Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

[UK] 2021.4.x

This site may earn commission on affiliate links.
i thought Tesla’s loved cones 😂
They do love cones but as it's Tesla, they're the 99% Flakey variety.

flake.jpg
 
  • Funny
Reactions: Somex
Disappointing - no sign of visual driver monitoring as rumoured elsewhere - possibly a regional thing now rather than a (radarless) Tesla Vision only vehicle.

Model 3/FSD 2021.4.21.2 release notes said:
Minor fixes
Miscellaneous improvements

We look to be the first UK Model 3 in UK in TeslaFi fleet - never had that privilege before. Lots of UK Model S and X though. I guess everyone else is still glued to the football...
 
Last edited:
Disappointing - no sign of visual driver monitoring as rumoured elsewhere - possibly a regional thing now rather than a (radarless) Tesla Vision only vehicle.



We look to be the first UK Model 3 in UK in TeslaFi fleet - never had that privilege before. Lots of UK Model S and X though. I guess everyone else is still glued to the football...
Interesting. You still on the “standard” interval too?
 
You still on the “standard” interval too?

Advanced. Having endured the .4.15.x quick fire updates, once back on .4.18.x I stuck with advanced (was hoping to get the visual driver monitoring) due to a chance in driving pattern. I would have been first to drive the car after an update so allowed me to test on regular routes.

I'll probably revert to standard if/when V11 starts making an appearance - it took a while for V10 to settle - we took 4 updates in 15 days.
 
  • Like
Reactions: CMc1
I think tesla have changed their release schedules so 4 updates in 15 days isn't going to happen again.. they used to be rapid turnover, fix bugs on the fly (release early, release often). But since January they've gone to a more cautious release schedule and aren't throwing everything out there only stuff that's been through some testing - even the FSD beta testers haven't had an update for a while.
 
  • Like
Reactions: AndrewGR
I think tesla have changed their release schedules so 4 updates in 15 days isn't going to happen again.. they used to be rapid turnover, fix bugs on the fly (release early, release often). But since January they've gone to a more cautious release schedule and aren't throwing everything out there only stuff that's been through some testing - even the FSD beta testers haven't had an update for a while.

There have been loads of updates to the .4 base, some of us got the experience a much better version about 6 weeks ago.
The reason for apparent stability is this codebase is a dead-end with most development work going into last Thanksgivings day somewhat delayed release ;-)
 
First drive on 2021.4.21.2 - a regular route but done as a circular trip instead of out and back. Skipped a few miles of urban/town driving at far end though.

All very subjective. From the release notes I wasn't expecting much and TBH much of my focus was on performance of forward collision warnings, TACC and AP/NoA. To be taken with a thick dose of confirmation bias knowing that there is a transition to Tesla Vision (no Radar), visual based driver monitoring and changes in auto wipers.

tl;dr - TACC speed control felt smoother, pressure needed to satisfy AP nag felt much reduced, no sign of TACC warning beep

A couple of edge cases encountered.

First section. Undulating and 50mph twisting A road. Driven fully manual. Nothing to report. Plenty of opportunity for rogue forward collision warning as occurred during .4.15.x branch.

Second section. 40mph twisting A road with some side roads/houses. Done on TACC @39mph. Not quite as busy as normal. TACC felt very smooth even on curves with oncoming traffic. Forward collision warnings on this section are not uncommon, especially with 4.15.x. Really nice ride today.

Third section. 30/40/50 A road with couple of tight bends that I normally do not start TACC until I hit the 50 and disengage ahead of a couple of corners. The 30 section is also parked up on one side so whilst the road has centre chevrons and islands, it can feel a little busy. As it was slightly lighter traffic and TACC felt more positive, I chose to engage TACC in the 30. No issues. On entering the 40, I was following another vehicle at quite a distance (not close enough to initially affect TACC) and on a bend it slowed and came to a stop. Edge case #1. At first it felt like TACC did not see the car slow, but once stopped TACC slowed. It was all safe, but it just all felt a little late and more abrupt than I would have handled the situation. It was a temporary traffic light controlled construction zone and our lane was coned off. As nothing was behind me, I decided to leave car on TACC to see how it handled pulling away and me having to move into the opposite lane. Car was a little slower to pull away than I would have, but not too bad and fine had there been following vehicles. On getting close to the end of the construction zone, I had to pull sharply back to the proper lane. Faced with a stationary car at the oncoming lights, TACC slowed sharply and sounded a warning. It was all a bit of a blur, but I think I spotted the forward collision warning out of the corner of my eye. I suspect that I also adjusted the accelerator to keep on going as unlike TACC, I knew where I was going. TBH, I think the car behaved properly - it didn't know what steering actions I was or was not going to take so handled it as if I was going to continue head on into the waiting car.

The remainder of the 50 section was uneventful. TACC remained smooth even when a slight adjustment in speed as a car turned off left at a Y junction on a corner. I still disengaged TACC for some of the corners. By this point, a few light spots of rain. I helped the wipers on one occasion and it handled the other couple of times that a wipe was necessary. For my liking, the wipers could have been a fraction more sensitive - here hoping that one day Tesla will put a user sensitivity preference on the Model 3. I think that would make a big difference in most situations.

Next significant section was M25/M3. M25 was stop/start at 40mph speed limit. AP/NoA was engaged whilst still on the on ramp. My lane required a merge at the end of the on ramp. A couple of cars in front merged probably where I would have done so, but AP/NoA continued past to the ever closing apex. As a car in the other lane left a nice big gap I decided to see how the car would behave. This was all very slow speed. I decided to take over when the car failed to merge into the nice big space, even whilst indicating right. At no point did it seem to want to do anything other than take the car to the very end of the on ramp, and I wasn't going to see what it did when it got there. #fail.

Back on AP/NoA, as we came up to the M25/M3 intersection, the speed limit lifted. The intersection and significant stretch of M3 following the merge is a 50. As usual, TACC/AP set speed to 55 which I dialed down to 50. M3 was very busy and a couple of slow moving vehicles. It was a bit slower moving than normal. I ended up following a car towing a caravan. Edge case #2. The caravan failed to speed up coming out of the 50 zone and I didn't take the opportunity to get into the next lane - it wasn't that much faster and I was not in a hurry. Then the brake lights came on the caravan and it slowed to a crawl then started slowly pulling left into the safety refuge. We slowed right down but by this time the speed differential between me and next lane was getting much bigger. I tried indicating to move right but car was not having any of it. I'm pretty sure that there was no nag or vehicle in blind spot detected. There was a nice big gap so I took over and accelerated into next lane. It was all drama free and quite safe to that point, but I think the car could have handled the situation better especially as I and caravan in front was travelling so slowly.

For rest of trip, even with busy motorway, auto lane changes worked very well. Which got me concentrating on the nag. I had got one earlier in the drive which confirmed that holding the wheel still appeared to be necessary, but I wasn't convinced that it was quite normal. Suspicions were raised that the nag would appear but you could satisfy it visually as the nags seemed to be satisfied without as much pressure on the wheel as normal. I totally relaxed my grip, got the nag, but couldn't clear it visually, however the slightest of touch of the wheel cleared the nag, and I mean slightest of touch. Confirmation bias at work here, but I think the amount of torque needed on the wheel to satisfy driver in charge may have changed. Frequency seems no different though as I had a couple in fairly quick succession.

Rest of trip uneventful.

Another thing of note is that there is no TACC engagement warning chimes like those that appeared recently in China. Also nothing new in controls that I could see to enable it as an option.
 
here hoping that one day Tesla will put a user sensitivity preference on the Model 3. I think that would make a big difference in most situations.

TBH with Tesla all being about AI or even Machine Learning; you would think this wouldn't need a setting. Just gradually adjust to your preference.

Then again they call Alexa/Siri 'smart' assistants; but each time I ask them to turn off the living lights it's confused because it's mixed up with living room. Instead of asking me, which lights are in the living and remembering for next time; it plays dumb.