AP1 owners take heart

[Apricot]

From another thread: My feelings are mixed of course. I took the car on Sept 11 and it is fantastic. So of course it hurts to miss the refresh by six weeks. But AP1 is already so far beyond other cars.

Now I was thinking about whether to buy the new car and trade in the current. But I'm going to wait because:

• AP2 isn't even planned to be better until 2017.
• Given Elon's optimism, it probably will still be catching up well into 2017.
• And given Tesla's takeover of the whole software, i imagine they'll be working out significant bugs for a while compared to AP1. I mean, AP1 has its quirks, but I know them and accept them. AP2 quirks could be something less acceptable to me like "don't use it in fog".
• Also, AP2 will take a while to be significantly, trustably better. I'm not even sure what features would make me want to upgrade. I want to see a lot of data before trusting AP2 to handle intersections and lights. So we're already talking about a year plus.
• And level 5 autonomy? That will take years for approval, and every accidental death is going to set back the process.
• The whole door-to-door thing sounds nice, but I don't trust the current nav to get me anywhere nontrivial. If Tesla improves that, it will improve for all cars.
• Auto-Find-parking would be a KILLER feature, but I really have a tough time believing this feature will ever work. I mean, how will it know where to look? Current auto park doesn't even find most spots, and often offers to park in driveways. Lots of parking signs are messed up. And time limits? Paying for a ticket?

So in the end, I mainly want AP1 features to be better! More stable, handle more situations. As long as I have to be alert to intercede, AP2 will only be an incremental upgrade. And it will be a long time, maybe forever, before I trust full autonomy, as in sleeping the whole way or working on a laptop.

So in the end, I'm a little sad, but also glad to not be alpha testing AP2 either. And I'll be watching the trade-in market of course...

-

Having thought a bit more about auto park , I really don't think it's going to be possible. The car just can't deal with the huge range of possible situations and human interactions. And what about bozos throwing themselves in front of driverless cars looking for a lawsuit? This feels much further away than hands-free coast to coast driving w backup driver in seat.

 

[AWDtsla]

• And level 5 autonomy? That will take years for approval, and every accidental death is going to set back the process.

Not a chance in hell that the AP 2 suite will be actually able to do level 5. Just like AP 1 can't read stop signs or look at traffic lights, or "bring you from onramp to offramp without touching anything."

If you're lucky, maybe a CPU upgrade will do it, but I bet there will be an upgraded sensor suite.

 

[MP3Mike]

And given Tesla's takeover of the whole software, i imagine they'll be working out significant bugs for a while compared to AP1. I mean, AP1 has its quirks, but I know them and accept them. AP2 quirks could be something less acceptable to me like "don't use it in fog".

Doesn't AP1 already have a "don't use it in fog" restriction? How would it see the lines? (Assuming heavy fog, and you weren't following another car running you off a cliff.)

 

[Canuck]

AP2 quirks could be something less acceptable to me like "don't use it in fog".

This is from the press release:

"A forward-facing radar with enhanced processing provides additional data about the world on a redundant wavelength, capable of seeing through heavy rain, fog, dust and even the car ahead."

 

[MP3Mike]

This is from the press release:

"A forward-facing radar with enhanced processing provides additional data about the world on a redundant wavelength, capable of seeing through heavy rain, fog, dust and even the car ahead."

Sure that helps it see obstacles in front of you, but it doesn't help it see the lanes that it needs to stay in or corners that it needs to navigate.

 

[WannabeOwner]

Do you really want to give up the driving control to the computer and just be a passenger for your trip?

I would ... I would rate myself as a keen driver, I never mind driving, but I regard it as the biggest waste of time in my whole life. I'd prefer to be doing something else - reading, working, or even just catching up on some sleep.

hire a personal chauffeur

I'm a good driver (using my own personal scale of 1-to-10 !!). A professional chauffeur will be no better, maybe worse ... Lady Diana's to take an example ... I'd prefer a computer system that improved mile-on-mile over a human any time. Right now, on AP1, I'm happy that I have AP + Human and the best of both worlds, as soon as AP2 gets to the point of negligible failures then, for me, there's no point me, the Human part, getting involved at all.

 

[Pruitt]

For some reason I seem to be the only one that finds it's funny that AP 2.0 is all done in-house by Tesla. The same Tesla that can't get a USB music player to work.

I don't find it funny - I find it a bit disquieting.

 

[jgs]

Having thought a bit more about auto park , I really don't think it's going to be possible. The car just can't deal with the huge range of possible situations and human interactions.

I suspect it'll be able to park itself eventually but it'll be greatly restricted in the spaces it can use compared to a human driver. As I noted in my earlier post, if the fleet of autonomous cars explodes, I think it's likely the driving environment (including parking) will be adapted to accommodate the limitations of the cars, but until then (for example) your autonomous car isn't going to be parking in the structure I generally use because... it won't be able to pull a ticket to cause the gate to open. This, even though that structure is otherwise a very well-behaved space from the point of view of navigation and traffic flow. I'm sure there are millions of similar examples to be had.

 

[AWDtsla]

AP 2.0 owners are already getting overpromised undelivered. Look at where the camera's are. No mechanism to wipe and probably no mechanism to heat? There goes level 4/5 autonomy in all winter weather areas. "told you so"

Can you spot the self-driving cameras on this Tesla Model S?

 

[Neuromancer]

Sure that helps it see obstacles in front of you, but it doesn't help it see the lanes that it needs to stay in or corners that it needs to navigate.

This is why Elon has referenced much higher resolution maps being necessary for full autonomy. With high resolution maps, and high accuracy GPS, then all you need is the radar to make sure there's not something in that lane that shouldn't be there.

Driving in fog too heavy to see is dangerous for human drivers. I am pretty confident that high res maps, white lists, long range sonar, and cloud point radar will be better driving in the fog than Aunt Ethyl.

This is part of what he is referring to when he says enhanced 'software' is a requirement for full autonomy.

 

[MP3Mike]

AP 2.0 owners are already getting overpromised undelivered. Look at where the camera's are. No mechanism to wipe and probably no mechanism to heat?

On the call Elon said that all cameras have heated surfaces.

 

[AWDtsla]

On the call Elon said that all cameras have heated surfaces.

Good. Now they need wipers. Takes about 30 seconds on a salted road for that to become crusted over.

 

[MP3Mike]

Good. Now they need wipers. Takes about 30 seconds on a salted road for that to become crusted over.

He said they were positioned/designed such to avoid them getting covered/blocked. I guess time will tell how well they did on that.

 

[AWDtsla]

He said they were positioned/designed such to avoid them getting covered/blocked. I guess time will tell how well they did on that.

LOL

 

[golfpilot]

My thinking was that this announcement was also intended to get Bolt prospective owners to ask themselves if they might want to be able to upgrade to a self-driving car, without having to sell and buy a different car

I'm looking at new cars now and the Bolt isn't even a question. I drive 200 mile trips very often and in the event that I do need a recharge, it is a slow experience with a Bolt. Tesla, if it is looking close I can pull into the supercharger, relieve my bladder, and have 40 extra miles.

I really don't think the Bolt is going to hurt Tesla sales at all. Leaf sales, yes.

 

[jya*]

LOL

He said they were positioned/designed such to avoid them getting covered/blocked. I guess time will tell how well they did on that.

Yeah, road salt and mud have a way of only going where you want it to go.
I wouldn't be ok with my car notifying me that it needs to be rescued because it sloshed in a big puddle and can't see where it's going. Anything on the outside of the car is susceptible to being blocked or vandalized. If the car can limp home using non-visual sensors, that would be ideal.
Individual wipers could be done; presumably could be a retrofit. This is why I'm a bit hesitant to buy until they've achieved full L5, no matter how confident they are that the current hardware is sufficient.

 

[Xenius]

I thought I'd be pissed about AP2.0, and then I remembered how long it took them to release AP 1.0. So while I am a little jealous, not going to lie, I'm more or less OK with this announcement. AP 1.0 wont be outdated for a few more years, at best.

That and I plan to get a Model 3, so this makes it more likely that the wife gets the S and I keep the 3 .

In the same situation. Now I'm questioning whether I'll want to give the wife the S and use the 3 myself. Decisions decisions. It also highly depends on if we get a Model Y around the time of the 3 release (a compact suv) because my wife prefers that style. But between now and then I'm sure she could be convinced to take a sedan, and we can keep her stupid Subaru around for hauling junk.

 

[jgs]

This is why I'm a bit hesitant to buy until they've achieved full L5

Yeah, if L5 is a firm requirement for you, you should not buy until it's actually shipping. I bought my S after AP was announced but before it shipped... well before, as it turned out, because Elon time. Although I was very happy to get the AP features when they arrived, my purchase decision was based on the car as delivered. OTOH if I didn't already have an S and were making a buying decision today, I'd pull the trigger without hesitation for the same reasons -- the car is already pretty great even if they don't fully deliver on the promises.

I do think it's pretty cheeky (and risky) for Tesla to promise some of the stuff they do, and I think "caveat emptor" applies.

 

[mediocreguy]

I have a recent S60 - I am really pleased with the car - I love BEV cars and the way they drive.

I have AP 1.0, but I'm not really comfortable with it yet:

1. The auto-steer has a difficult time maintaining lanes over a crest. I have a few spots on my commute that I have to take over after the car starts moving from side to side. I generally drive the speed limit, so the car usually doesn't have a car to follow - this would probably improve the problem.
2. The auto-steer hugs the lane division marker too close for my comfort - especially when other cars are passing (I'm an edge-hugger). I wish they could provide some sort of bias parameter to adjust that.
3. The car brakes too hard and too far away from cars ahead turning right - even when there's a turn lane. I'm always afraid that someone is going to rear-end me when this happens.

Maybe I just need more time to get used to the AP - although I'm sure they will continue to improve AP 1.0 software for the foreseeable future.

I'm looking forward to the progress Tesla will undoubtedly make in the coming months and years - I will definitely consider trading for an AP 2.0 car once the technology is proven. Meanwhile I'm definitely going to enjoy one of the most technologically advanced production vehicles on the market today!

 

[eye.surgeon]

Good. Now they need wipers. Takes about 30 seconds on a salted road for that to become crusted over.

I'm sure they never even thought of that