Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Phantom Braking

This site may earn commission on affiliate links.
Another guy who posts without reading any other posts :)

The NHTSA story was already posted here- like 3 posts above yours.

Nor is your claim based on...anything in reality...It exists in all brands of adaptive cruise, as discussed pretty exhaustively now in this very thread.
But may be after NHTSA "incidence", perhaps Tesla could offer a dumbed down version of AutoPilot so that you can (manually) flip between AP and DumbAP depending on the type of road you are on? I would be happy with this setup until FSD becomes more stable...

You are on a 2-lane road and a big truck approaches. You click, speed stays constant unless you have a car close right in front of you, and you take over the wheel until the truck passes by. Then you click back and continue...
 
  • Like
Reactions: 2101Guy
But may be after NHTSA "incidence", perhaps Tesla could offer a dumbed down version of AutoPilot so that you can (manually) flip between AP and DumbAP depending on the type of road you are on? I would be happy with this setup until FSD becomes more stable...

You are on a 2-lane road and a big truck approaches. You click, speed stays constant unless you have a car close right in front of you, and you take over the wheel until the truck passes by. Then you click back and continue...
Now that’d be nice, it’d be set to “dumb” all the time in mine🤣
 
I don’t see any reason Tesla couldn’t build something like that in and make it intuitive, just having regular dumb cruise control available would likely resolve a lot of the complaints. It might reduce the miles driven on AP, but maybe not so much considering many people seem to avoid using it because of the issue.

The evolving nature of this is also why it’s not helpful to harp on no investigation existing yet: we’re at the bleeding edge of this, people complaining is what creates investigations and people frequenting this thread are already moving more quickly than the regulators can.
 
Are 2021+ Models 3 and Y the only models experiencing phantom braking? Is that the reason they have limited the probe, or is it just the start?
Phantom braking exists across all models (and all similar systems from other manufacturers as well), but there's a massive uptick in reports to the NHTSA against the 2021 and 2022 model years. For example the Y, just focusing on items tagged specifically to Forward Collision Avoidance:

1645114610287.png

1645114632866.png


Almost 10x as many complaints and in a newer model

The 2022 Model Y already has 3x as many complaints as the 2020 model year and we're only 2.5 months in. So there appears to be some difference in the 2021+ models that is producing far more complaints, we can only speculate what that might be
 
Last edited by a moderator:
So there appears to be some difference in the 2021+ models that is producing far more complaints, we can only speculate what that might be

I don't think we need to speculate on most of the difference.



Tesla sold ~25,000 Model Ys in 2020 in the US. (and keep in mind Tesla now changes model year in ~October, so at least 1/4 of THOSE are 2021s by VIN)

Tesla sold ~161,529 Model Ys in 2021 in the US (plus the roughly 1/4 of those sold in late 2020 VINed as '21s)


So yes I'd certainly expect a LOT more complaints from 2021 Ys, even if the actual occurrence of the issue is exactly the same in both years.
 
I don't think we need to speculate on most of the difference.



Tesla sold ~25,000 Model Ys in 2020 in the US. (and keep in mind Tesla now changes model year in ~October, so at least 1/4 of THOSE are 2021s by VIN)

Tesla sold ~161,529 Model Ys in 2021 in the US (plus the roughly 1/4 of those sold in late 2020 VINed as '21s)


So yes I'd certainly expect a LOT more complaints from 2021 Ys, even if the actual occurrence of the issue is exactly the same in both years.
Are those the cumulative numbers of 2020 vs 2021 Model Ys that have been sold? The NHTSA has or will be able to get all these figures

There could be a huge range of things influencing why more reports are coming in, from sales numbers to earlier adopters possibly being less inclined to report issues versus the broader consumer base. That's stuff for the NHTSA to dig into.
 
So yes I'd certainly expect a LOT more complaints from 2021 Ys, even if the actual occurrence of the issue is exactly the same in both years.
Agreed, and the numbers are pretty low to draw conclusions, but it's also worth noting that in 2020, 23.9% of all Model Y complaints were about "Forward Collision Avoidance" and in 2021, 61.1% of all Model Y complaints were about "Forward Collision Avoidance."

So from a percentage-of-reports perspective, it increased. No conclusions can be drawn from that alone, but it is worth consideration.
 
Agreed, and the numbers are pretty low to draw conclusions, but it's also worth noting that in 2020, 23.9% of all Model Y complaints were about "Forward Collision Avoidance" and in 2021, 61.1% of all Model Y complaints were about "Forward Collision Avoidance."

So from a percentage-of-reports perspective, it increased. No conclusions can be drawn from that alone, but it is worth consideration.
Yup and the people who work for the NHTSA, specifically in a corner of the industry dedicated to investigating automotive issues etc, will surely consider all of this and likely many things we haven't yet thought of. And they can work with Tesla to get data to analyze.
 
We can also consider that volumes sold will increase the likelihood of something bad happening and can be a trend worth investigating by itself. Some issues don’t become issues until you have sufficient numbers exposing the risk of bad things happening.

Where were the 2020 models sold? Were most sold in California or something where the system operates more smoothly? Not to say that‘s the case, but it’s just another layer of nuance to look at and I’m sure it’s one of many.
 
Where were the 2020 models sold? Were most sold in California or something where the system operates more smoothly? Not to say that‘s the case, but it’s just another layer of nuance to look at and I’m sure it’s one of many.
CA has always accounted for nearly 50% of EVs sold since 2010. That is true for Tesla too.

CA share has come down in recent years as sales in other states like TX and in NE has picked up. Still, CA accounts for 40%+ of EV registrations.

Ofcourse Tesla is now selling a LOT more cars than it did in 2017 or even 2019. Though most of the increase is in China, it is quite a bit more in US as well.

But the most important thing is - what is the number of phantom braking per 1 million miles driven on AP. Only Tesla has that number, unfortunately.

1645117945275.png



 
Last edited:
...Where were the 2020 models sold? Were most sold in California or something where the system operates more smoothly? Not to say that‘s the case, but it’s just another layer of nuance to look at and I’m sure it’s one of many.
2020 still uses radar. Radarless has started on 4/28/2021 for Model 3 an Y in North Americqa.

Radar version also have Phantom Brakes. I experienced that with both my Model 3 and Model X. However, as soon as I switched my Model 3 to Pure Vision, I experienced NEW MORE WAYS of slowdowns that I have never experienced with radar and with my current radar Model X.
 
Last edited:
A friend of mine just texted me asking what's going on with Teslas "literally stopping in the middle of the highway....the cars stop operating". He said he just heard it on the news. I wonder how many other people heard it that way. Amazing how people don't fully listen to the news and that's what they hear. I had to explain phantom braking to him and that the cars don't stop in the middle of the highway dead. SMH
 
Just for me to understand as a soon to be owner - the maps issue is caused by incorrect speed limit information? Does TACC not have a simple speed setting? Does it continually have to adjust speed based on what it thinks the speed limit is?
Speed limits seem to be a combination of information stored with the maps (and correlated with GPS) and visual detection of speed limit signs. It’s not clear to me how the car decides.

Typically it’s fairly accurate but there’s a highway near our cabin that has a 55mph sign which the car recognizes and speeds up appropriately, then about 200 feet down the road it thinks the speed limit drops to 35 for no clear reason. That’s the only place I’ve experienced that. There are other places where it get confused by a school speed limit sign or some state highway signs that look/are shaped somewhat like speed limit signs. These are relatively infrequent, though.
 
My AP1 Tesla very rarely does this.

When it does unnecessary slowdown, I can always see the reason.
How well does it stop for automatic emergency braking ?

It is extremely easy to eliminate phantom braking by increasing the error on AEB.

But, if you want to reduce the possibility of AEB not kicking in when needed, you also increase phantom braking.

Remember all those news stories of Tesla hitting stopped vehicles or AP1 tesla crashing into a crossing truck killing the driver (which eventually led to Mobileye Tesla divorce) ? Phantom braking is a result of that.

How many phantom brakings are "ok" if it stops one head-on collision and death ?
 
  • Like
Reactions: DrGriz and COS Blue
Bringing up the issue has caused zero reported accidents (NHTSA prelim doc just posted confirms 0 BTW) is directly relevant to if it's an actual safety issue versus just an annoyance.
1. Absence of evidence is not absence of effect
2. the question has never been whether it's caused accidents, the question is whether it's a crappy, annoying system that doesn't function as it should. whether it's caused accidents is irrelevant to this.
3. I long ago quit reading past the first line of your nonsensical rants and ramblings because they were never worth the effort.