-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
2023 Model 3 without USS and proximity functionality [park assist / summon not available]
- Thread starter mpgxsvcd
- Start date
Don't worry, you have a good shot at learning all about phantom braking too. There is clearly a pattern with Tesla and starting to de-content useful things. I was one of the first people who got one of the Y's last year that didn't have radar. What a cluster that was. My car with radar has MUCH better AP and cruise control (TACC) than my cars where they gutted the radar.
I am sure this will take a year, or more, for Tesla to sort out given their past track record. Take something away from you and then sell it as a feature.
PianoAl
Active Member
I'm hopeful it will be reimplemented soon. We're talking 4 MPH here in a car that negotiates roads at 85 MPH. The main task of the computer is to figure out where things are. 
Read pages 4 and 5 of your original order. There is a mention of Arbitration and Small Claims court.
Msjulie
Active Member
Speechless (well nearly 
) with frustration with Tesla. Ok they are cheaping out the new cars but on existing cars, just leave well enough alone. The software to handle them exists and there's no good reason to break it.. yeah it is Tesla though
) with frustration with Tesla. Ok they are cheaping out the new cars but on existing cars, just leave well enough alone. The software to handle them exists and there's no good reason to break it.. yeah it is Tesla though
Speechless (well nearly
I don't think anyone has said that existing cars will disable the ultrasonic sensors. And even if it does, as long as the new software meets or exceeds the performance of the low resolution sensors, what's the issue with that?
I don't consider removing all those sensors as cheaping out -- not at all. It's optimization. The best part is no part. This change removes at least 3 dozen discrete parts from the car. A dozen sensors, a dozen mounts, several wiring harnesses and at least 14 wire connectors. All failure points, all removed. This'll reduce cost and labor - significantly - for the two most commonly-replaced body parts. Part manufacture, ship, installation all simplified. Stocking, simplified. Supply chain, simplified.
And even after the car is in the field -- this GREATLY improves the process for repair as well. Much, much quicker, many less parts, less wiring -- that's all goodness.
CQI - Continuous Quality Improvement. Building a better product. Removing these parts is a huge process improvement. Bigger than it appears on the surface, for sure.
That's an effort I can get behind.
I can’t disagree with your sentiments, however one would expect that they have the replacement technology ready to go before implementing any changes. It’s shameful that they have adopted this approach.
Shameful is the wrong word... My suspicion (and I don't have ANY inside information, to be clear ...) is that this was a decision made quite a while ago. So that means shifting suppliers - developing the new molds for the bumper covers to eliminate the USS holes, and even possibly a new VCRIGHT body controller to remove the now-unused connectors. Those things take time. It also means depleting stock of the sensors, brackets, wire harnesses and VCRIGHT modules.
You'd have to plan this long in advance - months, if not a year+.
And that plan has to coincide with the software development. There's 3 possible outcomes:
1) Best possible case: Both are ready at same time.
- Great. Change software, change hardware, customers don't notice a thing. Perfect scenario.
2) Next worst case: Hardware is ready, software is not.
- Not so great. Now I've got all new parts, I've stopped ordering the old parts, so can't install any more sensors. Customers will have to wait for parking sensor functionality until software is ready.
3) Worse yet: Software is ready, hardware is not.
- A Very Bad Day: We have to stop the assembly line because now I'm out of the old parts, and don't have the new parts yet to install. As soon as I get a shipment of parts, we turn the lines back on.
4) Worst of all: Software & hardware both not ready.
- Armageddon: Assembly line is stopped and no idea how the hell to get it going again.
And if I were a betting person, that's where we're at. Scenario #2. The software took longer to develop than the plan called for, and now they're out of supply of the sensors or bumper covers or brackets or wire looms or VCRIGHT modules or any number of parts that they weren't restocking for the old setup. So to keep things moving, they have to start shipping the new parts - they're ready, and out of stock on the old stuff. Unfortunately supply chains don't turn on a dime, and so you can't just call up and say "I need another million USS's - oh, and in all 5 colors still, and I need them next week because I'm just about out."
SO if the software team doesn't hit the commitment date (for whatever reason - development took too long, they hit a major bug, management set an unrealistic date) -- then we end up where we are right now. Who to blame? I'm sure there's plenty of finger-pointing, but the reality doesn't matter... Thing is, assembly has to switch, now, and the developers get whipped until the software is ready.
It's not pleasant, but the reality of the world is that getting the software to finish on the same timeline as when the hardware orders switch over -- which had to be planned probably a year in advance -- is a really tough proposition.
So we wait. And hopefully it's only a month or two, at which time everyone forgets all about this, moves on with a great functioning car, and all the benefits of removing the sensors.
Again, this is all speculation, but having spent my career on various sides of the tech industry - customer, manufacturer, supplier - I've seen this scenario time and time again.
I posted them above -- 2023 Model 3 without USS and proximity functionality [park assist / summon not available]
Reduction in parts, basically. Ease of manufacture, inventory and repair.
"The best part is no part."
I don't know if it's because my car has the newer cameras, but I haven't had any phantom braking yet. With all the talk I would have expected it to surface by now. Granted I only have about 2000 miles on mine so far, but 90% or more of my driving is on autopilot.
stopcrazypp
Well-Known Member
The Vision version have improved greatly from when it first started. It also varies depending on where you are. If you are in areas where Teslas are most popular it also seems to do better.
5) Hardware and software are both ready, but the software is mediocre and leads to small fender benders in perpetuity.
--
In my experience, the software seems likely to be problematic. Auto wipers still don't work well after 5 years of production... climate controls are laughable bad in some ways. The software is littered with poor UX choices. The software department is failing hard, and the CEO sleeps at Twitter HQ.
I'm almost crossing fingers that Herbert Diess takes over. Perhaps some of the long standing issues will improve.
So having no car seats would also be better? Removing wipers would be an improvement? We can maybe remove the batteries, huge cost reduction and all our range anxiety would dissappear!
I also agree with your sentiment, but only when you have a correct working alternative.
The simple truth now is that Tesla is delivering a car without parking sensors. That's prehistoric, even our 8year old, 10.000€ Peugeot has this!
Msjulie
Active Member
There have already been tests showing that the existing front cameras can not see directly in front of the car where the sensors can. I know, I know, back in the old days when cars were not the size of semi trucks, we didn't need no stinking sensors to keep track of our kids/pets/whatever to avoid running them over.
Claims of no-part-is-better in this regard seems to come up short; so I'm not buying it here. Just like I disagree with vision only vs vision + radar. Certainly not with the current cameras.
The real truth, regardless of happy sounding justifications, is it makes the cars cheaper to build, just like haptic buttons and/or touch screens for everything vs real controls. That's all well and good except when it has a negative impact on the actual functionality/usability of the car.
Claims of no-part-is-better in this regard seems to come up short; so I'm not buying it here. Just like I disagree with vision only vs vision + radar. Certainly not with the current cameras.
The real truth, regardless of happy sounding justifications, is it makes the cars cheaper to build, just like haptic buttons and/or touch screens for everything vs real controls. That's all well and good except when it has a negative impact on the actual functionality/usability of the car.
zumoney515
Member
Can't wait. I saw interviews of ppl saying it'll be even more accurate than USS when it's all done.
stopcrazypp
Well-Known Member
USS were not designed for keeping track of kids/pets/whatever in the first place. See limitations in manual, most relevant portions:
Model 3 Owner's Manual | Tesla
The core functionality the replacement occupancy network needs to support is judging distance to large objects like walls and cars, which is something it can do by object permanence, even with that blind spot. It remains to be seen however how much depth resolution it has (will it still measure by 1 inch increments?), but I have little doubt it can do it in general.
I think the part you are referring to is saying it is "better" for Tesla, not necessarily for the customer (other than getting their car delivered in a more timely manner).
2022.40.4 does not include solution for the missing USS. I have installed 40.4 on my wife brand new M3 and no proximity warnings. In addition, Tesla's proximity warnings is the worst I have ever seen. The warnings sounds are not dynamic. They have the same tone of you are 70cm from a wall or if you are 1mm from it. Systems 20 years old have sounds which are changing according to the distance from an object making a strong and constant beep when you are about to hit something. I have a TM3 long range for the last 18 months and I find the Tesla proximity system useless !!!
I drive a Tesla model 3 long range for the last 18 months and I have to say that the proximity warning system from Tesla is the worst I have ever used in the last 20 years.
The warning sounds are not dynamic therefore it does not matter if you are 70cm from an object or just about to hit it. A 100$ ultrasonic system from AliExpress is doing a better job !! You get dynamic sound notification the closer you are to an object and a very loud sound when you are about to hit something !!
I expect Tesla to do better in this field. The correct system is simply useless. I keep looking in the mirrors as having the same beep all the time adds nothing!!
The warning sounds are not dynamic therefore it does not matter if you are 70cm from an object or just about to hit it. A 100$ ultrasonic system from AliExpress is doing a better job !! You get dynamic sound notification the closer you are to an object and a very loud sound when you are about to hit something !!
I expect Tesla to do better in this field. The correct system is simply useless. I keep looking in the mirrors as having the same beep all the time adds nothing!!
oreobbq
Member
Sounds like you hit something huh. 
With the mirrors and camera on every side of the car, why do you even need a beep? And the visualizer shows how many inches away you are from an object when you get close enough. All of that seems sufficient unless you're trying to park blindfolded.
With the mirrors and camera on every side of the car, why do you even need a beep? And the visualizer shows how many inches away you are from an object when you get close enough. All of that seems sufficient unless you're trying to park blindfolded.Similar threads
