Autopilot lane keeping still not available over 6 months after delivery

[MarcG]

Enabled is a pretty clear to me. If it needs to be enabled with OTA updates, that means the car currently does not have it. At the very least, it would justify a question from a potential buyer.

Yes, and that question should be "when will the update that enables lane keeping and self parking be installed on my Model S if I purchase one". The problem is that, given the present tense used in other parts of the website, the "will be enabled" part could be interpreted as being available upon delivery. In my opinion, Tesla is walking a very fine marketing line here.

 

[breser]

I was sticking to the arguments as breser made them. He argued, (paraphrasing) that Musk is the over-reaching optimist, and that once the promises he has made can't be met, it's up to the "Tesla employees" (meaning everyone else at Tesla, presumably with Musk's OK) to "put things right." His conclusion was that there was little that could be done, short of correcting the website and hoping for better communication. My post was a counter-point to that argument.

I was thinking more of various employees going above and beyond to resolve specific issues with specific customers. I was trying to highlight that there are a lot of people within the organization that have the best of intentions. I wasn't suggesting that if an employee couldn't do something directly within their power that absolutely nothing could be done. In fact many times the resolutions that happen end up happening because the employee acts as an advocate for the customer.

However, I think there is a huge difference from employees working to do something for someone that has an unusual experience and essentially the majority of the last 9 months of production (I'm presuming the majority have paid for the Autopilot convenience features). I can't see them doing something that applies to all the people who have paid for Autopilot. At best a handful of really upset and pushy people might get something (maybe a free annual service or something like that). But I just don't see the masses getting anything. Even ignoring that, given the very real risk of legal action around this I really can't see Tesla doing anything that would look like an admission of guilt. I think the warranty extension falls very much into that area. It pretty much says "Look we wronged you by not having Autopilot available so we're going to extend your warranty so as to compensate."

I was also trying to explain that these sorts of compensation things really don't solve the issue. You guys want your cars to auto steer. Everything else that anyone has suggested is just a way to try and calm you down. But once you've gotten that compensation to calm you down at the end of the day it doesn't make your car steer itself.

I was trying to focus the effort into something that I thought would be constructive. Getting the website changed and working to get Tesla to be more clear in the future is something you can do. Arguing about the meanings of "several" and what the best way to compensate you for waiting longer than you think you should have to is not something that is likely to bear much fruit.

 

[gizmoboy]

Yes, and that question should be "when will the update that enables lane keeping and self parking be installed on my Model S if I purchase one". The problem is that, given the present tense used in other parts of the website, the "will be enabled" part could be interpreted as being available upon delivery. In my opinion, Tesla is walking a very fine marketing line here.

Agreed. Their verbiage here is clear if you already understand it, but subject to misinterpretation if you don't.

Their copy elsewhere on the site describing AP features is flat-out misleading and I think it does Tesla a disservice. However, precisely how damaging it is to their reputation and sales is something less clear to me.

 

[Lump]

Tesla doesn't need to resort to marketing hyperbole or puffery to sell its cars. The web site should not have the headline price "including gas savings" and should not use future autopilot features in the present tense.

Then why does it continue to happen?

I would assume a Director on the Board would object to these tactics but is being overruled.

 

[tomas]

Hmmm... lemme see. I bought a car in 2012 that was to be supported by a supercharger network to enable long distance travel. As I recall, the material was present tense, but it was pretty obvious that meant "the car was technically able, but the network needed to be built out." Most of the people who buy a $100k+ car are not dummies - they made a few good business decisions along the way in order to have the disposable income to make that move. I was going to say it is ludicrous, but since the word is taken, I'll use amazing. It is amazing to me that people would complain about being hoodwinked by the tense of market materials when the fact that this is evolutionary was quite obvious. Now, you can complain about how long it is taking... but anyone who has tracked Tesla since inception knows 3 things: Musk states aspirational and aggressive timelines, Tesla is usually late, but generally over-delivers in the end. Plus you get a lot of good stuff you were not expecting.

 

[gizmoboy]

Then why does it continue to happen?

I would assume a Director on the Board would object to these tactics but is being overruled.

I can easily see people being okay with rolling gas savings into the marketing material. It's more compelling, and if you believe it, then you'd have no qualms. If you don't believe it but it's legally defensible, then anything you do to juice sales is good for the company (unless you believe it will do more harm than good).

 

[wk057]

but anyone who has tracked Tesla since inception knows

And therein lies the problem. Not everyone has tracked Tesla since inception. Specifically the gentleman I spoke to yesterday didn't even know Tesla existed until recently (this year), like many other people out there.

 

[Lump]

I can easily see people being okay with rolling gas savings into the marketing material. It's more compelling, and if you believe it, then you'd have no qualms. If you don't believe it but it's legally defensible, then anything you do to juice sales is good for the company (unless you believe it will do more harm than good).

Totally OK using it for marketing, I believe most here would prefer if it was removed from the order tab.

 

[gizmoboy]

Totally OK using it for marketing, I believe most here would prefer if it was removed from the order tab.

That's still marketing, in my book.... until you click "submit," it's all marketing.

 

[Zzzz...]

Tesla should create opt in(in mytesla or anywhere) for beta test of 7.0. Just give warnings that using new AP features owners should be extra careful. But the features themselves are working. And I bet they are working better already compared to similar features from competitors.

 

[TexasEV]

That's still marketing, in my book.... until you click "submit," it's all marketing.

I agree it's totally appropriate to show how gas savings lowers cost of ownership over time, but it's not appropriate to subtract an assumed gas savings from the headline price of the car-- the first price someone sees on the web site. The price you see should be the price you would write the check for.

 

[rickgt]

Really like your suggestion!!! There are many of us that would sign releases to be able to test the 7.0 software and autopilot... add me to the list of volunteers!

Tesla should create opt in(in mytesla or anywhere) for beta test of 7.0. Just give warnings that using new AP features owners should be extra careful. But the features themselves are working. And I bet they are working better already compared to similar features from competitors.

 

[RyanT]

Tesla should create opt in(in mytesla or anywhere) for beta test of 7.0. Just give warnings that using new AP features owners should be extra careful. But the features themselves are working. And I bet they are working better already compared to similar features from competitors.

I like this idea. People who care enough could sign up while the casual user will just wait until the official update.

 

[kuttakamina]

Tesla should create opt in(in mytesla or anywhere) for beta test of 7.0. Just give warnings that using new AP features owners should be extra careful. But the features themselves are working. And I bet they are working better already compared to similar features from competitors.

Good idea, Roll it out as an opt-in beta. I'll sign up, as long as I can take control immediately if Autopilot doesn't behave.

 

[Max*]

I don't see Tesla doing this. The average person will get complacent while relying on a beta product, leading to accidents and bad press.

 

[commasign]

Tesla should create opt in(in mytesla or anywhere) for beta test of 7.0. Just give warnings that using new AP features owners should be extra careful. But the features themselves are working. And I bet they are working better already compared to similar features from competitors.

+1. Happy to sign on the dotted line if needed. Even if they want both hands on wheel once every 15 seconds, I'm fine with that.

 

[GlmnAlyAirCar]

Hmmm... lemme see. I bought a car in 2012 that was to be supported by a supercharger network to enable long distance travel. As I recall, the material was present tense, but it was pretty obvious that meant "the car was technically able, but the network needed to be built out." Most of the people who buy a $100k+ car are not dummies - they made a few good business decisions along the way in order to have the disposable income to make that move. I was going to say it is ludicrous, but since the word is taken, I'll use amazing. It is amazing to me that people would complain about being hoodwinked by the tense of market materials when the fact that this is evolutionary was quite obvious. Now, you can complain about how long it is taking... but anyone who has tracked Tesla since inception knows 3 things: Musk states aspirational and aggressive timelines, Tesla is usually late, but generally over-delivers in the end. Plus you get a lot of good stuff you were not expecting.

Thanks for weighing in. It's good to see a little rationality injected into this thread.

 

[donv]

This is a great post, but I disagree about some things.

First of all, Google and Urmson are talking their book. They have made choices about how to do things, and have a serious business (not technology) problem. Their problem is that many of their autonomous car efforts have been leapfrogged by actual automotive suppliers, with more experience at real time computing than Google has.

Why would any car company buy an autonomous system from Google, when they could buy it from Bosch, Delphi, Mobileye-- all companies with proven histories as automotive suppliers, who know how to work within the supply chain, have massive experience with real time software development?

So Google has to somehow change the game, which is why they a) make it seem really hard, and b) have decided to build their own cars, going the full autonomous route (which, to be fair, the companies I mentioned above have so far shied away from).

Secondly, in regard to regulators, there is certainly some regulatory peril here. However, it is pretty clear that, as of now, the features Tesla has proposed are nothing more than an enhanced form of cruise control. The driver is still responsible for the car. This is clearly legal in all states, with the possible exception of New York (where a hand has to be on the wheel).

I have good reason to believe this a significant reason for the delay. Ever noticed that in all the slick demos Tesla really didn't demo the interface? How does the car know you're actually paying attention? The truth is that Google is at least partially right. The better the driver assistance features get the worse drivers will be. See the video below for Google's argument against what Tesla is doing.

I wouldn't be surprised if there wasn't serious discussions ongoing with regulators about what Tesla had planned. If they can't convince the regulators that what they're giving us is safe then there's a very good chance that they'll put rules in place. Rules that very well might force Tesla to take the feature away or make it not as nice to use. So I'm guessing they're probably taking their time to make sure they don't give and then take it back. I'm sure we all remember how people reacted to the air suspension height being raised. I'm sure the reaction to such a change would be even worse.

Someone will again point out that other cars have these features. I'm just going to point out that Elon has a different philosophy about how this feature works, see his comments about keeping your hand on the steering wheel.

- - - Updated - - -

Also, I should point out that this fear of the steering portion of autopilot causing all these accidents seems rather odd to me. TACC is just as likely, maybe more likely, to cause an accident by inappropriately accelerating or braking as the car is to steering into something. Frankly, not steering into objects, given a camera and radar system, seems fairly easy to me.

It might not hold the center of the lane as smoothly as we would like, or it might disengage more often than we would like, but I suspect that the likelihood of it causing some sort of accident is rather low.

 

[breser]

First of all, Google and Urmson are talking their book. They have made choices about how to do things, and have a serious business (not technology) problem. Their problem is that many of their autonomous car efforts have been leapfrogged by actual automotive suppliers, with more experience at real time computing than Google has.

You're jumping to a lot of conclusions about what I meant that I didn't say and entirely missing the point I was making in sharing the Google video. I don't really agree with Google on everything. For instance I think his argument about driving assistance features only having to do the right thing once every 100,000 miles versus a self driving car needing to do the right thing a 1000 times a second is distorted. He's comparing two unrelated things. So I agree that he's making it seem harder than it is.

I'm also not saying it's hard from a technical perspective. Rather I think it's hard from a human interface perspective. Which is part of the reason that I think Elon is underestimating the difficulty. He's looking at it from the technology perspective. But as long as it's a driver assistance feature and not a self driving car there are pesky humans involved.

Also, I should point out that this fear of the steering portion of autopilot causing all these accidents seems rather odd to me. TACC is just as likely, maybe more likely, to cause an accident by inappropriately accelerating or braking as the car is to steering into something. Frankly, not steering into objects, given a camera and radar system, seems fairly easy to me.

It might not hold the center of the lane as smoothly as we would like, or it might disengage more often than we would like, but I suspect that the likelihood of it causing some sort of accident is rather low.

The point I was making is that Google is right about their comments about driver assistance features making people worse drivers because as they get more and more advanced people pay less and less attention and the driver assistance feature is absolutely not able to deal with every scenario. It needs to be actively monitored.

Adding Auto steering isn't inherently difficult. The problem is that once we add Auto steering you have removed the last active input that the driver has to make on a regular basis. They are now just monitoring and only stepping in when the car is doing something wrong. As people put miles on the car without anything bad happening while operating in this mode they are going to become less diligent about monitoring the car. Because a well implemented system is going to only need occasional involvement from the driver.

The combined effect of the current features and auto steering are what makes this hard. It's not even really a technical implementation problem so much as a human interface problem.

 

[Soolim]

Tesla is usually late, but generally over-delivers in the end. Plus you get a lot of good stuff you were not expecting.

I would buy that since I continue to trust that EM will do the right thing. Did I speculate better H/W?:wink: