[UK] 2024.8.x

[KennethS]

At this point - 4 years in - if Tesla aren’t making a big song and dance about auto wipers having been improved, explicitly in the release notes, then my feeling is that people’s perceptions of updates improving them is basically psychological, i.e. wish fulfillment.

And yet all the reported changes in AP behaviour, never challenged here, are similarly never mentioned in the release notes.

I'm not contending the auto wipers have or have not been tweaked ever in any software release. Just find it curious that claims of perceived change in behaviour are routinely dismissed as fantasy.

 

[MrBadger]

And yet all the reported changes in AP behaviour, never challenged here, are similarly never mentioned in the release notes.

I'm not contending the auto wipers have or have not been tweaked ever in any software release. Just find it curious that claims of perceived change in behaviour are routinely dismissed as fantasy.

A significant number of the reported AP changes are from multiple independent sources in clusters affecting repeat drives over a specific set of firmware versions.

The majority of auto wipers reports come from spurious reports gleaned from a single drive.

If the wiper behaviour changes also came from multiple independent reports then I think they would be taken more seriously.

As it happens, many people report the same auto wiper issues in the same, often common, long standing reproducable scenarios, yet there are no reports that I can think of that the behaviour has changed in those scenarios. I think many, such as myself, will only start taking note of an improvement when those traits have been remedied.

I’m not saying that there are no changes because some Tesla staff publicise that they are making changes, but I don’t think what they changing makes us any closer to remedying the major issues.

 

[Utumno]

As soon as Tesla claim “wipers are fixed” or “phantom braking is fixed” there will be the usual onslaught of people claiming it isn’t.

The classic software developer approach to release notes” text is therefore “Bugfixes” and making sure the duct tape adhesive on Elon mouth and fingers is still strong enough.

They only escaped this approach with matrix headlights because it’s a new feature they had no choice but to announce quietly in release notes.

The auto wiper performance is variable, situational, individual and always will be.

 

[Dilly]

The auto wiper performance is variable, situational, individual and always will be.

It’s not unlike “the wrong sort of leaves on the line”

 

[MrBadger]

The auto wiper performance is variable, situational, individual and always will be.

Yet there are several repeatedly documented scenarios that are almost guaranteed to cause issues and thus make a good scenario for any change in behaviour. Situational yes, but not variable or individual.

Spray on motorway being one example and this trait had been there even before ‘deep rain’ and still there years later.

 

[init6]

I appreciate if you've had variable performance then it's hard to say if an update has made any difference, but when you've had 3 years of good performance and then they suddenly start to dry wipe, then it's unlikely to be a change in the weather.

 

[Skie]

I recognise that some people perceive change when hoping for change. But the notion that absolutely no tweaks to the wiper algorithm are ever made when the firmware is updated seems unfounded. Yet on this forum it is verboten to suggest it. At the same time people often suggest that AP behaviour has changed and that is accepted. Why is one accepted fact and the other dismissed as imagination? I find some aspects of the forum lore here quite amusing.

Conversely, there is no way in hell that the wiper algorithm is changed after every update (including every point release and bug fix). And yet each release has people saying they got better or worse after it.

 

[Skie]

Yet there are several repeatedly documented scenarios that are almost guaranteed to cause issues and thus make a good scenario for any change in behaviour. Situational yes, but not variable or individual.

Spray on motorway being one example and this trait had been there even before ‘deep rain’ and still there years later.

Lighting, headlights/brakelights, spray volume, cleanliness of glass will all be variables. Camera based rain detection is hard because there are so many variables.

 

[KennethS]

Conversely, there is no way in hell that the wiper algorithm is changed after every update (including every point release and bug fix). And yet each release has people saying they got better or worse after it.

And yet when claims are made that AP has changed in each one of those same releases there is no pushback.

I will stop now.

 

[M1tch]

To summarise, the wipers are fixed right?

 

[squerble]

No, to summarise, I'm still waiting for this update

 

[M1tch]

It will come. I got offered it within minutes of landing overseas and not seeing the car for months. My wife thinks it’s hilarious as I said my next car would defo have matrix headlights and although I did - I also didn’t until two weeks ago. And now I can’t play with it

 

[andyorkney]

I love my new to me Model 3, but I remain disappointed by poorly performing auto wipers and the nighttime repeater glare from indicator blind spot cameras. Guess I’ll have to spend a bit to fix the camera issues, but my 2003 rover 75 tourer had the rain detection nailed. Is this all about camera quality? Still waiting for the update, also on .2.8 so guess this iteration puts me in the back of the queue….

 

[M1tch]

Ah your wipers aren’t great, don’t worry, its fixed in the next update

 

[Drew57]

I've always experienced good, bad and appalling wiper behaviour happening at random & today I had to use the stalk button numerous times on a screen that was almost completely obscured by light rain.

Tweaking the software is most likely pointless because the focal distance of our relatively simple windscreen cameras needs to be set well beyond the surface of the screen & on that basis I doubt Tesla ever have or ever will solve this issue.

 

[Foxsbiscuits]

nighttime repeater glare from indicator blind spot cameras.

Very hard to fix this yourself, there's a gap in the mechanism that lets light through but it's not accessible to patch without breaking open the unit. There's a YouTube of someone doing it. It was clearly a fault in the electric board as it isn't found on later black trim models.

 

[GRiLLA]

And yet when claims are made that AP has changed in each one of those same releases there is no pushback.

I will stop now.

I gave up pointing out that Autopilot drives today just like it did at the end of 2019, and before that when they actually made changes, they were quite noticeable and generally in the release notes. The group has a preference to believe there are incremental minor changes that add up to nothing at all on an annual scale.

 

[Zilla91]

I gave up pointing out that Autopilot drives today just like it did at the end of 2019, and before that when they actually made changes, they were quite noticeable and generally in the release notes. The group has a preference to believe there are incremental minor changes that add up to nothing at all on an annual scale.

Except that’s easy to disprove because they would have had to make changes to Autopilot when they disabled the radar units.

Or do you think they did that without making any alterations to it?

 

[Roy B]

Is this all about camera quality?

Not exactly. It's more about the use of cameras at all.

No one other than Tesla has attempted to do auto wipers with cameras - everyone else uses a rain sensor.

 

[Skie]

Except that’s easy to disprove because they would have had to make changes to Autopilot when they disabled the radar units.

Or do you think they did that without making any alterations to it?

Autopilot also gets "updates" via map/route changes. Plenty of tales of weird behaviour in the same location caused by dodgy map info that suddenly goes away when the route is fixed. And disabling radar fixed a very reproducable bit of phantom breaking I had. Pretty sure it was a tree and steep downward slope combining to cause a radar return that made the car crap itself. As soon as radar was disabled it stopped happening and has never happened since at that location.

Autopilot does have loads of random issues caused by the unique combination of circumstances so has some similarity to Autowipers, but there are other things in the mix too. I don't think the underlying logic regularly though, but the other factors do contribute to it getting better (or worse).