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

How much are you willing to pay for FSD V12?

This site may earn commission on affiliate links.
My only problem with the proposal is that it makes it easier to accidentally interrupt software control. Without a torque sensor, just bumping the wheel could be enough to make the car go ballistic. I'd want a pretty obvious indication of whether the software was in control or not so that I'd subconsciously know. A small blue steering wheel in the corner of the display isn't nearly enough.

There wouldn't really be software control in the proposed mode, it's just you pointing the steering and the computer holding it unless you are "turning" too much and in such situation it would not hold. This sould be measured relative to the curvature of the road such that in a "bend" following the bend would be considered driving "straight" and it would hold along the bend.
 
There wouldn't really be software control in the proposed mode, it's just you pointing the steering and the computer holding it unless you are "turning" too much and in such situation it would not hold. This sould be measured relative to the curvature of the road such that in a "bend" following the bend would be considered driving "straight" and it would hold along the bend.
Perhaps if you presented a scenario or two where this would provide some unique value then I would understand better. What problem is this system addressing, or what functionality does it provide? All I'm getting is that the steering wheel doesn't turn when the car turns, and that the only time I use the steering wheel is when I want to do something other than follow the lane I'm in.
 
  • Like
Reactions: pilotSteve
Perhaps if you presented a scenario or two where this would provide some unique value then I would understand better. What problem is this system addressing, or what functionality does it provide? All I'm getting is that the steering wheel doesn't turn when the car turns, and that the only time I use the steering wheel is when I want to do something other than follow the lane I'm in.

Try letting go of the steering wheel while driving in any continuously consistent consistent (straight or nearly straight) direction, and you will understand the benefit of direction hold. )
 
Try letting go of the steering wheel while driving in any continuously consistent consistent (straight or nearly straight) direction, and you will understand the benefit of direction hold. )
How is this different from Autosteer?

Separately, I ran across the following text in the description for Full Self Driving on the Autopilot page.

From Home​

All you will need to do is get in and tell your car where to go. If you don’t say anything, your car will look at your calendar and take you there as the assumed destination. Your Tesla will figure out the optimal route, navigating urban streets, complex intersections and freeways.

To your Destination​

When you arrive at your destination, simply step out at the entrance and your car will enter park seek mode, automatically search for a spot and park itself. A tap on your phone summons it back to you.
I thought they had dropped those sorts of statements, but I guess they're still saying stuff in print about these near-robotaxi scenarios. Will they get there with HW3, will they just keep things under development until all HW3 cars are off the road, or will they upgrade HW3 cars? That scenario is pretty impressive.
 
  • Like
Reactions: pilotSteve
Perhaps if you presented a scenario or two where this would provide some unique value then I would understand better. What problem is this system addressing, or what functionality does it provide? All I'm getting is that the steering wheel doesn't turn when the car turns, and that the only time I use the steering wheel is when I want to do something other than follow the lane I'm in.
I'm not following the benefit of this, either.
 
How is this different from Autosteer?

The problem with autosteer is that it does not drive straight, try for example a heavy rain situation, it will be driving all over. This is just an extreme example, but not the only one. My preference would be a mode (which can be switched on or off) where it simply drives straight into the direction chosen by the driver, probably more closer to a drive by wire "feel".
 
The problem with autosteer is that it does not drive straight, try for example a heavy rain situation, it will be driving all over. This is just an extreme example, but not the only one. My preference would be a mode (which can be switched on or off) where it simply drives straight into the direction chosen by the driver, probably more closer to a drive by wire "feel".
So if I wanted to simulate this today, I would point the car and then "lock" the steering wheel in place with my knee, right? That would keep the car going straight or following a curve or whatever.
 
  • Like
Reactions: Tiger
The problem with autosteer is that it does not drive straight, try for example a heavy rain situation, it will be driving all over. This is just an extreme example, but not the only one. My preference would be a mode (which can be switched on or off) where it simply drives straight into the direction chosen by the driver, probably more closer to a drive by wire "feel".
Your premise is untrue. Autopilot does a great job of lane centering. It does not drive 'all over the place' when it is engaged.
 
  • Like
Reactions: pilotSteve
So if I wanted to simulate this today, I would point the car and then "lock" the steering wheel in place with my knee, right? That would keep the car going straight or following a curve or whatever.

Yes, exactly, I would like the autopilot to have a "[x] do just the knee driving for me please, but without the wiggle wobble" option.
 
Last edited:
Yes, exactly, I would like the autopilot to have a "[x] do just the knee driving for me please, but without the wiggle wobble" option.
Thanks for sticking with me through the explanation. Part of my difficulty in understanding was that my mind couldn't accept that you were really after such a feature :) If conditions are so challenging that Autosteer cannot handle it, then I'm not going to be holding a line with my knee, or even an automated knee.
 
  • Like
Reactions: GalacticHero
Thanks for sticking with me through the explanation. Part of my difficulty in understanding was that my mind couldn't accept that you were really after such a feature :) If conditions are so challenging that Autosteer cannot handle it, then I'm not going to be holding a line with my knee, or even an automated knee.

It's not the "conditions", it's the wide sticky summar tires that pull the steering on regular uneven paved surface. With studded winter tires the problem does not exist, they have less friction and steering is feather light.
 
The problem with autosteer is that it does not drive straight, try for example a heavy rain situation, it will be driving all over. This is just an extreme example, but not the only one. My preference would be a mode (which can be switched on or off) where it simply drives straight into the direction chosen by the driver, probably more closer to a drive by wire "feel".
I was in such a downpour a couple weeks back and was really impressed with how well the car did. It was one of those downpours where most cars slowed from 65 to 20/25. FSD kicked out and AP downgraded but the car kept in the lane better than probably would have. It also limited the speed quite severely. But then if I were driving, I'd probably do about the same or maybe a little lower.

Maybe I don't understand the issue either but I think it does a good job of driving when you have no destination and just double tap the stalk. I think that still counts as FSDb but I'm not sure.
 
I think Tesla will eventually need to split the service.

1. As a private drivers aid - probably $5k (or maybe on a per km subscription)

2. As a rideshare service, integrated into a platform (Tesla's or Uber/Lyft), but still requiring driver for supervision (for regulatory reasons) or to charge/ handle difficult areas. - maybe $20k

3. As a fully autonomous integrated rideshare service, including automated charging and a team of technicians who can take over - $$$$