Model X Crash on US-101 (Mountain View, CA)

[Peteski]

I’m not saying that you’re MORE likely to depart the lane with AS.. just that it’s probably negligibly better than LKA systems alone. I suppose if the driver were unconscious AS might be better (for a while) whereas LKA (GM at least) would eventually ping pong you out of the lane. But if lane excursions aren’t noticeably reduced (compared to LKA) but driver inattentiveness is mildly increased.. well, hard to say the net impact on over all safety.

I still use AS all the time.. I like to think I remain attentive even though my system doesn’t require hands on the wheel. But if the system doesn’t require intervention for tens of minutes there is always the chance I might have a lapse of better judgement :/

Okay, I see your point now. But here's a thought. Every car I've owned or driven with LDW I've invariably switched it off because of all the annoying bleeps/vibrations etc. when not wanted. Of course I could switch it on when appropriate (motorways journeys etc) but I rarely did because you either had to dig through system menus or grope around for a badly positioned switch. I also switched off Volvo's steering assist because it would ping-pong between lane markings and if you corrected the steering manually it just felt like 2 people were trying to drive the car simultaneously. So off it went!

So all those systems were all pretty useless for me in the real world. Tesla AS on the other hand is competent and convenient enough to use on many roads, therefore I often use it. It's actually the only steering aid that has been of any real benefit to my driving.

 

[JRP3]

Look. 131 pages on one car fire. There have been Roughly 14 Tesla that burned, mostly after very high speed collisions like the one discussed here. Icev are 64x more likely to burn. The collisions are so devastating that the people on the cars are not likely to survive those injuries regardless of the fire. The packs generally burn very slowly and very hot over a long period of time. But I can assure you, it's safer for a firefighter to put out 1/64th the fires even if it takes 10x the time and water. Stop the madness, this conversation is of the rails

Since you missed the entire point of the discussion you aren't qualified to judge it's rail status

 

[greghanssen]

Okay, I see your point now. But here's a thought. Every car I've owned or driven with LDW I've invariably switched it off because of all the annoying bleeps/vibrations etc. when not wanted. Of course I could switch it on when appropriate (motorways journeys etc) but I rarely did because you either had to dig through system menus or grope around for a badly positioned switch. I also switched off Volvo's steering assist because it would ping-pong between lane markings and if you corrected the steering manually it just felt like 2 people were trying to drive the car simultaneously. So off it went!

So all those systems were all pretty useless for me in the real world. Tesla AS on the other hand is competent and convenient enough to use on many roads, therefore I often use it. It's actually the only steering aid that has been of any real benefit to my driving.

Valid point... I’ve heard others say the same. I let my exGF drive my Volt and the LKA freaked her out.. I kind of like the Volt LKA because it’s more or less silent yet puts enough torque on the wheel to mildly nudge me back towards lane center. The Toyota LKA on the other hand I found very irritating as it always beeps and provides little actual corrective torque (not sufficient to re-center the vehicle when your hands are on the wheel).

Ironically the Toyota system might be safer (if one opted to use it!) Less likely to let the car reach the edges of the lane to avoid the dreaded beep but a subtle/silent nudge from the Volt doesn’t bother me at all... yet it may also contribute to me paying LESS attention)

 

[Barklikeadog]

Okay, I see your point now. But here's a thought. Every car I've owned or driven with LDW I've invariably switched it off because of all the annoying bleeps/vibrations etc. when not wanted. Of course I could switch it on when appropriate (motorways journeys etc) but I rarely did because you either had to dig through system menus or grope around for a badly positioned switch. I also switched off Volvo's steering assist because it would ping-pong between lane markings and if you corrected the steering manually it just felt like 2 people were trying to drive the car simultaneously. So off it went!

So all those systems were all pretty useless for me in the real world. Tesla AS on the other hand is competent and convenient enough to use on many roads, therefore I often use it. It's actually the only steering aid that has been of any real benefit to my driving.

The first thing I do in every vehicle I purchase is open the manual to the warnings page and go through the motions to turn off the seatbelt alarm.
turn car 3/4 on, turn off, honk horn twice, press 'ok' turn car on 3/4....

 

[Peteski]

Valid point... I’ve heard others say the same. I let my exGF drive my Volt and the LKA freaked her out.. I kind of like the Volt LKA because it’s more or less silent yet puts enough torque on the wheel to mildly nudge me back towards lane center. The Toyota LKA on the other hand I found very irritating as it always beeps and provides little actual corrective torque (not sufficient to re-center the vehicle when your hands are on the wheel).

Ironically the Toyota system might be safer (if one opted to use it!) Less likely to let the car reach the edges of the lane to avoid the dreaded beep but a subtle/silent nudge from the Volt doesn’t bother me at all... yet it may also contribute to me paying LESS attention)

Current cars I have with LDW are a Nissan and of course the Model X. LDW on the Nissan is in the annoying beep category and I don't see much point in the Tesla LDW when using AS. But the biggest problem I have with LDW is that most of my driving is mixed motorway, town and narrow country winding roads where you often cross the lane markings in normal safe driving. LDW on those roads is unbearable, so I have to switch it off. Of course I could switch it on for motorway use when relevant, but it's not that convenient to toggle on/off in the Nissan (early cars didn't even have a physical button and then they added one as an afterthought out of sight under the dashboard next to the bonnet release and fuel filler flap release buttons!). That's what I really like about Tesla AS, it's so easy to switch on/off without taking your hands off the wheel.

 

[TEG]

Small update:

I had seen Caltrans doing some work at that location including extended lane closure.
It looks like they cleaned things up somewhat, such as removing those misleading old signs (leftover from construction last year) that had said "Carpool Lane Closed".
It also looks like they did some repainting on some of the lane markers.
Still no sign of any road chevrons, candlestick delineators or other more obvious warnings of the upcoming gore point.

 

[Brando]

Small update:

I had seen Caltrans doing some work at that location including extended lane closure.
It looks like they cleaned things up somewhat, such as removing those misleading old signs (leftover from construction last year) that had said "Carpool Lane Closed".
It also looks like they did some repainting on some of the lane markers.
Still no sign of any road chevrons, candlestick delineators or other more obvious warnings of the upcoming gore point.

Is the crash barrier "reset" ??

 

[TEG]

Yes, currently it looks to be in the proper state.

I still wish they would mark the gore area more clearly.

 

[mongo]

Preliminary report
https://www.ntsb.gov/investigations/AccidentReports/Pages/HWY18FH011-preliminary.aspx

• The Autopilot system was engaged on four separate occasions during the 32-minute trip, including a continuous operation for the last 18 minutes 55 seconds prior to the crash.
  
• During the 18-minute 55-second segment, the vehicle provided two visual alerts and one auditory alert for the driver to place his hands on the steering wheel. These alerts were made more than 15 minutes prior to the crash.
  
• During the 60 seconds prior to the crash, the driver’s hands were detected on the steering wheel on three separate occasions, for a total of 34 seconds; for the last 6 seconds prior to the crash, the vehicle did not detect the driver’s hands on the steering wheel.
  
• At 8 seconds prior to the crash, the Tesla was following a lead vehicle and was traveling about 65 mph.
  
• At 7 seconds prior to the crash, the Tesla began a left steering movement while following a lead vehicle.
  
• At 4 seconds prior to the crash, the Tesla was no longer following a lead vehicle.
  
• At 3 seconds prior to the crash and up to the time of impact with the crash attenuator, the Tesla’s speed increased from 62 to 70.8 mph, with no precrash braking or evasive steering movement detected.

PURE OPINION WARNING:
4 seconds at 62-70.8 MPH is ~400 ft before barrier. Based on Google maps, that appears to be within the gore triangle.
Sounds like the same sort of scenario as the other "leading vehicle changes lanes and a stationary object in the path of travel" with the added factor of the 'lane' of travel being the gore point.

 

[TEG]

Contemplating why the car started to steer left, and then accelerate seems like it could be useful in reconstructing what happened.
Did "autopilot" do that, or the driver?

Autopilot handled that area on many other days including for many other cars, so there was something different that fateful day.
Perhaps it was the exact timing of the "lead" car changing lanes that tripped things up.
Perhaps the lead car started to head into the gore area and then swerved away at that last moment and so lead/directed the Tesla into the "gore area" (not supposed to be a lane).

I wonder how much autopilot considers following a lead vehicle vs what it thinks are the lane markings.
Is the car in front only used for speed / distance, or a guide for steering control too?

 

[mongo]

Contemplating why the car started to steer left, and then accelerate seems like it could be useful in reconstructing what happened.
Did "autopilot" do that, or the driver?

Autopilot handled that area on many other days including for many other cars, so there was something different that fateful day.
Perhaps it was the exact timing of the "lead" car changing lanes that tripped things up.
Perhaps the lead car started to head into the gore area and then swerved away at that last moment and so lead/directed the Tesla into the "gore area" (not supposed to be a lane).

I wonder how much autopilot considers following a lead vehicle vs what it thinks are the lane markings.
Is the car in front only used for speed / distance, or a guide for steering control too?

Cruise was set to 75, so likely auto-accelerating.

Probably confirmation bias, but this fits into my scenario of the Tesla following closely into the gore section, then the lead car moving out of the 'lane' leaving the Tesla in it.

from Tesla blog:

In the moments before the collision, which occurred at 9:27 a.m. on Friday, March 23rd, Autopilot was engaged with the adaptive cruise control follow-distance set to minimum. The driver had received several visual and one audible hands-on warning earlier in the drive and the driver’s hands were not detected on the wheel for six seconds prior to the collision. The driver had about five seconds and 150 meters of unobstructed view of the concrete divider with the crushed crash attenuator, but the vehicle logs show that no action was taken.

 

[EVie'sDad]

"It serves as a tragic reminder that drivers need to always pay attention when using Autopilot and be ready to take control at all time."

Indeed.

NTSB releases preliminary report on fatal Tesla crash on Autopilot

 

[alcibiades]

Preliminary report
https://www.ntsb.gov/investigations/AccidentReports/Pages/HWY18FH011-preliminary.aspx



PURE OPINION WARNING:
4 seconds at 62-70.8 MPH is ~400 ft before barrier. Based on Google maps, that appears to be within the gore triangle.
Sounds like the same sort of scenario as the other "leading vehicle changes lanes and a stationary object in the path of travel" with the added factor of the 'lane' of travel being the gore point.

Tesla's statement on this includes:

"the only way for this accident to have occurred is if Mr. Huang was not paying attention to the road, despite the car providing multiple warnings to do so."

But the report says that there were no alerts in the last 15 minutes?

 

[mongo]

Tesla's statement on this includes:

"the only way for this accident to have occurred is if Mr. Huang was not paying attention to the road, despite the car providing multiple warnings to do so."

But the report says that there were no alerts in the last 15 minutes?

There were 3 warnings in the last 18 minute activation window, along with the static warnings in the manual and autopilot enable screen, but no nags the last 15 minutes.

 

[EVie'sDad]

What about his cell phone records? Was he using it during the drive?(notably illegal and dangerous)

 

[TEG]

Interesting that the report shows the damaged crash attenuator removed from the site.

I guess they must have replaced it with a new one after the crash.

 

[TEG]

What about his cell phone records? Was he using it during the drive?(notably illegal and dangerous)

There are any number of ways someone could be distracted and not watching the road, including reading a book, or running an app (not phone call) on a phone.
Would there be any record if he was doing anything with the car's touchscreen?

 

[mongo]

Link is down apparently,
NTSB releases preliminary report on fatal Tesla crash on Autopilot
Has it in article
I'd upload screenshots, but that seems broken lately.

 

[MorrisonHiker]

Contemplating why the car started to steer left, and then accelerate seems like it could be useful in reconstructing what happened.
Did "autopilot" do that, or the driver?

Autopilot handled that area on many other days including for many other cars, so there was something different that fateful day.
Perhaps it was the exact timing of the "lead" car changing lanes that tripped things up.
Perhaps the lead car started to head into the gore area and then swerved away at that last moment and so lead/directed the Tesla into the "gore area" (not supposed to be a lane).

I wonder how much autopilot considers following a lead vehicle vs what it thinks are the lane markings.

That's what it sounds like to me. The lead car wasn't paying attention and drove into the gore point. The X followed and when the lead car moved into the correct lane, the X probably locked onto the lane line that was still visible and increased speed before hitting the attenuator.

 

[mongo]

That's what it sounds like to me. The lead car wasn't paying attention and drove into the gore point. The X followed and when the lead car moved into the correct lane, the X probably locked onto the lane line that was still visible and increased speed before hitting the attenuator.

There is video somewhere in this thread that demonstrates that scenario.
With the worn right gore line, it is not intuitive that the car is going the wrong way. Especially to a center mounted camera.