Autopilot had no involvement in fatal Texas Tesla crash, NTSB says

[Dewg]

To counter some of the other threads blaming AP/FSD for crashes, here is an article that was just released after NTSB finished their investigation:

Autopilot had no involvement in fatal Texas Tesla crash, NTSB says

Alcohol and sedating antihistamines were to blame, according to crash investigator.

arstechnica.com

 

[Dan D.]

Good that they concluded that AP was not used and TACC was not engaged. The car accelerated from rest at a high rate and crashed in seconds at a final speed of 57mph. The whole controversy started when the police chief (I think) said it looked like nobody was in the driver's seat.

The driver and passenger both died from blunt force trauma, fire, and additionally from smoke inhalation for the driver. The driver being capable of making it from the driver's seat to the rear seat where he was found.

It is unfortunate that the NTSB report failed to investigate the post-crash survivability. Almost like it was unnecessary to investigate.

For example, why did the passenger not move from his seat? At 57mph into a solid tree, was there sufficient intrusion and damage to the car that the airbags' action & seatbelt systems were insufficient to prevent incapacitating injuries. Quite likely but they did not mention this. As the passenger did not have reported smoke inhalation, we might infer that he did not survive long enough to be killed by the later fire, but that the injuries were the cause of his death. Or possibly that his respiration was insufficient to receive smoke injury or that the autopsy could not prove it one way or another. Whether he could or could not open the door (jammed against trees?) appears to be a moot point, or at least the NTSB did not mention this.

The driver managed to reach the rear seat. Was he unable to open the door/window to escape or did he fail to try? We can't know as there were no witnesses to any attempt. He also succumbed to injury, smoke inhalation, and fire.

The NTSB report mentioned that damage to the 12V battery can prevent electrical opening of doors and that there is a manual system in the rear (though tricky to locate). They concluded that the 12V system was destroyed. They did not consider whether the driver attempted to manually open the doors. Or course it would be impossible to prove or disprove whether the driver tried, as he did not survive and there were no witnesses.

Nonetheless. I wish the NTSB report had called out Tesla for not having an easy method of opening the rear doors or opening the windows in the event of 12V failure. After all it is likely in my opinion that the driver survived the crash, was physically mobile for another period of time, and survived for the period of time it took for the car to burn. In that time he managed to undo his seat belt and crawl to the back seat (implied).

If the rear door was open or easily manually opened, I believe he would have at least managed to fall out of the car and perhaps survived the smoke inhalation and fire. If the window was open or easily manually opened he may have been able to climb out or at least survive the smoke inhalation.

All suppositions but I think worthy of mentioning in terms of survivability, after all his failure to be able to get out of the car may indeed have been the cause of his death, people do survive blunt force trauma injuries. The NTSB did not seem to consider this. They mentioned his alcohol and prescription drugs, and the reckless speed, but that did not make his death inevitable. He may still have survived had he been able to get out of the car.

In my opinion the NTSB should advise that car makers improve the post-crash survivability of cars, especially ones with electrical-only door & window systems. (Hidden and tricky-to-find manual systems are poor safety systems).

 

[Dewg]

Good that they concluded that AP was not used and TACC was not engaged. The car accelerated from rest at a high rate and crashed in seconds at a final speed of 57mph. The whole controversy started when the police chief (I think) said it looked like nobody was in the driver's seat.

The driver and passenger both died from blunt force trauma, fire, and additionally from smoke inhalation for the driver. The driver being capable of making it from the driver's seat to the rear seat where he was found.

It is unfortunate that the NTSB report failed to investigate the post-crash survivability. Almost like it was unnecessary to investigate.

For example, why did the passenger not move from his seat? At 57mph into a solid tree, was there sufficient intrusion and damage to the car that the airbags' action & seatbelt systems were insufficient to prevent incapacitating injuries. Quite likely but they did not mention this. As the passenger did not have reported smoke inhalation, we might infer that he did not survive long enough to be killed by the later fire, but that the injuries were the cause of his death. Or possibly that his respiration was insufficient to receive smoke injury or that the autopsy could not prove it one way or another. Whether he could or could not open the door (jammed against trees?) appears to be a moot point, or at least the NTSB did not mention this.

The driver managed to reach the rear seat. Was he unable to open the door/window to escape or did he fail to try? We can't know as there were no witnesses to any attempt. He also succumbed to injury, smoke inhalation, and fire.

The NTSB report mentioned that damage to the 12V battery can prevent electrical opening of doors and that there is a manual system in the rear (though tricky to locate). They concluded that the 12V system was destroyed. They did not consider whether the driver attempted to manually open the doors. Or course it would be impossible to prove or disprove whether the driver tried, as he did not survive and there were no witnesses.

Nonetheless. I wish the NTSB report had called out Tesla for not having an easy method of opening the rear doors or opening the windows in the event of 12V failure. After all it is likely in my opinion that the driver survived the crash, was physically mobile for another period of time, and survived for the period of time it took for the car to burn. In that time he managed to undo his seat belt and crawl to the back seat (implied).

If the rear door was open or easily manually opened, I believe he would have at least managed to fall out of the car and perhaps survived the smoke inhalation and fire. If the window was open or easily manually opened he may have been able to climb out or at least survive the smoke inhalation.

All suppositions but I think worthy of mentioning in terms of survivability, after all his failure to be able to get out of the car may indeed have been the cause of his death, people do survive blunt force trauma injuries. The NTSB did not seem to consider this. They mentioned his alcohol and prescription drugs, and the reckless speed, but that did not make his death inevitable. He may still have survived had he been able to get out of the car.

In my opinion the NTSB should advise that car makers improve the post-crash survivability of cars, especially ones with electrical-only door & window systems. (Hidden and tricky-to-find manual systems are poor safety systems).

Good points - but makes me think of the tools that are/were popular pre-Tesla designed to break the glass so people could crawl out of the window in various scenarios, usually if the car is submerged.

I know that my MY has manual door releases in the front seats, but not in the back seats. This is likely by design for child safety. If it's easy to open the back doors via a manual release, children could open the doors even with child-locks enabled.

 

[2101Guy]

i keep one of these in all of my vehicles. Just in case..

Car Safety Hammer Window Glass Breaker Seat Belt Emergency Escape Cutter Tool | eBay

The sharp blade with good corrosion resistant has a detached cover to avoid unnecessary hurting. Lightweight & Small size, easy to carry around, with a base can be placed anywhere in the car. The classic black color and simple shape design build a fashion appearance.

www.ebay.com

 

[Mardak]

NTSB should advise that car makers improve the post-crash survivability of cars

Including situations where the driver was already dealing with a crash, airbag deployment, fire and smoke, and maybe most relevant here, "impairment from alcohol intoxication in combination with the effects of two sedating antihistamines?"

 

[jsmay311]

My goodness, the NTSB sure works slowly.

IIRC correctly, it was known within a few days or weeks that AP wasn't involved in this crash.

 

[Dan D.]

I know that my MY has manual door releases in the front seats, but not in the back seats. This is likely by design for child safety. If it's easy to open the back doors via a manual release, children could open the doors even with child-locks enabled.

Consider every car until Tesla electric doors came along. You can open the rear doors manually, unless child-safety locks are engaged. It's the child-safety locks that prevent escape.

What Tesla did is remove all manual releases. It's got nothing to do with child-safety locks.

If Tesla had manual-release handles in the rear that would be fine. Sure, have a child-safety interlock if they want. Even better, have a 12V failure fail-safe so the child-safety feature disengages, such as in a crash.

They can't say having no easy manual release is a child-safety feature, that's what the child-safety lock is for, it blocks the manual release. Every modern car has a rear manual release with a child -safety lock. Well every car did until Tesla removed manual releases. Now there is essentially no door release.

And if a child pulled Tesla's hidden release then the door would open, while driving even. So in some respects Tesla's system is worse then. It's got nothing to go with child-safe.

 

[Dan D.]

i keep one of these in all of my vehicles. Just in case..

Car Safety Hammer Window Glass Breaker Seat Belt Emergency Escape Cutter Tool | eBay

The sharp blade with good corrosion resistant has a detached cover to avoid unnecessary hurting. Lightweight & Small size, easy to carry around, with a base can be placed anywhere in the car. The classic black color and simple shape design build a fashion appearance.

www.ebay.com

Those are essentially useless with newer Tesla all-laminated windows.

 

[Dewg]

If Tesla had manual-release handles in the rear that would be fine. Sure, have a child-safety interlock if they want. Even better, have a 12V failure fail-safe so the child-safety feature disengages, such as in a crash.

An excellent suggestion - you should present this to Tesla and perhaps it makes it into future designs.

 

[Dan D.]

An excellent suggestion - you should present this to Tesla and perhaps it makes it into future designs.

I'll take that as praise, not sarcasm. It's hard to tell sometimes

I'd really like it to come from the NTSB and the NHTSA to suggest and mandate proper safety egress if car-makers won't do it willingly.

 

[Dewg]

I'll take that as praise, not sarcasm. It's hard to tell sometimes

I'd really like it to come from the NTSB and the NHTSA to suggest and mandate proper safety egress if car-makers won't do it willingly.

It was sincere. However, as you noted, other cars' manual releases won't work with child-locks enabled. I wonder if the other manufacturers have your idea of a 12-V fail-safe? The last car I checked with child-locks (10-12 years ago) it was a physical switch inside the door mechanism, so it would have prevented opening the door in a crash as well.

 

[Tam]

...NTSB...

That means the dude never used Autopilot when he first bought it: Neither before the accident or during the accident.

"Review of the data indicated no use of the Autopilot system at any time during this ownership period of the vehicle, including the timeframe up to the last transmitted timestamp on April 17, 2021"

 

[Dan D.]

It was sincere. However, as you noted, other cars' manual releases won't work with child-locks enabled. I wonder if the other manufacturers have your idea of a 12-V fail-safe? The last car I checked with child-locks (10-12 years ago) it was a physical switch inside the door mechanism, so it would have prevented opening the door in a crash as well.

Another point, if you have child-safe locks engaged on most cars, in the event of an emergency you can open the rear doors by unlocking them, lowering/breaking the window, and pulling the door handle from the outside.

Thus an adult, and not likely a child, can do this as it requires a longer arm and some strength.

However.. you can see where I'm going, a 12V failure on the Tesla means the door cannot be manually unlocked, nor can you lower the window without power, you can't break through the laminated glass, nor can you open the door manually from the outside anyways. Even if someone wanted to get you out they can't.

Sure other cars have the inability to lower the window without power, but Tesla's are more unique in that you often can't smash the rear laminated windows and there is no external manual door handle.

I mean they reality have cheaped out by removing safety features that most other cars have always had.

 

[JHCCAZ]

It was sincere. However, as you noted, other cars' manual releases won't work with child-locks enabled. I wonder if the other manufacturers have your idea of a 12-V fail-safe? The last car I checked with child-locks (10-12 years ago) it was a physical switch inside the door mechanism, so it would have prevented opening the door in a crash as well.

I don't think it's obvious what to do about ensuring an easy method to open the doors or windows in a true emergency, vs the long-established need to prevent children from operating the doors. Ideally, we want the easiest possible way to operate the exits - from inside or from outside - after a crash or in a medical emergency. Certainly not some tricky secret cable-pull under the seat, or anything else that requires education, a cool head and dexterity. we also want to stay protected from intrusion by hostile outsiders in unsafe situations, and we still want to protect children from themselves.

I think there's a good argument for releasing the locks if one of the crash sensors is activated and/or the 12 volt system fails. However, such a feature is certainly not infallible. It's not hard to imagine various theft or carjacking scenarios in which the criminal could activate the emergency release mechanism by bumping the car hard enough, or by learning the ways that the electrical system can be tampered with from the outside. These kinds of clever defeats become tools of the trade, scumbag lore.

In trying to think it through, the best end simplest idea I can come up with is to couple the reliable and simple mechanical front door emergency handle, to also enable mechanical opening of the back doors. I'm thinking that operating the emergency front lever would also spring open the child latch via the intervening door jamb pillar (you don't want a complex cable routing through the front door hinge and then to the back). So there would be an easy mechanical emergency release in back, but if child-locked it would release upon activation of the front emergency lever. And it doesn't require actual opening of the front door; it would work even if the front door is jammed shut - but of course there are no guarantees if the car body is totally smashed and deformed.

This is just a thought, but I think the car should be equipped with something along these lines. And it still doesn't address the problem of Good Samaritan access from the outside while protecting from hostile criminals. That requires further thought and may not have any logical solution.

Bottom line, I do agree that the lack of a mechanical release, or a tricky hidden release, is not the right solution for the rear doors and hatch. I'm also not saying any of this was critical in the Texas incident, but the uncertain possibility highlighted the need for better emergency-exit features.

 

[kabin]

Those are essentially useless with newer Tesla all-laminated windows.

That's good to know. I guess we found another good reason to carry a personal protection device.

 

[Dan D.]

I don't think it's obvious what to do about ensuring an easy method to open the doors or windows in a true emergency, vs the long-established need to prevent children from operating the doors. Ideally, we want the easiest possible way to operate the exits - from inside or from outside - after a crash or in a medical emergency. Certainly not some tricky secret cable-pull under the seat, or anything else that requires education, a cool head and dexterity. we also want to stay protected from intrusion by hostile outsiders in unsafe situations, and we still want to protect children from themselves.

I think there's a good argument for releasing the locks if one of the crash sensors is activated and/or the 12 volt system fails. However, such a feature is certainly not infallible. It's not hard to imagine various theft or carjacking scenarios in which the criminal could activate the emergency release mechanism by bumping the car hard enough, or by learning the ways that the electrical system can be tampered with from the outside. These kinds of clever defeats become tools of the trade, scumbag lore.

In trying to think it through, the best end simplest idea I can come up with is to couple the reliable and simple mechanical front door emergency handle, to also enable mechanical opening of the back doors. I'm thinking that operating the emergency front lever would also spring open the child latch via the intervening door jamb pillar ( you don't want a complex cable routing through the front door hinge and then to the back). So there would be an easy mechanical emergency release in back, but if child-locked it would release upon activation of the front emergency lever. And it doesn't require actual opening of the front door; it would work even if the front door is jammed shut - but of course there are no guarantees if the car body is totally smashed and deformed.

This is just a thought, but I think the car should be equipped with something along these lines. And it still doesn't address the problem of Good Samaritan access from the outside while protecting from hostile criminals. That requires further thought and may not have any logical solution.

Bottom line, I do agree that the lack of a mechanical release, or a tricky hidden release, it's not the right solution for the rear doors and hatch.

Oh my, what a world.

So we have to prevent hostile outsiders from wrenching open the door, or reaching in to unlock the door and then opening it.
Also we have to allow emergency Good Samaritan access to do exactly the same thing. Somehow the car has to be able to tell the difference, assuming we are incapacitated - crash or no crash.

We have to consider electronic solutions and manual no-12V solutions.
Solutions where the front is unoccupied, front is inaccessible/on fire, side is damaged, all windows are unsmashable.
Children/incapacitated people inside who cannot assist with opening the doors.
Doors are stuck, doors are locked, car is on a steep slope, gull-wing doors are too heavy to lift manually.
Pointlessly weak door handles that break when pulled hard.

Now this is not all exclusive to Tesla, some other cars have dangerous systems too. But Tesla is the topic here.

If all the features prevent a person inside from getting out, then you need external help. In theory after a 12V failure some Tesla models/years give you no obvious options from the rear, if say the front is inaccessible/destroyed/on fire. The doors won't unlock if auto-unlock didn't function, the doors won't open anyway without power, the windows won't move, the windows can't be broken, the manual door release is hidden. You can't reach out to open the door from the outside because the external handle doesn't work either.

There are also some features that don't allow you to use them even if you have 12V power. Child-lock doors, child-lock windows, auto-lock doors, retractable handles. These may have crash-overrides, but they may not have worked, or you may not have crashed, it could be a fire.

The initial Good Samaritan thing to do after an accident is to pull the external handle, or then gesture to unlock it, or then reach in to unlock the door, or then break the window to unlock the door, and then try to open the interior handle to open the door. If the car has NONE of those features that actually work... I mean, if the 12V is broken and everyone is trying but failing to get someone out of a car, I think the design sucks.

What's my solution? Well, I'd like to see the problem actually posed first. I don't see any agency really asking.

At the end of the day, I say - learn how to go out the trunk.

 

[sleepydoc]

Wait - you mean that all the hysteria in the press and people jumping to conclusions without any data were wrong?????

This is a great example of how the initial stories are wrong, yet they are what people remember. Unfortunately, the NTSB took almost 2 years to issue the final report, so from Tesla’s perspective, the damage is done.

 

[Tam]

... the damage is done...

So how dismantling the public relations would help to contain the damage early rather than waiting for the free report department NTSB 2 years later?