume, for better decision making and minimizing time, to avoid accidents at intersections etc.

 

[Stavros Kiri]

I got my driver's license, but I didn't dare to drive to the highway, especially when the traffic was heavy

Welcome to PF and to the driving world.
I would try to take the "risk" for the highway ... before/or considering once and for all to settle with a self-driving car. Highways are not that bad, if you have a safe vehicle, and they tend to be necessary ... . And in any case, improving driving is nessecary too.

 

[Spinnor]

(Sometimes, I first look right, then left. Or 3 times [left (to make initial decision), right, left (again, for final decision)].)
+ after have entered the intersection and have passed safely left side, check right side again for last time, to avoid high speed incomers ...

My head swivels a bit as well! Right, left, right, left,...

Look left last and live is my motto, well kind of. A driver side impact by another car will put your body at most risk. Now if you have a passenger you have to recalculate but I still think it makes sense to look left last. When my wife is in the car I always ask for help, While I make my last look left she is updating me if I can go, "good?, ...go, go". There are caveats to "Look left last and live" rule, if for example you can see very far to the left but to the right the road curves out of view and so may hide oncoming traffic then you would want to look right last.

 

[jack action]

My head swivels a bit as well! Right, left, right, left,...

Look left last and live is my motto, well kind of. A driver side impact by another car will put your body at most risk. Now if you have a passenger you have to recalculate but I still think it makes sense to look left last. When my wife is in the car I always ask for help, While I make my last look left she is updating me if I can go, "good?, ...go, go". There are caveats to "Look left last and live" rule, if for example you can see very far to the left but to the right the road curves out of view and so may hide oncoming traffic then you would want to look right last.

You should logically look left last all the time before engaging as it is the first lane you will engage in. Once engaged in the left lane, you should take a last look at the right lane before engaging in that lane. If there is no traffic in the left lane, it is easy to safely stop for any unexpected car coming in the right lane. If there is traffic in both lanes once engaged (mistakes happen), you usually should step on the throttle to get out of the way as fast as possible. People coming out at you can obliviously see you and slow down. If you stop in the middle of the road, people have to stop (as opposed to slow down) and if they don't hit you, they might be rear-ended. To do this kind of emergency maneuver, you need to have confidence in other drivers (starting by not assuming all other drivers are incompetent is a good start), otherwise you might freeze and do what you shouldn't.

In short, your last look is at traffic coming from the lane you are crossing before engaging in that lane.

 

[Stavros Kiri]

My head swivels a bit as well! Right, left, right, left,...

Look left last and live is my motto, well kind of. A driver side impact by another car will put your body at most risk. Now if you have a passenger you have to recalculate but I still think it makes sense to look left last. When my wife is in the car I always ask for help, While I make my last look left she is updating me if I can go, "good?, ...go, go". There are caveats to "Look left last and live" rule, if for example you can see very far to the left but to the right the road curves out of view and so may hide oncoming traffic then you would want to look right last.

& @jack action (both well said!)
But in any case, minimizing and totally eliminating the risks for an accident is not an easy thing, if not impossible, even for an A.I. ! ...

I drive by one golden rule:
"Adjust speed, decisions and all actions (while driving) according to visibility and configuration of the driving relevant environment ... ". [+ Be alert, think and decide on your feet (counting all possibilities) and let your mind and body control your vehicle (and not vice-versa) ...]
(Only malfunction and/or other people's faults can get you in an accident then ...)

I was taught the first part of that rule (highlighted part) by an experienced driver, long time ago. Fortunately or unfortunately that veteran driver retired recently from driving (with no serious major accidents so far) because (his words)
"not that I am not good or careful driver anymore, but ... because the other drivers aren't! ..."
and "after some point, one doesn't want to push their luck any further! ..."

Because driving "alive" in real life driving conditions is, more or less, a matter of luck! ...
May be A.I. cars will improve that ...

 

[OCR]

...you might froze...

I think I might freeze, first, Jack... . .

You should logically look left last all the time before engaging as it is the first lane you will engage in.

.

 

[Spinnor]

Because driving "alive" in real life driving conditions is, more or less, a matter of luck! ...

Some things you can't control but you can improve your odds. You sound careful, when you are at a stoplight and it turns green do you look left and then right to look for red light runners?

 

[Spinnor]

should take a last look at the right lane before engaging in that lane.

I think I normally do that as well, once moving a quick glance to the right.

 

[Stavros Kiri]

when you are at a stoplight and it turns green do you look left and then right to look for red light runners?

Always. Sometimes they are more than the green ones ...

 

[Stavros Kiri]

Always. Sometimes they are more than the green ones ...

It also depends on the country, of course. In some countries driving is as crazy as it can get! ... (One could mention Italy, Greece, France, Spain, India etc. , I think, more or less ...)

Moving violations however, are very common in the "driving world of the human species", statistically, in most (if not all) countries around the globe. [I had posted a nice video from YouTube in 'Lame Jokes' (a while ago) about statistics in crossing stop signs, but one can find many such videos about red lights too and other moving violations, showing that sometimes they are more than the non-violations ... . So, what I said was not totally exaggeration neither just a figure of speech.]

I think that's one of the biggest problems with human live driving, that puts our lives at risk everyday, which machines wouldn't even consider (i.e. to make moving violations). However, that doesn't mean that the statistics of accidents are similar. A relatively low percentage of the overall moving violations lead to an actual accident, because of the abilities of the human species (reflexes, intuition etc.), and I agree with @jack action etc. here. Things are getting better too! ... for more and more humans.
But can we count on that?

I am too a fan of the human species, very much so! And you can quote me on the following:
"We, humans, are a very important and promising species, with a great potential of evolving. And, since humans made the machines (limited by the human potential etc.), and not vice versa, we have to acknowledge that now humans and machines evolve together for an even greater and promising potential! ..."

So I still vote yes to the poll.

 

[yungman]

I absolutely do not feel safe with autonomous cars. 40 years after PC becomes popular, you still have blue screen, freezing and all. When I PC froze, you swear, you kick and you restart it. When you are in on the road and the car goes crazy, you can get killed.

Also more importantly, anything that connect wireless to anything, people can hack, hijack. It happened already. There is no way out of this. If they can totally prevent hacking, we won't have virus problem.

Actually this is my biggest frustration. I am buying a car, It is hard to find one without those. I don't want parking assist, I don't want accident avoidance. I don't want anything other than stability control and antilock break that does not control the steering wheel. If you buy a nicer car, you'll quickly find it's not as easy as you think to get a strip model.

 

[Greg Bernhardt]

When I PC froze, you swear, you kick and you restart it. When you are in on the road and the car goes crazy, you can get killed.

That is not a fair comparison. PC operating systems are many times more diverse than a specific and closed system as automated cars and less regulated. Also many PC crashes are due to poor user care. Car riders will not be able to install malware ridden software. I think a better comparison is airplane auto-pilot mode. How many planes go crazy when auto pilot is turned on? Can you think of a single instance?

 

[Stavros Kiri]

+ We put our trust on technology all the time ...

When you are in on the road and the car goes crazy, you can get killed.

What about regular general mechanical failure? That never happens? Then we should avoid all transportation etc.
With similar thinking and fears we would still be moving with horses ...
(Which is not that bad, I could even agree on that, but still I voted yes to the poll ...)

The security and hacking issue has been addressed earlier too.

(Edit) + even horses can fail and go crazy too! ...

 

[yungman]

Yeh, I know, a lot of people just amazed by new technology, I am just not one of those. All I know is people died already because of the self driving car failed, I think it's a Tesla. It was on the news. There were ransom on like Mercedes and other brands from hackers. You like to take the chance, by all means.

I have been a design engineer and manager of EE for 30 years, I designed processor control circuits ( many of them throughout the years), say they never fail is very idealistic. They fail.BTW, it is not a fair comparison between auto pilot in planes and cars. There are so so fewer planes than cars. Planes has a wide open space and FAA has regulation planes cannot get within a certain distant ( big distance) between planes. Try having plane traffic jam like cars. This is like if you have auto drive cars in a desert or in Alaska, you are the only one on the road, then it is safe. Try bumper to bumper, with lousy drivers.