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

Tesla replacing ultrasonic sensors with Tesla Vision

This site may earn commission on affiliate links.
But those who report properly functioning wipers prove that it is not impossible for them to work correctly.
I never said it was: it’s certainly possible for them to work correctly in some or even most situations. Most people, like me, have problems at night when on poorly lit or unlit roads. You’ve doubtless seen messages here from people frustrated that auto wipers must be enabled to use NOA or FSD since in their experience the wipers work poorly in some situations.

I think the standard automotive rain sensor works better in a wider variety of situations. Like many of Tesla’s recent moves— ditching steering wheel stalks and gear shifts for touch controls, for example— I suspect the main driver for these changes is simply lower production cost.
 
I never said it was: it’s certainly possible for them to work correctly in some or even most situations. Most people, like me, have problems at night when on poorly lit or unlit roads. You’ve doubtless seen messages here from people frustrated that auto wipers must be enabled to use NOA or FSD since in their experience the wipers work poorly in some situations.

I think the standard automotive rain sensor works better in a wider variety of situations. Like many of Tesla’s recent moves— ditching steering wheel stalks and gear shifts for touch controls, for example— I suspect the main driver for these changes is simply lower production cost.
Oh, but you did say you thought it was 'impossible' to achieve parity. To refresh your memory:

"I don't think they will ever reach parity with a physical rain sensor, because it's impossible."

I can assure you that I never give my wipers a second thought and have driven through all sorts of weather, day and night. You state that "most people" have problems with them under certain conditions. From what do you conclude that >50% of Tesla owners have problems with their automatic wipers? Do you have a link to a well conducted survey?
 
The concern about auto wipers is lack of consistency. How come some people are super happy about them, while others are super unhappy about them? And we are not talking about single individuals in either group.
Making something “work” is very different than making something work consistently in a wide variety of scenarios. Yes, camera based solution does work but it does not have parity in terms of consistency and spectrum of use cases when compared to the traditional sensor.
In fact, this is very common with Tesla product design philosophy - very narrow use cases, no optionality.
 
Oh, but you did say you thought it was 'impossible' to achieve parity. To refresh your memory:

"I don't think they will ever reach parity with a physical rain sensor, because it's impossible."
Yep, you're right, I did in fact say that. Mea culpa. I don't _know_ that it's impossible for Tesla Vision Auto Wipers to work as well as the standard rain sensors. But they'd been working on it since 2016 and it isn't yet, so that's certainly where I'd put my money. I mean, how long do you think it should take for "rain sensing wiper" software to come out of beta? 7 years seems...excessive
I can assure you that I never give my wipers a second thought and have driven through all sorts of weather, day and night. You state that "most people" have problems with them under certain conditions. From what do you conclude that >50% of Tesla owners have problems with their automatic wipers? Do you have a link to a well conducted survey?
Many, many people here on the forums have not had your sanguine experience with the current auto wipers. Did you perchance try the search I suggested? But let me amend my statement to satisfy the pedant in you: "Most people on these forums who post about their experiences with TV auto wipers don't like them".

Honestly not sure why you want to wrestle this to the ground. That the current auto wipers don't work well is hardly a controversial position here on the forums-- at least I'd never thought it was. But if they work well for you, great. They don't work well for a lot of people, and after 6+ years of no perceptible improvements, I don't think they ever will.

And that's all I have to say on the subject, You may have the last word if you wish.
 
Tesla announced today that they are transitioning away from ultrasonic sensors and replacing them Tesla Vision:




View attachment 861249
Future software update is the kicker. We have seen that before. Stupidest move ever. How can camera see tbrough the frunk
 
Yep, you're right, I did in fact say that. Mea culpa. I don't _know_ that it's impossible for Tesla Vision Auto Wipers to work as well as the standard rain sensors. But they'd been working on it since 2016 and it isn't yet, so that's certainly where I'd put my money. I mean, how long do you think it should take for "rain sensing wiper" software to come out of beta? 7 years seems...excessive

Many, many people here on the forums have not had your sanguine experience with the current auto wipers. Did you perchance try the search I suggested? But let me amend my statement to satisfy the pedant in you: "Most people on these forums who post about their experiences with TV auto wipers don't like them".

Honestly not sure why you want to wrestle this to the ground. That the current auto wipers don't work well is hardly a controversial position here on the forums-- at least I'd never thought it was. But if they work well for you, great. They don't work well for a lot of people, and after 6+ years of no perceptible improvements, I don't think they ever will.

And that's all I have to say on the subject, You may have the last word if you wish.
Since you’re done with your opinion. I would remind that few if Anyone come to a forum like this to say “wow I love this random X feature” ever. So if your assessment of majority is based on a forum then your misguided.
 
  • Like
Reactions: JHCCAZ and Supcom
Hi all. In an attempt to get any information from Tesla and some focus one the issue, I have written the following one Twitter
First time ever posing a tweet (and don’t have any followers) so the chances of getting any kind of response are slim to non🙈. If anyone else have a more active account, maybe they can do the same. Ore just like this one
 
  • Funny
Reactions: AlanSubie4Life
I can assure you that I never give my wipers a second thought and have driven through all sorts of weather, day and night.
It rarely rains in San Diego, but I’ve not had any problems or even thought about the wipers with recent rains. Really all seems to work fine. I think I occasionally push the button in to request a wipe in very light rain, but it is rare even in that case and I don’t think I am even conscious of doing so any more.

However, I’m sure there are certain situations where they don’t work well. Also I suspect some people have different preferences for when/how to wipe the window.
 
come to a forum like this to say “wow I love this random X feature”

Evidently this thread is full of potential USS lovers!

I would say that actually, TMC is predominantly made up of people who have a close relationship with their car(s) and who are generally disposed to love every feature (or even future feature). When their disposition is challenged by events, they start off looking for affirmation and reassurance that all will be well. In the event that this isn't forthcoming, their tone becomes more negative sounding. But I do not think that means they are not ready to shift outlook at any time given hard evidence.
 
I've only got a few week's experience but that includes night and day and all kinds of rain, hail and snow. I find them pretty OK to be honest. Never leave me blind, rather tend to wipe faster than necessary but not offensively. Only a few times been flapping away on a clean screen. Perfectly useable but not as good as my 10 year old Ford or my 20 year old BMW. Still, as an engineer, I just can't help shaking my head that they are wasting their time developing a specialised neural network AI solution for a task that was solved decades ago by a simple, cheap dedicated sensor!
 
...Still, as an engineer, I just can't help shaking my head that they are wasting their time developing a specialised neural network AI solution for a task that was solved decades ago by a simple, cheap dedicated sensor!
I think it makes more sense if you consider that there are two drivers: the human is one, and arguably would be better served by the tried and true IR sensor that everyone here wants. But the other is the computer, looking through the front camera assembly. The optics and POV of the camera, right up against the upper windshield, mean that rain and dirt have a different effect on the computer's vision then on the human's vision.

So, even if the industry standard solution were adopted (even if free), it would still be just as necessary to develop the AI response network to keep the computer vision adequately clear and functional.

There is obviously a fair amount of commonality between the computer's windshield-clearing needs and the human's (which is why you and many others think it's "pretty OK"), but not full commonality (hence the frequent complaints).

Remember also that it's not a matter of "I'm not even using Autopilot, so don't bother me". Tesla cannot choose to disable the camera clearing needed for AEB or other accident-avoidance features, so logically it's a moot point whether AP/FSD is engaged.

Considering all that, I conclude that the camera-driven AI auto-wiper functionality is not at all superfluous or misguided, but in fact required - unavoidable. The next and separate question is, should Tesla have kept the IR sensor anyway, just for the bio-driver?

Maybe, but the reasoning above would dictate that those who think the wipers are too aggressive would still be annoyed - it's what the computer vision needs, in situations where a few sprinkled droplets right in front of the camera are occluding its vision.

The only aspect that would become better with a backup IR sensor, would be situations where the computer is satisfied but the human wants More Faster Wipers (aka "run the MF Wipers" 😀). These are the situations that require the driver to periodically goose the wipers with a manual button-press. Not the end of the world, but I do think a little algorithm could be written (I think I could do it but then it wouldn't be an NN) to pick up on "N wiper requests over the prior M seconds" and to adjust accordingly.
 
  • Like
Reactions: stopcrazypp
It rarely rains in San Diego, but I’ve not had any problems or even thought about the wipers with recent rains. Really all seems to work fine. I think I occasionally push the button in to request a wipe in very light rain, but it is rare even in that case and I don’t think I am even conscious of doing so any more.

However, I’m sure there are certain situations where they don’t work well. Also I suspect some people have different preferences for when/how to wipe the window.
I’ve had numerous (dozens+) cases where my windshield is perfectly clear, there has been no moisture or other debris and indeed I’m not even following a car when suddenly, 10 minutes into the drive the wipers come on. No rhyme or reason, just because.

Conversely, I’ve had numerous occasions while driving in light rain where the windshield is obscured enough that I’m not comfortable with it yet the automatic wipers feel no need to kick in.

Now, one can argue that the field immediately in front of the camera may be dirty/clear in these cases, however when the entire windshield is otherwise clear or obscured it’s quite difficult to argue that the small area in front of the camera is somehow dramatically different.
 
I’ve had numerous (dozens+) cases where my windshield is perfectly clear, there has been no moisture or other debris and indeed I’m not even following a car when suddenly, 10 minutes into the drive the wipers come on. No rhyme or reason, just because.

Conversely, I’ve had numerous occasions while driving in light rain where the windshield is obscured enough that I’m not comfortable with it yet the automatic wipers feel no need to kick in.

Now, one can argue that the field immediately in front of the camera may be dirty/clear in these cases, however when the entire windshield is otherwise clear or obscured it’s quite difficult to argue that the small area in front of the camera is somehow dramatically different.
I have experienced full speed wipers while driving down the road, sunny day, simply turned a corner and they slow and then turn off. Keep driving, turn again to the direction of rapid wiping, and the wiping doesn't resume. This is what makes no sense to me. Glare could be the problem, but why does it then come and go, essentially same conditions. It can't be just something on the windshield over the camera, because if that were the case, it would stop with a turn and then resume when I face the original direction. I regard it as a glitch.
 
  • Like
Reactions: Boza and sleepydoc
So, even if the industry standard solution were adopted (even if free), it would still be just as necessary to develop the AI response network to keep the computer vision adequately clear and functional.
Exactly, a NN for rain/dust detection is always required because the windshield cameras need to know when it is obscured and act to clean it. That's likely why they ditched the separate rain sensor.

Your point about wipe preferences is true also, it's very much subjective. That's why it was trivial for me up thread to find complaints about convention rain sensors even when they are presented by most here as if they are perfect.

And as I mentioned, a relatively easy way to address preferences would be to allow toggling sensitivity (like many other cars allow). There's probably at least two settings:
1) activation threshold (when to activate wipers at all)
2) activation speed (how fast to wipe for a given amount of rain)

And at least when TACC/AP is off, there are probably heuristics they can use to predict if the windshield might have rain even when camera is relatively clear.
 
Exactly, a NN for rain/dust detection is always required because the windshield cameras need to know when it is obscured and act to clean it. That's likely why they ditched the separate rain sensor.

Your point about wipe preferences is true also, it's very much subjective. That's why it was trivial for me up thread to find complaints about convention rain sensors even when they are presented by most here as if they are perfect.

And as I mentioned, a relatively easy way to address preferences would be to allow toggling sensitivity (like many other cars allow). There's probably at least two settings:
1) activation threshold (when to activate wipers at all)
2) activation speed (how fast to wipe for a given amount of rain)

And at least when TACC/AP is off, there are probably heuristics they can use to predict if the windshield might have rain even when camera is relatively clear.
NN, heuristic analysis… vs $5 chip? Sounds to my like using microscope instead of a hammer to drive nails.
 
I’ve had numerous (dozens+) cases where my windshield is perfectly clear, there has been no moisture or other debris and indeed I’m not even following a car when suddenly, 10 minutes into the drive the wipers come on. No rhyme or reason, just because.

Conversely, I’ve had numerous occasions while driving in light rain where the windshield is obscured enough that I’m not comfortable with it yet the automatic wipers feel no need to kick in.

Now, one can argue that the field immediately in front of the camera may be dirty/clear in these cases, however when the entire windshield is otherwise clear or obscured it’s quite difficult to argue that the small area in front of the camera is somehow dramatically different.
Actually it's fairly trivial for situations to come up where the camera/sensor might be dirty, but the windshield is spotless. One is dust or dried liquid that leaves a residue. When the light hits it at just the right angle, that can cause activation even when windshield is dry.
Just a quick Google found examples with a conventional rain sensor (an example where it happens consistently is car in front using their wiper fluid and spray hitting your car, usually near the top where sensors are).

The flip side of it not activating, I was able to observe easily too: when there is a lot of road spray, the lower part of windshield can become obscured but topmost part remains largely clear.

It is for this same reason that NN based on camera is required regardless, to eliminate possibilities of sensor being clear and camera not (and vice versa).
 
NN, heuristic analysis… vs $5 chip? Sounds to my like using microscope instead of a hammer to drive nails.
As mentioned, the $5 chip doesn't solve it either, it has the same problem that it only observes a small section of the windshield (although usually they are positioned a bit lower than where a camera would be). A true solution that would work in all situations would be a camera observing the entire windshield, but beyond that, it's all heuristics (usually assumption wetness of windshield is uniform, which many times it's not, examples pointed out elsewhere).
 
Actually it's fairly trivial for situations to come up where the camera/sensor might be dirty, but the windshield is spotless. One is dust or dried liquid that leaves a residue. When the light hits it at just the right angle, that can cause activation even when windshield is dry.
Just a quick Google found examples with a conventional rain sensor (an example where it happens consistently is car in front using their wiper fluid and spray hitting your car, usually near the top where sensors are).

The flip side of it not activating, I was able to observe easily too: when there is a lot of road spray, the lower part of windshield can become obscured but topmost part remains largely clear.

It is for this same reason that NN based on camera is required regardless, to eliminate possibilities of sensor being clear and camera not (and vice versa).
If it happened on occasion I might buy your excuse, but it’s an incredibly regular occurrence. To the point that I have to stop using cruise control because of the wipers. It also happens on cloudy days where there is no glare or direct sunlight to suddenly cause glare. On several occasions when this has happened I’ve looked and confirmed that the glass in front of the camera is clear. Further more, when I resume my trip, cruise control or FSD is perfectly content with the glass that it was obsessed with cleaning just a few minutes prior.

The examples of it not activating were not unevenly distributed road spray - they were rain that very evenly distributes across the windshield To the point that I, as a human using the entire windshield was uncomfortable with the poor view.

So yes, it is abundantly clear that the issues is simply due to failure of Tesla’s approach.