Blog NTSB “Unhappy” With Tesla’s Comment on Fatal Model X Crash

The National Transportation Safety Board (NTSB) said it is “unhappy” that Tesla released information about a March 23rd crash that killed a driver, the Washington Post reported Sunday.

The driver was on a busy stretch of Highway 101 when his Model X drifted out of its lane and crashed into a concrete rail. Tesla published a blog post last week on the crash, saying that Autopilot was in use when the accident occurred.

But, it seems the NTSB didn’t appreciate Tesla commenting on the specifics of the accident during an active investigation.

“In each of our investigations involving a Tesla vehicle, Tesla has been extremely cooperative on assisting with the vehicle data,” an agency spokesman told the Washington Post. “However, the NTSB is unhappy with the release of investigative information by Tesla.”

Tesla’s blog post didn’t offer any explanation as to why the car drifted out of its lane, but suggested that the driver ignored warnings to grab the wheel. “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 company said in the post.

The NTSB said not to expect further comment until the completion of a preliminary report, which generally occurs within a few weeks of completion of field work, the Washington Post reported.

Update (2:50 p.m. ET): In response to a tweet from TMC linking to this story, Tesla Chief Executive Elon Musk defended the automaker’s decision to release information about the crash.

“Lot of respect for NTSB, but NHTSA regulates cars, not NTSB, which is an advisory body,” Musk tweeted. “Tesla releases critical crash data affecting public safety immediately & always will. To do otherwise would be unsafe.”

https://twitter.com/x/status/980876926830039041

“NHTSA” refers to the National Highway Traffic Safety Administration, a division of the U.S. Department of Transportation with a stated mission to “save lives, prevent injuries and reduce economic costs due to road traffic crashes, through education, research, safety standards and enforcement activity.”

Photo: NBC Bay Area

 

[jeewee3000]

Often, with hands on the steering wheel while in Autopilot mode, I get warnings to put my hands on the wheel. The steering wheel sensors require a pretty heavy hand. The unfortunate driver of the car who died in this accident may very well have had his hands on the steering wheel, even though the electronic record states that he did not.

Adding to sundaymorning's excellent point: the infamous 'orange trick' is all the proof you need to realize the steering wheel sensors don't require a very heavy touch.

(Disclaimer: using the orange trick is very irresponsible, dangerous and stupid.)

 

[Ulmo]

NTSB should shut up and serve us, the people.

 

[Ulmo]

On Monday, ABC Channel 7 reported that the victim's brother stated:

"...7-10 times the car would swivel toward that same exact barrier during auto-pilot. Walter took it into dealership addressing the issue, but they couldn't duplicate it there."

Not to speak ill of the dead, but if that is true, why did the driver not have his hands on the steering wheel during the fatal approach?

Possibly auto pilot forced the car over there. Back when I had my AP2.0 Model S and didn't even have Auto Pilot software (I didn't purchase it), the older programming would sometimes force my car over into barriers, such as driving up Highway 152 over the hill passes while passing trucks. It usually shoved me left into the barrier while I was driving past a truck. I ping-ponged back and forth between the barrier and the truck, fighting the car autopilot emergency steering, never hitting anything, but so far, I've been very good at emergency steering.

I reported it as bugs every time, and it seemed to happen less often with time, but it still happened. If I wasn't super-alert about it, I would have crashed at freeway speeds into barriers, but thankfully, most of them were not head-on collisions with sword tip concrete walls as in the fatal Model X incident; I would have scraped at an angle on the wall, possibly causing the tail end to wedge against the truck, which would have caused the tail end to slow down due to the slower truck speed, which would have dragged the tail end backwards, probably straightening out the car but then shoving the front into the truck, which would have grabbed the hood and twisted the car at a 90º angle underneath the truck's trailer, and killed me. Not good.

AP2 does some damned things at highway speeds on hills. Maybe the latest version fixed that; I wouldn't know; I already sold my Model S back in July 2017.

 

[Ulmo]

Often, with hands on the steering wheel while in Autopilot mode, I get warnings to put my hands on the wheel. The steering wheel sensors require a pretty heavy hand. The unfortunate driver of the car who died in this accident may very well have had his hands on the steering wheel, even though the electronic record states that he did not.

When a Tesla AP2 car is in a self-perceived "emergency" maneuver, would it register a driver's hands on wheel as a "hands on wheel" event, even though the car is busy vibrating the wheel to notify you of the emergency? When I experienced that in my AP2 Model S, the steering wheel's vibration was pretty intense as I attempted to override its treacherous murderous emergency steering. My input into the steering ended up being mere suggestions to the emergency algorithms, that seemed to want to do their own things, always trying to shove me into concrete barriers or trucks at highway speeds. Luckily, my "suggestions" were enough to stop crashes from happening, because I was able to put about 75 pounds torque pressure into the steering wheel with my arms. I'm pretty alert and medium strong (quite enough to lift 100#), though, and not at all trusting of technology that is not thoroughly worked out; a typical Apple employee might be quite the opposite, mentally and physically.

 

[McRat]

...
Ford Pinto > ENGINEERING.com

"During design and production, however, crash tests revealed a serious defect in the gas tank. In crashes over 25 miles per hour, the gas tank always ruptured. To correct it would have required changing and strengthening the design."
...

That is the worst written article on the subject I've read to date. The Pinto had to compete against the VW Beetle to fight off Japanese imports? What? The VW Beetle was made in Germany, and the Japanese cars were made... uh... not in Germany?

From Wiki: " A subsequent analysis of the overall safety of the vehicle indicated its safety was comparable to other 1970s subcompact cars. The Ford Pinto has been cited widely as a business ethics as well as tort reform case study."

The NHTSA found 27 design related fire deaths in 2,200,000 cars. Plaintiff lawyers found 280. The discrepancy was due to the fire deaths unrelated to the fuel tank design and non-survivable crashes.

This urban myth gets worse as the decades go on. When you remove the plaintiff's 'expert witnesses' and legal team funded media marketing, and look at the real numbers and facts, you find that the Pinto actually had and average number of fire deaths. The myth that a self-sealing tank is a $5 bump is a joke also. They are rare except in things like military/aviation equipment that is susceptible to hot projectile punctures and race cars.

The Pinto was a cheap POS that saved gas that low income or ecology-minded people could afford. In the US, there is no such car for the masses anymore. We have outlawed them both by over-the-top safety mandates and lack of tort reform efforts. There are $6000 cars made. You just don't get to make the decision to buy one. You are not responsible enough to decide such things.

 

[Peter Lucas]

He was likely not paying attention and hands were elsewhere.

The sensors that detect hands on the steering wheel are not very sensitive. Your hands have to apply a significant torque to the wheel before the sensors acknowledge that your hands are indeed on the wheel.
So, while the driver was certainly not paying attention in the last few seconds, he may have had his hands on the steering wheel.

 

[Peter Lucas]

Is this true? That under some circumstances the newer Tesla's automatic steering is difficult to overpower? If so, that is a serious problem. My early 2016 Model S certainly does NOT have this problem.

When a Tesla AP2 car is in a self-perceived "emergency" maneuver, would it register a driver's hands on wheel as a "hands on wheel" event, even though the car is busy vibrating the wheel to notify you of the emergency? When I experienced that in my AP2 Model S, the steering wheel's vibration was pretty intense as I attempted to override its treacherous murderous emergency steering. My input into the steering ended up being mere suggestions to the emergency algorithms, that seemed to want to do their own things, always trying to shove me into concrete barriers or trucks at highway speeds. Luckily, my "suggestions" were enough to stop crashes from happening, because I was able to put about 75 pounds torque pressure into the steering wheel with my arms. I'm pretty alert and medium strong (quite enough to lift 100#), though, and not at all trusting of technology that is not thoroughly worked out; a typical Apple employee might be quite the opposite, mentally and physically.

 

[sundaymorning]

The sensors that detect hands on the steering wheel are not very sensitive. Your hands have to apply a significant torque to the wheel before the sensors acknowledge that your hands are indeed on the wheel.
So, while the driver was certainly not paying attention in the last few seconds, he may have had his hands on the steering wheel.

I don’t buy this. He’s got 5 full seconds to react, for driving that is a lifetime. When you see your vehicle approaching a stationary object, it’s a no brainer what you should do. In the end, it’s driver negligence.

The AP’s steering wheel IS sensitive, but it won’t make sense for it to be ultra reactive to the tiniest of touch, if this were to occur then a inadvertent touch of your hand would mean automatic disengagement (likely more dangerous). For this reason, the car will only react to disengagement if it senses the driver is consciously trying to retake control, hence, some pressure is necessary. As AP becomes more advance, driver negligence and dependence will also increase IMO.

 

[Tam]

Often, with hands on the steering wheel while in Autopilot mode, I get warnings to put my hands on the wheel.

That was my experience when I first drove with Autopilot for the first time. However, I quickly learned that it wants a constant and slight torque on the steering wheel which I have been providing for more than 1 year ever since I bought it and I have never got any more warning alerts.

For example, I drove more than 200 miles on freeway and I got no "keep hands on" warning. I only had to correct the system twice very easily and very seamlessly for that freeway trip (1--Autolane Change aborted for missing markers, 2--Autolane Change aborted due to honking.)

It's so seamless because at least one of my hands was on the steering wheel and I could feel when it was about to make an error that needed my manual correction.

 

[bonnie]

That was my experience when I first drove with Autopilot for the first time. However, I quickly learned that it wants a constant and slight torque on the steering wheel which I have been provide for more than 1 year ever since I bought it and I have never got any more warning alerts.

For example, I drove more than 200 miles on freeway and I got no "keep hands on" warning. I only had to correct the system twice very easily and very seamlessly for that freeway trip (1--Autolane Change aborted for missing markers, 2--Autolane Change aborted due to honking.)

It's so seamless because at least one of my hands was on the steering wheel and I could feel when it was about to make an error that needed my manual correction.

This. I rarely get the warning, because I always have at least one hand (usually both) on the steering wheel & I can immediately feel/correct any movement that needs correction. The last two trips to Portland (60+ miles each way), I've had my X in AP the entire time until exiting the freeway, zero nags. I am far more likely to inadvertently disengage AP than I am to get a nag.

 

[Brando]

My mother owned a 1964 Corvair Monza. The car I learned to drive with a 4-speed.
Wonderful in the snow of the Rocky Mountains. Great fun to drive. over 26-29 mpg.
Early rear engine car designs (actually most independent rear suspension including Corvette)- Beetle, Porsche, Corvair did not have a stablizer bar between the rear wheels. Not good. Corvair in the 4th year of productions added bar in 1964 (Porsche and Beetles did the same, not sure of the years). In many ways the 1960-1963 Corvair was the unwitting start of people trying to get the auto manufacturers to take auto design safety seriously. Below is a good quick intro to the story.

How the Corvair’s rise and fall changed America forever

 

[Brando]

That is the worst written article on the subject I've read to date. The Pinto had to compete against the VW Beetle to fight off Japanese imports? What? The VW Beetle was made in Germany, and the Japanese cars were made... uh... not in Germany?

From Wiki: " A subsequent analysis of the overall safety of the vehicle indicated its safety was comparable to other 1970s subcompact cars. The Ford Pinto has been cited widely as a business ethics as well as tort reform case study."

The NHTSA found 27 design related fire deaths in 2,200,000 cars. Plaintiff lawyers found 280. The discrepancy was due to the fire deaths unrelated to the fuel tank design and non-survivable crashes.

This urban myth gets worse as the decades go on. When you remove the plaintiff's 'expert witnesses' and legal team funded media marketing, and look at the real numbers and facts, you find that the Pinto actually had and average number of fire deaths. The myth that a self-sealing tank is a $5 bump is a joke also. They are rare except in things like military/aviation equipment that is susceptible to hot projectile punctures and race cars.

The Pinto was a cheap POS that saved gas that low income or ecology-minded people could afford. In the US, there is no such car for the masses anymore. We have outlawed them both by over-the-top safety mandates and lack of tort reform efforts. There are $6000 cars made. You just don't get to make the decision to buy one. You are not responsible enough to decide such things.

 

[Brando]

That is the worst written article on the subject I've read to date. The Pinto had to compete against the VW Beetle to fight off Japanese imports? What? The VW Beetle was made in Germany, and the Japanese cars were made... uh... not in Germany?

So since other cars aren't safe FORD doesn't need to design safe cars??

This urban myth gets worse as the decades go on. When you remove the plaintiff's 'expert witnesses' and legal team funded media marketing, and look at the real numbers and facts, you find that the Pinto actually had and average number of fire deaths. The myth that a self-sealing tank is a $5 bump is a joke also. They are rare except in things like military/aviation equipment that is susceptible to hot projectile punctures and race cars.

Ford Pinto > ENGINEERING.com

"The other side of the equation, the alleged $11 cost of a fire-prevention device, is also a misleading estimation. One document that was not sent to Washington by Ford was a "Confidential" cost analysis Mother Jones has managed to obtain, showing that crash fires could be largely prevented for considerably less than $11 a car. The cheapest method involves placing a heavy rubber bladder inside the gas tank to keep the fuel from spilling if the tank ruptures. Goodyear had developed the bladder and had demonstrated it to the automotive industry. Ford Motor Company ran a rear-end crash test on a car with the rubber bladder in the gas tank. The tank ruptured, but no fuel leaked. On January 15, 1971, Ford again tested the bladder and again it worked. The total purchase and installation cost of the bladder would have been $5.08 per car. That $5.08 could have saved the lives several hundred people."

I encourage people to read the article for themselves. Why McRat finds the above paragraph a myth, I'm not sure. I won't characterize that paragraph as a myth.

Anyway, I find US auto makers approach to safety often "lacking". Anyway, IF I considered a GM, FORD, FCA car I'd look at safety info very closely.

 

[McRat]

So since other cars aren't safe FORD doesn't need to design safe cars??




Ford Pinto > ENGINEERING.com

"The other side of the equation, the alleged $11 cost of a fire-prevention device, is also a misleading estimation. One document that was not sent to Washington by Ford was a "Confidential" cost analysis Mother Jones has managed to obtain, showing that crash fires could be largely prevented for considerably less than $11 a car. The cheapest method involves placing a heavy rubber bladder inside the gas tank to keep the fuel from spilling if the tank ruptures. Goodyear had developed the bladder and had demonstrated it to the automotive industry. Ford Motor Company ran a rear-end crash test on a car with the rubber bladder in the gas tank. The tank ruptured, but no fuel leaked. On January 15, 1971, Ford again tested the bladder and again it worked. The total purchase and installation cost of the bladder would have been $5.08 per car. That $5.08 could have saved the lives several hundred people."

I encourage people to read the article for themselves. Why McRat finds the above paragraph a myth, I'm not sure. I won't characterize that paragraph as a myth.

Anyway, I find US auto makers approach to safety often "lacking". Anyway, IF I considered a GM, FORD, FCA car I'd look at safety info very closely.

You are reading information provided by (probably payola too) the class action legal team. The Mother Jones source is The Center for Auto Safety, which is comprised of mostly lawyers who sue auto companies. Ralph Nader was paid by lawyers who were members of the CAS. Today, they are mainly a legal referral service skirting the Non-Profit Tax Exemption laws. Board of Directors - Center for Autosafety It seem they were questioned in 2009 if they were indeed a non-profit. They won.

Anyhow, it's not messenger, it's the message that is screwy.

A bladder style fuel tank is AKA a self-sealing tank. Go price them. Bull on that $5 aircraft tank. Complete and utter nonsense or all cars would have them, not just racecars, planes, and military vehicles. $5 for a $1000 tank? Shucks gimme 10!

You don't need to use Confidential sources unless you have something to hide. Mother Jones should have said "Our source is a Ford engineer with information on the project who we will not name, who stated the cost of the repair was actually $5" if that was a true story, but instead, they sorta say:

"Our sources want to be confidential, because if we say it's the plaintiff's attorney telling us a story with nothing to back it up, you'll think we are being paid to write this article. We are, but we don't you idiots to know that. Thanks!"

You want proof? Ford had to fix all the Pintos. Did they use the cheap $5 racing/aviation self-sealing gas tank fix. Are you insane???? They used:

A different filler neck.
A hose clamp.
2 pieces of plastic armor.

Did Goodyear offer auto companies self-sealing fuel tank liners for $5? If they did, why didn't even race car teams buy them? No manufacturer did. Very odd. Somebody hands you GOBS of money, and you turn it down. Millions and millions. As far as I know, Goodyear only made rubber liners for aircraft under a Uniroyal patent. If they have made them for production passenger cars, they don't want anybody to know.

It doesn't pass the smell test. Goodyear did not sell any of these to anyone, the source is confidential even though all the docs from Ford were accessible by the plaintiff's team. The source is a legal referral business.

 

[Brando]

You are reading information provided by (probably payola too) the class action legal team. The Mother Jones source is The Center for Auto Safety, which is comprised of mostly lawyers who sue auto companies. Ralph Nader was paid by lawyers who were members of the CAS. Today, they are mainly a legal referral service skirting the Non-Profit Tax Exemption laws. Board of Directors - Center for Autosafety It seem they were questioned in 2009 if they were indeed a non-profit. They won.

Anyhow, it's not messenger, it's the message that is screwy.

A bladder style fuel tank is AKA a self-sealing tank. Go price them. Bull on that $5 aircraft tank. Complete and utter nonsense or all cars would have them, not just racecars, planes, and military vehicles. $5 for a $1000 tank? Shucks gimme 10!

You don't need to use Confidential sources unless you have something to hide. Mother Jones should have said "Our source is a Ford engineer with information on the project who we will not name, who stated the cost of the repair was actually $5" if that was a true story, but instead, they sorta say:

"Our sources want to be confidential, because if we say it's the plaintiff's attorney telling us a story with nothing to back it up, you'll think we are being paid to write this article. We are, but we don't you idiots to know that. Thanks!"

You want proof? Ford had to fix all the Pintos. Did they use the cheap $5 racing/aviation self-sealing gas tank fix. Are you insane???? They used:

A different filler neck.
A hose clamp.
2 pieces of plastic armor.

Did Goodyear offer auto companies self-sealing fuel tank liners for $5? If they did, why didn't even race car teams buy them? No manufacturer did. Very odd. Somebody hands you GOBS of money, and you turn it down. Millions and millions. As far as I know, Goodyear only made rubber liners for aircraft under a Uniroyal patent. If they have made them for production passenger cars, they don't want anybody to know.

It doesn't pass the smell test. Goodyear did not sell any of these to anyone, the source is confidential even though all the docs from Ford were accessible by the plaintiff's team. The source is a legal referral business.

SORRY - MY BAD
Your careful reading caught the fact that the engineering.com article was apparently written using the Mother Jones article only.
I assumed [you know what they say about doing that] that some site called ENGINEERING.COM would use an engineers approach and not just use a news magazine article. AND you pointed out the "law suit witness" problems as well.

THANK YOU @McRat for being persistent enough to clear things up.

side note: Though I think Ralph Nader helped get seat belts, air bags and safety into cars, I also learned to drive in my Mother's 1964 Corvair Monza and find Nader's attack on that car out of line. Sadly, GM went for the messenger (Nader) rather than addressing the Corvair design improvements 1964 and the 1965 new model. I lusted after Yenko Stinger (see Jay Leno Garage). Also search Electrovair I, II, III and the concept Concept Cars of the Week: Corvair Sports Cars - Car Design News
and the Testudo worth a search. I must agree with Elon, I hate concept cars that never get produced.

Some how Chevy managed to produce about 1.8 million Corvairs and I hope Tesla makes at least that many of each of their models.

 

[McRat]

... I also learned to drive in my Mother's 1964 Corvair Monza and find Nader's attack on that car out of line. Sadly, GM went for the messenger (Nader) rather than addressing the Corvair design improvements 1964 and the 1965 new model. ...

The Corvair / Nader situation also was driven (har) by litigation from the CAS members. The Corvair was no worse than Porsches or other rear engine performance cars. Cars that are tail heavy are a bit tail happy when driven aggressively. GM improved the suspension, but there was nothing inherently dangerous about the car. If the car was judged fairly against other cars of it's time, it's one more victim of negative publicity for profit. Some Porsches today still have that same quirk if you turn off the stability control. They will oversteer easier than a car with slight front bias when trail braking.

If the Corvair was given a fair shake it would have been remembered as a great sports car, but instead it will be remembered as another Killer Car of Doom.

Some folk love to throw darts at auto industry 'disasters' without knowing many are exaggerated heavily for profit.

Trivia - The same trait Nader claimed made the Corvair unsafe at any speed, is the trait that caused the Carrera GT to come around on Rodas and killed him and Paul Walker. A judge ruled the car was not to blame. Imagine that. My theory on what caused Rodas to lose control was the 8 year old tires on the car. Tires lose their traction over time and develop a glazed surface. Rodas was a experienced racer, but he probably only realized the tires were glazed when it was too late. I nearly did the same thing once in a Vette. 5 years max on performance cars, then get new rubber.

 

[malcolm]

The NTSB understands autopiloting in the same way that the Senate understands the Internet.

Tesla and NTSB tried and failed to hash the issue out over Autopilot fatal crash investigation

 

[AceSkywalker]

The NTSB understands autopiloting in the same way that the Senate understands the Internet.

Tesla and NTSB tried and failed to hash the issue out over Autopilot fatal crash investigation

Maybe the NTSB is scared of becoming irrelevant when all transportation, from land vehicles to ships and planes becomes 100% automated.

 

[malcolm]

Maybe the NTSB is scared of becoming irrelevant when all transportation, from land vehicles to ships and planes becomes 100% automated.

Agreed. The NTSB has evolved alongside automakers that only revise their products on a yearly basis, if that.

In that context, an investigation which takes 12 months is a reasonable match.

But as cars become more and more like computers, this is simply too slow.

 

[chillaban]

Agreed. The NTSB has evolved alongside automakers that only revise their products on a yearly basis, if that.

In that context, an investigation which takes 12 months is a reasonable match.

But as cars become more and more like computers, this is simply too slow.

And the NTSB and to some extent even the NHTSA is not well equipped to Tesla's unique model of being able to react literally overnight in the form of OTA updates. If they identify anything to improve, they can roll it out to their fleets before government agencies can even manage to schedule a press conference.

We are already seeing evidence suggesting 2018.12 is playing around more with radar braking.