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

[Nosken]

Just curious.. did AP1 have the same "must have hands on wheel at all times" byline? When did the message change? Just since the Florida accident?

Yes, AP1 introduced the "Must have hands on wheel" later, after initial implementation of AutoSteer didn't have the nag. The nag seemed to be timed to a bunch of miscreants posting YouTube videos "Driving" from the back seat. AP1 was decent from the beginning, but there is no way I would not be hands on almost all the time. I had AP! active the first day it was offered to me on my March 2015 P85D. I am now on AP2, and it finally has parity to AP1. Recent updates has made driving so much smoother.

 

[Icer]

Right... But ....

I hear ya. Really.

But I like that the people have been given the latitude, freedom to do this.
Riding the bleeding edge.

Literally.
*
There is now a sign up for new releases.
Be careful people, more stupid is not needed.
*
BTW, I finally figured it out, with some help from
Miss Marple's good looking
and pleasantly overdeveloped
granddaughter.

The unfortunate Mr Wong had died.
Heart attack.

The impact acted as super CPR,
got his ticker going again.
*
That is stupid.
How could a gal named "Miss" have a granddaughter, buddy?

Some things remain a mystery.

 

[buttershrimp]

Does anyone else want to see Tweak from Southpark using Autopilot?

 

[MrAustraliaTax]

Just curious.. did AP1 have the same "must have hands on wheel at all times" byline? When did the message change?

The initial release of AP (7.0) effectively had no nags. You could easily go 30 minutes without any messages at all.

7.0 also had a habit of failing catastrophically and then flashing "take over immediately" whilst trying to kill you. Whilst labelled as "BETA" there was a strong implication of "go for it" rather than "hands on wheel" at the time. All the Tesla demos were prolonged hands-off at the time.

7.1 introduced the nags at 5 minute intervals in response to the various You-tube videos of people sitting in the passenger seat or back seat on public roads with AP activated.

8.0 brought in 1 minutely nags. Reading slightly between the lines, these were probably forced "voluntarily" on Tesla during the Florida crash investigation where it was noted that torque monitoring was inadequate.

(Edit: Nosken beat me to it).

 

[Peteski]

Okay this a long thread as you might expect and I've lost track of what we may have learned so far about this accident. I'm especially wondering why AEB didn't avoid this crash as you would presumably expect AEB to activate if driving head-on toward a large stationary object? Was AEB perhaps disabled? Or simply ineffective at highway speed? I expect we don't know the details of the car settings, but thought I would ask.

 

[mongo]

Okay this a long thread as you might expect and I've lost track of what we may have learned so far about this accident. I'm especially wondering why AEB didn't avoid this crash as you would presumably expect AEB to activate if driving head-on toward a large stationary object? Was AEB perhaps disabled? Or simply ineffective at highway speed? I expect we don't know the details of the car settings, but thought I would ask.

AEB is designed around not hitting traffic (v.s. totally stationary things). FCW is the warning system.
The object was not really that large, less that half the width of the car.
AEB is likely setup up to reduce false positives and to give the driver maximum opportunity to maneuver themselves.

For more detailed discussions from this thread put:
site:Model X Crash on US-101 (Mountain View, CA) AEB
In Google

 

[AWDtsla]

AEB is designed around not hitting traffic (v.s. totally stationary things). FCW is the warning system.
The object was not really that large, less that half the width of the car.
AEB is likely setup up to reduce false positives and to give the driver maximum opportunity to maneuver themselves.

For more detailed discussions from this thread put:
site:Model X Crash on US-101 (Mountain View, CA) AEB
In Google

AEB NEVER stops a crash. It only reduces crash velocity if traveling above a certain minimum speed.

 

[mongo]

AEB NEVER stops a crash. It only reduces crash velocity if traveling above a certain minimum speed.

True, I did not work that well. It is designed to reduce speed by 25 MPH per this posting: Thoughts on why AEB is designed to only reduce speed by 25mph?.

 

[bhzmark]

AEB NEVER stops a crash. It only reduces crash velocity if traveling above a certain minimum speed.

I posted a link above to the thread with a video of aeb bringing the car to a stop to avoid hitting the stopped car ahead after the Prius turned away.

Probably stopped the car because speed was relatively low.

Also see

 

[Az_Rael]

I had a thought - do we think this was an AP2 car? The reason I wonder is because the force required to override the wheel in AP2 (at least before 10.4) is a lot higher than with AP1. In my 3 I had a couple of scary disengages at 70mph where I corrected the car with the wheel, AP2 wrestled with me for a bit, then let go, causing me to jerk the wheel. It was scary enough that I stopped using the wheel to disengage because I was worried about losing control at freeway speeds.

There must be something wrong with your car. Get it checked. I find the torque required to disengage AP2 to be the same as AP1. It's minimal & sometimes I kick myself out of AP2 just because of the minor pressure of my hands on the steering wheel.

Just as a follow up on this - I did take my car to the service center for the high force required to overcome Autosteer. My car was checked out and determined it was normal and I should just use the stalk or the brake instead (owner's manual instructions notwithstanding). I asked if the wheel break-away torque was adjustable and the technician told me that they could not change that setting, it was set by the engineers and they didn't have access to it.

So it is what it is, which is unfortunate, as it does seem to vary quite a bit between different cars, just based on my loaner experiences.

 

[Economite]

But I've seen no data or fact that supports the idea you need for your hypothesis that the tweaks to AEB after October 2015 were more material and affected air bag deployments more than the massive change in the availability of auto steering.

First... It's not just improvements to AEB, but also improvements to TACC, ACW, BSM, and other things.

Second... I agree that there is no published data that disaggregates AS from these other functions. But I see no reason to assume that just because AS was new (while the others were improved), we must assume that AS positively contributed towards the safety increase.

The other features are clearly ones that directly serve a safety purpose. Even if it works reliably, the only real safety improvement that would come from AS is a potential decrease in cars weaving within their lanes or drifting over lane lines. In my experience, this sort of bas driving rarely causes accidents with other vehicles, since the other cars just avoid the collision. So, what you're left with from AS is a feature that enables some drivers to pay less attention to driving (which has to have a negative safety benefit) and may allow some drivers to be less tired by routine driving and therefore be more aware of surroundings (which might have a positive safety benefit). As compared with TACC and ACW (which mitigate reduce rear end collisions) and BSM (which reduces lane changes that cut off other cars), even a perfectly functioning AS would seem to have primarily convenience, rather than safety benefits.

I know that many AS users perceive it as "making them better drivers." But that perception may not map towards actual likelihood of having an airbag-activating crash. These crashes are rare-- on the order of one crash per million miles. That means that if every driver drives 15,000 miles per year; only one in sixty-six drivers would experience. And the average driver will have only zero or one such accident in their lifetimes. Thus most of the "AS saves" and "near misses" that people are experiencing are probably incidents where, without AP, there would also only have been a "last minute save" or "near miss" rather than an actual accident.

For rare events, only statistical analysis can really determine causes and mitigations. Anecdotes are only somewhat useful. And we just don't have statistical proof in either direction with AS.

 

[bhzmark]

Anecdotes are only somewhat useful. And we just don't have statistical proof in either direction with AS.

You never have statistical proof. You just have more or less statistically relevant correlations based on the data as outlined in the nhtsa report. That correlation based on their experienced analysis of the data is much more persuasive than your unfounded hypothesis. It's possible you're right. But probably not. More likely that both contributed to increasing safety.

 

[Economite]

You never have statistical proof. You just have more or less statistically relevant correlations based on the data as outlined in the nhtsa report. That correlation based on their experienced analysis of the data is much more persuasive than your unfounded hypothesis. It's possible you're right. But probably not. More likely that both contributed to increasing safety.

I've never claimed that I have statistical support. I've only stated, correctly I think, that my theory is not disproved by the statistics and that your theory is not proved by the statistics. I don't believe that that statistic presented by NHTSA provides support for the statement "More likely that both contributed to increasing safety." The statistic simply doesn't say anything about how the improvement in crash statistics can be allocated amongst the various changes that happened to Tesla features during the test period.

 

[bhzmark]

my theory is not disproved by the statistics and that your theory is not proved by the statistics.

That's like when tobacco manafacturers said the data/stats doesnt prove cigarettes are harmful.

Which is why i explained that waiting for proof and until then entertaining any hypothesis is fruitless. and why I keep saying, and you keep ignoring and returning to your quest for proof, the concept of more or less supported by the data. Your hypothesis is less supported; mine is more supported. And NHTSA chose to emphasize a correlation that they thought relevant.

I've never claimed that I have statistical support.

I agree you have no statistical support.

 

[Economite]

mine and nhtsas is more supported.

NHTSA did not conclude that AS provides a safety improvement. NHTSA only concluded that, as of the time of its report, there was not sufficient evidence for it to conclude that AS was a defective product because of its failure to react to cross traffic (as in Florida).

 

[bhzmark]

NHTSA did not conclude that AS provides a safety improvement. NHTSA only concluded that, as of the time of its report, there was not sufficient evidence for it to conclude that AS was a defective product because of its failure to react to cross traffic (as in Florida).

NHTSA only took the time to make this chart emphasizing a correlation that they thought relevant and worthy of being one of only two charts in their entire report.

 

[Economite]

Bertl, is that you?

NHTSA only took the time to make this chart

View attachment 295392

As we've been discussing, that chart compares bucket one and bucket two miles. It does not attempt to say that AS was the cause of the decrease or to isolate AS from other changes.

Here is the exact description NHTSA gives of the data in that chart:

"Crash Rates. ODI analyzed mileage and airbag deployment data supplied by Tesla for all MY 2014 through 2016 Model S and 2016 Model X vehicles equipped with the Autopilot Technology Package, either installed in the vehicle when sold or through an OTA update, to calculate crash rates by miles travelled prior to and after Autopilot installation. Figure 11 shows the rates calculated by ODI for airbag deployment crashes in the subject Tesla vehicles before and after Autosteer installation. The data show that the Tesla vehicles crash rate dropped by almost 40 percent after Autosteer installation. Autopilot Updates. Since it released Autopilot in October 2015, Tesla has made continuous updates to the system’s firmware that are made available to consumers as OTA updates. These updates have included changes to improve TACC, AEB and Autosteer performance, as well as adding new driver assistance safety features, such as In-Path Stationary Object (IPSO) braking and Pedal Misapplication Mitigation (PMM). In September 2016, Tesla released its 8.0 firmware update which included revisions in the driver monitoring strategy, as well as several enhancements to AEB, DBS, and TACC performance. "

The footnotes to that section state:
"21 Approximately one-third of the subject vehicles accumulated mileage prior to Autopilot installation.
22 The crash rates are for all miles travelled before and after Autopilot installation and are not limited to actual Autopilot use."

And here is the overall conclusion of the NHTSA report:

"Advanced Driver Assistance Systems, such as Tesla’s Autopilot, require the continual and full attention of the driver to monitor the traffic environment and be prepared to take action to avoid crashes. Automated Emergency Braking systems have been developed to aid in avoiding or mitigating rear-end collisions. The systems have limitations and may not always detect threats or provide warnings or automatic braking early enough to avoid collisions. Although perhaps not as specific as it could be, Tesla has provided information about system limitations in the owner’s manuals, user interface and associated warnings/alerts, as well as a driver monitoring system that is intended to aid the driver in remaining engaged in the driving task at all times. Drivers should read all instructions and warnings provided in owner’s manuals for ADAS technologies and be aware of system limitations. While ADAS technologies are continually improving in performance in larger percentages of crash types, a driver should never wait for automatic braking to occur when a collision threat is perceived. NHTSA’s examination did not identify any defects in design or performance of the AEB or Autopilot systems of the subject vehicles nor any incidents in which the systems did not perform as designed. AEB systems used in the automotive industry through MY 2016 are rear-end collision avoidance technologies that are not designed to reliably perform in all crash modes, including crossing path collisions. Tesla appears to have recognized HMI factors, such as the potential for driver distraction, in its design process for the Autopilot system. Tesla's design included a hands-on the steering wheel system for monitoring driver engagement. That system has been updated to further reinforce the need for driver engagement through a "strike out" strategy. Drivers that do not respond to visual cues in the driver monitoring system alerts may "strike out" and lose Autopilot function for the remainder of the drive cycle. A safety-related defect trend has not been identified at this time and further examination of this issue does not appear to be warranted. Accordingly, this investigation is closed. The closing of this investigation does not constitute a finding by NHTSA that no safety-related defect exists. The agency will monitor the issue and reserves the right to take future action if warranted by the circumstances." [italics added].

 

[Barklikeadog]

As we've been discussing, that chart compares bucket one and bucket two miles. It does not attempt to say that AS was the cause of the decrease or to isolate AS from other changes.

Here is the exact description NHTSA gives of the data in that chart:

Yea... I don't understand why this is promoted like it is unique. This is happening all across the auto industry with blind spot, AEB and lane departure warnings.

New report shows how many accidents, injuries collision avoidance systems prevent
The next time you are driving and a lane departure warning system or blind spot alert keeps you from steering into another vehicle, count your blessings.

A new study by the Insurance Institute for Highway Safety found collision avoidance systems dramatically cut the number of accidents and injury-related crashes.

"These systems are saving lives," said Jessica Cicchino, IIHS vice president for research. "The numbers show warning systems work."


Cicchino analyzed more than 5,000 accidents in 2015 involving the types of collisions that lane departure and blind spot warning systems are designed to prevent. She also looked at what happened in those vehicles that had the warning systems.

She found the rate of single-vehicle, sideswipe and head-on crashes was 11 percent lower in vehicles with the warning systems. More importantly, the collision avoidance technology cut the rates of injury crashes of the same type by 21 percent, according to the study.

What does that mean in terms of drivers and passengers avoiding injuries?

The IIHS says if all passenger vehicles had been equipped with lane departure warning systems in 2015, more than 55,000 injuries would have been prevented.

"These systems have shown they prevent some of the deadliest accidents on the road because they give the driver time to avoid a crash," said Cicchino.

While the data should drive home the importance of collision avoidance systems, it also shows many drivers may be turning off warning systems in their vehicles.

Why do researchers suspect this is happening?

The IIHS compared the results of this report with two similar studies conducted in 2015. One study focused on trucking fleets in the U.S., the other on Volvo cars in Sweden. They found lane departure warning systems cut crash rates by roughly 50 percent.

 

[Economite]

A few other interesting footnotes in the NHTSA report (https://static.nhtsa.gov/odi/inv/2016/INCLA-PE16007-7876.PDF):

fn 19. "An unreasonable risk due to owner abuse that is reasonably foreseeable (i.e., ordinary abuse) may constitute a safety-related defect."

fn 20. "Driver misuse in the context of semi-autonomous vehicles is an emerging issue and the agency intends to continue its evaluation and monitoring of this topic, including best practices for handling driver misuse as well as driver education."

fn 23. "While drivers have a responsibility to read the owner’s manual and comply with all manufacturer instructions and warnings, the reality is that drivers do not always do so. Manufacturers therefore have a responsibility to design with the inattentive driver in mind."

 

[Economite]

The next time you are driving and a lane departure warning system or blind spot alert keeps you from steering into another vehicle, count your blessings.

This is a really good point. Seems to me like LDW delivers essentially all of the lane-keeping safety benefits of AS, but without AS's steering mistakes and possibility of encouraging driver inattention. And most AS updates probably also have resulted in similar updates to LDW.