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

OFFICIAL BUTTON WATCH

This site may earn commission on affiliate links.
I agree that would seem to be the conclusion. Having said that, I gotta say this whole thing makes no sense then. The whole thing about FSD is that the car is driving while the driver needs to pay attention. Yet they're measuring our driving habits/performance only without AP engaged. I don't believe the latter is necessarily predictive of the former. For instance the rich boy idiot (malignant narcissist) who insists on buying new Teslas then riding around in the back seat could drive like a granny for a week, get the download, then turn right around and get in the backseat. Seems like the thing to measure would have been an "attentiveness score" and it would be the inverse -- only active during AP. Pay attention: keep your FSD. Exhibit lack of attention: have your FSD revoked. Feels like the whole thing is driven (no pun intended!) by image and legality more than logic.
I also agree - Attentiveness would be the most important moving forward once deployed. I think think this process helps align individual driving styles to that of FSD.
 
I want to have a freedom of choice. If I want to be safe, slow, and careful…. I turn on AP/FSD and let it do it’s granny driving, if I want to do some spirited driving on the twisties with my 600hp car I spent 90k on, I should be able to do so without being penalized.

I don’t expect anyone to agree, because well, that’s just my personal opinion and it’s mine, you can have your own.
Oh god, this is going to be another anti-mask debate! 🍿
 
There’s always going to be bad actors but mitigating risk to reduce the number of them is within teslas best interest because the moment a terrible fsd related accident occurs, the media is going to unfairly pounce on Tesla and hammer them for it.
Yes, it only takes one incident to nullify the "no accidents on FSD" claim oft cited by Elon. And yes, the media will pick up the rhetoric on the evils of automated driving.

Still, I would have to say FSD-related accidents should reflect poorly on Tesla. Which is why they are trying to recruit 'safe' test drivers to catch problems. FSD beta is not ready for the mainstream I-paid-for-it-I-wanna-use-it type person quite yet.
 
... FSD-related accidents should reflect poorly on Tesla. ...
A few years ago, an insider told me there are significant number of autopilot accidents each week. I was somewhat amazed few knew about them. Now the NHTSA is requiring Tesla to report. I wouldn't be surprised if some minor FSD accidents go under the radar. There is of course an incentive to under report. If you are in an accident then FSD is going to be taken away. No! not my $10K toy!
 
Using AP is good for your Safety Score. Since the safety score won't record any AP fails but still track the miles, it will skew the average in your favor. The only time using AP will hurt your safety score is if AP is going to cause a crash and you have to disengage and jerk the wheel or slam on the brakes to avoid the accident. But for the most part, just cruising on AP will help increase your score.

Thanks for the explanation!

I think the blog's phrase "Autopilot will not be included in Safety Score calculations" has confused some readers as if Tesla would penalize owners for using AP and as if Tesla wants us to "hand drive" because normally the words "not included" sound too negative:

"Driving on Autopilot will not be included in Safety Score calculations, but the miles driven while on Autopilot are included in the total."

I tested the drives on Autopilot and it turns out your explanation is true: If we want to be graded well, use Autopilot!
 
Last edited:
A few years ago, an insider told me there are significant number of autopilot accidents each week. I was somewhat amazed few knew about them. Now the NHTSA is requiring Tesla to report. I wouldn't be surprised if some minor FSD accidents go under the radar. There is of course an incentive to under report. If you are in an accident then FSD is going to be taken away. No! not my $10K toy!
Conspiracy theory.
 
What’s gonna get me some time is when AP slams on the brakes way too freaking hard and i disengage so the guy tailgating me doesn’t plow into me…and then have to brake heavy myself. Freaking AP right now has chirped my tires slamming on the brakes while I’m following a car on 7 distance…
 
I suspect this is going to go poorly, depending on how actively Tesla is going to monitor folks once they are in the Beta and how quickly they kick them out if they are misusing the system. I just saw a post on Reddit from a person who had requested beta access and in that same post admitted to using their phone while on AP and not bothering to look up to satisfy a wheel nag.

I am still debating if I want to participate or not. I don't know if the highway code has changed, as that is the only portion I would be interested/willing to test. I probably wouldn't provide enough useful data to them to keep me in the program since I wouldn't use it on city streets.
 
I suspect this is going to go poorly, depending on how actively Tesla is going to monitor folks once they are in the Beta and how quickly they kick them out if they are misusing the system. I just saw a post on Reddit from a person who had requested beta access and in that same post admitted to using their phone while on AP and not bothering to look up to satisfy a wheel nag.

I am still debating if I want to participate or not. I don't know if the highway code has changed, as that is the only portion I would be interested/willing to test. I probably wouldn't provide enough useful data to them to keep me in the program since I wouldn't use it on city streets.
Your reticence given the reasoning that you did is understandable. Please understand this, the more data Tesla is given by driving feedback on FSD whether that data is gotten on highway driving or city driving, the sooner the software can come to fruition. So your participation, no matter how small or where you drive, adds to the sum total of data banked.
 
  • Like
Reactions: rxlawdude
There's plenty of cases where I've disengaged Autopilot not really to prevent a crash but because it was going to take a curve too quickly, doesn't yield for waiting pedestrians, would stop too far, etc. and these often require braking that would likely negatively impact the Safety Score with hard braking.
Maybe Tesla is testing for people who are aware of Autopilots current shortcomings and manually take over in circumstances where Autopilot isn’t the best. It seems like they would definitely want people who can drive better than Autopilot otherwise that could be used to validate FSD beta.

Thinking about it a little more, if you let Autopilot get you into bad situations, FSD Beta could do even more damage since it can depart the lane.

Autopilot/FSD failure is what they want people to watch for so getting a ding in your score when you let Autopilot drive into a scenario it doesn’t handle well Is probably good.
 
Last edited:
A few years ago, an insider told me there are significant number of autopilot accidents each week. I was somewhat amazed few knew about them. Now the NHTSA is requiring Tesla to report. I wouldn't be surprised if some minor FSD accidents go under the radar. There is of course an incentive to under report. If you are in an accident then FSD is going to be taken away. No! not my $10K toy!
The NHTSA requires reporting of crashes which "must have resulted in an injury, fatality, or property damage". This appears different from Tesla's safety reporting style "We also receive a crash alert anytime there is a crash ... we count all crashes in which the crash alert indicated an airbag or other active restraint deployed". Tesla seems to rely on a "crash alert" being sent from the car. The NHTSA requires a report when there is a crash. Some crashes may not result in a "crash alert" being sent from the car.

It's not clear how Tesla will report to the NHTSA when there is no "crash alert" - for example running over someone's foot causing injury (if that meets the NHTSA's criteria of a crash). Same goes for the other manufacturers. Do they have to troll the body shops, local news and police reports for crashes that may not have been reported as Autopilot-related? Or do they not bother to look. How much due diligence do they have to do?

I doubt Tesla will immediately take away FSD Beta for having an accident. That's just more likely to make you bitch about it to the media, IMO.
 
Last edited:
Your reticence given the reasoning that you did is understandable. Please understand this, the more data Tesla is given by driving feedback on FSD whether that data is gotten on highway driving or city driving, the sooner the software can come to fruition. So your participation, no matter how small or where you drive, adds to the sum total of data banked.
True. I might throw my name into the hat anyways. I probably won't try to game the 7 day trial to get in though. Too much hassle. If my (and my husband's) driving style is deemed too "unsafe", oh well.