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.
Thanks, yes. I had already corrected that bad mistake, if you go back and look at my edit, about 10 minutes after I posted it. I reread it and spotted it myself. I felt a little sheepish, so I acknowledged the error instead of simply replacing it. Just in case someone was already replying as I think you did. (The edit window is now closed but if you see anything else let me know!)
Ah, I hadn't refreshed in a while... 😅
 
  • Like
Reactions: JHCCAZ
My concern is that USS is a proven technology that works extremely well. Absolute best case scenario with using vision instead -- is that it turns out to not be worse. For consumers there will be zero benefit. If there is any savings related to this, Tesla will be the one pocketing the profit.
Yeah! Dam that company trying to make money and profit! Stop taking away things that are proven and stop that innovating. And tell Apple I want the 40 buttons and plastic pencil back on my phone! Those stinken profiteers!
 
Yeah! Dam that company trying to make money and profit! Stop taking away things that are proven and stop that innovating. And tell Apple I want the 40 buttons and plastic pencil back on my phone! Those stinken profiteers!
Innovation requires change, but change is not automatically innovation, improvement, or progress (it is often not). Windows 8 was a big change, but only a few weirdos thought it was innovation or progress. Tesla has had a lot of 'Windows 8' moments in the last few years. But at lest Microsoft was willing to fix some of those issues eventually and 10 was much better than 8. Even Apple sometimes learns from their mistakes, for example the return of many ports and physical function keys to recent laptops, where they had previously been removed for 'innovation' reasons. I don't see Tesla learning from or correcting many of their mistakes any more even in the face of abundant owner feedback (auto wiper still suck, manual wiper controls have not improved, V11 UI still not as good as what came before, etc, etc, etc).
 
Innovation requires change, but change is not automatically innovation, improvement, or progress (it is often not). Windows 8 was a big change, but only a few weirdos thought it was innovation or progress. Tesla has had a lot of 'Windows 8' moments in the last few years. But at lest Microsoft was willing to fix some of those issues eventually and 10 was much better than 8. Even Apple sometimes learns from their mistakes, for example the return of many ports and physical function keys to recent laptops, where they had previously been removed for 'innovation' reasons. I don't see Tesla learning from or correcting many of their mistakes any more even in the face of abundant owner feedback (auto wiper still suck, manual wiper controls have not improved, V11 UI still not as good as what came before, etc, etc, etc).
Some “innovations” from Tesla only benefits their bottom line but nevertheless narrated as the greatest beta for the customers .. even the version number is just to feed one’s ego.
 
  • Like
Reactions: E90alex
Tesla is pursuing Vision Only as their philosophy for the future of ADAS. We've known this for some time, and when radar was removed from cars, it further cemented the trajectory. With the removal of ultrasonics from the cars, the company is completely on vision going forward.

Tesla cars are still excellent EVs, and at one point were the only game in town. The landscape has changed recently, with new EVs coming out from established companies and new, innovative companies as well.

If you like Tesla, and agree with their Vision Only philosophy, then you have voted with your dollars. If you think that ADAS features demand a sensor fusion with multiple types of sensors, then Tesla is not for you.

Tesla has a large head-start on ADAS features, but other car manufacturers are coming with large arrays of sensors (like NIO), and will no doubt give us ADAS functions on par with Tesla in the near future.

It's an exciting time for EVs and ADAS, and we get to be a part of it. Do your research, explore what's coming, and vote with your dollars.
 
  • Like
Reactions: Yelobird and DrGriz
If you like Tesla, and agree with their Vision Only philosophy, then you have voted with your dollars. If you think that ADAS features demand a sensor fusion with multiple types of sensors, then Tesla is not for you.

Going vision only would be great if they had great vision sensors: full 360 degrees coverage, high resolution, stereoscopic front and back. But they don't. Sensors aren't competitive in vision even vs Chinese.

XPeng P7: 13 cameras outside, one inside. 12 ultrasonics, 5 millimeter wave radars.
 
  • Like
Reactions: Mark II and E90alex
What's stopping them from doing so? HW4, HW5, etc. We'll likely see new vision sensors on the cars as time progresses.
Well.......they are building and delivering HW3 cars that don't have USS and can't do Vision (Parking and Summon) features today. So HW4 with more/better cameras may be in the future and make Vision complete but that is of no use on a new ModelY/3. Also they will probably (aspirational) remove USS support from all cars.
 
Well.......they are building and delivering HW3 cars that don't have USS and can't do Vision (Parking and Summon) features today. So HW4 with more/better cameras may be in the future and make Vision complete but that is of no use on a new ModelY/3. Also they will probably (aspirational) remove USS support from all cars.
You're coming from a very pessimistic viewpoint. Why can't parking be done with vision on current hardware? The new occupancy network can render down to cm blocks at close range. I'll just say that I hope you'll be pleasantly surprised when they update summon and autopark features with new FSD NN capabilities.
 
You're coming from a very pessimistic viewpoint. Why can't parking be done with vision on current hardware? The new occupancy network can render down to cm blocks at close range. I'll just say that I hope you'll be pleasantly surprised when they update summon and autopark features with new FSD NN capabilities.
I'm NOT being pessimistic but realistic. Tesla states as a fact that the car's will NOT ship with Summon and AutoPark.

I don't want to "HOPE" I'm "surprised" since I'm already surprised every year about FSD L5 being released. How about this idea that makes a HELL of a lot more sense. Update the software to were Vision can do Summon and AutoPark and THEN remove the USS from production.

Removing USS first and then working on Vision to replace them is a$$ backwards.
 
Last edited:
I'm NOT being pessimistic but realistic. Tesla states as a fact that the car's will NOT ship with Summon and AutoPark.

I don't want to "HOPE" I'm "surprised" since I'm already surprised every year about FSD L5 being released. How about this idea that makes a HELL of a lot more sense. Update the software to were Vision can do Summon and AutoPark and THEN remove the USS from production.

Removing USS first and then working on Vision to replace them is a$$ backwards.
Ah, I see where you're coming from. Got it.
 
How about this idea that makes a HELL of a lot more sense. Update the software to were Vision can do Summon and AutoPark and THEN remove the USS from production.
From what we could tell that was the plan. It seems like they thought that the software would be ready at the end of September, so they cancelled their USS orders starting 10/1. (Order changes likely have to be done months ahead of time.) They likely didn't want to wait until the software was ready to cancel, as that would mean buying months of USS that they weren't going to use.

Hopefully their predication isn't far off and the software will ship soon.
 
From what we could tell that was the plan. It seems like they thought that the software would be ready at the end of September, so they cancelled their USS orders starting 10/1.
Lol, Elon told US the timeline was end of September...you really think the whole business should be making arrangements based on that? I think most people at Tesla are at least smarter than us who know to take any self-driving timelines and multiply it by 3 or more.

If the software was expected to be ready by the end of September on Elon time, they should've cancelled USS orders at the earliest 12/31
 
Last edited:
when adding feature the timing is in months. when removing it's immediate.
Note they needed new bumpers for this change (see the recently ones on the trucks, they have no holes). Unlike, the radar case, this isn't a change where they just leave an item out.

So putting aside supplier agreements this couldn't possibly have been an immediate change.
 
Note they needed new bumpers for this change (see the recently ones on the trucks, they have no holes). Unlike, the radar case, this isn't a change where they just leave an item out.

So putting aside supplier agreements this couldn't possibly have been an immediate change.
Bumpers have a steel frame that will stay the same that has a composite/plastic molded on. And molding changes are relatively easy. This is a quick change over and any made pre USS can just be moved to the Parts Department.
 
Check out the Tesla Used Car market for selling your existing car so you can get into a car you'd enjoy. Prices are hot right now, so you may not lose much of your value.
The only reason I keep my current car is the drivetrain. No one else is at 400mi+ range. Otherwise, that would have been the shortest ownership I had. I am fed up with constantly be worried what crazy idea they would spring on me. I love cars but I also have a life.