• The TUGBBS forums are completely free and open to the public and exist as the absolute best place for owners to get help and advice about their timeshares for more than 30 years!

    Join Tens of Thousands of other Owners just like you here to get any and all Timeshare questions answered 24 hours a day!
  • TUG started 30 years ago in October 1993 as a group of regular Timeshare owners just like you!

    Read about our 30th anniversary: Happy 30th Birthday TUG!
  • TUG has a YouTube Channel to produce weekly short informative videos on popular Timeshare topics!

    Free memberships for every 50 subscribers!

    Visit TUG on Youtube!
  • TUG has now saved timeshare owners more than $21,000,000 dollars just by finding us in time to rescind a new Timeshare purchase! A truly incredible milestone!

    Read more here: TUG saves owners more than $21 Million dollars
  • Sign up to get the TUG Newsletter for free!

    60,000+ subscribing owners! A weekly recap of the best Timeshare resort reviews and the most popular topics discussed by owners!
  • Our official "end my sales presentation early" T-shirts are available again! Also come with the option for a free membership extension with purchase to offset the cost!

    All T-shirt options here!
  • A few of the most common links here on the forums for newbies and guests!

Autonomous Uber accident

Elan

TUG Member
Joined
Jun 6, 2005
Messages
4,455
Reaction score
409
Points
468
Location
Idaho
I'm sure most of us have read about the fatal accident involving a self-driving Uber and a pedestrian.
I'm curious as to what most here think? As much as I'd like this to point to the possible shortfalls of autonomous vehicles, I more strongly feel that had this same accident occurred in a human piloted vehicle, we'd have never even been aware of it's occurrence. The newly released video makes it seem as though the accident was virtually unavoidable.
So, what say you? And, to extrapolate, what are your thoughts regarding the mindset that, while autonomous vehicles might not avoid some accidents we would normally deem avoidable, they will ultimately result in fewer accidents overall? In other words, are you willing to trade more human caused accidents for fewer computer caused accidents?

Sent from my Moto G (5S) Plus using Tapatalk
 
Last edited:

bluehende

TUG Review Crew: Veteran
TUG Member
Joined
Jun 6, 2005
Messages
4,507
Reaction score
3,967
Points
598
As a former research scientist this fascinates me. I am sure more study is needed on this particular accident and hopefully it can improve the software. Autonomous driving as weird as it would be is almost certainly safer now and will continue to improve. There is no way something that heavy, moving that fast, can handle any situation that stupid humans can get themselves into. I am sure it can handle a lot more than I can at 62. I also see great uses for those that cannot drive. I can imagine in the future being able to go to sleep in my car and wake up in time to go skiing when I wake up after traveling 400 miles. The only negative is the jobs that it will replace. But progress always displaces employment. We no longer have telephone operators. The key is finding the next generation of jobs not trying to save the last generations jobs.
 

Roger830

TUG Member
Joined
Dec 11, 2013
Messages
1,456
Reaction score
587
Points
323
Location
CT
The video that's on dailymail.com also shows the driver looking down until the instant of impact.

The woman with the bike is crossing the road from left. It seems likely that an alert driver might have spotted the woman and bike out of the corner of his eye prior to a camera with software and hit the brake.
 

davidvel

TUG Member
Joined
May 9, 2008
Messages
7,432
Reaction score
4,477
Points
648
Location
No. Cty. San Diego
Resorts Owned
Marriott Shadow Ridge (Villages)
Carlsbad Inn
The video that's on dailymail.com also shows the driver looking down until the instant of impact.

The woman with the bike is crossing the road from left. It seems likely that an alert driver might have spotted the woman and bike out of the corner of his eye prior to a camera with software and hit the brake.
That may or may not be true. I don't think one can say it is likely or not, without knowing more about all the cameras, sensors etc on board. For all I know (not much) the vehicle had 360 degree "vision." My car sees things I can't all the time, and it's not an autonomous vehicle.
 

ottawasquaw

TUG Member
Joined
Mar 20, 2006
Messages
313
Reaction score
60
Points
238
The video that's on dailymail.com also shows the driver looking down until the instant of impact.

The woman with the bike is crossing the road from left. It seems likely that an alert driver might have spotted the woman and bike out of the corner of his eye prior to a camera with software and hit the brake.
Why didn't she see the vehicle? Honestly I haven't the video. I am guessing she was actually riding her bike as opposed to walking it across the street. This happened in my neighborhood. I see these vehicles nearly daily and especially on weekends, Sunday mornings. I was nearly hit a few weeks ago when one nearly drove right into me as it turned left at an intersection. Scared the h*ll out of me. I don't recall that ever happening with a driver at the wheel.
Outside of this accident, I've never seen or heard of any incident involving these test vehicles. For the number of test miles logged their record has got to be better than other vehicles on the road.
AZ and the Phoenix area regularly has auto/bicycle fatalities in broad daylight. I am a confident road rider. I know the risks. Seeing pedestrians and cyclists at night is not easy out here. Many communities promote low light pollution (of which I am a fan), so it's as dark as a country road. Yet, many of us have our windows darkly tinted because of the daily bright sunlight. I do. I had to be super careful last night to spot a pedestrian crossing the street.
 

Roger830

TUG Member
Joined
Dec 11, 2013
Messages
1,456
Reaction score
587
Points
323
Location
CT
Here's the video:
http://www.dailymail.co.uk/news/art...-pedestrian-killed-self-driving-Uber-car.html

We know two things:
1. The computer controlled car made no attempt to stop.
2. The driver was looking down rather than at the road.

The woman is clearly visible in the headlights walking her bike across the road from the left.
There is some probability that an alert driver could have seen the woman and brake the car.
Even if hit while braking, she could have survived, the car was traveling 38 mph.
 

x3 skier

TUG Review Crew: Veteran
TUG Member
Joined
Apr 17, 2006
Messages
5,266
Reaction score
2,295
Points
649
Location
Ohio and Colorado
Resorts Owned
Steamboat Grand, The West,
Raintree and, formerly, The Allen House
Every time I read about autonomous vehicles, I’m reminded of the movie, 2001 with HAL and Dave.

Dave: “HAL, I want to get out of the car”

HAL: “I’m sorry Dave, I can’t do that”

Cheers
 

Elan

TUG Member
Joined
Jun 6, 2005
Messages
4,455
Reaction score
409
Points
468
Location
Idaho
I've watched the video multiple times since it was released. My opinion is that the accident was largely unavoidable, meaning that with a human controlled car, the pedestrian would still get hit >90% of the time. And the other times would require extraordinary measures that could possibly result in driver fatality. There's little doubt that a human could have possibly reduced the impact speed. Would that have mattered? We'll never know.
As an engineer, it's easy to think of ways to make the autonomous system more "safe", but at what expense? For instance, what if that was a large paper leaf bag blowing across the road? Should the car come to a panic stop, risking a massive pile up? Do we add infrared profiling to the autonomous systems? Point being that there will never be a foolproof system, which is why I asked my original question -- are we willing to accept some possibly "avoidable" accidents in exchange for a lower overall accident rate?
 
Last edited:

littlestar

TUG Member
Joined
Oct 17, 2005
Messages
2,647
Reaction score
367
Points
468
Location
Midwest
Resorts Owned
Disney Vacation Club, Marriott & Wyndham pts
I worked for an insurance company around claims for about 30 years - makes me wonder if a human operating the vehicle totally would have seen a shadow off to the side and at least braked and slowed down before impact.
 
Last edited:

JulieAB

TUG Review Crew: Expert
TUG Member
Joined
Apr 27, 2006
Messages
903
Reaction score
92
Points
388
Resorts Owned
Lifetime in Hawaii, Disney's Grand Californian
We just got a hybrid pacifica 5 months ago with all sorts of fancy advanced driving features like lane departure warnings/correction, etc. Occasionally when driving my husband's old car, I've noticed how accustomed and dependent I've become on those features. I know they're not supposed to be a substitute for being an alert driver, but I think they've made me lazy! :( When I watch that video, I agree, it was likely unavoidable, but I also think the driver was being careless/lazy by looking down for so long at a time. She was too trusting of the tech. I think she could've at least started to break before hitting the bike.
 

bluehende

TUG Review Crew: Veteran
TUG Member
Joined
Jun 6, 2005
Messages
4,507
Reaction score
3,967
Points
598
No way a driver would have been able to avoid this. By stopping the film there is 1 second between when I can see the bike and impact. With no tenths it could be 1.5 +- .5. Here is reaction time

Reaction times vary greatly with situation and from person to person between about 0.7 to 3 seconds (sec or s) or more. Some accident reconstruction specialists use 1.5 seconds. A controlled study in 2000 (IEA2000_ABS51.pdf) found average driver reaction brake time to be 2.3 seconds.

Not a chance for significant slowing. Now maybe someone in real life with younger eyes could have seen it, but I doubt it.

The question is that if the car didn't brake and should have. I do not know what the autonomous actually means. Not knowing the parameters of what the software was supposed to see I would think that some breaking should have taken place. Like Elan said if infrared or only visible recognition was being used is important. Also I cannot see what the "biker" was doing before the accident. I would think that if the biker was moving at a consistent speed and infrared was being used this accident would have been avoided.


One thing to point out is that everyone thinks they are good drivers. I do although with my age I know that my eyes and reaction times have to be slowing. Everyone also sees the video and figures out what they could do. This is great when you see a video you know what is going to happen. Processing the info in real time and make those decisions in real time and you will realize how few times we do the perfect thing with perfect timing. My take on this accident is that had it been a human at the wheel the bike would have been hit nearly 100% of the time and it would have been done at significantly higher velocity.
 

ace2000

TUG Member
Joined
Dec 17, 2006
Messages
5,032
Reaction score
152
Points
498
No way a driver would have been able to avoid this. By stopping the film there is 1 second between when I can see the bike and impact. With no tenths it could be 1.5 +- .5. Here is reaction time

You're assuming that the camera here and the human eye would be equal, which I'm doubtful that is the case.
 

bluehende

TUG Review Crew: Veteran
TUG Member
Joined
Jun 6, 2005
Messages
4,507
Reaction score
3,967
Points
598
You're assuming that the camera here and the human eye would be equal, which I'm doubtful that is the case.

We need a lot more info to actually know what happened. With other factors in the mix (my computer, the video quality,and the edit) the video is not the best for drawing hard and fast conclusions. I would assume that the camera used would be a good one. I know I have used my cheap point and shoot to take a picture then go to my computer to blow up to see details my eyes cann't make out. Wouldn't that camera be much better than the eye. I also would put my safety in the hands of average software as opposed to most of the brains out there.
 

Elan

TUG Member
Joined
Jun 6, 2005
Messages
4,455
Reaction score
409
Points
468
Location
Idaho
I also would put my safety in the hands of average software as opposed to most of the brains out there.

I agree with this. Again, no system is going to be able to handle all situations effectively and efficiently.

It's the mental exercise that intrigues me -- whether we're willing to give up control, knowing that there's a small probability that the system will allow for an otherwise avoidable accident, in exchange for greater overall safety. In that context, we willingly hop in big aluminum tubes that are subjected to tremendous forces to fly around the world. We have no control there, and there's definitely a finite probability that we'll be in an accident, yet...
 

bluehende

TUG Review Crew: Veteran
TUG Member
Joined
Jun 6, 2005
Messages
4,507
Reaction score
3,967
Points
598
I agree with this. Again, no system is going to be able to handle all situations effectively and efficiently.

It's the mental exercise that intrigues me -- whether we're willing to give up control, knowing that there's a small probability that the system will allow for an otherwise avoidable accident, in exchange for greater overall safety. In that context, we willingly hop in big aluminum tubes that are subjected to tremendous forces to fly around the world. We have no control there, and there's definitely a finite probability that we'll be in an accident, yet...

I think it all comes down to whether the rational side of the brain overcomes the emotional. This is one of the great challenges of mankind. Look at your example. Many people refuse to fly even though we have decades of statistics showing it is the safest way to travel. I can give you another. How many people will not go into the ocean because of fears of sharks, yet the overwhelming danger is drowning.
 

Roger830

TUG Member
Joined
Dec 11, 2013
Messages
1,456
Reaction score
587
Points
323
Location
CT
When software controls aircraft it's not interpreting images from a camera.

Software is created by people, it's not infallible. Look at the Wyndham website experience.

I don't see how anyone can say that the accident was unavoidable by looking at the video. It seems more likely that the software was too slow in interpreting the camera's image and it's highly likely that an alert human could have done at least some braking.
 

taterhed

TUG Member
Joined
Sep 28, 2011
Messages
4,536
Reaction score
1,901
Points
399
Location
Virginia
Resorts Owned
Westin WKORV OFD
Marriott's Grande Vista
Worldmark x2
SVV Bella 81k
I, for one, do not think the driver would have avoided the crash...unless hyper-vigilant and extremely quick-reacting.

God rest her soul.....a women walked a bike across a four lane highway, in the dark, out of a crosswalk, directly in front of a large and well-illuminated vehicle moving at nearly 40 MPH. The 'safety' driver was anything-but, and the car wasn't designed (as best I know) for high-speed pedestrian avoidance.

Sad, but there you have it.

BTW: Don't think the Volvo is designed for pedestrian-avoidance at these speeds....not sure what other systems the car might have as well.

Pedestrians or cyclists crossing the path of your vehicle or suddenly swerving in front of it. Depending on the situation, Volvo City Safety can help mitigate the effects of the collision or avoid it altogether if relative speeds are 28 MPH or below.
 

Elan

TUG Member
Joined
Jun 6, 2005
Messages
4,455
Reaction score
409
Points
468
Location
Idaho
When software controls aircraft it's not interpreting images from a camera.

Software is created by people, it's not infallible. Look at the Wyndham website experience.

I don't see how anyone can say that the accident was unavoidable by looking at the video. It seems more likely that the software was too slow in interpreting the camera's image and it's highly likely that an alert human could have done at least some braking.

Highly unlikely that the software/system was "too slow" at least relative to human response. Look how fast airbags deploy. I work in high tech, where micro-seconds (1 millionth of a second) are an eternity to a computer system. We work in pico-seconds (1 trillionth of second).

I don't know this for an absolute fact, but I would presume that the level of testing of autonomous vehicles far exceeds most anything else we've ever dealt with as consumers. They're not turning these cars loose on the open road just to see what happens. Most of the tech is well established. It's the integration that's new.

Again, my opinion is that an ultra-responsive human could have possibly braked slightly before impact. But a human might have been travelling at a higher rate of speed, as well, as I presume autonomous vehicles adhere to the posted limits? I'm not sure I envision any scenario where the collision would be avoided, that wouldn't also result in a rollover.

All speculation, of course, based on 1 video. :shrug:

And again, if this accident hadn't involved a self-driving car, would we be faulting the driver? Or, simply lamenting that someone chose to push her bike across an unlit road in the dark?
 

SmithOp

TUG Review Crew
TUG Member
Joined
Jun 17, 2010
Messages
7,610
Reaction score
3,403
Points
499
Location
Huntington Beach, CA
Resorts Owned
HGVC King's Land 2BR Premier 23.040K Points.
The pedestrian pushing the bike had on dark clothing, I had trouble seeing her until fully illuminated by the headlights. I didn’t even see the reflectors that are usually in the bike spokes.

If I was ever stupid enough to ride a bike at night I would wear a reflective vest.

I’m putting the blame on a jaywalking pedestrian in dark clothes at night, I see very little to convince me it was avoidable.


Sent from my iPad using Tapatalk Pro
 

heathpack

TUG Review Crew: Veteran
TUG Member
Joined
Oct 22, 2008
Messages
4,651
Reaction score
3,750
Points
598
Location
Los Angeles
Resorts Owned
Hyatt High Sierra and Highland Inn
Disney’s Grand Californian and Hilton Head Island
Marriott Barony Beach and Mountainside
MVC Points
Sheraton Broadway Plantation
The pedestrian pushing the bike had on dark clothing, I had trouble seeing her until fully illuminated by the headlights. I didn’t even see the reflectors that are usually in the bike spokes.

If I was ever stupid enough to ride a bike at night I would wear a reflective vest.

I’m putting the blame on a jaywalking pedestrian in dark clothes at night, I see very little to convince me it was avoidable.


Sent from my iPad using Tapatalk Pro

I ride my bike all the time in the dark (before dawn rather than at night). On my bike I have an extremely bright rear blinking light, a bright front white light (to illuminate the road, and I take great care to position it so as not to blind drivers), a white helmet, reflective booties over my shoes (winter only) and ankle lights to delineate the pedaling motion and ID me as a cyclist. In the dark, I’ll only wear bright colored jerseys- yellow or pink or bright orange. Plus a bright orange cycling vest if it’s cold.

In these things, I’m pretty much always on the side of the cyclist.

But I saw this video and think I would have probably hit her too, as a motorist. She just stepped right out of the shadows into the road, not at an intersection or place where you’d expect a pedestrian to be.
 

T-Dot-Traveller

TUG Member
Joined
Jun 10, 2015
Messages
4,645
Reaction score
3,592
Points
348
Location
Canada
Resorts Owned
Mayan Palace Regency
Taranova
I worked for an insurance company around claims for about 30 years - makes me wonder if a human operating the vehicle totally would have seen a shadow off to the side and at least braked and slowed down before impact.
The pedestrian pushing the bike had on dark clothing, I had trouble seeing her until fully illuminated by the headlights. I didn’t even see the reflectors that are usually in the bike spokes.

If I was ever stupid enough to ride a bike at night I would wear a reflective vest.

I’m putting the blame on a jaywalking pedestrian in dark clothes at night, I see very little to convince me it was avoidable.


Sent from my iPad using Tapatalk Pro

Insurance and $$$$ is the issue regardless of "avoidibility "

If I hit a pedestrian - sue me & my insurance coverage .- get 2 + $ million
If a self driving car hits the pedestrian - sue the mfg ,software company. Uber , etc - get 2 +$ BILLION

Deep $$ pockets means Uber stopped all usage ( for now)

When I was 17 / 18 years old - in New York State the minimum required vehicle liability insurance was $ 12,000 .
Lots of friends had the minimum .
The common " 18 year old (ironic)folk wisdom " was - " if you hit a pedestrian - back up and run them over again - your liability insurance was enough to pay for the burial .
but if they were badly injured for life - your wages would be garnished for a long time . PS :< I know of NO ONE who did this >

Different situation for Uber , their insurance coverage & supplier(s)& and all related sueable entities .

And someone died -
 
Last edited:

Roger830

TUG Member
Joined
Dec 11, 2013
Messages
1,456
Reaction score
587
Points
323
Location
CT
It's my understanding that it's not a question of wither the pedestrian to blame, but if the accident could have been prevented or not resulted in death by an alert driver.

It's obvious that the camera saw the woman and bike, then why didn't the super fast computer software react to the image and attempt to brake the car.
 

bluehende

TUG Review Crew: Veteran
TUG Member
Joined
Jun 6, 2005
Messages
4,507
Reaction score
3,967
Points
598
It's my understanding that it's not a question of wither the pedestrian to blame, but if the accident could have been prevented or not resulted in death by an alert driver.

It's obvious that the camera saw the woman and bike, then why didn't the super fast computer software react to the image and attempt to brake the car.

I watched the video again. There is very little reference to be able to tell whether the car applied the brakes. It doesn't look like it, but I am not sure it would be obvious in this video as it is so quick. I am sure they are avoiding the contact for various reasons,but if the video went on for a few seconds it would be obvious if it is slowing. I am sure the investigation would have that info. It would be nice to know.
 

ace2000

TUG Member
Joined
Dec 17, 2006
Messages
5,032
Reaction score
152
Points
498
We need a lot more info to actually know what happened. With other factors in the mix (my computer, the video quality,and the edit) the video is not the best for drawing hard and fast conclusions. I would assume that the camera used would be a good one.

The quality of the camera is not the issue, and we also know 100% that their technology did not work properly. The issue is whether or not a human driver would've reacted in time to prevent the accident. My point is that you can't watch the video to duplicate what the human eye could've seen in those conditions, and therefore make a proper judgement. Anyone that is involved with photography knows what I'm describing.
 

bluehende

TUG Review Crew: Veteran
TUG Member
Joined
Jun 6, 2005
Messages
4,507
Reaction score
3,967
Points
598
The quality of the camera is not the issue, and we also know 100% that their technology did not work properly. The issue is whether or not a human driver would've reacted in time to prevent the accident. My point is that you can't watch the video to duplicate what the human eye could've seen in those conditions, and therefore make a proper judgement. Anyone that is involved with photography knows what I'm describing.

I agree with everything you said except that we know 100% that their technology did not work properly. We do not have enough info to judge that. That is the easiest explanation, but far from the only one.
 
Top