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.
Nobody is considering that this could be partly supply chain issue. Probably not but it could be.

Second mixing input sensor type into the AutoPilot computer is probably extremely complex. We do everything with our eyes. Why shouldn’t the AutoPilot computer do the same? It may have a few dips during the transition. But it the long run I suspect it will all work better.

And eventually, like Radar, they will stop using the sensors on cars equipped with them.

I saw no negative impact going radarless.
What a terrible take...

1. Has been speculated multiple times in this thread that the decision is supply chain driven, just like removal of radar was (regardless of what Elon tells you)
2. Human eyes are situated on this thing called a head, which is connected to an awesome support column that can rotate, bend, and stretch. The cameras on your car are in static locations and are dirty and low resolution. Comparing static cameras to eyes that can be adjusted depending on the situation is completely asinine
3. My radar 2019 model 3 was perfect both on AP & NoA. My wife's 2022 model 3 without radar is atrocious and brakes for shadows and overpasses to the point we literally do not use it at all.

Stop the fanboyism
 
12 Bosch sensors, color matched
12 brackets
Bumper fascia punch outs
2 wire harnesses each with 6x 3 pole and 1x 12 pole connector, 8 circuits, 2 which are multisplice
Mating harness (unless direct to ECU) each with 2 connectors and 8 circuits.
Possibly 2 ECUs, at a minimum interface circuitry.
Shipping, handling, supplier markup, installation, warranty.

At $8 a sensor, that is $100 on their own. Aftermarket puts then at $50-$60, at 4x markup, that still $12 a piece.

Even at a single hundred, that's $100 million a year in savings, which is a lot of engineering. Equivalent to ~4% of Q2's R&D annualized.
Don't you think they should have solved the problem first?

(This is a trick question -- There is no solution with the current camera spec / layout)
 
12 Bosch sensors, color matched
12 brackets
Bumper fascia punch outs
2 wire harnesses each with 6x 3 pole and 1x 12 pole connector, 8 circuits, 2 which are multisplice
Mating harness (unless direct to ECU) each with 2 connectors and 8 circuits.
Possibly 2 ECUs, at a minimum interface circuitry.
Shipping, handling, supplier markup, installation, warranty.

At $8 a sensor, that is $100 on their own. Aftermarket puts then at $50-$60, at 4x markup, that still $12 a piece.

Even at a single hundred, that's $100 million a year in savings, which is a lot of engineering. Equivalent to ~4% of Q2's R&D annualized.
With the speed in which this was announced, I doubt that they had time to re-engineer the wiring harness/connectors and brackets. Chances are, its just like the radar removal and all the HW exists to plug everything in if you can find the parts on ebay.
 
  • Like
Reactions: Boza
Don't you think they should have solved the problem first?

(This is a trick question -- There is no solution with the current camera spec / layout)
Not sure how serious to take someone who spells tesla motors backwards for a user name...

Human eyes are situated on this thing called a head, which is connected to an awesome support column that can rotate, bend, and stretch. The cameras on your car are in static locations and are dirty and low resolution. Comparing static cameras to eyes that can be adjusted depending on the situation is completely asinine

So your two eyes inside the cabin see more than Tesla's 8 outside? Interesting take.
The eye has an angular resolution of 1 arcminute (0.02 degrees). A Tesla 1.2MP camera (1280x960) uses raw photon count from the 4 sub-pixel sensors, so it is more like 5MP. HW4 will have 5MP (pixel groups) from Samsung.

But ignoring that for the moment, 1289x960 pixels at 0.02 degrees is 25.6x19.2 degrees for human eye equivalency, so 1/6 eye resolution if it's a 180 degree field of view (which most aren't), 1/3 eye resolution for 90 degree.

Heck, let's say the resolution is only 1/10 of the human eye or 0.2 degree. At 20 feet, that's still resolves object less than one inch. Call it two if you are concerned about pixel overlap.

Where's the impossibility?
 
With the speed in which this was announced, I doubt that they had time to re-engineer the wiring harness/connectors and brackets. Chances are, its just like the radar removal and all the HW exists to plug everything in if you can find the parts on ebay.
Brackets hold sensors to fascia, harnesses were USS only separate parts, both straight decontent. Chassis wiring may have mating connectors until existing stock runs out, but that's a easy supplier change, and interim partial-switchover cost is not the point anyway.

Speed of announcement is to reduce Osborning/ attemps to get old version. Change would have been planned for a while and coordinated with suppliers.
 
  • Like
Reactions: DrGriz
I keep seeing statements like this... Is 2018 back in the day? Because my 2018 Mazda3 doesn't have ultrasonics. Even having a back-up camera was a big change for me. I'm sure many higher-end cars have had sensors like these for quite a while (and with their price range, this absolutely should include Tesla vehicles), but I don't think you can realistically claim that even a simple majority of drivers have these sensors in their cars.

That being said, I do agree with many of the points in this thread. I think it's a mistake to remove these sensors, especially when they have to remove functionality because the intended replacement isn't ready (reserving judgment on whether that's even feasible). It's undeniably a downgrade at this point. I got my VIN on Wednesday and found out about this change minutes later. When I asked, the SA confirmed that my car will come without sensors. I did find that disappointing, but it's still an enormous increase in driver assistance features for me, so I don't see this as a dealbreaker. I certainly never had any intention of relying on these sensors to the extent that I'd be running over children and puppies if one of them failed. The number of those kinds of statements that I've seen related to this change really concerns me....

@Mark II, with regard to your comment about the hood geometry -- is the visibility really that much worse than any other car? The only visibility issues I noticed on the test drive were with the mirrors, which I determined could be alleviated by looking more thoroughly over my shoulder if the camera failed for whatever reason. I've never had a great view of what's in front of the bumper of any car, beyond where the hood drops off.

I'll let you know about any compensating features that may have been added. I think they would have mentioned something like that to alleviate the backlash, though.
Did they mention any other changes or is the only 2023 model year change the removal of parking sensors?
 
Speed of announcement is to reduce Osborning/ attemps to get old version. Change would have been planned for a while and coordinated with suppliers.
In a typical economy where every supplier has what they need to deliver to all OEMs, sure.

2022 models of many OEMs were sold lacking the same safety features as comparable 2020/2021 models.

Tesla's response to the global parts shortages faced by many OEMs was to push Vision as a means to continue car sales. Had they held cars waiting on Continental’s radar, they would have gone bankrupt. Seeing as many OEMs use the Bosch USS, and and Bosch sees shortages carrying into 2024... doesnt take a rocket scientist to realise Tesla cant afford to suspend deliveries waiting on parts forcasted to be in shortages for years.

 
In a typical economy where every supplier has what they need to deliver to all OEMs, sure.

2022 models of many OEMs were sold lacking the same safety features as comparable 2020/2021 models.

Tesla's response to the global parts shortages faced by many OEMs was to push Vision as a means to continue car sales. Had they held cars waiting on Continental’s radar, they would have gone bankrupt. Seeing as many OEMs use the Bosch USS, and and Bosch sees shortages carrying into 2024... doesnt take a rocket scientist to realise Tesla cant afford to suspend deliveries waiting on parts forcasted to be in shortages for years.

Radar was pulled in May/April 2021, was there a shortage then? Continental was making 30 million a year in Texas as of August 2020 and 2021 had total company sales up by 6%.

I wouldn't expect USS to be hit as hard by the chip crunch but it could be. Seems strange timing though based on your linked article.
“The semi industry’s most pressing challenges may have peaked. More respondents expect semi shortages to ease by the end of 2022, as compared to the February survey. Respondents also expect continued improvement from here, rather than exhibiting further deterioration prior to a rebound. Our findings corroborate management teams at many automotive (manufacturers), that expect improving semiconductor availability in Q3 and Q4 2022,” the report said.
 
Not sure how serious to take someone who spells tesla motors backwards for a user name...



So your two eyes inside the cabin see more than Tesla's 8 outside? Interesting take.
The eye has an angular resolution of 1 arcminute (0.02 degrees). A Tesla 1.2MP camera (1280x960) uses raw photon count from the 4 sub-pixel sensors, so it is more like 5MP. HW4 will have 5MP (pixel groups) from Samsung.

But ignoring that for the moment, 1289x960 pixels at 0.02 degrees is 25.6x19.2 degrees for human eye equivalency, so 1/6 eye resolution if it's a 180 degree field of view (which most aren't), 1/3 eye resolution for 90 degree.

Heck, let's say the resolution is only 1/10 of the human eye or 0.2 degree. At 20 feet, that's still resolves object less than one inch. Call it two if you are concerned about pixel overlap.

Where's the impossibility?
Um yes, my eyes which can be manuevered by my spinal column to see in any direction I need at any time can see more than 8 static cameras that are exposed to the elements.

Nevermind of all of the peripheral items my eyes can use such as rear view and side view mirrors.

We get it, you love Elon Musk (who is a scum bag), but don't pretend that removing a very helpful sensor is not detrimental to the driving experience.
 
Um yes, my eyes which can be manuevered by my spinal column to see in any direction I need at any time can see more than 8 static cameras that are exposed to the elements.

Nevermind of all of the peripheral items my eyes can use such as rear view and side view mirrors.

We get it, you love Elon Musk (who is a scum bag), but don't pretend that removing a very helpful sensor is not detrimental to the driving experience.
Lol, have you Ever posted anything positive in your 3 years here. Impressed by your 8 camera view eyes though. Lol
 
  • Funny
Reactions: DrGriz
Radar was pulled in May/April 2021, was there a shortage then? Continental was making 30 million a year in Texas as of August 2020 and 2021 had total company sales up by 6%.

I wouldn't expect USS to be hit as hard by the chip crunch but it could be. Seems strange timing though based on your linked article.
Radar was pulled for cost reasons as much as the blind faith pursuit that vision only will cure all issues.
 
March 9th, 2021
Radar was pulled in May/April 2021, was there a shortage then? Continental was making 30 million a year in Texas as of August 2020 and 2021 had total company sales up by 6%.

I wouldn't expect USS to be hit as hard by the chip crunch but it could be. Seems strange timing though based on your linked article.
March 9th, 2021:

 
  • Helpful
  • Like
Reactions: Mikecm1 and mongo
It may take couple of weeks for cars with no ultrasonic sensors to be delivered. They have to be manufactured in such configuration and then delivered to customers. I think similarly to new features like door trim, etc next we will see a mix of cars with or without sensors being delivered. They have to start making bumpers differently, at least stop cutting holes in them.
I'm not trying to say that lack of sensors is a new feature though :D
If you want to sue for $250, you don’t need hundreds of thousands of dollars in attorney fees. Just file yourself in small claims court. Tesla won’t show, you’ll win a judgement by default, send it to Tesla, and they’ll pay it. Your just out court costs.
This is all 100% correct
 
Last edited by a moderator:
Radar was pulled for cost reasons as much as the blind faith pursuit that vision only will cure all issues.
Tesla has established a pattern:
- passenger side lumbar was removed in many mfrs cars last year and in some cases (ford), customers notified in advance and given credit re: price reduction at purchase. Tesla? Customers found out by other Tesla customers posting on the internet after they found out themselves after buying. No credit from Tesla.

- chip shortages impacted many mfrs ability to source components for radar. Other car mfrs publically announced the impact to their lineup. Tesla touts “vision is superior” for them removing radar.

- chip shortage impacting parking sensors. Cadillac/Chevy issue statements that they are having to remove park assist in some cars due to chip shortage. Tesla states they are removing parking sensors solely because vision is better.

See the pattern?
 
And this... very exhaustive list of 2022 models...

Note the several instances of "parking sensor" or "parking assist" that were pulled. If you correlate the OEM to supplier of the sensor... Bosch...

 
  • Like
Reactions: texas_star_TM3
Radar was pulled for cost reasons as much as the blind faith pursuit that vision only will cure all issues.
That may be true. But Tesla had to include cameras to identify pedestrians, animals, traffic cones, and other things that radar could not see. So it kinda makes sense for them also use it for TACC. Once that is done, it is no longer necessary to include it.
 
A better more honest statement would have been “due to global chip shortages, we are forced to remove parking sensors from our vehicles in the near future. Unfortunately This will result in the removal of the features shown in the chart below. We hope to restore these features using Tesla vision in the near future.”


5CBBBB45-DB2C-4D9C-9B15-D6D364B922F8.jpeg
 
12 Bosch sensors, color matched
12 brackets
Bumper fascia punch outs
2 wire harnesses each with 6x 3 pole and 1x 12 pole connector, 8 circuits, 2 which are multisplice
Mating harness (unless direct to ECU) each with 2 connectors and 8 circuits.
Possibly 2 ECUs, at a minimum interface circuitry.
Shipping, handling, supplier markup, installation, warranty.

At $8 a sensor, that is $100 on their own. Aftermarket puts then at $50-$60, at 4x markup, that still $12 a piece.

Even at a single hundred, that's $100 million a year in savings, which is a lot of engineering. Equivalent to ~4% of Q2's R&D annualized.
We’ll probably have to just agree to disagree since neither one of us has any actual data on the costs, but I think you are underestimating the extreme discounts auto manufacturers get from component suppliers, and the high quantities that auto manufacturers purchase that enable those extreme discounts.

Regardless, this decision was purely a supply chain/logistics decision. It had nothing to do with component cost. Just like the radar removal decision.
 
  • Like
Reactions: Cheburashka
We’ll probably have to just agree to disagree since neither one of us has any actual data on the costs, but I think you are underestimating the extreme discounts auto manufacturers get from component suppliers, and the high quantities that auto manufacturers purchase that enable those extreme discounts.

Regardless, this decision was purely a supply chain/logistics decision. It had nothing to do with component cost. Just like the radar removal decision.
I never said it was a component cost decision, I said the component cost savings would more than offset the extra software engineering expense needed in reply to:
In a few years they could probably get Tesla Vision somewhere close to what other manufacturers already do using various sensors. But at what cost? Instead of fixing real issues (the list is long) they will spend time, money and efforts to get back to where they started (and I am being optimistic here - in a few areas V11 is still worse than V10). How does that make sense?!
And it doesn't matter if the price is hundred or hundreds. Say the sensors and wire harness only cost $1 each totaling $14. At their current build rate, that's over $14 million a year in savings. That pays for a lot of full time engineers.