FSD beta expansion to new beta testers June 2022

[starship42]

Teslafi doesn't show anyone downgrading to the FSD Beta from 2022.16, so it hasn't happened yet. I would be surprised if they downgraded you to get to the beta, but what do I know?

Knowing what I know now from the forum, I feel like I shot myself in the foot upgrading from 2022.12.3.2 to 2022.16.1.1, when it appears unlikely (at the moment) of getting a “downgrade” to 2022.12.3.20 to get the beta.

2020 M3 SR+ with FSD monthly subscription. Once I clicked the “request” button, I have about 350 miles driven over 6 days so far, mixed AP, NOA, and manual driving, with 99SS. Looks like I may be waiting until the next FSD release before I pass and get admitted.

Now it looks like I’m waiting that mythical “two weeks” (per Elon) or longer for FSD 10.13. ¯\_(ツ)_/¯

I would like to have FSD for my drive to Austin for August 4

 

[Dewg]

Knowing what I know now from the forum, I feel like I shot myself in the foot upgrading from 2022.12.3.2 to 2022.16.1.1, when it appears unlikely (at the moment) of getting a “downgrade” to 2022.12.3.20 to get the beta.

2020 M3 SR+ with FSD monthly subscription. Once I clicked the “request” button, I have about 350 miles driven over 6 days so far, mixed AP, NOA, and manual driving, with 99SS. Looks like I may be waiting until the next FSD release before I pass and get admitted.

Now it looks like I’m waiting that mythical “two weeks” (per Elon) or longer for FSD 10.13. ¯\_(ツ)_/¯

I would like to have FSD for my drive to Austin for August 4

Most of your trip will be on freeways/highways, so AP and NoA will still take care of you. FSD Beta would only help you on city streets, and even then it's a commitment to use it safely and report issues.

 

[kbecks13]

Hey all - Been trying for beta for a while and never had much luck. If I'm already down to a 93 score, should i just reset and try again or do you think its worth trying to keep driving to pull it back up? Super hard in LA to keep it high.

 

[Beachybumm]

I have a M3 in LA with a safety score of 81, which I feel is pretty good for LA. There are certain areas where I get a warning even when there is no issue. Will I get FSD beta with this score?

 

[VeeKay]

I have a M3 in LA with a safety score of 81, which I feel is pretty good for LA. There are certain areas where I get a warning even when there is no issue. Will I get FSD beta with this score?

My suggestion would be to opt out, wait for the safety score to disappear in the app(not sure if it will if you also have Tesla auto insurance), opt back in, and then drive 100+ miles on NAP (off peak hours if traffic is a concern). This should do it.

 

[kbecks13]

My suggestion would be to opt out, wait for the safety score to disappear in the app(not sure if it will if you also have Tesla auto insurance), opt back in, and then drive 100+ miles on NAP (off peak hours if traffic is a concern). This should do it.

Haha off peak hours in LA means midnight runs essentially is it 100 miles on NAP + 100 driven regularly?

I'm at the point where I'm not driving the Model 3 anytime besides nighttime with great conditions to ensure I can finally get this stupid beta after years of waiting.

 

[EVNow]

I have a M3 in LA with a safety score of 81, which I feel is pretty good for LA.

How are so many others in your area getting 95%+ scores and are in Beta testing ?

 

[LowlyOilBurner]

So, no one on 2022.16.1 has gotten the beta as of yet?

 

[Hamilton237]

So, no one on 2022.16.1 has gotten the beta as of yet?

Doesn't seem like it. We can't know for sure that there isn't a car out there that had 2022.12.3.20 pushed after 2022.16.1 but none of the trackers show cars with that upgrade path and generally Tesla hasn't pushed branch 'downgrades' like 2022.16 back to 2022.12.

 

[bradbrok]

So, no one on 2022.16.1 has gotten the beta as of yet?

We're going to have to wait for 10.13.

 

[kbecks13]

How are so many others in your area getting 95%+ scores and are in Beta testing ?

It's not impossible, but its tough. In Los Angeles if you leave a 1-2 second gap in front of you on the highway, someone will quickly jump in front of you because they think that lane is faster. Following distance is tough for me to do well on and i now try to find empty highway to drive manually on (since it doesn't count AP miles for that one right?)

 

[LowlyOilBurner]

We're going to have to wait for 10.13.

You mean 10.13.2….

 

[TS.42]

I finally got in a week ago after a 1 month wait and 98-99% score. My thoughts are running through the same routes in my testing, the car does not behave the same each time, even though the traffic is basically the same each time I tested. In one test it handled turns with no problem. The next test it tried to run into curbs and gates. I will keep at it but I am reporting these strange behaviors. I am looking for consistency on a route to prove we are learning.

 

[bradbrok]

It's not impossible, but its tough. In Los Angeles if you leave a 1-2 second gap in front of you on the highway, someone will quickly jump in front of you because they think that lane is faster. Following distance is tough for me to do well on and i now try to find empty highway to drive manually on (since it doesn't count AP miles for that one right?)

Driving manually in this instance will continue to nuke your score, keep it on AP or TACC on freeway like all the time if possible. Also you need 100 AP miles minimum to get in now, so manual driving is pointless in general.

 

[VeeKay]

Haha off peak hours in LA means midnight runs essentially is it 100 miles on NAP + 100 driven regularly?

I'm at the point where I'm not driving the Model 3 anytime besides nighttime with great conditions to ensure I can finally get this stupid beta after years of waiting.

Minimum 100 miles on AP/NAP

 

[Hamilton237]

Driving manually in this instance will continue to nuke your score, keep it on AP or TACC on freeway like all the time if possible. Also you need 100 AP miles minimum to get in now, so manual driving is pointless in general.

Sort of... If you know you have an incident of close following, hard braking, or aggressive turning it may be advantageous to drive manually for the rest of the day. All of those factors are measured as a percentage of the "safe" driving behavior. (i.e. % of all braking greater than .3g) If you already have 20% hard braking due to an incident, adding manual driving with some normal braking will decrease that percentage for the day assuming you keep it under the .3g limit.

Relevant quote from the safety score page:
"The percentage shown in the app is the percentage of manual braking that is done with excessive force when driving and Autopilot is not engaged."

AP/TACC will add miles to the daily score but they won't add to the 'safe' duration for some of the individual factors. Starting a new day or if you are already at 100 for the day... definitely use AP as much as possible.

To fix unsafe following on the same day, you are right that you don't want to just drive and add miles, you need to find a car to follow at a safe distance while manually driving. Those miles will help improve your % in that factor.

This stuff matters less now that they are down to mid 90s. Keeping 100/99 was tough.

 

[Dewg]

I finally got in a week ago after a 1 month wait and 98-99% score. My thoughts are running through the same routes in my testing, the car does not behave the same each time, even though the traffic is basically the same each time I tested. In one test it handled turns with no problem. The next test it tried to run into curbs and gates. I will keep at it but I am reporting these strange behaviors. I am looking for consistency on a route to prove we are learning.

You need time and experience with Beta between updates over time. You'll see progress in some areas and regression in others, then the flip the next update. But over time (6 months+) you'll see a positive progression.

 

[Silicon Desert]

Driving manually in this instance will continue to nuke your score, keep it on AP or TACC on freeway like all the time if possible. Also you need 100 AP miles minimum to get in now, so manual driving is pointless in general.

Uh, keep it on AP. The 100 AP miles is actually not new...... just that it was not properly communicated by Tesla in the past.

 

[kbecks13]

Relevant quote from the safety score page:
"The percentage shown in the app is the percentage of manual braking that is done with excessive force when driving and Autopilot is not engaged."

AP/TACC will add miles to the daily score but they won't add to the 'safe' duration for some of the individual factors. Starting a new day or if you are already at 100 for the day... definitely use AP as much as possible.

To fix unsafe following on the same day, you are right that you don't want to just drive and add miles, you need to find a car to follow at a safe distance while manually driving. Those miles will help improve your % in that factor.

Thank you (and others) for adding the details, i have a way better understanding of how they are averaging and calculating things and feel i have a MUCH better shot at getting good scores now. It's a shame it requires this sort of gamification to get the beta but i guess it is what it is, just get better at playing the game

 

[BnGrm]

Thank you (and others) for adding the details, i have a way better understanding of how they are averaging and calculating things and feel i have a MUCH better shot at getting good scores now. It's a shame it requires this sort of gamification to get the beta but i guess it is what it is, just get better at playing the game

Once I internalized the rules of the game my score shot up considerably. From like 84 to 98. BUT I didn’t realize I needed 100 miles of AP. I was doing it all manual because I wanted complete control and didn’t trust what would be scored against me during the transitions. So now I’m trying to get the AP miles….