|
Getting your Trinity Audio player ready...
|
The driver who died in a fiery crash on Highway 101 last week had his Tesla’s Autopilot engaged before the crash, according to a statement released by the company Friday night.
The National Transportation Safety Board, which is investigating the crash and its aftermath, expressed its displeasure with Tesla’s surprise announcement.
San Mateo resident Wei Huang, 38, crashed into a highway barrier separating southbound Highway 101 from the Highway 85 carpool flyover lane shortly before 9:30 a.m. on Friday, March 23. The Tesla struck the median at freeway speeds, triggering a three-vehicle accident and causing the car to catch fire. Huang was transported to Stanford Hospital with major injuries, where he later died.
Tesla retrieved the vehicle logs from Huang’s vehicle, a Model X, and determined that the Autopilot system was engaged at 9:27 a.m., moments before the collision with the “adaptive cruise control follow-distance set to a minimum,” according to a blog post from the company.
As Huang approached the barrier, he had received “several visual” warnings and an audible warning to take control of the vehicle again, according to the post. But data from the vehicle shows that 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,” according to the statement.
ABC7 News reported that Huang’s family said Huang had frequent trouble with the Autopilot system on the Model X, and took it took his dealer on multiple occasions claiming that the Autopilot veered towards the same Highway 101 barrier that his vehicle collided with on March 23.
A Tesla spokesperson disputed that claim, telling the Voice in an email on April 4, “We’ve been doing a thorough search of our service records and we cannot find anything suggesting that the customer ever complained to Tesla about the performance of Autopilot. There was a concern raised once about navigation not working correctly, but Autopilot’s performance is unrelated to navigation.”
Tesla officials argued in the blog post that while the company’s Autopilot system doesn’t prevent all accidents, it has maintained a strong safety track record since its rollout over a year ago. The blog post claims that there is only one fatality for every 320 million miles traveled by vehicles with the function — significantly lower than the average in the U.S. of one death per 86 million miles traveled.
“Tesla Autopilot does not prevent all accidents –- such a standard would be impossible — but it makes them much less likely to occur,” according to the blog post. “It unequivocally makes the world safer for the vehicle occupants, pedestrians and cyclists.”
The National Transportation Safety Board (NTSB) announced on Tuesday, March 27, that it was stepping in to investigate the crash as well as the subsequent emergency response.
Battery fires in electric vehicles can reach temperatures of 900 degrees and require thousands of gallons of water to extinguish. Tesla engineers came out to the scene of the crash to assist the Mountain View Fire Department in getting the battery temperature under control.
Telsa apparently posted Friday night without warning the NTSB, which is “unhappy with the release of investigative information by Tesla” after the company confirmed that the Model X involved in a deadly crash in Mountain View last month did in fact have automated driving activated, a spokesman said.
NTSB spokesman Chris O’Neil said that while Tesla has been “extremely cooperative” in assisting with the vehicle data aspect of investigations involving their vehicles in the past, the NTSB expects parties involved in their investigations to inform them of releases before making information public.
In an initial blog post about the accident on March 27, Tesla officials stated that the attenuator barrier, a buffer designed to cushion a collision with the cement median of the Highway 85 flyover, had either been “removed or crushed” without an adequate replacement, which added to the severity of the crash.
“We have never seen this level of damage to a Model X in any other crash,” according to the initial post.
Bay City News Service contributed to this report.
Sea lion found on Mountain View street returns to ocean after months of recoveryMarch 28, 2026 9:12 am
Shame on Tesla for calling their cruise-control feature “autopilot”, which is clearly overstating the capabilities of the car. How many people have to die before they back off that claim? Hopefully no pedestrians have been killed yet.
First of all, let me apologize in advance for the length of this post.
I’ll start with a link to a car that seems to experience a very similar phenomenon to the one that crashed in Mountain View killing its driver. https://youtu.be/6QCF8tVqM3I
I read the press release from Tesla, and I was honestly quite appalled at how cynical and manipulative it was, and how it has impacted the coverage of the story. To quote the most salient portion from the press release:
’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.’
These several sentences are PR and legal genius. By stitching together several disjoint events, Tesla describes a narrative that cynically lays the blame for the accident at the drivers feet; however, if we read the above sentences in isolation we can deconstruct this narrative:
’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’
All this says is that the driver had Autopilot engaged with a minimum follow distance prior to the crash. It doesn’t say anything more or anything less. The last point about the adaptive cruise control being set to minimum follow distance has no bearing on the crash as no manufacturers adaptive cruise control will detect a stationary object, as proven by a Tesla rear ending a stationary fire engine recently
‘The driver had received several visual and one audible hands-on warning earlier in the drive…’
Quite a cleverly written sentence fragment. It tells us that at one or more points during the drive the driver had received several visual and one audible hands-on warning. When exactly during the drive? Did they happen all at once? Were there multiple warning over a period of time? What was the status immediately prior to the crash?
‘…and the driver’s hands were not detected on the wheel for six seconds prior to the collision.’
This is very cunning phraseology. At first blush it reads as if the driver didn’t have their hands on the wheel for the six seconds prior to the collision; however, it could just as equally be read that at some point prior to the collision the driver didn’t have his hand on the wheel for a total of six seconds. Again those six seconds could have been concurrent or disjoint. I would suggest that if Tesla wanted to be specific they would have said: ‘…and the driver’s hands were not detected on the wheel for six seconds IMMEDIATELY prior to the collision.’. They didn’t. Was this an omission or was it intentional?
‘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.’
This is essentially a non-statement. Basically it says that for 150 meters before the crash the driver had an unobstructed view of the concrete divider. It makes no mention of what was happening during those five seconds. The driver could have just as well been happily driving along in the far left HOV lane with autopilot engaged and at the last second the vehicle could have veered into the median, in which case he would have no time to take action.
I know this particular stretch of road, and I would be surprised if the driver wasn’t in the far left HOV lane that becomes a flyover onto 85S and heads towards Cupertino and the Apple campus. Given that this was likely his commute, he would not be likely to swerve over to continue on 101S. I would also suggest that since the driver had had prior issues with Autopilot he would likely be quite attentive at this point in his drive.
I would further suggest that a question to ask is if this vehicle has the latest version of Tesla’s autopilot installed. This version involved an extensive rewrite of the autopilot code. Could this have been a factor in the accident?
I think there are a lot of unanswered questions here. I don’t think that Tesla should be allowed to simply put out an ambiguous press release that effectively blames the driver for his death with hard questions being asked of them.
It was a one car collision. If the driver had been paying attention, how did this happen? How fast was he going? If he had taken the car to the dealer because it was drifting to the side why didn’t he correct? Was he texting, or sleeping? If your car is drifting, you correct the drift. Why didn’t he? I think the phone records should be checked to see if he was texting or talking on his phone at the time of this accidents. We don’t want to blame the victim generally, but in a one car accident, what else is at fault here? Drivers who speed, text, talk on their phones, eat, drink, put on make-up, etc., endanger us all. We see them everyday and they need to be stopped.
Tesla is talking out of both sides of their mouth regarding their “autopilot”. Either it’s capable of safely driving the car or it’s not.
If the driver has to be paying full attention and ready to take over in a split second, it’s not really an “autopilot”.
While I feel for his family, I cannot see how anything the owner did was reasonable. He repeatedly chose to risk his life in a manner he knew was highly dangerous and he knew could have easily been avoided and eventually, the predictable outcome happened.
Since this spot was on his commute, I understand why he would pass this spot daily, but not why he would repeatedly risk his life and the lives of the people in other cars by driving through this dangerous spot in the manner he knew was so dangerous?
Why didn’t he simply use a different lane at that spot?
Why didn’t he simply keep his hands on the steering wheel at that spot?
Why didn’t he simply not have Autopilot on while he approached that spot?
It’s like sticking a metal fork in an electric toaster to pull out your stuck bagel and seeing an arc of electricity the first time. Then instead of learning that this is a bad idea and never doing that again, you go ask your wife to do the same thing. And then you just keep on using a metal fork to get your stuck bagel out until eventually you get electrocuted to death and then the family sues the makers of the toaster and the bagel.
If your bagel gets stuck and you really have no other tool that a metal object to remove the bagel, then at least unplug the toaster first!
Just like it says in the manual that comes with all toasters!
Or am I being to reasonable?
@Mike Laursen
“Tesla is talking out of both sides of their mouth regarding their “autopilot”.”
Tesla named it “Autopilot” for a reason, the problem is that the general public does not even understand what an autopilot in an airplane actually can and cannot do. Blame Hollywood movies for this, and many other examples, of how the general public ended up being seriously misinformed.
People with an agenda, or out of ignorance are miss-quoting Tesla and miss-characterizing what Tesla has always said about it’s “Autopilot Driver Assist” system can and cannot do and how it must be used. It’s not a driver-less car system, it’s a safety improvement system, when used as directed.
“Either it’s capable of safely driving the car or it’s not.”
Tesla never claimed the “Autopilot” was capable of “safely driving the car” by itself.
That’s not what “Autopilot” means for airplanes nor for Tesla.
Aircraft pilots only use their autopilot systems to hold a steady course in situations where nothing is changing, no traffic on their flight path, and they don’t just abandon the cockpit and trust that nothing will change.
Proper use of airplane autopilots is a safety feature that allows pilots to deal with other problems while holding a steady course is deemed safe for short periods while the pilot is dealing with another priority issue.
Even the most advanced airplane autopilots are not to be trusted to deal with unexpected situations. That’s the job of the pilot! That’s also why in most airplanes you need to have a pilot and co-pilot and one of them is always supposed to be keeping aware of things and ready to take back control from the autopilot.
“If the driver has to be paying full attention and ready to take over in a split second, it’s not really an “autopilot”.”
Actually, it is exactly like an airplane “autopilot”.
Correction, I don’t think an airplane autopilot can slam on the brakes to prevent a collision, so I guess the Tesla autopilot can do at least one thing an airplane autopilot cannot do.