Certainly for the non-technical, you're correct. I spent a career in software design and development. So I am getting access to a technology without incurring the cost of that technology once it is flushed out. I believe Tesla will have a level 4, perhaps 5, solution by end of next year, including neighborhood driving. I'm also using my professional determination of where they are and where they have left to go.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Don't take your hands off the wheel
- Thread starter smatthew
- Start date
patrick80132
Member
Yep, I agree. What I tend to do is compensate for centripetal force. Taking a clockwise turn centripetal force pushes the car to the outside (towards the opposing lane. With a counter clockwise turn the force tries to push the car out of the lane into the barrier or trees or some guy's lawn. So I tend to hug the divider going ccw and the curb going clockwise. But there are exceptions to every rule.
No, they get data even if you didn't pay for it or use it. They call it shadow mode. Every time you disengage AP that disengagement is recorded and forwarded. This is how they capture edge cases.
There is a new roundabout in Verdi, NV, put there about a year ago, that TACC and Auto steer completely ignore if you don't take over. It will drive you directly into the center island at 45 MPH.
I don't know that moderators monitor threads. However it appears the individual you directly referenced is one I don't see comments from at all. Maybe that's success! My only point here if on individual posts you can report it which would get the moderator's attention.
N5329K
Active Member
A life of flying has totally spoiled me to the whole concept of "performance car." You look down at the clogged freeways at cruise and you think, Performance? Are you joking?great, and come to think of it. I'd rather be in the air with AP than on the ground in a car AP
That said, I have had more than a few airplanes that could be outrun by a real high-performance automobile. My first one, an Ercoupe, cruised (landed, and took off) at about 90 MPH.
Robin
This is precisely why I shudder at the thought of flying cars. The thing about flying is virtually all pilots have a base level of training and recurring training. The higher your altitude the less worries you have about 'the other guy'. But with flying cars it'll be like a pinball machine.
Silicon Desert
Well-Known Member
Class A airspace yes. But I've spent a lot of time right below 12,000' whether IFR or VFR and it's pretty thin air there, no pun intended. But specifically, I was referring to and hoping flying cars are limited to class G airspace.
N5329K
Active Member
Even in “congested” airspace, it’s still hard to find something to hit in an airplane. Except at OSH.
Robin
We still need to get to level 3 first before we can start making estimations about higher levels of automation. In terms of when we'll get to future levels of automation, whether its five years out or even ten years, that's nothing (very little time) in the global scale of things. I don't believe there is much point debating whether it will be off by one or two years. As an aside, software developers are regularly off by more than 50% when it comes to estimations of generally available (GA) releases.
By this point, we will likely be pushing for alternate modes of transportation in congested areas. Unless we can reduce the number of vehicles traveling on surface roads, I'm not so sure this form of automation will be the soul-saving technology we hope it will be. Today I traveled more than three hours to cover 30 miles of range. The first hour of that commute covered 5 to 6 miles. Automation takes some of that edge off, but it still doesn't get us where we need to be more quickly. Eventually, smart cities will interface with smart cars to help control traffic congestion, but early studies are showing this is only a marginal benefit.
strider
Active Member
Elon said they would do a coast to coast FSD drive this year. That's way beyond feature complete. That means it works in the real world without human intervention. If the human has to do anything more than plug and unplug the charging cable then it's not FSD.
Just curious, are you in the software field? That's not what feature complete means. It means all the features of the software are complete, there is no planned stuff that hasn't been written. It doesn't mean GA, it doesn't mean Beta.
A car can't do coast to coast if, for no other reasons, they don't have snake chargers yet.
It's pretty well documented what FSD means. If you want to short TESLA, go for it.
Not the ones that know what they are doing, they pretty much can nail it. Now, that said, that does not mean the Product Mgr or VP of Dev have a clue. In this case I'd put money on Elon is pretty hands on with the developers.As an aside, software developers are regularly off by more than 50% when it comes to estimations of generally available (GA) releases.
The key to accurately estimating lies in problem decomposition. Ya don't stick your wet finger in the air. People aren't born knowing how to do it but the ones that advance learn how before they advance.
Not that I want to call you out on it, but why don't you explain how you came up with one year to FSD with no exposure to their internal product roadmap, engineering notes, bugs, QA, testing, feedback from regulators and regularly scheduled progress updates from hardware and software teams?
If you want to say you believe Elon Musk's timeline, that's perfectly OK to do - but if you're basing it off our own industry experience, you must know the risks of doing so without data?
When I hear Elon make a projection, I think, "OK, this is the best case scenario - perhaps more of a goal worth aspiring to." The idea of setting lofty goals is good, because otherwise if you set modest goals, you may not even reach those. As a rule of thumb, I try not to take these timelines at face value, because I'll set myself up for disappointment. This is because experience has shown that Tesla's delivery timeframes slip, and even Musk jokes about it. Under promise, and over-deliver is a mantra I prefer to see.
If you like, I'll bet you a dollar ;-) But here, we must agree to what FSD actually is, since there is plenty of FUD around that too.
Fernand
Active Member
OK, boys and girls, I want to tell you ALL about how
we nearly crashed tonight, Nicki (my Model 3) and I.
Hello, Ditch ? no, not this time.
I think it allowed me to understand the (few) cases
I've had of the "aiming for the edge of the road" type.
Needless to say, just as in all the horror stories
that the media like to spread, it was in AutoSteer,
nay, all the way in the dread Navigate On Autopilot.
On a familiar route.
So, what happened? Nicki wanted to do a lane
change to prepare for an upcoming freeway split,
and so s/he started. But then ... started crossing
two lanes going sharp right towards the shoulder
when I hurriedly grabbed the reins. Scary stuff.
[Back up to an hour earlier] That's when my little
90 gram bag of BBs held with velcro at the meet
of the cross-arm and the wheel fell off. No time
to re-strap it, I'll do it right later. It happens that
this 90 grams is just enough to prevent the blue
nags as long as I keep a hand on the wheel.
Without that counterweight, I was having a hard
time giving enough resistance in any comfortable
position without dropping out of AutoSteer.
So I was driving without the counterweight. And
this time I was watching. And so was my wife.
And that's exactly what happened. As lil' Nicki
started into the lane change, wheel turning right,
I resisted too much, or jerked, and I was OUT
of AutoSteer ... heading for the ditch. Doh.
Ever since 2019.12.1.2 when I installed the little
bag of BBs, I've been a very happy camper all
around, NO anomalies of this (or any other) sort.
Because I haven't dropped out of AutoSteer.
I know some people, most people, have worked
out ways to give the wheel just the right torque.
And that's the kosher way. More power to ya.
Not I, not me, not us.
I could curse or worse,
or badmouth the code,
but I'll put back the weights
and we'll be back on the road.
Lesson learned.
we nearly crashed tonight, Nicki (my Model 3) and I.
Hello, Ditch ? no, not this time.
I think it allowed me to understand the (few) cases
I've had of the "aiming for the edge of the road" type.
Needless to say, just as in all the horror stories
that the media like to spread, it was in AutoSteer,
nay, all the way in the dread Navigate On Autopilot.
On a familiar route.
So, what happened? Nicki wanted to do a lane
change to prepare for an upcoming freeway split,
and so s/he started. But then ... started crossing
two lanes going sharp right towards the shoulder
when I hurriedly grabbed the reins. Scary stuff.
[Back up to an hour earlier] That's when my little
90 gram bag of BBs held with velcro at the meet
of the cross-arm and the wheel fell off. No time
to re-strap it, I'll do it right later. It happens that
this 90 grams is just enough to prevent the blue
nags as long as I keep a hand on the wheel.
Without that counterweight, I was having a hard
time giving enough resistance in any comfortable
position without dropping out of AutoSteer.
So I was driving without the counterweight. And
this time I was watching. And so was my wife.
And that's exactly what happened. As lil' Nicki
started into the lane change, wheel turning right,
I resisted too much, or jerked, and I was OUT
of AutoSteer ... heading for the ditch. Doh.
Ever since 2019.12.1.2 when I installed the little
bag of BBs, I've been a very happy camper all
around, NO anomalies of this (or any other) sort.
Because I haven't dropped out of AutoSteer.
I know some people, most people, have worked
out ways to give the wheel just the right torque.
And that's the kosher way. More power to ya.
Not I, not me, not us.
I could curse or worse,
or badmouth the code,
but I'll put back the weights
and we'll be back on the road.
Lesson learned.