Tesla Data Leak Shows Complaints of Unintended Braking and Acceleration

[Dewg]

Tesla leak reportedly shows thousands of Full Self-Driving safety complaints

The leak contains 100GB of confidential data from Tesla.

www.theverge.com

Oddly there were more reports of phantom acceleration than phantom braking, which took me by surprise.

The complaints, which were reported across the US, Europe, and Asia, span from 2015 to March 2022. During this period, Handelsblatt says Tesla customers reported over 2,400 self-acceleration issues and 1,500 braking problems, including 139 reports of “unintentional emergency braking” and 383 reports of “phantom stops” from false collision warnings.

 

[Twiglett]

139 unintentional emergency stops in seven years? Those are really low numbers for something that's supposed to be widespread.

 

[EVNow]

139 unintentional emergency stops in seven years? Those are really low numbers for something that's supposed to be widespread.

I think all these can be clubbed under "phantom braking".

1,500 braking problems, including 139 reports of “unintentional emergency braking” and 383 reports of “phantom stops” from false collision warnings.

The " 2,400 self-acceleration issues" .... are usually user errors. People have complained about this again and again with so many manufacturers and in almost all cases these are user errors (pressing accelerator instead of brake etc). With AP there may be issues with AP being disengaged by steering wheel action by TACC is still on.

 

[DayTrippin]

I was told in this forum when I experienced severe phantom braking (PB) issues with my Y that basically I was smoking crack since none of my near misses, where I was almost hit but wasn't, were an issue. I maintained that just because it hadn't happened yet, doesn't mean it was an issue. I felt it was just a matter of time given how severe PB was on my car that it would sure to be if other experienced it as well.

A massive leak of internal Tesla data dropped and confirmed my fears. Over 1k accident reports of either PB caused event or acceleration issues. A quote from the article. "

"Customers from the U.S. and Europe told Handelsblatt Tesla wasn’t too interested in assisting with their issues, but seemed more intent on covering for the company. It turns out, this was explicit policy at Tesla:"

Whistleblower Drops 100 Gigabytes Of Tesla Secrets To German News Site: Report

The files contain over 1,000 accident reports involving phantom braking or unintended acceleration--mostly in the U.S. and Germany.

jalopnik.com

Apologies if this has been posted somewhere else here. I didn't see it.

 

[Midpack]

Not to discount your concern, but the title may be misleading? Self acceleration is the leading complaint. And most braking function complaints were NOT unintentional or phantom, presumably they were outright failure to brake? No correlation between complaints and accidents is provided, we can’t assume it’s a linear relationship.

Maybe there’s better data deeper in the report (I would have looked but it’s behind a paywall and in German), but the linked article does not provide a basis to conclude how many accidents were caused by phantom braking (lesser number of complaints).

1000 accidents over 7 years and X million cars and miles. Wonder what that compares to all vehicles?

2,400 self-acceleration complaints and more than 1,500 braking function problems, including 139 cases of unintentional emergency braking and 383 reported phantom stops resulting from false collision warnings.

 

[DayTrippin]

Your points are valid. Other forum members were inferring that no accidents would happen. There should be NONE of these types accidents. I plan to read through the German site when I have time.

My general thought is where there is smoke there is fire. These are the ones that actually made it to Tesla. Keep in mind a lot of the people aren't even aware of this issue when it happens. They may not make the connection and as a result, not report it. The fact that they have 383 PB reports is still a significant amount. It isn't a sample size of "1".

 

[Midpack]

Your points are valid. Other forum members were inferring that no accidents would happen. There should be NONE of these types accidents. I plan to read through the German site when I have time.

My general thought is where there is smoke there is fire. These are the ones that actually made it to Tesla. Keep in mind a lot of the people aren't even aware of this issue when it happens. They may not make the connection and as a result, not report it. The fact that they have 383 PB reports is still a significant amount. It isn't a sample size of "1".

To my knowledge, while NHTSA has acknowledged hundreds on phantom braking complaints, only 1 accident has been reported - “In November 2022, nine people were injured after an eight-car pileup was blamed on Tesla's phantom braking problem.”

If you have data on confirmed phantom braking accidents you should provide that as well? Or reserve judgement?

That said, it is a known issue that needs to be addressed - Tesla owners should not be expected to rely on following drivers being on alert.

 

[Fourdoor]

Not to discount your concern, but the title may be misleading? Self acceleration is the leading complaint. And most braking function complaints were NOT unintentional or phantom, presumably they were outright failure to brake? No correlation between complaints and accidents is provided, we can’t assume it’s a linear relationship.

Maybe there’s better data deeper in the report (I would have looked but it’s behind a paywall and in German), but the linked article does not provide a basis to conclude how many accidents were caused by phantom braking (lesser number of complaints).

1000 accidents over 7 years and X million cars and miles. Wonder what that compares to all vehicles?

Not sure if that is a low or high accident rate overall compared to other driver aids... the problem is the "this isn't a problem at all, your concerns are not valid" attitude from the company.

I no longer have severe PB problems myself, but back when I did it was infuriating that nothing seemed to be happening on coming up with a solution... and days after an NHTSA investigation was started a software update came out that fixed 95% of the problems I was having.

So if the OP is still having severe PB issues I sympathize... if he is still complaining about past issues I sympathize with the outlook that the company has a history of not taking care of problems until forced to do so.

Keith

 

[DayTrippin]

I used to stay on top of it closely. Keep in mind NHTSA will only be US data anyway. While this is a large portion of the base, it isn't the only one. I don't think we'll ever know the extent of the problem. Tesla clearly isn't transparent about it.

Once I dumped the Y, I never looked back and never had the issue to the extent I did with the Y, with my other cars. It was the cavalier attitude Tesla had every time I took it into to complain about it. Not to mention the number of near misses I had. I ultimately ended up so dissatisfied with the car, and I ended up selling it as PB drove my wife crazy.

It was a case where Tesla yanked the radar, and had no replacement for it. Then we were all treated like morons because we questioned what we were given by all-mighty Tesla. Should be thankful you even got a car and all that. I literally had about 40 PB evens on less than a 100 mile stretch of road on a 1400 mile trip one way. It was insane. My S with radar had basically none on the same exact trip, same time of year.

Let's just say my opinion of Tesla vastly decreased. They yank something out without having the replacement at a level where the transition if even remotely seamless. Just look at the HW4 cars and the yanking of USS.

I literally had semi jamming on their brakes and swerving past me so they wouldn't rear end me when my Y had a severe PB issue. Tesla didn't give a flying you know what. Best advice anyone gave me when I bought my Y was and I quote "Assume AP is out to kill you at all times." I added TACC to that list as well.

People getting a car more recently haven't lived through how terrible the rollout of TV was for some of us.

 

[DayTrippin]

Another take on the same data. Don't know if the brake problems below were PB related or not, but could be. All in all the data paints a less than positive picture about Tesla.

"Handelsblatt says there were more than 1,000 crashes linked to brake problems and more than 3,000 entries where customers reported safety concerns with the driver assists."

"According to Manfred Schon, he experienced something similar on the M14 highway. The former
 Bosch employee was on his way to a meeting in the US state of Michigan on June 1, 2019 when his Tesla "suddenly slammed on the brakes, as hard as you can imagine," Schon told the Handelsblatt. "I was pushed into the seat belt and the car almost came to a stop. Then another car hit me from behind."

I can relate to Manfred. Just I was lucky enough that several times when this happened the people behind me were paying attention. The data also talks about Phantom Stop complaints, not just PB.

Huge Tesla leak reveals thousands of safety concerns, privacy problems – Ars Technica

 

[Midpack]

Again not discounting the phantom braking concern at all, I am sure it’s a terrifying sensation.

Maybe because I’m not as current as Tesla owners, but I read a lot about Tesla and I’ve heard few if any concerns about self acceleration? That‘s the primary source of complaints.

 

[DayTrippin]

Let's call it about a 60/40 split; acceleration vs braking complaints.

The major difference to me is that if the car accelerates, you can out brake the cars ability to accelerate by using the brakes. Often times people confuse brakes with accelerators. With the PB/PS events, really nothing you can do until the car stops braking other than to try and steer away from getting rear-ended. Not to mention almost all cars can brake far harder than they can accelerate.

So if the car applies maximum braking, you are pretty much stuck until it stops braking. After I figured out what it could do, I would make sure my foot was never far away from the accelerator so if it did brake, I could punch it and hopefully not get rear-ended. I lost count the number of times I had to do that to avoid being hit.

We are not talking a light tap on the brakes. It can be full force, throw out the anchor, emergency stop type of braking. Imagine driving along at 70 mph and the car just randomly slams on the brakes as if it were trying to avoid a crash. That is what I would experience at times on the extreme end. The other would be to just drop a quick 10-20 mph. Even if I hit the accelerator immediately, nothing was going to happen until it let off the brakes. If it were just accelerating, I could override that by braking.

 

[Midpack]

I’m holding off on ordering a MYLR because I can wait for HW4. As I understand it, HW4 now comes with higher res cameras and HD radar at least on MS/X. Your experience (with no radar) just reinforces my decision to wait - thank you. Luckily it appears some MY w HW4 have begun to appear in delivery inventory so I may not have to wait as long as I thought. Yes, I’m making some hopeful assumptions based on the MS/X rollout - all HW4 cars also have/had HD radar?

Post in thread 'HW4 and Phoenix Radar Build Dates'
HW4 and Phoenix Radar Build Dates

 

[Fourdoor]

I used to stay on top of it closely. Keep in mind NHTSA will only be US data anyway. While this is a large portion of the base, it isn't the only one. I don't think we'll ever know the extent of the problem. Tesla clearly isn't transparent about it.

Once I dumped the Y, I never looked back and never had the issue to the extent I did with the Y, with my other cars. It was the cavalier attitude Tesla had every time I took it into to complain about it. Not to mention the number of near misses I had. I ultimately ended up so dissatisfied with the car, and I ended up selling it as PB drove my wife crazy.

It was a case where Tesla yanked the radar, and had no replacement for it. Then we were all treated like morons because we questioned what we were given by all-mighty Tesla. Should be thankful you even got a car and all that. I literally had about 40 PB evens on less than a 100 mile stretch of road on a 1400 mile trip one way. It was insane. My S with radar had basically none on the same exact trip, same time of year.

Let's just say my opinion of Tesla vastly decreased. They yank something out without having the replacement at a level where the transition if even remotely seamless. Just look at the HW4 cars and the yanking of USS.

I literally had semi jamming on their brakes and swerving past me so they wouldn't rear end me when my Y had a severe PB issue. Tesla didn't give a flying you know what. Best advice anyone gave me when I bought my Y was and I quote "Assume AP is out to kill you at all times." I added TACC to that list as well.

People getting a car more recently haven't lived through how terrible the rollout of TV was for some of us.

I purchased my Y right at the transition, I wasn't sure if mine would have radar or not on delivery... sadly, it did not. On my 30 mile each way daily commute, when it was daylight I would get 5 to 10 PB events, when it was dark I would get 25+ on the same route. That is each way, so on avearage I was having 30 PB events every day. I figured out a "work around" by setting the max speed to 60 mph and just planting my foot to the floor. No PB's in that situation, but also no safety interventions by the car if I had actually needed one.

I was seriously considering selling my Y when the NHTSA investigation started, and my problems were all solved over night.

Keith

 

[Yelobird]

Let's call it about a 60/40 split; acceleration vs braking complaints.

The major difference to me is that if the car accelerates, you can out brake the cars ability to accelerate by using the brakes. Often times people confuse brakes with accelerators. With the PB/PS events, really nothing you can do until the car stops braking other than to try and steer away from getting rear-ended. Not to mention almost all cars can brake far harder than they can accelerate.

So if the car applies maximum braking, you are pretty much stuck until it stops braking. After I figured out what it could do, I would make sure my foot was never far away from the accelerator so if it did brake, I could punch it and hopefully not get rear-ended. I lost count the number of times I had to do that to avoid being hit.

We are not talking a light tap on the brakes. It can be full force, throw out the anchor, emergency stop type of braking. Imagine driving along at 70 mph and the car just randomly slams on the brakes as if it were trying to avoid a crash. That is what I would experience at times on the extreme end. The other would be to just drop a quick 10-20 mph. Even if I hit the accelerator immediately, nothing was going to happen until it let off the brakes. If it were just accelerating, I could override that by braking.

Would only say that Assuming data doesn’t validate a specific concern. When (and it was real) there were PB events I would push that report button 50 plus times per week. As for the pile of complains I would like to know of ANY other maker by comparison that gives the owner a button to complain/record directly to the data base. There are none. For me, they collected data and have since fixed the issue. could be mad that it happened but it was resolved and life moves on. (for me). Regarding radar, in my view that is the grand go to placebo effect of a forum that can only assume change must be bad. As soon as the radar was disabled that was the end of PB events for me. Glad they dumped it personally and progressed forward.

 

[MP3Mike]

My general thought is where there is smoke there is fire.

I don't think that applies here. There has been a lot of smoke, reports of uncommanded acceleration, in the US, and NHTSA investigated and determined that there was no fire, in every case it was driver error. (i.e. it was unintended acceleration where the driver pressed the accelerator instead of the brake pedal.) I would assume it is likely the same for all of those cases reported to Tesla. (Which Tesla has supposedly passed all of those reports on to NHTSA as part of NHTSA's SUA investigation.)

 

[DayTrippin]

I am talking about the braking, not the acceleration.

 

[RichBenn]

Unintended acceleration due to operator error “could” be more prevalent in one pedal driving EVs. Without real data including other vehicles it’s not possible to surmise how many the complaints are actually due to software, if any. But with numbers like that reported, a good company would jump on it to determine the problem.

 

[MP3Mike]

But with numbers like that reported, a good company would jump on it to determine the problem.

And Tesla did. They performed a joint investigation with NHTSA and determined that every cases they had data for was pedal misapplication. Just as @wk057 has said, there is never any uncommanded acceleartion.

 

[clydeiii]

Let's call it about a 60/40 split; acceleration vs braking complaints.

The major difference to me is that if the car accelerates, you can out brake the cars ability to accelerate by using the brakes. Often times people confuse brakes with accelerators. With the PB/PS events, really nothing you can do until the car stops braking other than to try and steer away from getting rear-ended. Not to mention almost all cars can brake far harder than they can accelerate.

So if the car applies maximum braking, you are pretty much stuck until it stops braking.

Is this really true??? I use FSDb all the time and if my car is braking (phantom or no) at any time, for any reason, and I disagree with its braking, I simply hit acceleration pedal and it's fine. It stops slowing down while continuing to drive the car for me.