Navigate On Autopilot: automatic lane change results

[navguy12]

I went back to an older counterweight that's better sealed, smaller, sits right in the "hollow" (so a little further out on the cross-arm towards the wheel itself), and only weights 100 grams ~ 3.5 Oz. Seems as effective as the bigger one. No significant hunting, nothing I'd call "ping-ponging".

As I mentioned, with 2019.40 I think the car drives more decidedly, more dynamically, much more like a human, and isn't as oriented to always keeping dead center in the lane, but that doesn't bother me at all.
.

I think there is something to the argument that the software is no longer planting the car in the middle of a lane under all circumstances.

On a freeway drive yesterday, the car definitely gives a wide berth to large pickup trucks that are passing the car on the left.

However, this version is most definitely ping ponging on straight sections of road with no other traffic around.

Also, it wants to arbitrarily move to the left lane whilst approaching an interchange, which would make sense if there was any traffic in the area.....but even at deserted interchanges, this car wanted to always switch to the left lane.

So, I call this contextual lane positioning V1.0.

 

[Dithermaster]

I find I'm hoping for more improvements as well but when stopped on autopilot behind another car. I find it starts to move still too slowly and at least now guns it to catch up but it could just start earlier and be less heavy on the accelerator in most cases.

Agreed! I waits too long, and then guns it. I'm a fast driver and like acceleration, but Autopilot accelerates too quickly for my "daily driving" taste. I'd prefer riding with a chauffeur rather than a teenager on Redbull.

I have a technique to make it better: when the car ahead starts moving (for real, not just inching ahead) press slightly on the accelerator to start moving. A couple seconds later AP takes over and because you already rolled some, it doesn't need to accelerate quite as much, and it's an overall much smoother experience.

Personally I'd rather it sit tight for a second rather than rabbit along and rear-end or panic-stop to avoid rear-ending someone who changed their mind.

A second, sure, I agree. But 2 seconds is "too long" by human driver standards, and surely has the car behind almost ready to honk. And then (due to the delay) the amount of acceleration it uses to catch back up is too much. Tesla needs to hire a professional driver to teach them what good driving is like.

 

[Fernand]

It's great that we're at the point of discussing subtleties. When I think about what it was a year ago, and what it will be next year, I'm really happy to be witnessing this development. Like watching the first space-walks, the first CAT scans. We can remind our grandchildren how cars weren't always as smart as what they're driving, and we were there when a small team of engineers made it happen.
.

 

[Didgie]

Just reviewed 2019.40.50.1 - From my testing so far it's a definite AP/NoAP improvement, ping ponging gone (maybe a little in the bends), still holding passing lane a bit but it felt much more sorted overall

I touch on Enginerd's most recent test results, although holding passing lane saw a sharp decline everything else looks better

Is anyone else seeing it better or worse?

 

[Fernand]

2019.40.50.7 Looks very good to me. But I'm not counting marbles in a commute, I'm looking at "can I leave it in NOA all the time", with the subjective impression that it's passed the threshold of being better than average humans.

Humans abort lane changes, weave in and between lanes, and do all sorts of high risk stuff all the time. Nicki doesn't do much of that, s/he does have occasional moments of indecision, but I'm worse, so I'm fine with the way s/he drives.
.

 

[DopeGhoti]

Has anyone seen 40.50.x leave the passing lane on its own in order to leave the passing lane, rather than to follow the route, avoid cones, or get past a slower vehicle?

 

[navguy12]

Has anyone seen 40.50.x leave the passing lane on its own in order to leave the passing lane, rather than to follow the route, avoid cones, or get past a slower vehicle?

I have not, although I have NOA set to require me to confirm via the signal stalk so as to avoid a lot of these "Left Lane Larry" moves.

Even on a deserted, middle of the night scenario, when on a two lane section of (HWY 401) freeway, when approaching an interchange, this car wants to move to the left.

I view this as a software "cheat" where the system is preemptively getting out of any merging traffic scenarios even if no one is around......

........and when I stay in the right lane and enter the zone where the lane widens and splits off for an exit and car begins to wander as it searches for a lane center (as it has for 19 months of ownership).....

.......I now always take control of steering because once the lane width returns to normal after the split/merge, the over correction ping ponging as the car re aquires center is too uncomfortable for my passenger.

 

[Phil Seastrand]

Has anyone seen 40.50.x leave the passing lane on its own in order to leave the passing lane, rather than to follow the route, avoid cones, or get past a slower vehicle?

It happened to me today on version 40.50.7.

 

[DopeGhoti]

........and when I stay in the right lane and enter the zone where the lane widens and splits off for an exit and car begins to wander as it searches for a lane center (as it has for 19 months of ownership).....

This, at least for me, has gotten a lot better in the incremental versions since "V10" came out. I've also noticed AP being 'willing' to let me 'guide' the car in these situations. I'm one of those oddballs that actually keeps a grip on the wheel as instructed, and when the lane would widen and Autosteer would try to seek center, I would hold the wheel in place. After a while, Autosteer would stop disengaging and also stop even trying to pull away from the left lane line in these situations.

 

[DirtyT3sla]

Has anyone seen 40.50.x leave the passing lane on its own in order to leave the passing lane, rather than to follow the route, avoid cones, or get past a slower vehicle?

I've had it get out of the passing lane 2 times on the 2019.40.x versions, but it was long before 40.50... it's definitely still broken. Hopefully a week number update will fix it

 

[Didgie]

I've had it get out of the passing lane 2 times on the 2019.40.x versions, but it was long before 40.50... it's definitely still broken. Hopefully a week number update will fix it

Strange, I have been testing this with my Model S on 40.50.7, it is not holding the passing lane much at all, often highlighting the driving lane (in UK we don't get auto lane change) within 30 seconds.

 

[DirtyT3sla]

Strange, I have been testing this with my Model S on 40.50.7, it is not holding the passing lane much at all, often highlighting the driving lane (in UK we don't get auto lane change) within 30 seconds.

Are you on HW3? It doesn't have this bug.

 

[DopeGhoti]

Are you on HW3? It doesn't have this bug.

Some HW3 cars clearly do. The only things that will make my car exit the passing lane on its own are:

• Routing
• Slower traffic in the passing lane
• Cones

 

[DirtyT3sla]

Some HW3 cars clearly do. The only things that will make my car exit the passing lane on its own are:

• Routing
• Slower traffic in the passing lane
• Cones

You car sees cones and won't get out of the passing lane?

 

[DopeGhoti]

I know I have HW3 because I now see cones, stoplights, stop signs, turn lane markings, etc. on the MFD.

My car will happily park in the leftmost lane indefinitely. Not once have I seen "exiting passing lane" as the MFD-displayed reason for the lane change. The only reasons I have ever seen are: "Changing lanes to follow route", "Moving into faster lane", and "Moving away from traffic cones".

 

[DirtyT3sla]

I know I have HW3 because I now see cones, stoplights, stop signs, turn lane markings, etc. on the MFD.

My car will happily park in the leftmost lane indefinitely. Not once have I seen "exiting passing lane" as the MFD-displayed reason for the lane change. The only reasons I have ever seen are: "Changing lanes to follow route", "Moving into faster lane", and "Moving away from traffic cones".

Wow. First time I've heard of it happening on HW3

 

[Didgie]

Are you on HW3? It doesn't have this bug.

I’m on HW2.5 MCU2

 

[Didgie]

I’m getting good AP performance on 40.50.7. Just went from home to Shakespeare’s home town, 31mins driving, all on country roads, villages and town, 87% on AP (No lane changes). Pretty good going in my opinion.

Tesla Update 2019.40.50.7 - Take me to Shakespeare 100% on AutoPilot!

 

[DirtyT3sla]

Hey everyone, just wanted to put the final nail in the coffin for "HW2.5 doesn't exit passing lane, HW3 does exit passing lane," bug as I still have some doubters.

I picked up my Model 3 with HW3 yesterday morning, and SEVERAL times on the first drive using NoA I saw the exiting passing lane message. It was still 2019.40.50.7. The only difference was the HW3 upgrade. Here's even a pic I was so excited

 

[drtimhill]

I've not heard any reports from HW2.5 owners if 2020.4.1 makes any difference yet. I wonder if this has been addressed?

(I have 2020.4.1. but also HW3).