NOA with 2.5 hardware

[golfpilot]

I have EAP and hardware 2.5.
If I want to change lanes and actually hit the blinker the car moves over right away. When I am on NOA there is still a 3-4 blinker delay. My car also no longer moves out of the passing lane on its own.
Are these annoyances an attempt to get people to buy the FSD chip or do you think they will be fixed with updates?
All 3 other cars in my family have the hardware 3 chip and don’t have these annoyances.

 

[Tam]

...My car also no longer moves out of the passing lane on its own...hardware 3 chip and don’t have these annoyances.

That's quite interesting that you observe there's a difference between Hardware 3 VS 2.5 in terms of automatically moving out of the passing lane.

I doubt there's one.

In the beginning, the system used to be very enthusiastically switching lanes like crazy. It didn't make sense after passing a car then it immediately get out of the passing lane then immediately go back to the passing lane again because there's another car in front after it filled in that gap.

Now, after passing a car, and if there's a gap, it might wait and pass that gap because there's another lines of cars in front of that gap.

I think it is more dependent on the algorithm progression/maturity, firmware update versions and not the chips.

 

[thewishmaster]

Isn't the delay before moving on NoA (I counted 6 blinks) part of the programming for all cars? Even with HW2.5 it can definitely make quicker lane changes in certain conditions (e.g. short off-ramps) and like you said manually initiated lane changes are pretty much instant.

Not sure why they changed behavior with passing lanes though - mine also doesn't move out of the way when it did previously.

 

[Enginerd]

I've been tracking NOA lane change performance for a while now, and inability to exit the passing lane is definitely a new thing. It's clearly firmware-version dependent. Prior to 2019.40.*, it only happened rarely. I had not heard that HW3 is any different... are you absolutely certain? I've paid for FSD, and I'm waiting on my hardware upgrade.

In the NOA settings, I've selected no-confirmation lane changes, which also shortens the delay. The delay between turn signal and steering command was reduced in FW 2019.40.* from 6-7 blinks to 3-4 blinks. It still makes sense to announce your intentions before maneuvering, although I would prefer 2-3 blinks in most situations.

 

[DirtyT3sla]

That's quite interesting that you observe there's a difference between Hardware 3 VS 2.5 in terms of automatically moving out of the passing lane.

I doubt there's one.

There most definitely is a difference. HW2.5 will not auto leave the passing lane, but it does with HW3. I think I was the first to note it in this video:

but others on TMC have also documented it: Navigate On Autopilot: automatic lane change results

OP, as for the delayed Nav on AP changes vs instant initiated changes, I think this is the same for everyone. I just don't think Tesla is at a spot that they're comfortable with the car making that quick of a lane change on its own without the driver telling the car to do it.

 

[tsh2]

There most definitely is a difference. HW2.5 will not auto leave the passing lane, but it does with HW3.

HW3 didn't do this when I picked up my car in November (UK), but it does so now. Or it it did, it was not very keen to offer.

 

[WilliamG]

There most definitely is a difference. HW2.5 will not auto leave the passing lane, but it does with HW3. I think I was the first to note it in this video:

but others on TMC have also documented it: Navigate On Autopilot: automatic lane change results

OP, as for the delayed Nav on AP changes vs instant initiated changes, I think this is the same for everyone. I just don't think Tesla is at a spot that they're comfortable with the car making that quick of a lane change on its own without the driver telling the car to do it.

Pretty sad that I can tell that video was HW2.5 just by your backup camera frame rate. So nice in HW2.5!

 

[GalacticHero]

I've been tracking NOA lane change performance for a while now, and inability to exit the passing lane is definitely a new thing. It's clearly firmware-version dependent. Prior to 2019.40.*, it only happened rarely. I had not heard that HW3 is any different... are you absolutely certain? I've paid for FSD, and I'm waiting on my hardware upgrade.

In the NOA settings, I've selected no-confirmation lane changes, which also shortens the delay. The delay between turn signal and steering command was reduced in FW 2019.40.* from 6-7 blinks to 3-4 blinks. It still makes sense to announce your intentions before maneuvering, although I would prefer 2-3 blinks in most situations.

Yes, it is very annoying that my car now camps out in the passing lane.

 

[thewishmaster]

The delay between turn signal and steering command was reduced in FW 2019.40.* from 6-7 blinks to 3-4 blinks. It still makes sense to announce your intentions before maneuvering, although I would prefer 2-3 blinks in most situations.

Strange; mine always starts moving right after the 6th blink even on 40.50.x. I started counting turn indicator blinks as a guide for gauging whether the car might cut off someone while on NoA, though thankfully it's gotten good enough at moving over that it rarely puts me in that situation. I am however using Mad Max mode ever since the "more aggressive lane changes" update. I wonder if there's any difference on the other settings.

 

[Rogue one]

I have a model 3 with HW2.5 and it camps in the passing lane, it’s gotten rather annoying.

 

[Clinton]

I’m having the exact same problem. I also have hardware 2.5 with EAP. Now the car stays in the passing lane and won’t get out of it unless I turn on the turn signal.

I have EAP and hardware 2.5.
If I want to change lanes and actually hit the blinker the car moves over right away. When I am on NOA there is still a 3-4 blinker delay. My car also no longer moves out of the passing lane on its own.
Are these annoyances an attempt to get people to buy the FSD chip or do you think they will be fixed with updates?
All 3 other cars in my family have the hardware 3 chip and don’t have these annoyances.

 

[SunCharged]

I’m having the exact same problem. I also have hardware 2.5 with EAP. Now the car stays in the passing lane and won’t get out of it unless I turn on the turn signal.

My hw 2.5 stays in the passing lane also. But the most annoying part of NOA is when it prompts a lane change “to follow route” when that lane change is unnecessary. I live in a relatively small metropolitan area in MI and there are so many instances of this NOA is completely useless until you get out of town. And no matter how many times you cancel the lane change it just keeps popping up and activating the turn signal over and over again. Drivers behind me must think I’m completely crazy or intoxicated!
I have seen no improvement with this with continuous bug reports submitted. It’s actually becoming worse as time goes on.

 

[Tam]

...I have seen no improvement with this with continuous bug reports submitted...

To alleviate the frustration while working with a beta program, I just learn what are its good and bad behaviors and find a way to work with the system.

Thus, for NOA, I use it as a hybrid between Autopilot alone and NOA also.

When I am at an entrance ramp, I would turn on the NOA and let it do some initial lane changes. Once that is done, I turn the NOA off and let the Autopilot alone takes over. I then use my own judgment to toggle the turn signal as needed. Once I am about a mile or two from a desireable interchange or exit, I would then turn back the NOA on.

Very relaxing for me!

 

[SunCharged]

To alleviate the frustration while working with a beta program, I just learn what are its good and bad behaviors and find a way to work with the system.

Thus, for NOA, I use it as a hybrid between Autopilot alone and NOA also.

When I am at an entrance ramp, I would let turn on the NOA and let it do some initial lane changes. Once that is done, I turn the NOA off and let the Autopilot alone takes over. I then use my own judgment to toggle the turn signal as needed. Once I am about a mile or two from a desireable interchange or exit, I would then turn back the NOA on.

Very relaxing for me!

I have been doing this as well lately. It does make a big difference. I do wish NOA worked well enough to just use as intended . If I were buying a new Tesla today I would not purchase FSD and just use the standard autopilot that comes standard.

 

[NHK X]

There most definitely is a difference. HW2.5 will not auto leave the passing lane, but it does with HW3. I think I was the first to note it in this video:

but others on TMC have also documented it: Navigate On Autopilot: automatic lane change results

OP, as for the delayed Nav on AP changes vs instant initiated changes, I think this is the same for everyone. I just don't think Tesla is at a spot that they're comfortable with the car making that quick of a lane change on its own without the driver telling the car to do it.

I have HW 2.5 and it used to take me out of the passing lane, for quite a while. Like the OP, I’ve noticed it doesn’t do that anymore. I do have FSD purchased.

 

[Didgie]

I have FSD & HW2.5 on UK MS and use the data Enginerd produces in my videos (useful graphs, see 13:30 into the video). This video shows overtaking lane holding, speed matching cars on the inside lane and swift lane changes on a dual carriageway.

Tesla Software 2019.40.50.1 - Christmas Update

 

[kcdk99]

Any update on this from anyone? I have a late 2017 HW 2.5 Model X. Just installed 2020.12.5 and it still won’t automatically move out of the passing lane. This previously worked on this car a few versions ago. A bit aggravating.

 

[LowlyOilBurner]

I have a FSD Y HW3 running 2020.5.10.7 and here’s what I notice:

-When it decides to change lanes (from slower lane to passing lane) it puts the blinker on and takes a while to move over. Longer than I would wait to make the move. This should be fixed, and maybe it is on newer software versions?

-Once in the passing lane, sometimes it will move out of it quickly, other times it hangs out there forever. No rhyme or reason to it that I can tell?

 

[Enginerd]

Any update on this from anyone? I have a late 2017 HW 2.5 Model X. Just installed 2020.12.5 and it still won’t automatically move out of the passing lane. This previously worked on this car a few versions ago. A bit aggravating.

I'm pretty sure that for HW 2.5 cars, the "exit passing lane" feature broke with 2019.40.2.1. To the best of my knowledge and experience to date, nothing fixes it except the HW3 upgrade. I passed along my data and bug report to the service rep when I had my FSD upgrade. I'm hoping that the rumored core AP rewrite fixes this bug for everyone, including HW 2.5 cars.