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

When do you think the Model 3 will no longer require a driver?

When will the Model 3 no longer require a driver?

  • In less than a year.

    Votes: 1 0.9%
  • In between one and two years.

    Votes: 2 1.8%
  • In between two and three years.

    Votes: 5 4.4%
  • In between three and four years.

    Votes: 8 7.0%
  • In between four and five years.

    Votes: 11 9.6%
  • In between five and ten years.

    Votes: 45 39.5%
  • More than ten but less than twenty years.

    Votes: 17 14.9%
  • More than twenty years.

    Votes: 6 5.3%
  • Never.

    Votes: 17 14.9%
  • Other. Comment in a post.

    Votes: 2 1.8%

  • Total voters
    114
This site may earn commission on affiliate links.
Personally, before it can actually become a reality, I think they will need to adopt a system where specific roads are certified as autonomous ready, which will limit those roads to a strict set of rules (on markings, shoulders, how road construction is handled, etc) that the autonomous cars can rely on.
 
Personally, before it can actually become a reality, I think they will need to adopt a system where specific roads are certified as autonomous ready, which will limit those roads to a strict set of rules (on markings, shoulders, how road construction is handled, etc) that the autonomous cars can rely on.
No way, if they were going to do that they would have done it in the 80s when they were first testing autonomous cars.

These are all made to interact on the road with human drivers, otherwise they'd be useless for the actual task of full self driving.
 
No way, if they were going to do that they would have done it in the 80s when they were first testing autonomous cars.

These are all made to interact on the road with human drivers, otherwise they'd be useless for the actual task of full self driving.
You could also argue that if they were going to land rocket boosters, they would have done it decades ago when we were sending men to the moon.

Limiting roads to a specific set of rules designed for autonomous driving makes the job of autonomous driving a lot easier and require a lot less processing power. In my opinion, it's the right move, whether people go for it or not.

And in case you didn't fully understand my position, I wasn't saying that human drivers couldn't also drive on such roads -- just that the roads would have specific rules on markings/etc that autonomous systems could depend on. Let me give you an example -- there is a 2 lane road I drive on that was recently re-paved, and it's been open with absolutely no lane markings for weeks now. That's fine for biological thinking super-computers (known as humans), but would be inexcusable for a road intended for autonomous drive systems. Sure, you can make a high powered computer handle that, but why should they have to? Just paint the dang lines.

As is now, this software problem has poorly defined unbounded requirements/constraints. That rarely turns out well.
 
You could also argue that if they were going to land rocket boosters, they would have done it decades ago when we were sending men to the moon.

Limiting roads to a specific set of rules designed for autonomous driving makes the job of autonomous driving a lot easier and require a lot less processing power. In my opinion, it's the right move, whether people go for it or not.

And in case you didn't fully understand my position, I wasn't saying that human drivers couldn't also drive on such roads -- just that the roads would have specific rules on markings/etc that autonomous systems could depend on. Let me give you an example -- there is a 2 lane road I drive on that was recently re-paved, and it's been open with absolutely no lane markings for weeks now. That's fine for biological thinking super-computers (known as humans), but would be inexcusable for a road intended for autonomous drive systems. Sure, you can make a high powered computer handle that, but why should they have to? Just paint the dang lines.

As is now, this software problem has poorly defined unbounded requirements/constraints. That rarely turns out well.
Infrastructure costs billions and roads are eaten away every winter. We need vehicles that can handle anything.

We let 16 yr olds on the road with far less experience than current autonomous cars.
 
There's over 4 million miles of road in the US- much of it already poorly maintained. The chances of redoing them all to make FSD easy is roughly 0.

Now add all the vast amounts of private property you'd have to redo (every parking lot in the country for example) and the fact you'd need 100% of pedestrians, drivers, and bike riders to obey all laws relevant to said roads and lots- then even 0 is looking pretty high.

If you can reliably solve for vision all that isn't needed.
 
  • Like
Reactions: JeffK