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

Tesla.com - "Transitioning to Tesla Vision"

This site may earn commission on affiliate links.
Well, be prepared not just to jiggle the steering wheel, but also make a face at the camera to prove that you're not sleeping.

I am hoping that later, they will remove the steering wheel torque and just use the driver monitoring camera, to allow hands-free AP. Having both the wheel torque and the inside camera is probably just temporary while they validate Tesla Vision.
 
  • Like
Reactions: rxlawdude
Even though firmware is supporting this function, you'll need a vehicle that has a cabin camera like M3 and MY. Vehicles without cameras (older MS and MX) won't be getting this function ever.

I know --- DDdduuuhhhhh :rolleyes: --- but you know the usual nicks will latch onto this to whine and complain (FUD). There was a thread where something was misinterpreted as "see, FSD is dead" and I pointed out the obvious.

Likewise, indications are Tesla is going to add a in-cabin radar. Older vehicles that don't have an in-cabin radar will never get this functionality through a firmware update.

....etc.... Rinse and repeat.

From the other Tesla Vision thread. A tweet from a customer picking up a new camera-only car reports screenshots from the new software update that indicate the interior camera is now being used (for the first time in production mode, I think) to monitor and provide feedback on driver attentiveness. Hands on the wheels is still required.


View attachment 666975
 
Why would you want unnecessary appendages in your car weighing it down?
Nobody is going to have any preference for a 1846.87 kg car vs a 1847.00 kg car. 0.007%.

If Tesla is fully committed to vision, and there's nothing to indicate they aren't
Retention of the radar on the S/X and non-NA cars indicates they are not yet fully committed, by definition.
 
Retention of the radar on the S/X and non-NA cars indicates they are not yet fully committed, by definition.

This discussion is in the context of future resale value / preference. Tesla is definitely fully committed to vision. The timeline is likely within a few months. FSD V9 is vision-only and limitedly releasing Soon(TM). This means The Button(TM) is also vision-only.
 
The timeline is likely within a few months. FSD V9 is vision-only and limitedly releasing Soon(TM). This means The Button(TM) is also vision-only.
v8 didn't change anything on the highway. So unless v9 has a much larger scope than v8, this doesn't mean that cars with radar will completely stop using it. Highway and city codebases are separate, for good reasons I am told.

Do you believe that when v9 and the button roll out, it will go to all cars, or just FSD cars? If it's just FSD cars, who knows how long it will be before a non-FSD car stops using radar. Given Tesla/Elon timeframes, that could easily be years, way past when some cars will get sold. Most cars don't have FSD, and ironically for this conversation it appears to add little to resale.

And with that, I will re-ask the question that always gets ignored. Do you not find an issue with this which was advertised to all radar cars?:
A forward-facing radar with enhanced processing provides additional data about the world on a redundant wavelength that is able to see through heavy rain, fog, dust and even the car ahead.
 
Last edited:
  • Like
Reactions: YieldFarmer
And with that, I will re-ask the question that always gets ignored. Do you not find an issue with this which was advertised to all radar cars?

I see it as a benefit / risk balance. Apparently, Tesla sees the benefits of vision-only outweighing radar+vision, for reasons discussed by Elon and others here.

Yes, it used to be that radar was nice to have for those reasons you pointed out, but those reasons are less compelling as Computer Vision is rapidly improving.

The most pertinent question is if the vision can't see the lane lines, isn't radar not very useful for driving? Vision comes first for driving.
 
  • Like
Reactions: Snerruc
I see it as a benefit / risk balance. Apparently, Tesla sees the benefits of vision-only outweighing radar+vision, for reasons discussed by Elon and others here.

Yes, it used to be that radar was nice to have for those reasons you pointed out, but those reasons are less compelling as Computer Vision is rapidly improving.

The most pertinent question is if the vision can't see the lane lines, isn't radar not very useful for driving? Vision comes first for driving.
What benefit? Only benefit here is cost savings to Tesla.

They want to disable radar? Fine, do it when it’s at parity. Don’t remove components their own site deemed critical days ago and then deliver me a car with severely limited & insufficiently tested capabilities as a last minute surprise.

I really don’t understand people trying to justify this. Why is more sensors ever bad? If it were actually at parity, there wouldn’t be these severe limitations on the cars.
 
I see it as a benefit / risk balance. Apparently, Tesla sees the benefits of vision-only outweighing radar+vision, for reasons discussed by Elon and others here.
I don't think Tesla has ever explained how vision only is better than vision+radar. They just say it's sufficient, and cheaper. They used to have a post explaining how radar was a nice addition to vision, but I guess they don't want us to see that anymore. But I don't see a post saying that vision is better. Elon has hinted that it's hard to fuse things, but easier is not always better. I mean, it's better for Tesla's bottom line, but not for users.

The most pertinent question is if the vision can't see the lane lines, isn't radar not very useful for driving? Vision comes first for driving.
NOBODY is saying radar can drive a car by itself or that vision is not primary or a place that development should be focusing on. However, I personally really like radar because it has been a highly effective AEB / FCW system for me, I've had it beep before the car in front of me was braking. It served purposes that were not just when the car was on AP.

I still don't see how anyone can argue that Tesla is fully committed or 100% sure this will work when they haven't removed it from all cars/regions yet. They have to have a reason for not doing that. But even with that, nothing about Tesla is ever fully committed- this is the company that "fully committed" to AP2, and it took years to get to AP1 parity. They were fully committed to HW2, and then HW2.5, and then HW3. They were fully committed to radar, and now....

So about that question on if v9 is going to all cars or just FSD paid cars while they are on city streets...
 
Allowing AP autosteer up to 75mph would have required a lot of testing and validation actually.
As a prospective customer, he asked for parity with the previous feature set, not just that they tried hard and did "a lot".

I believe Tesla has done "a lot" of testing and validation on FSD v8 and v9 also. Doesn't mean they have delivered any value to a customer, so maybe they shouldn't be charging yet, because there is a difference between a customer and an investor.
 
I don't think Tesla has ever explained how vision only is better than vision+radar. They just say it's sufficient, and cheaper. They used to have a post explaining how radar was a nice addition to vision, but I guess they don't want us to see that anymore.

Tesla isn't known for posting technical descriptions of their AP decisions / technology though, except for patent applications and during Autonomy Day.


 
Tesla isn't known for posting technical descriptions of their AP decisions / technology though, except for patent applications and during Autonomy Day.
And in deleted blog posts.

You literally used by example where Elon says fusion is hard, so should just "double down" on vision. He does not say vision can do everything radar can, he just says it's more precise, which does not tell you about broad functionality.

The hilarious bit is that "double down" is literally rooted in betting and is based on lack of knowledge and risk taking. The definition is:

strengthen one's commitment to a particular strategy or course of action, typically one that is potentially risky.
I will agree all day long that Tesla has doubled down on vision, and that's a lot easier when the risk is carried by your customers, not you.
 
Last edited:
This discussion is in the context of future resale value / preference. Tesla is definitely fully committed to vision. The timeline is likely within a few months. FSD V9 is vision-only and limitedly releasing Soon(TM). This means The Button(TM) is also vision-only.

I don't understand. If they are fully committed, why are they planning to install it in the new MS/MX?
 
I am hoping that later, they will remove the steering wheel torque and just use the driver monitoring camera, to allow hands-free AP. Having both the wheel torque and the inside camera is probably just temporary while they validate Tesla Vision.
I don't believe that Tesla is capable of doing something right on the first 5 tries. So perhaps in 3 years?
 
  • Funny
Reactions: Cheburashka