Elon: "Feature complete for full self driving this year"

[emmz0r]

Well, the closest case I've been in has been those very sharp turns in an off ramp. Auto steer can handle those very sharp turns very well at low speeds. And I just hover a hand over the wheel. I don't need to actually hold the wheel at all. I could see the same with "automatic driving on city streets" where you get a few nags before a turn where you can gently tug the wheel but you don't need to hold the wheel during the turn. That would work I think.

Then it's not quite automatic since you have to sit there and hover your hands over the wheel. It's more stressful than relaxing. But I guess that's how it has to be in the start. I don't think it will become fully self driving in the city

 

[EVNow]

It actually doesn't. His paper is much more nuanced. Note e.g. this sentence:

"Also, if system reliability were to improve significantly over time, the levels of functional vigilance observed here may well change."

Already addresses this. The original comment was about L2.

Even rationally, it makes sense that if one is called upon to act in seconds but only once a week, it would be very difficult.

I'd say this is similar to how when you are driving on a barren freeway in Arizona and suddenly an animal runs onto the road from nowhere. Functional vigilance is similarly deteriorated and it is unlikely we can act very quickly.

Current Tesla EAP is such that I tend to be more attentive and pay more attention when using it rather than when not using it

 

[EVNow]

Well, the closest case I've been in has been those very sharp turns in an off ramp. Auto steer can handle those very sharp turns very well at low speeds. And I just hover a hand over the wheel. I don't need to actually hold the wheel at all. I could see the same with "automatic driving on city streets" where you get a few nags before a turn where you can gently tug the wheel but you don't need to hold the wheel during the turn. That would work I think.

Low speeds is the key - does the car slow down on the ramps ?I've not tried very sharp ramps yet - but it keeps showing 60 mph limit for me - I don't want it to try the sharp ramp at 60.

BTW, I can report that EAP handles sharp turns on roads quite well. We have a lot of those in our mountainous area.

 

[emmz0r]

Low speeds is the key - does the car slow down on the ramps ?I've not tried very sharp ramps yet - but it keeps showing 60 mph limit for me - I don't want it to try the sharp ramp at 60.

BTW, I can report that EAP handles sharp turns on roads quite well. We have a lot of those in our mountainous area.

It slows down but it slows down to grandma speed. The same behavior in city traffic would not be acceptable.

 

[diplomat33]

Low speeds is the key - does the car slow down on the ramps ?I've not tried very sharp ramps yet - but it keeps showing 60 mph limit for me - I don't want it to try the sharp ramp at 60.

Yes, AP/NOA automatically slows down before the ramp. It will take a very sharp ramp at about 25 mph.

 

[daniel]

On my long road trip last summer EAP did the driving maybe 95% of the time. But there was never a moment when I felt that I could relax my vigilance. The wonderful thing about EAP is that when my brain did not have to micro-manage staying in the lane (which we do subconsciously, but which uses brain power) I was able to be more vigilant about possible road hazards or erratic drivers.

EAP is so marvelously wonderful that I am unable express how much I love it, and how much more relaxing and safe it makes my driving. But what it is capable of does not point to the early appearance of even Level 3, much Level 4 or 5. Tesla may have a Level 4 system in development to make good on Elon's promises, but it would have to be as far beyond EAP as the Falcon Heavy is beyond a remote-controlled model glider.

Elon will keep his promises. But he will miss his self-imposed deadline by a decade. I suspect that his entirely unrealistic chrono-optimism is part of the reason he has accomplished so much. I would be willing to bet five inflation-adjusted dollars that the first car an ordinary consumer can buy with non-geofenced Level 4 (take a nap while it drives you anywhere passengers cars are allowed to drive on-road) is ten years from today.

 

[mongo]

. I would be willing to bet five inflation-adjusted dollars that the first car an ordinary consumer can buy with non-geofenced Level 4 (take a nap while it drives you anywhere passengers cars are allowed to drive on-road) is ten years from today.

So you're in the 'Tesla will stop selling cars once they start making robo-taxis' camp?

On my long road trip last summer EAP did the driving maybe 95% of the time. But there was never a moment when I felt that I could relax my vigilance. The wonderful thing about EAP is that when my brain did not have to micro-manage staying in the lane (which we do subconsciously, but which uses brain power) I was able to be more vigilant about possible road hazards or erratic drivers.

Agree with your feelings: our non-Telsa has adaptive cruise and weak lane assist, but those make driving so much better. Especially smart cruise, no need to keep checking speedometer and hold my foot up. Would love if it worked in stop and go...

 

[daniel]

So you're in the 'Tesla will stop selling cars once they start making robo-taxis' camp?

On the contrary. I don't think Tesla will want to own robo-taxis. It will be much more profitable to sell robo-taxi-capable cars to customers. The margins on ride-sharing are so thin that by some accounts, Uber drivers lose money if you factor in depreciation on the car.

 

[EVNow]

On the contrary. I don't think Tesla will want to own robo-taxis. It will be much more profitable to sell robo-taxi-capable cars to customers. The margins on ride-sharing are so thin that by some accounts, Uber drivers lose money if you factor in depreciation on the car.

Not really. You should checkout the model put out by ARK.

 

[ChrML]

Then it's not quite automatic since you have to sit there and hover your hands over the wheel. It's more stressful than relaxing. But I guess that's how it has to be in the start. I don't think it will become fully self driving in the city

I think this is a non-issue. When the steering wheel turns a lot, just keep a loose grip around the wheel and let it slide. If something goes wrong, you just tighten the grip and take over.

 

[rnortman]

Elon will keep his promises. But he will miss his self-imposed deadline by a decade.

Ah, but this is a contradiction. He cannot keep his promises to people who bought their cars in 2016, especially those who leased, if he only delivers L4/L5 autonomy in 2027. Very very few people who bought AP2.x cars will still own them, and their useful lives will be significantly reduced if they're still around at all. And, honesly, even if you do hold onto that car until 2027, you really think Tesla will have a retrofit option for AP6 onto your AP2 vehicle? Nope. That's a broken promise, even when Tesla does eventually deliver an L4/L5 system.

 

[EVNow]

Except for the folks that took rides after Autonomous Day.

Not even them ! Listen to Tesla specific podcasts (or even ARK analyst).

 

[EVNow]

If we are talking about L3/4 in all cases, I agree. But I think Tesla could do L3 now in limited situations if they really wanted to. For example, Tesla could make Autopilot L3 for traffic jams. And NOA could be L3 now for cases like open highways with no traffic. NOA is actually pretty close to self-driving now, there are just few cases, like with heavy traffic, where it is not reliable enough yet.

I think the big reason why Tesla does not do this is because they want their "general solution". They want FSD to work broadly on all roads, rather solve FSD in just a few narrow cases.

Well, its very difficult to explain to drivers in what specific traffic condition NOA is L3 and what it is L2. Confusing the drivers is the worst thing they can do in terms of safety. It has to be very simple and clear. Basically a jury of our "peers" should not find it confusing.

That is why I think Tesla will continue to insist driver has to be vigilant until they are sure they can manager the liability.

 

[daniel]

Not really. You should checkout the model put out by ARK.

That link does nothing for me but download a zip file. I choose not to open zip files from the internet. I trust you as you've been posting here since forever. But I have some rules I just never break.

Ah, but this is a contradiction. He cannot keep his promises to people who bought their cars in 2016, especially those who leased, if he only delivers L4/L5 autonomy in 2027. Very very few people who bought AP2.x cars will still own them, and their useful lives will be significantly reduced if they're still around at all. And, honesly, even if you do hold onto that car until 2027, you really think Tesla will have a retrofit option for AP6 onto your AP2 vehicle? Nope. That's a broken promise, even when Tesla does eventually deliver an L4/L5 system.

Well, that's why I didn't pay for FSD.

But, yes, you make a very good point. I need to modify what I said: Elon will deliver what he says he'll deliver, but not when he says he'll deliver it. This means a lot of people who paid for FSD will get something Tesla calls its "FSD Package" but won't be driverless operation while they own the car they paid for it on.

Who, knows: Maybe in five years Tesla will refund the money people paid for FSD if (as I believe) Tesla is not yet able to make their car into the promised robo-taxi. Or maybe Tesla will credit them for the FSD package on their next Tesla. In my opinion, it was a big mistake to promise to turn the 2019 Model 3 into a robo-taxi. But again, this is the same optimism that made Tesla possible in the first place. Elon is not the FSM and this is both his strength and his flaw.

 

[wcorey]

But wait...
Auto Park has a serious flaw. It determines a valid space to be between two parked vehicles not either an edge or a side of an empty rectangle.
______________
| | | |
| | | |
| | | |
_______________

or
_________________
| | |

For myself, I would not consider that feature/story complete unless one could stipulate farthest away no surrounding no immediately adjacent vehicles. Presumably 'come get me' summon would be working. This akin to the story drop me off at the mall then go park. When I signal come back to where you dropped me off. And it will have to handle vertical parking lots as well.. multi-floor parking garages.

OK, that didn't turn out near as well as I expected it would. Even if you reply to that post you'll see I added spacing to signify perpendicular spaces vs parallel spaces.

 

[wcorey]

Not even them ! Listen to Tesla specific podcasts (or even ARK analyst).

I did listen to the full Ark interview w/Musk. I don't recall anything specific on the post Investor Day test drives. I did just go looking for anything on that. It seems odd I didn't find any independent. I was going by 'reported' responses....which...well, you know. Frankly I vaguely recall they had to fill out paperwork to get that ride which likely was NDA. Just as the employees testing had to. BTW, the 2016 vid, wasn't a test drive it was an ad and from what I read it was likely a heavily 'learned' route. It's clear you don't see it coming. I believe we'll see it this year. As for what SAE level it is...who the F cares. If I can get into my car, tell it to go to Stop & Shop (a local grocery chain) and worst case all I have to do is put about 1 lb of torque on the steering wheel to signify I am paying attn, I'll call that FSD. If I travel from New England to NJ or MD or PA and it successfully navigates the George Washington Bridge Plaza I'll call that self driving too. It'll take regulators, be they state or if Congress preempts regulatory OK, to allpw Tesla to drop "Beta" nomenclature and drop the nags. SAE doesn't get a vote on that. SAE levels are great for those taking engineering classes.
If I am wrong, you're allowed to say I told you so. If you are, I will say I told you so. My believing it won't make it happen. But, there are only so many times he can go back to that same well and if December comes and goes and there is no non-interstate NoA with traffic sign/light recognition, speed limit sign recognition, yield sign recognition, automated turn at intersections or going straight through after stopping... it will be reflected in the price of the stock, shorts will say "We told you so".

 

[EVNow]

That link does nothing for me but download a zip file. I choose not to open zip files from the internet. I trust you as you've been posting here since forever. But I have some rules I just never break.

Indulging you by searching for the original link.

ARK Contributes Its Financial Autonomous Taxi Model to GitHub

ps : Tesla EAP - ok. Zip files - definitely not

 

[EVNow]

I did listen to the full Ark interview w/Musk. I don't recall anything specific on the post Investor Day test drives. I did just go looking for anything on that. It seems odd I didn't find any independent.

Checkout HyperTV interview with ARK analyst.

 

[Daniel in SD]

I did listen to the full Ark interview w/Musk. I don't recall anything specific on the post Investor Day test drives. I did just go looking for anything on that. It seems odd I didn't find any independent. I was going by 'reported' responses....which...well, you know. Frankly I vaguely recall they had to fill out paperwork to get that ride which likely was NDA. Just as the employees testing had to. BTW, the 2016 vid, wasn't a test drive it was an ad and from what I read it was likely a heavily 'learned' route. It's clear you don't see it coming. I believe we'll see it this year. As for what SAE level it is...who the F cares. If I can get into my car, tell it to go to Stop & Shop (a local grocery chain) and worst case all I have to do is put about 1 lb of torque on the steering wheel to signify I am paying attn, I'll call that FSD. If I travel from New England to NJ or MD or PA and it successfully navigates the George Washington Bridge Plaza I'll call that self driving too. It'll take regulators, be they state or if Congress preempts regulatory OK, to allpw Tesla to drop "Beta" nomenclature and drop the nags. SAE doesn't get a vote on that. SAE levels are great for those taking engineering classes.
If I am wrong, you're allowed to say I told you so. If you are, I will say I told you so. My believing it won't make it happen. But, there are only so many times he can go back to that same well and if December comes and goes and there is no non-interstate NoA with traffic sign/light recognition, speed limit sign recognition, yield sign recognition, automated turn at intersections or going straight through after stopping... it will be reflected in the price of the stock, shorts will say "We told you so".

In California at least the regulations refer to SAE levels. Anything that requires driver attention is not an autonomous vehicle (or FSD IMHO).

 

[wcorey]

In California at least the regulations refer to SAE levels. Anything that requires driver attention is not an autonomous vehicle (or FSD IMHO).

Huh! That's surprising. To lose nags and driver attn (lawyers, if no one else, would demand for legal reasons) would require regulator approval or Congress. As I said, for my purposes, if the car will get me from point A to point B without me being the physical driver, I am OK with attentiveness and I'd consider it FSD or close enough for government work. Consider this though. You and I might disagree on FSD or why but, IMHO, the real big drivers of what'll happen are UPS, FedEx and the long haul companies. There is money to be made by eliminating the trucker, limited by regulations to only so many hours / day...breaks, sleep etc. If they can encourage Tesla for a solution (their tractor trailer cab) they'll pay big. So, I think it's game-on. I think a lot will happen in the next 6 months. I could be wrong but there is serious money on the table now.