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

Tesla recalls 2 million vehicles to limit use of Autopilot

This site may earn commission on affiliate links.
The NHTSA tends be very utilitarian on driver assistance so I don't anticipate a restriction of use
Assuming Tesla and NHTSA are agreeing to remedies based on the same data, they probably know allowing broad usage of Autosteer even off highways can increase safety:


Reviewing the previous statement in light of this recall to "prevent driver misuse of Autosteer," it seems like one potential change is to trigger the big scary "take over immediately" alert when pressing the accelerator, e.g., maintaining 60 mph in 45 mph zone (similar to the existing alert when maintaining over 85 mph).
 
I only know of one person who did that. It was someone here on the forum. He said it was such a massive pain he didn't want to try it again.

There may be others, but I think the person who tried it had to disconnect from the Tesla mothership to get AP working. So the car would not be able to be updated anyway.
I've seen that there was another company offering the retrofit, I just can't find them right now. (I think 057 even ended up offering the retrofit because people were willing to pay for it.)

And from what I recall it didn't require disconnecting from Tesla. (Jason's P100++ upgrade for the Model S did require the disconnect.)
 
  • Informative
Reactions: wdolson
I have noticed previously that if you are on autopilot (AP) and you accelerate for whatever reason, I think past 85, even momentarily, then the car will "lock you out" of autopilot until you stop and turn the car off. I find this infuriating, and I feel like a parent has set me into the corner or something. If the new AP is even more restrictive, and even more likely to lock you out, this will really be a pain!

The only way I can see to improve this experience is ... tap the brakes to get out of AP anytime you want to do anything at all. Then, re-start AP after. That way you won't get locked out.

Hope this is helpful,

-TPC.
-------------------------------------------------
2023 Model Y with absolutely no upgrades, waited more than 6 months for it. Much to my surprise, turns out I like the darn thing! (And I LOVE passing by gas stations and not stopping hehehehe...)
 
Like I'm all for perosnal freedom but each one of these places are innapropriate for AP as it stands today and we all know it
Same could be said for "dumb" cruise control, yet most cars have this feature.

People are going to not pay attention to driving with or without ADAS. Limiting technology that small portions of the population cannot use correctly leads down a slippery slope.
 
I thought AutoPilot was basically just what we Ford folks call "lane centering" plus "adaptive cruise control." But hands have to stay on, as monitored by a torque sensor (and maybe cabin cam, too?) Whereas Full Self Driving is AutoPilot plus it changes lanes on the highway and actually makes turns and stops at signs and lights - basically "full self driving"
Your description of these terms is generally correct. On your next question, while FSD does enable car to make turns, stop at lights, etc. on city streets, driver attention is still required (via dashcam and steering wheel monitoring).

On your final point, the OTA is intended to make standard AutoPilot safer by enforcing more strict driver attention and perhaps restricting use on city streets (where it shouldn't be used). Importantly, standard AutoPilot can be enabled today on city streets but will not stop at lights / signs, etc. Therefore, poor driver behavior and inattention can lead to bad results.
 
  • Like
Reactions: FordMME
I have noticed previously that if you are on autopilot (AP) and you accelerate for whatever reason, I think past 85, even momentarily, then the car will "lock you out" of autopilot until you stop and turn the car off.
Considering that the fastest speed limit in the US is 85 MPH and that's only on a tollway near Austin, I can understand.
 
Same could be said for "dumb" cruise control, yet most cars have this feature.

People are going to not pay attention to driving with or without ADAS. Limiting technology that small portions of the population cannot use correctly leads down a slippery slope.
It really isn't. Like if GM's supercruise (I know the 30 sec limit on the data) showed 216 collisions on surface roads (including killing a pedestrian) in a 12month period there would be riots. Tesla "gets away" with this since it is pushing the limits of technology. But the reality is that AP/FSD is fundamentally no where near ready for public use on surface streets. The highway, sure, in fact I would contend AP is better than most human drivers (even this new pure-vision version) on the highway. But backroads and intersections, it's just not there yet. Especiallly since any Jim or Jane who has $30k to spend on a car can get one now. Tesla isn't some spunky startup anymore they are a real car company that has millions of regular people as customers
 
Last edited:
[data]

No one is trying to lie, I get those things. I dug thru the 2023 CSV the NHTSA posted. First GM and Subaru now post "telemetricly", not just Tesla. Also, look at these places people are using AP! Right before a crash! There are 216 of these in one year! Dozens of them into other passenger cars. Like I'm all for perosnal freedom but each one of these places are innapropriate for AP as it stands today and we all know it
Do you have a link where this data is from? I'd enjoy looking over it when I get chance. 216 crashes seems insanely little for a car as prolific as a Tesla. That's just 216 collisions around the time of a ADAS system being used? Not even fatalities.

I think a more useful way of measuring accidents is per miles driven. How many collisions were Teslas involved in around the time of AP being used per miles driven on AP or using an ADAS system. Then compare that to other manufacturers. My bet is that the number of miles for Teslas AP/FSD/TACC far dwarfs any other traditional manufacturer, but idk. It would be very fun to see that data.
 
Was just on NBC nationwide nightly news.

Of course had to bring up a crash from 2019 to make their point. And the dozen crashes into emergency vehicles even though that has already been mitigated with previous updates.

They also had to say that some people think it makes the car fully autonomous. Of course they leave out the fact that a driver is told to keep their hands on the wheel and pay attention EVERY SINGLE TIME it’s turned on.

At least they did make clear it is updated with a software over the air update rather than having to bring in the vehicle. And mentioned Tesla stated that overall it is safer.

 
Do you have a link where this data is from? I'd enjoy looking over it when I get chance. 216 crashes seems insanely little for a car as prolific as a Tesla. That's just 216 collisions around the time of a ADAS system being used? Not even fatalities.

I think a more useful way of measuring accidents is per miles driven. How many collisions were Teslas involved in around the time of AP being used per miles driven on AP or using an ADAS system. Then compare that to other manufacturers. My bet is that the number of miles for Teslas AP/FSD/TACC far dwarfs any other traditional manufacturer, but idk. It would be very fun to see that data.
Tesla could release coorisponding raw data to the public but that would require them to relate to the public.

Also, 216 is just non-highway collisions. There were nearly 1k in 2023, just for Tesla. Both Subaru and GM reported (at least some) of their collisions via "telemetrics" in 2023 and they both had fewer than 75.

Unfortuately the other 1.9mil Tesla owners who weren't early adopters and don't care about AP/FSD also use this tech. And a LOT of them abuse it, seemingly with fatal results.

 
  • Informative
Reactions: ArchHamster
I have noticed previously that if you are on autopilot (AP) and you accelerate for whatever reason, I think past 85, even momentarily, then the car will "lock you out" of autopilot until you stop and turn the car off. I find this infuriating, and I feel like a parent has set me into the corner or something. If the new AP is even more restrictive, and even more likely to lock you out, this will really be a pain!

The only way I can see to improve this experience is ... tap the brakes to get out of AP anytime you want to do anything at all. Then, re-start AP after. That way you won't get locked out.

Hope this is helpful,

-TPC.
-------------------------------------------------
2023 Model Y with absolutely no upgrades, waited more than 6 months for it. Much to my surprise, turns out I like the darn thing! (And I LOVE passing by gas stations and not stopping hehehehe...)
I always tap up on the shift selector to disengage whatever ADAS I am using whenever I assume control; invariably it is for passenger comfort.
 
Tesla could release coorisponding raw data to the public but that would require them to relate to the public.

Also, 216 is just non-highway collisions. There were nearly 1k in 2023, just for Tesla. Both Subaru and GM reported (at least some) of their collisions via "telemetrics" in 2023 and they both had fewer than 75.

Unfortuately the other 1.9mil Tesla owners who weren't early adopters and don't care about AP/FSD also use this tech. And a LOT of them abuse it, seemingly with fatal results.

Thank ya thank ya. I took a quick glance and wow:

1702513847074.png


It looks like basically no one is reporting data like tesla does. Maybe only honda has any significance a comparison and, on quick glance, this isn't collisions per miles driven on ADAS which is very unfortunate. It makes it difficult to use this data to compare adas systems across manufacturers.

For example, a Level 2 ADAS-equipped vehicle manufacturer with access to advanced data recording and telemetry may report a higher number of crashes than a manufacturer with limited access, simply due to the latter’s reliance on conventional crash reporting processes. In other words, it is feasible that some Level 2 ADAS-equipped vehicle crashes are not included in the summary incident report data because the reporting entity was not aware of them. Furthermore, some crashes of Level 2 ADAS-equipped vehicles with limited telematic capabilities may not be included in the General Order if the consumer did not state that the automation system was engaged within 30 seconds of the crash or if there is no other available information indicating Level 2 ADAS engagement due to limited data available from the crashed vehicle. By contrast, some manufacturers have access to a much greater amount of crash data almost immediately after a crash because of their advanced data recording and telemetry.

From: Standing General Order on Crash Reporting | NHTSA

This data seems like it's rife with selection bias. Mainly self selection and availability bias. Maybe instead of overregulating tesla, the NHTSA should be forcing other manufacturers with ADAS 2 systems to report as well as tesla does. In other words, don't shoot the messenger.

I'd love to get the accidents near AP per miles driven. Just a quick search reveals this: Tesla releases new Autopilot safety report, shows improvements despite limited data

“In the 4th quarter, we recorded one crash for every 4.31 million miles driven in which drivers were using Autopilot technology (Autosteer and active safety features). For drivers who were not using Autopilot technology (no Autosteer and active safety features), we recorded one crash for every 1.59 million miles driven. By comparison, NHTSA’s most recent data shows that in the United States there is an automobile crash every 484,000 miles.”

Which seems great. Why are we regulating AP more? I'd love to look at the tesla data, whatever they released directly for problems too. But tesla blocks my work VPN and I do need to get back to work. It might be here for those interested to dig deeper, again I can't tell because my work VPN is blocked: https://www.tesla.com/VehicleSafetyReport
 
  • Like
Reactions: gt2690b
I doubt this is a traditional recall. More likely, it will be an OTA update. As mentioned before, the word "recall" is misleading and does not really apply anymore since the cars don't need to be physically recalled back to the service center for the fix to happen. Tesla will simply push an OTA update to fix this issue.
As mentioned before, the word "recall" has legal significance. Not only does the manufacturer have to address the issue, but they have to address it at no charge to the owner. If for example, an early version computer was not large enough to run the newly released SW as a result of the recall, Tesla would have to pay to upgrade the computers as well.
 
  • Like
Reactions: advocate8
I have a ’23 Model 3, no EAP, no FSD. When using AP the car will warn it will not allow continued use after a number of times it has to warn for not holding the wheel. Also, it warns after every 25-30 seconds even if I have my hands lightly on the wheel. You have to move the wheel or hold it with some weight or else it will warn. I use the volume knob over and over again to stop the warning.

I’m all for safety, but at this point I’m thinking I will likely just stop using AP.
 
Thank ya thank ya. I took a quick glance and wow:

View attachment 999188

It looks like basically no one is reporting data like tesla does. Maybe only honda has any significance a comparison and, on quick glance, this isn't collisions per miles driven on ADAS which is very unfortunate. It makes it difficult to use this data to compare adas systems across manufacturers.

For example, a Level 2 ADAS-equipped vehicle manufacturer with access to advanced data recording and telemetry may report a higher number of crashes than a manufacturer with limited access, simply due to the latter’s reliance on conventional crash reporting processes. In other words, it is feasible that some Level 2 ADAS-equipped vehicle crashes are not included in the summary incident report data because the reporting entity was not aware of them. Furthermore, some crashes of Level 2 ADAS-equipped vehicles with limited telematic capabilities may not be included in the General Order if the consumer did not state that the automation system was engaged within 30 seconds of the crash or if there is no other available information indicating Level 2 ADAS engagement due to limited data available from the crashed vehicle. By contrast, some manufacturers have access to a much greater amount of crash data almost immediately after a crash because of their advanced data recording and telemetry.

From: Standing General Order on Crash Reporting | NHTSA

This data seems like it's rife with selection bias. Mainly self selection and availability bias. Maybe instead of overregulating tesla, the NHTSA should be forcing other manufacturers with ADAS 2 systems to report as well as tesla does. In other words, don't shoot the messenger.

I'd love to get the accidents near AP per miles driven. Just a quick search reveals this: Tesla releases new Autopilot safety report, shows improvements despite limited data



Which seems great. Why are we regulating AP more? I'd love to look at the tesla data, whatever they released directly for problems too. But tesla blocks my work VPN and I do need to get back to work. It might be here for those interested to dig deeper, again I can't tell because my work VPN is blocked: https://www.tesla.com/VehicleSafetyReport
Do you think Tesla numbers are high because the car reports that it had an accident?

If a Kia has an accident, does Kia find out?

Thats a question, not a statement.
 
  • Like
Reactions: ArchHamster