Just got new update: 2017.42 a88c8d5

[rnortman]

Agree, not changed. My car has behaved like this since march. All through summer (we have some cold summers).

It would be useful if you could mention the actual ambient temperatures where you start to see limited regen. As I mentioned earlier, I'm seeing it at 15C ambient when the car was recently charging at 6kW, so battery temperature should have been higher than that. This does not seem normal to me. Are you saying that it's normal for regen to be limited at 15C ambient with a battery that was warmed up within the past couple of hours? (And without a full battery, obviously... I only charge above 75% when I'm going on a long trip.)

I am a new owner, so I don't have a huge amount of experience to go on, but these aren't the first chilly days we've had this season, and I did not notice this before this firmware update.

 

[conman]

I just took this release out for my usual night time winding road test which gives me nice objective endpoints to determine progress. As per usual I will upload the video (which takes me ages so I'll link it here tomorrow) but... this is by far and away the best performing release ever on that run. It's still far from perfect - it has difficulty with cresting hills and when curves are very tight and go on for an extended period, it fails to stay within the lane. However, for the rest of the winding road it did a solid job of staying within the lane. It felt confident and was actually quite enjoyable during the less difficult stretches of road where previous ones felt apprehensive and made me driving feel very anxious.

Easier urban roads weren't all perfect. During straight intersections it confidently held its line when the markings would disappear, but on intersections in curves or across hill crests it would veer off and then abruptly head back into the lane once it rediscovered the lines. In well marked areas it was again very confident and seemed to have less phantom braking than the last release which did it a lot. Taking off at traffic lights still remains glacially slow for my liking even with car distance set to 1, but is probably slightly faster than the last release.

The highway behaviour (not filmed) was quite different again. 40.1 was already a lot better on curves on the highway and this continues the trend but adds a certain level of confidence to its movements. It seems to accelerate a lot more when cars move away from in front or if you change lanes. I suspect some people prefer the lazier behaviour of the old one whereas I much prefer this one. Again I was quite enjoying having it drive and sense the confidence in its motions. Pulling up to stationary cars after taking an exit lane was not great though with it still missing them and me braking before there was any indication it could see them.

All of this testing was done at night.

 

[daktari]

It would be useful if you could mention the actual ambient temperatures where you start to see limited regen. As I mentioned earlier, I'm seeing it at 15C ambient when the car was recently charging at 6kW, so battery temperature should have been higher than that. This does not seem normal to me. Are you saying that it's normal for regen to be limited at 15C ambient with a battery that was warmed up within the past couple of hours? (And without a full battery, obviously... I only charge above 75% when I'm going on a long trip.)

I am a new owner, so I don't have a huge amount of experience to go on, but these aren't the first chilly days we've had this season, and I did not notice this before this firmware update.

Other Norwegian members/ users have read battery temp to 17 C when regen is reduced. This was from November 2016.

 

[Veritas1980]

It would be useful if you could mention the actual ambient temperatures where you start to see limited regen. As I mentioned earlier, I'm seeing it at 15C ambient when the car was recently charging at 6kW, so battery temperature should have been higher than that. This does not seem normal to me. Are you saying that it's normal for regen to be limited at 15C ambient with a battery that was warmed up within the past couple of hours?

It is perfectly normal behavior from the car. I park in a heated parking garage where the temperature in winter ranges from 13-16°C and I get limitations when charging at 4-7 kW, when at 11 kW the pack usually heats up enough for them to be completely gone.
So I'm not sure if 6 kW is enough to remove them.

It's just part of having a Tesla in a climate, that is not California.

Anyway, there is a lot of info around the forum regarding this, so you can easily do more research on it. Also so we don't bring this thread too much off topic.

 

[MP3Mike]

It would be useful if you could mention the actual ambient temperatures where you start to see limited regen.

They can't because it isn't based on ambient temperature. It is based on the temperature of the battery pack. Which you can't find out unless you tap into the CAN bus. (Or you have a P100DL and you put it in ludicrous+ mode.)

 

[calisnow]

Just got .42 today after having .40 for 10 days or so. Not sure what is improved over

I just took this release out for my usual night time winding road test which gives me nice objective endpoints to determine progress. As per usual I will upload the video (which takes me ages so I'll link it here tomorrow) but... this is by far and away the best performing release ever on that run. It's still far from perfect - it has difficulty with cresting hills and when curves are very tight and go on for an extended period, it fails to stay within the lane. However, for the rest of the winding road it did a solid job of staying within the lane. It felt confident and was actually quite enjoyable during the less difficult stretches of road where previous ones felt apprehensive and made me driving feel very anxious.

Easier urban roads weren't all perfect. During straight intersections it confidently held its line when the markings would disappear, but on intersections in curves or across hill crests it would veer off and then abruptly head back into the lane once it rediscovered the lines. In well marked areas it was again very confident and seemed to have less phantom braking than the last release which did it a lot. Taking off at traffic lights still remains glacially slow for my liking even with car distance set to 1, but is probably slightly faster than the last release.

The highway behaviour (not filmed) was quite different again. 40.1 was already a lot better on curves on the highway and this continues the trend but adds a certain level of confidence to its movements. It seems to accelerate a lot more when cars move away from in front or if you change lanes. I suspect some people prefer the lazier behaviour of the old one whereas I much prefer this one. Again I was quite enjoying having it drive and sense the confidence in its motions. Pulling up to stationary cars after taking an exit lane was not great though with it still missing them and me braking before there was any indication it could see them.

All of this testing was done at night.

Oh so you're saying you found noticeable differences in driving performance between 40.1 and 42? I didn't realize there were any between those two releases. I got 42 now but 40 was so good I haven't been looking for changes with 42.

 

[chillaban]

Just got .42 today after having .40 for 10 days or so. Not sure what is improved over
Oh so you're saying you found noticeable differences in driving performance between 40.1 and 42? I didn't realize there were any between those two releases. I got 42 now but 40 was so good I haven't been looking for changes with 42.

Allegedly even .42 has a slightly tweaked neural net compared to .40 so yeah there is some difference. Crazy how much seems to be changing when we went months without a vision change.

 

[ishareit]

I have been pretty happy with my EAP performance recently. It performs very well in most of the situations.

 

[Gwgan]

Browser seems snappier in my X, 2 bars LTR

 

[conman]

Just got .42 today after having .40 for 10 days or so. Not sure what is improved over
Oh so you're saying you found noticeable differences in driving performance between 40.1 and 42? I didn't realize there were any between those two releases. I got 42 now but 40 was so good I haven't been looking for changes with 42.

Absolutely. I do the same difficult test in the same conditions to test it and video the results. There are hard endpoints I can see if there's a difference - how often it crosses or rides the lines, how often it crosses over to the other side of the road needing intervention, how many times it aborts entirely. Subjectively its movements also feel far more confident.

 

[daktari]

Hard endpoints for me is deaths. Good it is not for you! ;-)

 

[ecarfan]

Last night I received the latest firmware being discussed here. I have EAP. Today I tried AP and TACC in the following scenarios:

280 freeway from San Mateo to Page Mill Road (California, San Francisco peninsula area): Within a minute after activating AP on the freeway I realized that AutoSteer was noticeably more stable and was maintaining lane position better whether the road was straight or curved. This was at 0800 and sometimes straight into the sun which was fairly low, a challenging situation for the cameras.

Hwy 35 (Skyline Blvd) from Page Mill Road to Hwy 92: This is a narrow curvy two-lane road, I used TACC for most of it (not AP), set to 45 or 50mph, and was impressed by how the car automatically slowed down for tighter curves and then accelerated after the curve back up to the speed setting. It was really remarkable how good it was! It wasn’t perfect, but it was very good. I’ve driven this road over a hundred times (never used TACC on it before) and in most cases the car took the curves at speeds that I was comfortable with. Sure I would have taken some of the curves somewhat faster, but in general the car got the speed right in a curve.

But here is the really amazing part: multiple times, the car started to slow for a tight curve BEFORE the lane lines on the drivers display showed the curve. The lane lines shown on the display would be straight lines and the curve about a hundred feet away or more and at that distance the car would begin slow down to a speed that I felt was very appropriate for the curve.

I had not set a destination in the navigation so the car was not following a navigation route. Again, this was with only TACC on, not AutoSteer.

It was as if the car knew the turn was coming up and how severe the curve was without the drivers display indicating that the car was visualizing the curve through the forward cameras.

Is this evidence that Tesla is now using onboard high-precision mapping? Or is TACC using Google Maps to anticipate curves based on the known position of the car on the map? Or both? I don’t know.

Maybe others have already seen this TACC behavior before the latest firmware release (2017.42). I have seen hints of it on 2017.40 but nothing like what I experienced today.

 

[conman]

I had a new experience today with 42 on my AP2 car. I used it for as much of an urban trip as possible and it was really quite respectable at all times. The only time I would disengage autosteer would be for changing lanes since it will not recognise multilane roads at lower speeds. But here's the bit that was new: I was in the left most lane in a 3 lane road with a truck in the right most lane. I disengaged autosteer by turning the steering wheel to move right and as I moved into the middle lane and started approaching the truck in the right lane I got a warning I hadn't heard before - there was an orange warning box on the display (instead of the red one when autosteer disengages) saying "take over *something*" I couldn't make out the last word since I was busy watching the road but the car actually steered itself a little to the left again before the warning went away and it stopped. As I said, autosteer had been disengaged by this time. Has anyone else ever experienced this?

 

[TaoJones]

After the radar improvements post-Florida, and certainly with my currently antiquated version (*.34), I've noticed better curve recognition and negotiation - presumably as a function of that forward radar prior to seeing any IC display updates.

With that said, if ever I am blessed with *.42, I look forward to evaluating any potentially positive changes.

How helpful it would be to have a synopsis in the release notes so that we wouldn't have to crowdsource our suppositions.

 

[conman]

How helpful it would be to have a synopsis in the release notes so that we wouldn't have to crowdsource our suppositions.

It would indeed, but alas that's just now hot current Tesla works

As promised, here is the video of me doing the winding road course at night time.

 

[chillaban]

I had a new experience today with 42 on my AP2 car. I used it for as much of an urban trip as possible and it was really quite respectable at all times. The only time I would disengage autosteer would be for changing lanes since it will not recognise multilane roads at lower speeds. But here's the bit that was new: I was in the left most lane in a 3 lane road with a truck in the right most lane. I disengaged autosteer by turning the steering wheel to move right and as I moved into the middle lane and started approaching the truck in the right lane I got a warning I hadn't heard before - there was an orange warning box on the display (instead of the red one when autosteer disengages) saying "take over *something*" I couldn't make out the last word since I was busy watching the road but the car actually steered itself a little to the left again before the warning went away and it stopped. As I said, autosteer had been disengaged by this time. Has anyone else ever experienced this?

This isn’t new but it is rare to hit: the logic is, if you disengage AutoSteer and don’t make any significant steering input, then the car sees you drifting out of your lane, it will sound an alarm while still effectively giving you AutoSteer until it senses you taking over.

It was introduced around 8.0 to address accidental disengagement leading to accidents.

 

[Cyclone]

For those with the limited regen and temperature questions, please note that if your car is in “Range Mode”, it changes the heating targets of the charger and significantly reduces battery heating in cool climates (though very cold climates are typically cold enough to make no noticeable change).

 

[Cyclone]

It was introduced around 8.0 to address accidental disengagement leading to accidents.

Yup. An owner tried to blame AutoPilot when he wrecked (I think it was an X) because the vehicle didn’t curve with the road and instead went straight. Tesla came back with the logs that show AutoPilot had disengaged like 30-60 seconds prior and he was just going straight because of good alignment.

 

[calisnow]

But here is the really amazing part: multiple times, the car started to slow for a tight curve ... about a hundred feet away or more and at that distance the car would begin slow down to a speed that I felt was very appropriate for the curve.

I had not set a destination in the navigation so the car was not following a navigation route. Again, this was with only TACC on, not AutoSteer.

It was as if the car knew the turn was coming up and how severe the curve was without the drivers display indicating that the car was visualizing the curve through the forward cameras.

Is this evidence that Tesla is now using onboard high-precision mapping? Or is TACC using Google Maps to anticipate curves based on the known position of the car on the map? Or both?

I wondered the same thing when driving .40 in very high glare, low contrast freeway cruising last week that would have tripped up .34 so badly I'd have to take over so folks wouldn't think I was drunk. The fact that the car tracked through intersections so well also made me wonder if perhaps there is some mapping going on.

When AP1 started behaving much better back in mid-2016 on similar into-the-sun-high-glare-low-contrast-concrete-with-poor-lane-markers on the 10 free way in So Cal I also began to suspect that the crowd sourced precision maps Tesla claimed the AP1 cars were gradually building - had been activated in some way. I became more convinced when one day my AP1 car slowed way down for a sharp left 90 degree freeway interchange (10 east to 110 north in downtown L.A.) well before the car could see the curve.

Exciting times!

 

[conman]

Exciting times!

Agreed. After months of negative comments, aggressive vitriol, threats of litigation, alleging lies and deceit from Tesla, I can't recall a more positive set of changes to the ethereal component that is autopilot. There is every sign of it maturing in the correct direction at last. All they need now is to actually start delivering on some more of the features it is supposed to be capable of and the roar of opposition will die away into a whimper at last (but let's not bring up FSD shall we?)