Welcome to DU! The truly grassroots left-of-center political community where regular people, not algorithms, drive the discussions and set the standards. Join the community: Create a free account Support DU (and get rid of ads!): Become a Star Member Latest Breaking News General Discussion The DU Lounge All Forums Issue Forums Culture Forums Alliance Forums Region Forums Support Forums Help & Search
103 replies = new reply since forum marked as read
Highlight: NoneDon't highlight anything 5 newestHighlight 5 most recent replies
Self-Driving Uber Car Kills Arizona Pedestrian. (Original Post) elleng Mar 2018 OP
with a human "safety" driver in the car. unblock Mar 2018 #1
Agreed. Hassin Bin Sober Mar 2018 #26
Well that should end that stupid concept Cartoonist Mar 2018 #2
Which stupid concept? /nt tonedevil Mar 2018 #3
Really. elleng Mar 2018 #6
Did we ban normal cars the first time Codeine Mar 2018 #8
+1000, people are clueless about innovation. Look at early plane flights! Risky! nt USALiberal Mar 2018 #76
'Cause human drivers NEVER kill pedestrians. NT Adrahil Mar 2018 #9
Well, that's totally different. Somehow. nt. Mariana Mar 2018 #38
Cars of any kind are killing people and the planet gyroscope Mar 2018 #64
People dont WANT public transit. Codeine Mar 2018 #85
Not necessarily gyroscope Mar 2018 #92
That depend on what part of the public GWC58 Mar 2018 #102
That's why more of it needs to be built gyroscope Mar 2018 #103
I think it is going to work. NCTraveler Mar 2018 #14
You're kidding, right? gratuitous Mar 2018 #17
It is nice not to be visually handicapped jberryhill Mar 2018 #18
The number of pedestrians killed by human driver cars is off the charts in AZ grantcart Mar 2018 #54
Expecting perfection isn't reasonable. fescuerescue Mar 2018 #62
Wow, how short sighted are you......... USALiberal Mar 2018 #70
Another perspective: Self-driving cars work well in normal traffic situations... brush Mar 2018 #78
In the 30s and 40s planes crashed a lot...... USALiberal Mar 2018 #79
Perhaps. But that point has not been reached yet. brush Mar 2018 #80
And will never be reached without continued testing. 30,000 die a year now. it wil reduce deaths. nt USALiberal Mar 2018 #81
I agree with continued testing. brush Mar 2018 #84
Small Wonder The River Mar 2018 #4
How this plays out in court will set the course for self-driving cars Lee-Lee Mar 2018 #5
I don't think this will be the case that sets the tone. NCTraveler Mar 2018 #15
sorry this one was a jaywalker Egnever Mar 2018 #28
This message was self-deleted by its author Proud Liberal Dem Mar 2018 #50
No way to know for sure yet Egnever Mar 2018 #51
A person stepping into moving traffic? alphafemale Mar 2018 #58
Prrrrretty much. nt Codeine Mar 2018 #86
The video is out now and it's not looking good for the car/attendant/company. brush Mar 2018 #99
I would say it's a mixed bag Egnever Mar 2018 #100
It's pretty well established right now fescuerescue Mar 2018 #63
Maybe she just walked in front of it Drahthaardogs Mar 2018 #7
The first of many such deaths to come, I'm afraid. MineralMan Mar 2018 #10
I agree Sunsky Mar 2018 #13
Sorry but that is just not true. Egnever Mar 2018 #29
And now this one, which killed someone. MineralMan Mar 2018 #30
This one was a jaywalker Egnever Mar 2018 #32
So? A machine killed a woman. MineralMan Mar 2018 #37
no a woman killed herself by walking into traffic. Egnever Mar 2018 #40
No. She absolutely did NOT kill herself. MineralMan Mar 2018 #43
She almost certainly did Egnever Mar 2018 #44
the relevant question is really whether or not a human driver would have been JCanete Mar 2018 #55
Pedestrians DIE daily from human drivers. nt USALiberal Mar 2018 #74
So NO pedestrians have been killed by drivers before this one???? USALiberal Mar 2018 #73
For being an tech writer you seem pretty clueless about innovation.... USALiberal Mar 2018 #72
The blind community is going to have a hard time with this underpants Mar 2018 #41
you seem to think these cars just drive with no awareness of what is around them. Egnever Mar 2018 #46
The blind LOVE Uber underpants Mar 2018 #49
I do not understand what the problem is. Egnever Mar 2018 #53
Yes at the normal crossing points underpants Mar 2018 #56
You have NO IDEA what the situation was. nt USALiberal Mar 2018 #71
Where is your concern about pedestrials being killed by manned drivers? former9thward Mar 2018 #87
I am not a fan of these cars. Too many variables involved. Tipperary Mar 2018 #11
I feel the same about human drivers. LanternWaste Mar 2018 #22
30,000 die a year by human mistakes. nt USALiberal Mar 2018 #82
Now we will find out why they have a human co-pilot. lpbk2713 Mar 2018 #12
The pedestrian hit was jaywalking Egnever Mar 2018 #31
People jaywalk all the time gyroscope Mar 2018 #65
Was Arizona chosen as a state with low population or very weak personal injury laws? TheBlackAdder Mar 2018 #16
Neither. It's because AZ agreed to cooperate marybourg Mar 2018 #19
Arizona's weather is also a factor gratuitous Mar 2018 #23
I don't think you have a very solid grasp of these things Egnever Mar 2018 #34
Yep. My buddy works for one of the big auto carriers. Hassin Bin Sober Mar 2018 #39
That makes no sense to me. Mariana Mar 2018 #45
the risk doesn't go down a little Egnever Mar 2018 #47
Thank you, I think I understand. Mariana Mar 2018 #60
Risk, you say? gratuitous Mar 2018 #61
That right there will be a HUGE HUGE driver for driversless cars fescuerescue Mar 2018 #66
Statistically, 100+ were killed yesterday Thew Mar 2018 #20
Out of how many cars? exboyfil Mar 2018 #25
The metric is miles driven, not number of cars. Hassin Bin Sober Mar 2018 #52
This message was self-deleted by its author fescuerescue Mar 2018 #67
There's not enough information to say anything meaningful. Jim__ Mar 2018 #21
From my uderstanding the sensors act more like radar than like cameras Egnever Mar 2018 #33
I think you're right. They do have vision systems. Jim__ Mar 2018 #36
I have a difficult time seeing how this is going to work exboyfil Mar 2018 #24
In general yes it can Egnever Mar 2018 #27
Yes 30,000 die a year in auto accidents. So it is not working well either. nt USALiberal Mar 2018 #75
Contributory negligence by a jaywalker FarCenter Mar 2018 #35
I wonder (assume?) if there will be a video? Probably? Hassin Bin Sober Mar 2018 #42
I am quite sure there will not only be a video Egnever Mar 2018 #48
The sensors on the vehicle will make this the best documented traffic fatality ever FarCenter Mar 2018 #57
If it could stop instantly it would kill the passengers. alphafemale Mar 2018 #59
The topic of self-driving cars is a great example of Dunning-Krueger fescuerescue Mar 2018 #68
Yes. I know I know nothing, elleng Mar 2018 #69
Actually knowing nothing about the subject, means you don't have any basis mythology Mar 2018 #88
Annual Global Road Crash Statistics elleng Mar 2018 #91
I promise you, my post wasn't about you. fescuerescue Mar 2018 #93
Thanks. I took no offense. elleng Mar 2018 #97
No such thing as bug-less software. rickford66 Mar 2018 #77
People freeze up far more than computers. Codeine Mar 2018 #89
Great until the divide by zero happens. rickford66 Mar 2018 #90
Agree, computers are fundamentally dumb, radius777 Mar 2018 #94
Self driving cars will save 1000s by 2025. Guaranteed...... USALiberal Mar 2018 #95
I saw the dashcam and I'm not sure it would have made a difference with a human driver inwiththenew Mar 2018 #83
It would not have. Nt USALiberal Mar 2018 #96
Warning: Video goes right up to just before the accident. SomethingNew Mar 2018 #98
Why self driving cars before self driving lawn mowers? Or Zambonis? Freelancer Mar 2018 #101

unblock

(52,188 posts)
1. with a human "safety" driver in the car.
Mon Mar 19, 2018, 01:20 PM
Mar 2018

i don't think that concept works. it's hard enough to react quickly and properly to avoid an accident when you're actually engaged in the driving.

i would think it would be very, very difficult to *engage* in driving and then properly avoid an accident when you're in a mode of just watching a machine drive itself -- all in the fraction of a second necessary.

Hassin Bin Sober

(26,324 posts)
26. Agreed.
Mon Mar 19, 2018, 02:53 PM
Mar 2018

Our new Fusion has a few new gadgets like "adaptive cruise control" -- it's pretty cool on the open highway with one or two cars around you. But I still feel a little weird letting it do the braking and acceleration around heavier traffic. I'm watching the two or three cars ahead of the car in front of me, not just the car in front of me.

I still feel the computers will ultimately drive better than humans. Even more so when ALL cars are being driven by computerand they can communicate.

 

Codeine

(25,586 posts)
8. Did we ban normal cars the first time
Mon Mar 19, 2018, 01:33 PM
Mar 2018

somebody caused an accident? They’re still going to be safer than human-piloted cars, and all the Luddism in the world won’t stop them from coming.

 

gyroscope

(1,443 posts)
64. Cars of any kind are killing people and the planet
Wed Mar 21, 2018, 06:24 PM
Mar 2018

with their climate changing emissions.

public transit is safer and cleaner than any automobile, self-driving or not.

 

gyroscope

(1,443 posts)
92. Not necessarily
Wed Mar 21, 2018, 11:23 PM
Mar 2018

in North America cars are the only way to get around because good transit options are almost non-existent.

because of that people need cars more than they want them.

GWC58

(2,678 posts)
102. That depend on what part of the public
Thu Mar 22, 2018, 06:09 AM
Mar 2018

one needs to get to. Public transportation doesn’t go everywhere.

gratuitous

(82,849 posts)
17. You're kidding, right?
Mon Mar 19, 2018, 02:10 PM
Mar 2018

This stupid concept will never die. There are too many grifters getting in on the action for this vaporware. By golly, if billions and billions of taxpayer dollars aren't funneled into the pockets of these latter day flim-flam artists, it will mean we want people to keep dying because of automobiles! Is that what you want? Huh? Huh?

Now, if you'll excuse us, it's time for the next step, legislation absolving "self-driving" automobile companies of any liability for the people they kill. "Well that should end that stupid concept." As if!

grantcart

(53,061 posts)
54. The number of pedestrians killed by human driver cars is off the charts in AZ
Mon Mar 19, 2018, 04:31 PM
Mar 2018

Can't find recent statistics but this will give you a feel

https://www.kgun9.com/news/local-news/pedestrian-deaths-surge-in-tucson

Every time we cross at a corner with a "walk" sign we don't enter because almost every time a driver talking on the phone goes through without looking.

I would much rather chance it with a driverless car than those with drivers.

Two months ago a friend I haven't heard from in 40 years was visiting PHX and got killed riding his bike.

fescuerescue

(4,448 posts)
62. Expecting perfection isn't reasonable.
Wed Mar 21, 2018, 06:17 PM
Mar 2018

Cars are inherently dangerous. 3300 people EVERY SINGLE DAY day from them. An accident with a self driving car was bound to happen eventually.

hopefully, they will be safer than what we have now, but the jury is still out.

USALiberal

(10,877 posts)
70. Wow, how short sighted are you.........
Wed Mar 21, 2018, 09:15 PM
Mar 2018

I assume people like you would of wanted to keep the horse and buggy's because those darn horseless carriages killed someone.

And those new newfangled airplanes crashed! STOP FLYING!!!

LOL, wow!

brush

(53,764 posts)
78. Another perspective: Self-driving cars work well in normal traffic situations...
Wed Mar 21, 2018, 09:48 PM
Mar 2018

in good weather.

The problematic situations are the accidental ones when something happens on the road that requires a quick response different from the programmed actions. This is where the awareness and quick response of a human driver come into play and where the response of a computer making the decisions is quite another matter. And this is the skill that differentiates race-car drivers from the rest of us – and computers from all of us.

USALiberal

(10,877 posts)
79. In the 30s and 40s planes crashed a lot......
Wed Mar 21, 2018, 09:52 PM
Mar 2018

Now they can fly themselves and land on their own.

Technology will solve these issues.

And your "human drivers" kill 30,000 people a year.

brush

(53,764 posts)
80. Perhaps. But that point has not been reached yet.
Wed Mar 21, 2018, 10:06 PM
Mar 2018

Emergency situations, like jaywalkers happen.

And the back-up human, emergency attendant even was no help.

All this in sunny Arizona not in a nor'easter back east.

Back to the drawing board...er, ah...back to the coder and sensor and view cameras.

USALiberal

(10,877 posts)
81. And will never be reached without continued testing. 30,000 die a year now. it wil reduce deaths. nt
Wed Mar 21, 2018, 10:11 PM
Mar 2018
 

Lee-Lee

(6,324 posts)
5. How this plays out in court will set the course for self-driving cars
Mon Mar 19, 2018, 01:28 PM
Mar 2018

Who gets sued and who is liable.

The owner?

The “safety driver”?

How about the company who put it all together as a car?

How about the company that made the sensors?

And the big one- how about the coders who wrote the program that failed to protect the safety of the pedestrian? Do you sue the company that wrote the code, or the individual coders?

How this plays out will affect a lot about who gets into this game down the road. If you are a coder who can program these things but will get sued relentlessly unless your code is 100% perfect will you ever go into the business doing that?

 

NCTraveler

(30,481 posts)
15. I don't think this will be the case that sets the tone.
Mon Mar 19, 2018, 01:50 PM
Mar 2018

I do think your point is going to be one of the more interesting ones that has to be ironed out as we move forward.

Response to Egnever (Reply #28)

 

Egnever

(21,506 posts)
51. No way to know for sure yet
Mon Mar 19, 2018, 04:23 PM
Mar 2018

but the data will come out and I would be surprised if the pedestrian is not found to be at fault.

 

Egnever

(21,506 posts)
100. I would say it's a mixed bag
Thu Mar 22, 2018, 02:28 AM
Mar 2018

The car should have seen her. That said a human wouldn't have either.

Something definitely went wrong with the car. It should have seen her even though a human probably wouldn't have.

Will be interesting to see what the failure was.

fescuerescue

(4,448 posts)
63. It's pretty well established right now
Wed Mar 21, 2018, 06:23 PM
Mar 2018

The car owner is absolutely liable. (In this case a corporation)

The safety driver -may- share some.

The coders? Well I suppose it's possible, but highly unlikely for a very practical reason.

Coders - They don't have deep pockets. The company that employees them? Deep pockets. Nobody wastes their time establishing liability for people who can't pay.

Think about this. Remember all the buzz around Toyota and stuck accelerators? Toyota got sued. Not the guy who designed the accelerator, nor any other engineer.

Drahthaardogs

(6,843 posts)
7. Maybe she just walked in front of it
Mon Mar 19, 2018, 01:33 PM
Mar 2018

Self driving or not, the laws of physics still demand a minimum braking distance. It's not machine error if she violated that distance.

MineralMan

(146,284 posts)
10. The first of many such deaths to come, I'm afraid.
Mon Mar 19, 2018, 01:34 PM
Mar 2018

Pedestrians aren't used to driverless cars, and assume that someone's watching out for them.

Uber cars expect people to do what is expected, but people often do the unexpected. That's going to lead to accidents like this one.

Sunsky

(1,737 posts)
13. I agree
Mon Mar 19, 2018, 01:44 PM
Mar 2018

I cannot count the number of times that pedestrians put their lives in my hands. I'm all the technological advancements but I'm not so sure about this one.

 

Egnever

(21,506 posts)
29. Sorry but that is just not true.
Mon Mar 19, 2018, 03:04 PM
Mar 2018

These things way out think any human. They are constantly monitoring every moving object in sight and predicting movement based on trajectories and speed. A human can't even begin to match these things.

In almost a decade of testing there has been one accident so far that was the fault of the antonymous vehicle and it happened at 2mph

MineralMan

(146,284 posts)
30. And now this one, which killed someone.
Mon Mar 19, 2018, 03:11 PM
Mar 2018

So, now there are two accidents. This one was fatal.

I can't tell you how many times, I've anticipated some pedestrian stepping out in front of me from between cars and have taken action to prevent an accident. I've been driving for 55 years now, and have never had even a single accident. I drive completely engaged at all times, and try to anticipate what is going to happen.

Can the Uber car spot the head of some woman standing between two cars and about to step into the street? I can. I have. I've stopped and avoided the accident.

People do the unexpected. My brain can anticipate such things. My brain is smarter than the Uber car's brain. I guarantee that.

As I said, this is just the first of many such accidents that take someone's life. So far, the self-driving vehicle doesn't have all that much time on the road. It's also not programmed with the same kind of heuristics my own brain is capable of. We haven't advance AI research that far or even close to it.

 

Egnever

(21,506 posts)
32. This one was a jaywalker
Mon Mar 19, 2018, 03:15 PM
Mar 2018

and it is very rare for a driver to be found at fault when hitting a jaywalker and then usually only when the driver was breaking the traffic laws say by speeding.

As someone below said physics still apply and there will be a minimum stopping distance at any speed, you can not prevent an accident where someone breaks those rules of physics. That said an autonomys vehicle has a way better shot of mitigating even these situations as again it is constantly tracking moving objects and their speed and trajectory relative to the car, something just not even remotely possible for a human to accomplish.


edited to add that yes anything you can spot the driverless vehicle can see and track far more efficiently than you.

MineralMan

(146,284 posts)
37. So? A machine killed a woman.
Mon Mar 19, 2018, 03:39 PM
Mar 2018

People jaywalk. Humans know that and notice them, if they're paying attention. So, you appear to be saying that it's OK because the self-driving car just killed a jaywalker because that's OK. She shouldn't have jaywalked? Feh!

I have stopped for jaywalkers hundreds of times. I'm aware of jaywalkers and other people who do stupid things. You know, kids chasing a ball into the street. Dogs running into the street. Even the odd deer.

You're wrong too about my heuristic abilities compared to the computer in an autonomous vehicle. Way wrong.

 

Egnever

(21,506 posts)
40. no a woman killed herself by walking into traffic.
Mon Mar 19, 2018, 03:56 PM
Mar 2018

And I am sorry but while I have no doubt you are as safe a human driver as possible you just do not posses the ability to take into consideration all of the variables these things are able to.


https://www.ted.com/talks/chris_urmson_how_a_driverless_car_sees_the_road


This is a three year old video but there are several examples in it of what the cars take into account when driving down the road and there is a specific example that gets almost exactly to your point. in one of the examples there is a cyclist on the left hand corner that can not be seen by the naked eye yet the car identifies through it's sensors as a cyclist before the human can even see it and it starts to predict that cyclists actions all while also monitoring and predicting all the other objects around it. In this example the cyclist proceeds to run a red light and while the human driven cars fail to see him and almost hit him the autonymous vehicle predicts his crossing and does not move till he is clear.

You might be the safest driver in the world but you just can not compete with this level of awareness.

As a side note the woman chasing the duck in an electric vehicle is hilarious and worth watching the video just to see the car observe this.

 

Egnever

(21,506 posts)
44. She almost certainly did
Mon Mar 19, 2018, 04:03 PM
Mar 2018

and I would be willing to bet my next mortgage payment mario andretti himself would not have been able to avoid her.

 

JCanete

(5,272 posts)
55. the relevant question is really whether or not a human driver would have been
Mon Mar 19, 2018, 04:39 PM
Mar 2018

more likely to react under the circumstances. There are situations where nobody could possibly have time to avoid a disaster, whereas others where a human may have the sense that a computer does not to be wary of people lingering or moving near the street. Then there's the possibility that this thing glitched entirely.

USALiberal

(10,877 posts)
72. For being an tech writer you seem pretty clueless about innovation....
Wed Mar 21, 2018, 09:21 PM
Mar 2018

Cars killed people when horse and buggy went away.

Planes crashed a lot early on. You would want them banned. Yet now air travel is safer than driving.

Get a clue about how tech works.

underpants

(182,745 posts)
41. The blind community is going to have a hard time with this
Mon Mar 19, 2018, 04:01 PM
Mar 2018

Crossing the street on a light is one thing but a stop sign is another ballgame.

I work with blind people and get enough grief about driving a hybrid that shuts off at stoplights. Trust me.

 

Egnever

(21,506 posts)
46. you seem to think these cars just drive with no awareness of what is around them.
Mon Mar 19, 2018, 04:08 PM
Mar 2018

if anything this will be a huge boon to the blind as not only will they be seen by the car but they will also be able to ride in them anywhere they like.

this blind guy loved it.

underpants

(182,745 posts)
49. The blind LOVE Uber
Mon Mar 19, 2018, 04:19 PM
Mar 2018

Trust me. Much cheaper than taxis and quicker than mass transit.

The problem is when they aren’t in the car. The NFB is pushing on this but so far nothing has changed.

 

Egnever

(21,506 posts)
53. I do not understand what the problem is.
Mon Mar 19, 2018, 04:26 PM
Mar 2018

in general these things see pedestrians before most people can see them and track them in in real time and predict their movements based on speed and trajectory.

If any thing these things should prevent a lot of injuries for blind people.

underpants

(182,745 posts)
56. Yes at the normal crossing points
Mon Mar 19, 2018, 05:00 PM
Mar 2018

Blind people, like everyone else, should cross at stoplights but stop signs are another game. Hybrid cars are an issue the NFB has tried to address but that adds cost.

Programming will have to change for these to be truly safe. Yeah I’m all for it. You also have”the Pittsburgh left turn” - I’m told - in which the first car taking a left at an all green intersection is alllowed to go and then everyone goes on as usual.

Roundabouts- loved by civic planners, for good reason - are something the American blind community is just learning to deal with.

former9thward

(31,972 posts)
87. Where is your concern about pedestrials being killed by manned drivers?
Wed Mar 21, 2018, 10:45 PM
Mar 2018

Never seen you post about that.

 

LanternWaste

(37,748 posts)
22. I feel the same about human drivers.
Mon Mar 19, 2018, 02:35 PM
Mar 2018

I feel the same about human drivers.

Too many variables, indeed.

lpbk2713

(42,751 posts)
12. Now we will find out why they have a human co-pilot.
Mon Mar 19, 2018, 01:40 PM
Mar 2018



They need someone to be a goat for all the lawsuits that will result
claiming the human did not over ride the controls soon enough.

 

Egnever

(21,506 posts)
31. The pedestrian hit was jaywalking
Mon Mar 19, 2018, 03:11 PM
Mar 2018

There are very few exceptions where the driver is found at fault when hitting a jaywalking pedestrian. Those are usually when the driver is speeding. Guarenteed not to be the case in a driverless vehicle or when the driver was intoxicated also not likely to be a factor here.

 

gyroscope

(1,443 posts)
65. People jaywalk all the time
Wed Mar 21, 2018, 06:28 PM
Mar 2018

and I slow down for them when they do. it doesn't give me or anyone else the right to run them down

many streets don't have convenient crosswalks, or any crosswalk and there may be no other way to cross.

marybourg

(12,611 posts)
19. Neither. It's because AZ agreed to cooperate
Mon Mar 19, 2018, 02:25 PM
Mar 2018

in testing by not passing regulations preventing or circumscribing driverless cars. Since then, I believe some regs have been proposed.

gratuitous

(82,849 posts)
23. Arizona's weather is also a factor
Mon Mar 19, 2018, 02:40 PM
Mar 2018

If you've ever taken a digital photo of a snow storm, you know that it's hard to get a good shot. The autofocus lens zeroes in on the first snowflake, and you get a beautifully dark photo of that one snowflake. Same goes for a hard rainfall or a hailstorm. Arizona has a lot of nice, clear weather days that don't confuse the onboard systems. Even at that, obviously, there are problems.

When affordable, mass-produced digital lenses are developed that overlook those insignificant snowflakes or hailstones, there then becomes the problem of what else is deemed insignificant, like pedestrians of small stature, animals in the road, and so forth.

You mention personal injury laws, which brings up another point: Who's liable for the death of this woman, if anyone? The company that put the car on the road? Unlikely. The safety driver who was in the car? Who's going to volunteer for this job, and who will insure the driver? Farmers may cover lots of things because they've seen lots of things, but underwriting this risk might be a step too far for most insurance companies.

 

Egnever

(21,506 posts)
34. I don't think you have a very solid grasp of these things
Mon Mar 19, 2018, 03:28 PM
Mar 2018

The insurance companies fear these things not because they are afraid of the risk but understand that it will reduce the risks significantly and therefore eliminate a large portion of their profits.

The auto insurance industry faces upheaval in the next 25 years as the migration to autonomous safety features — and ultimately a self-driving car — shifts more of a car's accident risk from the driver to the vehicle, analysts said.

The number of accidents is expected to drop sharply because currently more than 90% of accidents are caused by driver error. That could lower insurance bills for consumers. The U.S. market for personal auto insurance policies, which currently generates $200 billion in premiums a year, could shrink substantially, some experts predict.

Hassin Bin Sober

(26,324 posts)
39. Yep. My buddy works for one of the big auto carriers.
Mon Mar 19, 2018, 03:53 PM
Mar 2018

I almost fell out of my chair when he told me they were freaking out about driverless cars. It makes complete sense. I just never thought of it that way.

The entire auto, and related, industry is going to be turned on its head.

Mariana

(14,854 posts)
45. That makes no sense to me.
Mon Mar 19, 2018, 04:07 PM
Mar 2018

If they reduce premiums and pay out much less in claims due to fewer accidents, don't they still make loads of money?

 

Egnever

(21,506 posts)
47. the risk doesn't go down a little
Mon Mar 19, 2018, 04:13 PM
Mar 2018

it goes down a ton.

90% of accidents are driver error. Driverless cars will mitigate a huge portion of that, especially when the Majority of cars are driverless as they will all communicate with each other.

Given a model of say a 20% profit as a whole of all policies written just the scale of the reduced risk alone will take a huge bite of the over all pie.

20% of anything is still a nice profit but 20% of a million is a lot more than 20% of 200k

Mariana

(14,854 posts)
60. Thank you, I think I understand.
Mon Mar 19, 2018, 08:36 PM
Mar 2018

Oh well, nothing lasts forever, does it? Once upon a time there was no such thing as automobile insurance. They've had a very good run, with most adults having little or no choice but to buy their expensive product for decades. Every golden goose dies sometime.

gratuitous

(82,849 posts)
61. Risk, you say?
Wed Mar 21, 2018, 12:31 PM
Mar 2018

Well, there is a risk in driving, no doubt. There are about 12.5 fatalities per billion miles driven in the United States in 2016:

https://en.wikipedia.org/wiki/Transportation_safety_in_the_United_States

Currently, self-driving vehicles are averaging . . . slightly more. About 333 deaths per billion miles:

https://www.forbes.com/sites/bizcarson/2017/12/22/ubers-self-driving-cars-2-million-miles/#f0dc01ea4fe4

I wouldn't say the risk "goes down a ton." But perhaps I'm not using the "right" math.

fescuerescue

(4,448 posts)
66. That right there will be a HUGE HUGE driver for driversless cars
Wed Mar 21, 2018, 07:24 PM
Mar 2018

When it cost $2000 a year to insure a normal car vs $200 a year to insure a driverless. That will drive lots of people to driverless.

This is especially true with younger drivers, who are more sensitive to insurance prices AND have the highest rates.

Thew

(162 posts)
20. Statistically, 100+ were killed yesterday
Mon Mar 19, 2018, 02:31 PM
Mar 2018

by vehicles with drivers. I would not be surprised if driverless proves far safer.

exboyfil

(17,862 posts)
25. Out of how many cars?
Mon Mar 19, 2018, 02:44 PM
Mar 2018

See my other post. Even with the idiots out there, there are millions of defensive drivers that are assessing risk all the time. Maybe AI will get to that point, but I don't it is anywhere near there now.

Hassin Bin Sober

(26,324 posts)
52. The metric is miles driven, not number of cars.
Mon Mar 19, 2018, 04:24 PM
Mar 2018
But how safe is it for people to ride in autonomous vehicles? The qualitative answer is “pretty darn safe.” Waymo has logged over two million miles on U.S. streets and has only had fault in one accident, making its cars by far the lowest at-fault rate of any driver class on the road— about 10 times lower than our safest demographic of human drivers (60–69 year-olds) and 40 times lower than new drivers, not to mention the obvious benefits gained from eliminating drunk drivers.


https://www.huffingtonpost.com/entry/how-safe-are-self-driving-cars_us_5908ba48e4b03b105b44bc6b

Response to Thew (Reply #20)

Jim__

(14,074 posts)
21. There's not enough information to say anything meaningful.
Mon Mar 19, 2018, 02:34 PM
Mar 2018

How do the AI vision systems work in the dark? Did the car try to stop? Did the human co-pilot see the woman? How far from the woman was the car when she stepped into the street?

I'm sure that information - and more - will play into the investigators decision as to how to assign blame for the accident.

 

Egnever

(21,506 posts)
33. From my uderstanding the sensors act more like radar than like cameras
Mon Mar 19, 2018, 03:18 PM
Mar 2018

they don't depend on light at all.

Jim__

(14,074 posts)
36. I think you're right. They do have vision systems.
Mon Mar 19, 2018, 03:35 PM
Mar 2018

But, it looks like they are not used to avoid collisions:

From VisionSystems:

...

A subset of vehicles from Waymo—formerly the Google self-driving car project—will operate in fully autonomous mode in the Phoenix metro region, and over the time, the fleet will cover a region larger than the size of Greater London, with more vehicles being added over time.

...

Waymo’s vehicles, according to the company, are equipped with the necessary features, including sensors and software, to provide full autonomy, including backup steering, braking, computer, and power that can bring the vehicle to a safe stop, if needed. The vehicles’ sensor suite includes the following:

  • LiDAR (Light detection and ranging): This sensor beams out millions of laser pulses per second in 360° to measure how long it takes to reflect off a surface and return to the vehicle. Waymo’s system includes three types of LiDAR developed in-house: a short-range LiDAR that provides an uninterrupted view directly around it, a high-resolution mid-range LiDAR, and a next generation long-range LiDAR that can see almost three football fields away.

  • Vision systems: The vision system includes 360° field of view cameras that detect color in order to spot things like traffic lights, construction zones, school buses, and the flashing lights of emergency vehicles. The system is comprised of several sets of high-resolution cameras that operate at long range in daylight and low-light conditions.


In addition to LiDAR and a vision system, the vehicles feature a radar system with a continuous 360° view and supplemental sensors including audio detection and GPS.

...

exboyfil

(17,862 posts)
24. I have a difficult time seeing how this is going to work
Mon Mar 19, 2018, 02:41 PM
Mar 2018

without an extreme change in expectations (or least around where I live).

When I drive I am immediately on alert for anything around me. Can this software determine whether what is next to the road is a trash can or a kid.? I know and adjust my driving accordingly (slow down/move over/). When it comes to hitting a pedestrian I really don't give a crap about who is at fault. I personally feel the obligation is on me because I won't get hurt, but the pedestrian may die. A big one is a kid playing with a ball in the front yard. The ball goes into the street I am immediately slowing.

 

Egnever

(21,506 posts)
27. In general yes it can
Mon Mar 19, 2018, 02:59 PM
Mar 2018

and far better than you and it never gets distracted.

This was someone jaywalking people are killed every day jaywalking by motorists. My bet in the end it will be found this was an unavoidable accident baring the pedestrian not jaywalking.

That said this was an uber car and I am not sure they don't cut corners. The google cars have had only one accident since they began testing that was their fault and that happened at 2mph after hundreds of thousands of miles driven.


 

FarCenter

(19,429 posts)
35. Contributory negligence by a jaywalker
Mon Mar 19, 2018, 03:34 PM
Mar 2018
The Uber vehicle was in autonomous mode with a human safety driver at the wheel when it struck the woman, who was crossing the street outside of a crosswalk, the Tempe police said in a statement. The episode happened on Sunday around 10 p.m. The woman was not publicly identified.
 

Egnever

(21,506 posts)
48. I am quite sure there will not only be a video
Mon Mar 19, 2018, 04:16 PM
Mar 2018

but there will be a real time mapping of 360 degrees around the car for at least a 100 yards. All of this data is logged by these vehicles.

For an example of the type of data these things log this video shows a few samples of it.

https://www.ted.com/talks/chris_urmson_how_a_driverless_car_sees_the_road

 

alphafemale

(18,497 posts)
59. If it could stop instantly it would kill the passengers.
Mon Mar 19, 2018, 08:20 PM
Mar 2018

I'm sorry she died but you don't step into traffic.

fescuerescue

(4,448 posts)
68. The topic of self-driving cars is a great example of Dunning-Krueger
Wed Mar 21, 2018, 07:30 PM
Mar 2018

Everyone thinks they know far far more about it than what they do.

 

mythology

(9,527 posts)
88. Actually knowing nothing about the subject, means you don't have any basis
Wed Mar 21, 2018, 10:49 PM
Mar 2018

for thinking it's a bad idea.

This is the first death ever involving an automated car, and was the fault of the person who stepped out into traffic.

It's currently the 80th day of the year. On average there are 3,287 deaths from car accidents a day. That means there have been 262,960 deaths from car accidents for far this year. Subtract the one from the automated vehicle and that means the count is 1 compared to 262,959. But yes, it's clearly a bad idea to get human drivers out of the way even though it wouldn't have saved the idiot who stepped out into traffic

fescuerescue

(4,448 posts)
93. I promise you, my post wasn't about you.
Wed Mar 21, 2018, 11:31 PM
Mar 2018

And I really mean no one offense. We are all entitled to our opinions.

But, taking a strong position on something that you don't know much about is an example of Dunning-Krueger.

It's ok. I know I do it to. We all do.

I think driving is a bit of a special area though. It's a skill that virtually all of us have, use on a regular basis and we all "know" that we are above average at it. It's practically tailor made for high emotion when it's suggested that a computer could do it better than us.

rickford66

(5,523 posts)
77. No such thing as bug-less software.
Wed Mar 21, 2018, 09:45 PM
Mar 2018

Sooner or later a number of inputs will present a condition not anticipated. That's why you reboot your PC when it freezes up. Hardware failures are another problem.

 

Codeine

(25,586 posts)
89. People freeze up far more than computers.
Wed Mar 21, 2018, 10:54 PM
Mar 2018

They process information faster, react more quickly, and can handle more inputs than we can.

radius777

(3,635 posts)
94. Agree, computers are fundamentally dumb,
Wed Mar 21, 2018, 11:41 PM
Mar 2018

and only as smart as the people who program them, and programming is a mess, millions of lines of code where a single bug or unforeseen scenario can lead to disaster:

The Coming Software Apocalypse - The Atlantic
The 911 outage, at the time the largest ever reported, was traced to software running on a server in Englewood, Colorado. Operated by a systems provider named Intrado, the server kept a running counter of how many calls it had routed to 911 dispatchers around the country. Intrado programmers had set a threshold for how high the counter could go. They picked a number in the millions.

Shortly before midnight on April 10, the counter exceeded that number, resulting in chaos. Because the counter was used to generate a unique identifier for each call, new calls were rejected. And because the programmers hadn’t anticipated the problem, they hadn’t created alarms to call attention to it. Nobody knew what was happening. Dispatch centers in Washington, California, Florida, the Carolinas, and Minnesota, serving 11 million Americans, struggled to make sense of reports that callers were getting busy signals. It took until morning to realize that Intrado’s software in Englewood was responsible, and that the fix was to change a single number.

Computers are good at rote tasks and in assisting humans with such tasks - but the human brain is adapted for things like unpredictability and intuition that cannot easily (if at all) be programmed.

Driving is one of those tasks that demand an ability to deal with the unexpected.

Flying is not comparable to driving, as flights take a fairly predictable path, with mostly rote takeoffs and landings guided by on-the-ground tech as well. It's a controlled dance involving professionals, whereas driving is mostly not.

USALiberal

(10,877 posts)
95. Self driving cars will save 1000s by 2025. Guaranteed......
Wed Mar 21, 2018, 11:57 PM
Mar 2018

A single error has crashed planes also. And collapsed walkways.

Bugs are fixed. Will self driving cars kill some people, sure they will. But in the long run will they save 1000s, yes they will.

You can’t stop progress!

inwiththenew

(972 posts)
83. I saw the dashcam and I'm not sure it would have made a difference with a human driver
Wed Mar 21, 2018, 10:16 PM
Mar 2018

That lady on the bike crossed at night on a poorly lit street outside of a crosswalk. Really bad move.

SomethingNew

(279 posts)
98. Warning: Video goes right up to just before the accident.
Thu Mar 22, 2018, 12:48 AM
Mar 2018

I think this makes it perfectly clear that the pedestrian was at fault and a human driver would have hit her too. If anything, the self driving car had a better chance of "seeing" her before she was visible.

https://mobile.twitter.com/TempePolice/status/976585098542833664

Freelancer

(2,107 posts)
101. Why self driving cars before self driving lawn mowers? Or Zambonis?
Thu Mar 22, 2018, 04:01 AM
Mar 2018

After a few years of seeing that the mowers didn't slash the begonias, or kill the garden gnomes, we could move on to bigger stuff. Instead, it's the most dangerous stuff that's being roboticized first.

Where I live, in the Midwest, huge agro machines without drivers are available. Soon they're going to be rolling up to the edge of my yard. One glitch, and they could wipe out my house. NO THANK YOU!

Latest Discussions»General Discussion»Self-Driving Uber Car Kil...