Safety Score

[serendipitous]

Same thing happened to me this morning on the way to work. 4:00am no cars on the street to follow, used NoA for 9 miles on freeway. Got off freeway, no cars to follow, pulled into parking lot and checked my score, 50% unsafe following. Clearly Tesla is counting AP and NoA in this bogus scoring system

It's not counting, but your usage of it is counterintuitively hurting you. If the only time you are following someone is while on AP, you're not getting any "following seconds" racked up. If you're on AP following somebody at 0.9s, then you disengage AP, it only just now starts counting your seconds and you're getting dinged. If you immediately get over a lane and take an exit, now you're not following anyone, so even if you were behind that guy for an hour and a half on AP at any distance you like, your score is going to be "1 second of following too closely" divided by like "2 seconds of total following" - 50% unsafe. You need to drive a bunch of non-AP miles WHILE FOLLOWING someone at 1-3s to rack up time spent following safely. Similarly, if your commute has a single turn, you can go over the hair trigger threshold on that one turn, then drive safely for 100 miles and park your car, and you had 100% "turning with extreme ( ) force". You have to game the system unless you drive like a grandma or have an oddly straight and uncongested route.

 

[momo3605]

Freeway or city AP ? I think most agree freeway AP is not counted.

There is nothing in the documentation that suggests using AP in the city or freeway would make any difference.

 

[EVNow]

There is nothing in the documentation that suggests using AP in the city or freeway would make any difference.

Which documentation ? They do say when "appropriately" used, AP won't count. The question is what exactly is "appropriate".

There are also several reports of being dinged for stuff happening when using AP. So, definitely its not all AP they are excluding.

 

[Robocheme]

I've basically stopped using Autopilot this past week unless the highway is straight with nobody close to follow. I normally have it active for 99% of miles even on city streets, but after getting the max hard braking score on my first trip, I've realized that any potential scoring events are excluded when on Autopilot. This means in the 1% where I drove without Autopilot for braking on a residential street on a hill with no lane lines, the denominator was basically only that braking scenario. So by not using Autopilot, I have more "normal" braking patterns throughout the day to maximize the denominator, so if there was actually anything in the numerator, it'll be easier to fix up.

I thought that we weren't even supposed to use AutoPilot on city streets.

I used it on the freeway today and I got dinged for hard braking. The only time that I could think of where I braked hard is when I purposely disengaged Autopilot when I exited the freeway. I didn't have my foot on the accelerator so the regen took over and dropped the speed pretty fast.

 

[FalconFour]

My example of using a quick split second engagement of AP to mask a hard braking event for a yellow light is an example of city AP usage. In my experience all AP is masked. I really think folks that think AP is what dinged them are really unaware of the events around their engagement that led to it. You can't get a score without turning the car on, driving a bit, enabling AP, disengaging AP, driving a bit more, then parking. The manual driving on both ends is what'll get ya, at minimum

But since the score is only shown per drive, there's no way to accurately sort it out. It's at least documented, and well proven, that AP usage masks all faults and thus can't create any of its own.

It'll also not count as positives in your stats either... and the rules can be confusing and overly specific -- having to be in specific windows, like "1-3 seconds but not more than 3 seconds following", "0.1 to 0.3g but not less than 0.1g braking", and "0.2 to 0.4g turning but not less than 0.2g".

 

[Charliek]

I think the gaming level you need ( I read this some where in a non EV board) was driving like a "grandma on valium". I got it up to a 98 ( the score not the speed) then on a trip I had to go across three lanes to get to HOV ( left lane) at rush hour and it dinged me for the lane changes, and then for coming out of HOV to my exit. it wasnt doing any thing unsafe just normal heavy traffic . if I had let NOA do it it would have put me in the slow lane 2 miles before and annoyed the heavy traffic slowing down and waiting for an opening. I did move my spacing number in autopilot from 3-4 and it seems to help the score for too close following significantly . but I could be getting paranoid , like a teenager with the driving instructor/ driving test sitting there taking notes.

 

[momo3605]

Which documentation ? They do say when "appropriately" used, AP won't count. The question is what exactly is "appropriate".

There are also several reports of being dinged for stuff happening when using AP. So, definitely its not all AP they are excluding.

Safety Score Beta | Tesla Support

The Safety Score Beta is an assessment of your driving behavior based on several metrics called Safety Factors. These are combined to estimate the likelihood that your driving could result in a future collision. We combine your daily Safety Scores (up to 30 days) to calculate the aggregated...

www.tesla.com

Yes, it does say "appropriately" but that's probably just to cover their a**es. There's plenty of people claiming all sorts of things on this page. That AP in the city has no effect...or that AP on the freeway is dinging them left and right. You have people stating basically every variation possible lol.

If you don't overthink it, the safety score page says that the only thing you can do wrong on AP is a forced disengagement. So driving on AP on a straight road with traffic lights in the city or on a freeway won't make any difference.

 

[EVNow]

You have people stating basically every variation possible lol.

If you don't overthink it, the safety score page says that the only thing you can do wrong on AP is a forced disengagement. So driving on AP on a straight road with traffic lights in the city or on a freeway won't make any difference.

Let's add the last sentence to one of the variations ...

 

[momo3605]

Let's add the last sentence to one of the variations ...

yea...i don't know how to reconcile everyone's opinions. I haven't been dinged on anything by AP in the city or on the highway, but two posts before this @Robocheme says he was. I dunno, all i know is, maybe I have it easy lol. Max regen + AP is not an issue on my car.

 

[EchoDelta]

For those that want to have fun instead of sitting it out or complain, I have found a crummy but ok app that you can set to make noises upon hitting preset lateral and front/back g thresholds.
like a lot of sensor apps it’s no beaut but it adds a fun ding ding when hitting high enough braking gs (I have mine set at 0.25)

G FORCE

G FORCE makes your iPhone or iPod Touch a handy vehicle G-Force meter. A nice looking enhancement to the dash of your car, too. Quantify the G-Force you experience in the vehicle (e.g. during acceleration, braking, drifting, cornering etc) or perform a quick dyno. Its unique "Sound Alert feat…

apps.apple.com

No affiliation and caveat emptor

 

[AlanSubie4Life]

There are also several reports of being dinged for stuff happening when using AP. So, definitely its not all AP they are excluding.

I would definitely not say definitely.
We’ll probably never know, until someone has the latitude to properly experiment, which no one has right now.

 

[FalconFour]

I'll add that my "complete suite of tricks" is adding up pretty well... might be on track to rounding my 99 into a 100 after all. I got my first sight of 98 (for the day, not overall) earlier today, was scary! But now it's back to 99 for the day again. haha.

Very important point: actually driving (not AP) is what accumulates "good" points. Since AP masks everything, it masks both the good and the bad. So you kinda have to practice being a robot in order to get that perfect 100. I'm still getting the hang of when to use AP to mask other drivers' stupidity, vs. accumulating points by "flintstoning" (driving with your feet). lol

Yeah sex is great and all, but have you ever let fully off the accelerator at exactly the perfect time to fully regen down to a stop just behind stopped traffic at an offramp, without needing to hit neither the accelerator nor the brake? Oof yasss gimme those "safe braking" seconds~~

 

[fiveten]

Happy to say both my cars are now at 99s…if tomorrow I can do 250 miles on one of the cars at a 100 I can get to 100…on the other I would need to do like 2500 miles tomorrow, lol

 

[FalconFour]

Also, it's in this realization (that every stat is "windowed") that leads me to suggest switching off "chill" mode... lemme explain.

In chill mode, the throttle isn't weakened, but it's just smoothed... like the car is high, man. So when you slam it, you still get {{POWER}}, but it just doesn't happen with a wallop, it happens gradually. Same with regen - when you let off the pedal, it ~gradually~ gives you full regen power. Yeah, that's not great in many situations. If I want smooth regen, I'll smoothly move the pedal. So I turned off Chill mode and it seems easier to drive now.

Crappy mspaint skills time.

Green filled zone = "the happy region" that braking positively affects your score.
Red filled zone = "u fukt up son region" that braking negatively affects your score.

With normal driving mode, regen comes to full power near instantly, and you slow down at a constant rate (within the "happy window" of G's for Safety Score to count your seconds):

With chill driving mode, you end up wasting a lot of braking time outside the "happy window" where it doesn't count, and it takes longer to come to a stop:

So you're spending precious "good braking" seconds on nothing, and taking longer to stop. Worse still, if you misjudge your stop, you now need to hit the brake pedal as well:

Freehand drawing plot lines using a touchpad is not the easiest thing but I hope it gets the idea across.

 

[Driver08]

3. Gauge your daily activity or lack there off - If you are busy drop off your kids, running thru drive thru. You are probably will not score 100. Those unsafe following dings happens often when you are bumper to bumper but could not use your AP for the purpose. This weed out the busy bees and give the ones that has ample of time to actually doing the FSD testing.

No - unsafe following dings only happen when you are going over 50mph.

 

[mhan00]

I wonder why speeding wouldn't be one of the metrics for the safety score?

Speeding honestly doesn’t matter much, imo. Much more important is speed differential between you and the rest of traffic. So long as you’re going with the flow of traffic (which would be presumably reflected in the follow distance and hard braking metrics), you’re far safer than someone who is going much faster or slower than traffic.

 

[SarahsDad]

I feel sorry for you guys who have to commute in a busy urban environment. I consider myself a good driver and was rocking a 99 after 4 days but I had to drive into the city during rush hour today (30 miles took 70 minutes) and despite being on autopilot 80-90% of the time and driving as cautiously as possible, I got a score of 81, which dropped my overall to 93. And it was one of the most stressful drives I've had in a Tesla because I was trying so hard keep my score up. Two cars ahead of me both ran a red light, autopilot tried to usher me through but I braked late and got hit with Hard Braking. A woman was walking close to the road, all lanes bumper to bumper so I couldn't give her any room, and got dinged for FCW. And the few times I wasn't on Autopilot, several cars cut in front of me (because I was trying to maintain the 3 second rule), and I think that triggered Unsafe Following. Getting a very high score (>98) is likely impossible in heavy rush hour traffic. Jeez this is too much work. I'm going back to just driving safely and not caring about the score or when (or if) I ever get FSD Beta. Best of luck to the rest of you.

 

[mltv]

I feel sorry for you guys who have to commute in a busy urban environment. I consider myself a good driver and was rocking a 99 after 4 days but I had to drive into the city during rush hour today (30 miles took 70 minutes) and despite being on autopilot 80-90% of the time and driving as cautiously as possible, I got a score of 81, which dropped my overall to 93. And it was one of the most stressful drives I've had in a Tesla because I was trying so hard keep my score up. Two cars ahead of me both ran a red light, autopilot tried to usher me through but I braked late and got hit with Hard Braking. A woman was walking close to the road, all lanes bumper to bumper so I couldn't give her any room, and got dinged for FCW. And the few times I wasn't on Autopilot, several cars cut in front of me (because I was trying to maintain the 3 second rule), and I think that triggered Unsafe Following. Getting a very high score (>98) is likely impossible in heavy rush hour traffic. Jeez this is too much work. I'm going back to just driving safely and not caring about the score or when (or if) I ever get FSD Beta. Best of luck to the rest of you.

View attachment 716447

I feel your pain. I just spent some early morning time just to boost my score up. I absolutely disagree that Unsafe Following is only apply to > 50 miles. Got plenty of this and I rarely do > 50 without on AP.

 

[fiveten]

I feel sorry for you guys who have to commute in a busy urban environment. I consider myself a good driver and was rocking a 99 after 4 days but I had to drive into the city during rush hour today (30 miles took 70 minutes) and despite being on autopilot 80-90% of the time and driving as cautiously as possible, I got a score of 81, which dropped my overall to 93. And it was one of the most stressful drives I've had in a Tesla because I was trying so hard keep my score up. Two cars ahead of me both ran a red light, autopilot tried to usher me through but I braked late and got hit with Hard Braking. A woman was walking close to the road, all lanes bumper to bumper so I couldn't give her any room, and got dinged for FCW. And the few times I wasn't on Autopilot, several cars cut in front of me (because I was trying to maintain the 3 second rule), and I think that triggered Unsafe Following. Getting a very high score (>98) is likely impossible in heavy rush hour traffic. Jeez this is too much work. I'm going back to just driving safely and not caring about the score or when (or if) I ever get FSD Beta. Best of luck to the rest of you.

View attachment 716447

Scroll wheel reset is your friend on tough drives

 

[stonelance]

How do you think they measure turning G forces? I'd bet ya anything there's a simple accelerometer in there. They're so dirt cheap and in every phone, it'd cost pennies to just throw one into one of the ECUs.

Yeah I'm sure they have an accelerometer in there somewhere, but you can easily calculate cornering force based on wheel angle\corner radius and speed (assuming the tires are not slipping)