Over the weekend - there appears to be a bunch of tweets from Elon (after about 3 months of silence, from the original statements).
As a MS & MX owner, and a Tesla well-wisher - it appears Tesla is using their 'computer/iphone brain' to address a problem that should be handled by their 'car brain'.
Some head scratchers as well - "more than 300+ parts" - huh, even the work done by wk057 did not have these many parts for a Non-AP -> AP1.0 upgrade. Sounds more like 'alternative facts'
Anyhow - the AP2.0 retrofit is looking bleak with these latest tweets... wonder how AP2.0 owners would feel when AP3.0 comes online, with this strategy in place.
The summarizing quote from the article says it all:
"When asked on Twitter whether the company would allow owners to pay for a retrofit of their first-generation Autopilot hardware to Tesla's October-launched improved new sensors for Autopilot, Musk said this won't be happening."
Oh, not again...
I think there will be even more outrage as they actually have you buying FSD. AP 1.0 has almost met the things it was said it would do - the only thing I can think of is it would read your calendar, pull out and be waiting for you at the curb. No where did it ever discuss full self driving, or to my knowledge, that it would even stop at stop signs or stop lights. In fact, EAP does not even mention it will do those things - only FSD mentions the stopping at stop lights and stop signs. So for AP 1.0 - they got pretty much what was offered. (Maybe I am missing a tweet or something, open to be corrected on that). As far as HW 2 - yeah, they are offering FSD and touting it. So when they don't deliver (now if it is government regulations stopping it, then they are off the hook) there is room for people to be mad. Me? Not sure yet - I bought FSD knowing full well that who knows if it will come to pass, but do feel Tesla will at least owe me my money back if they can't do it, as I purchased something and then they don't give it to me :) How long will it take? I don't know and realize it may be 2+ years, and I am ok with that too.
Just my thoughts on your last statement.
I don't believe many current AP2 owners expect to be able to retrofit their cars to AP3. The more natural way would be to trade in. Of course maybe skip a few major revs but it will make the buying of the new car all the more exciting when that time comes.
Why would you need AP3 if you have FSD? Perhaps AP3 hardware will make FSD2.0 even better, but if FSD1.0 ever makes it using HW2.0, thrn I don't see a need/want to upgrade to AP3/HW3 even if the upgrade is offered.
I think that is the point - a lot don't think HW2 has enough to truly be full self driving. Even though I bought it, I am a little apprehensive and skeptical. Time will tell though :)
FSD is essentially guaranteed pending regulations/approval. If regulations allow FSD and Tesla doesn't deliver, they are on the hook for a lot more than $3K FSD upgrade. Lots of people bought Teslas based on the promised capability of FSD. As an example, and please don't flame me, without Tesla's FSD capability, I could've been swayed into getting Cayenne instead.
True on your point philip, but that just allows you to be more furious at them, not sure you could get more out of them, as the car what you buy, the FSD is an option you purchase :) I am splitting hairs here and understand what you are saying. But, the car delivered everything else. I don't know cars very well but I know many would say even without FSD, you have the better car. (I can't give opinion, I came from a Nissan Altima :)
I think the issue becomes though, if one bought solely for the FSD - they have a little of themselves to blame as Tesla gave some pretty heavy disclaimers in there :) Yes, they would like it to be by end of 2017, but there are some things out of their control too. So those jumping just for FSD should have been a little cautious and bought first for other things, and then the added FSD. Just my thoughts to go along with what you said :)
I agree with you philip, besides being on the hook for $3k and an angry lynch mob, their entire reputation is at stake.
In my non tech unsubstantiated opinion, these guys are braniacs. FSD, is not Rocket Science, and oh yeah, they do that too.
So do you really believe in all conditions this current hardware configuration can truly drive itself with no one in the vehicle? have you ever looked at your backup camera? what about the side ones? yeah, right now, mine has gunk on it from the slushy snow. Ultrasonic sensors getting blocked from weather conditions?
So are we talking - yes what we have will do it in good weather, but not all. Can it really handle every nuance that is out there in the real world that will come up - with just this hardware suite? Again, i bought it, so not trying to discount it, just trying to say I am a little skeptical that this will be completely send your car from LA to pick you up in NY. Elon says things and then has to back down a little sometimes. When the brains says on stage - this is the first time my engineers are hearing this - great. Engineers can't always make things that sound good come true :)
I have to believe that Elon thought at some point in time the cameras would get gunk, and the weather would not always be ideal, maybe I'm just being naive , but I do believe :)
Yeah - I am not trying to be a downer or doubter, but have seen that Elon has had to change plans from what he said (AP 1.0 - the car will read your calendar and meet you at the curb warm and ready to go at that time for our appointment) - Did not happen, and we really don't ever see that happening with AP 1.0 the way it is equipped :)
So yes, I would think he would have thought about the gunk but as someone has posted, there could have been a way to have micro bursts of air across the camera to clean it, but that is not there. I don't see anyway those cameras stay clean - so now you are just relying on the three forward facing cameras and hoping the B-pillar cameras as well don't get stuff up there. But then I think of road conditions. Snow packed roads? I get the whole GPS but things change and GPS is not 100% accurate, at least not right now. So to me, this FSD is not as easy as we are being led to believe from their demo. I really just think there are too many variables and he made a huge jump from where we are now to what he claims this can do. Hence the skeptic that there will still need to be more for it to truly be FSD - at least all the time. But I wouldn't trust sending my car from here to NYC in the winter :)
I am not a brain in this area so could be way off, but just going off things others have brought up and lets face it, the world of driving is very complex, easy to say we can do it, hard to in the end really have everything in place without knowing if the software truly can do it too :)
The FSD description says in "almost all the circumstances"....may be we should ask them to list the problematic circumstances.
What do you think the engineers thought when the brain said we're going to send a rocket into space then land it back on earth ( although not every time :)...just saying.
Seriously though, I personally don't need or want my FSD MX driving around without me, to consider my extra $3k money well spent. If it can do what's shown in video on highway and city streets, stopping for pedestrianS ect, that's good enough for me.
burdogg, I agree with you that one should never buy based off of a promise. Now generally if it's some random company, I would follow that advice. This is Tesla's promise though, with its hard-earned stature and reputation on the line. I certainly did my homework and am ordering an MX not expecting FSD anytime soon. I am very excited about the car as a whole, with FSD as icing on the cake when it comes out. Having said that, if Tesla never releases FSD on HW2, no way to retrofit to make FSD work, and yet they release a new car with HW3 for FSD, I would rightly be very upset, as should you. At that point, whether I got a better car than say a Cayenne, with or without FSD, does not matter. It is fraud in advertising, period.
I agree :) I would be upset. I am in the same boat as you with purchasing :) I do think Elon believes fully that it will do FSD. Here is the problem though as some have stated. What if regulations require LIDAR :) Tesla does not control government regulations. The states could put such restrictions on FSD that require Tesla to have HW 3 to make their FSD legal. Now they can say they accomplished FSD with HW 2, but that does you and I no good as the government will not allow them to use it because it does not comply with their regulations. :)
Anyway, yes I will be upset, but I was actually buying the car regardless of FSD - I love the car. I placed my order the week before the announcement, knowing the announcement was coming, hoping it was AP 2.0. I knew full well that it might not be and if it wasn't I had to decide if I canceled my order or confirm it. I was really planning on confirming with AP 1.0 if need be (I have one and it does well enough for my needs). So, while disappointed and upset - yes, but if they gave my $3,000 back, I would have not problems - especially if it is regulations that require it.
@Saxman - I am right there with you - I just need it to take over driving every so often on long trips when I get a little sleepy. It needs to work so that I can trust it to do so. I don't need my car to drop me off and come pick me up :) I am really hoping that they enable the FSD stuff even without regulations as long as you remain in your seat and even hold the wheel if need be like AP 1.0.
Back atcha :)
Burdogg, you are a much nicer man than I :-)
If LIDAR is required for FSD, Tesla better figure out a way to retrofit LIDAR in current HW2 cars to make it compliant. $3K back is just not going to cut it with me. But who knows, maybe hanging out here with you nice folks will make me play nice as well :-)
You know, we all have different personalities and that is ok :) I do look at the order page though and note the Disclaimer again:
Please note that Self-driving functionality is dependent on extensive software validation and regulatory approval.
So does that mean that if the extensive software validation doesn't pan out we are SOL too? Just saw that and could read it that way too. The hard part to me is this is the Chicken or the egg scenario. They come out with Hardware first and HOPE they can get the software to utilize that hardware the way they see it doing it. What if they just can't get it? Or do you make the software first, but how do you really get your software to work without the hardware :)
We are into such complex systems that I really hope Tesla can pull this off. I do think they can. After having time thinking about it and seeing other posts, it has added a little skepticism to my though. The excitement of it wore off and reality set in and it may happen, it may not. I obviously though put my money on it happening :)
My understanding is though, Elon is pretty against LIDAR. I don't know if it is just the cost of it, or also some limitations that it is showing too.
By the way, i do think they will be successful with it in good weather conditions :)
Perhaps it's just me, but I'm actually very confident in Tesla's ability to get FSD off the ground. FSD is the future and a given, no way to resist it. Tesla will continue innovating hardware/software until FSD is achieved. Whenever that happens, Tesla will retrofit HW2 with new hardware to make FSD work. We bought FSD as a complete package! We didn't buy 8 cameras, 1 radar, etc. If new HW is what is needed to complete the package, Tesla will be on the hook to retrofit our cars free of charge.
I agree good weather FSD is good enough for me using HW2. I don't have a problem with having to pay for a new car with new HW to have FSD in bad weather too.
I agree HW2 FSD will be good enough for me as well, no need to get AP3 or AP4 as long as this AP2 gets FSD :) I don't care much about new upgrade like HUD or more things. Anyway really even cars with AP1 are fun to drive... AP2 advertised as FSD makes it beautiful. AP1 is kind of in the middle and I wonder where this goes after some time... however AP1 is also great and cars are fun to drive :)
We are proud and satisfied owners of a Model S (P11599) and have had over 80k miles of great driving. We’ve used the Super Charger network on 15k miles, and find it has completely eliminated any range anxiety, al beit we are retired and don’t have an aggressive schedule to meet when traveling. Our new order for a Model X 75D has been confirmed on 1/7/17, and I read with interest the many comments from the owners, whether having owned an X for months or anticipating delivery. Getting under the wire for unlimited SC was a major concern, and AP 2.0 was the tipping point for pulling the trigger. My thoughts are tempered with years of Tesla experience, knowing that by the time we take delivery in early April that many concerns will be addressed in the tightening up of the production line. My hopes are that in the next few months we will also see some small fixes go into my vehicle as issues arise from those on the road at this time. I cross my fingers that our vehicle will be enabled with the increased ‘Supercharger V3’, too. We have PUP but did not purchase FSD, figuring it will take the Feds and States many months, maybe years, to get the regs into some workable format Yes, it will cost an extra $1k for our MX when/if it becomes available, but that was considered under much thought before we hit the “Purchase” button on the site. On the other hand, if one has to repurchase to acquire AP 3.0 to get an acceptable HW3 vehicle, some of the sting is relieved.. I don't think Tesla will be obligated to retro fit your HW2 if the Feds/Sates require substantially more hardware to the existing suite.
I think when you take a look at AP2.0 today, it is reminiscent of how the original AP (remember, there was no AP1.0/2.0 back then). It's a belief that it will solve all usecases... and IMHO, Tesla is taking a risk here which will be worse as subsequent releases come online.
If you take a look at comparable approaches to self driving - you will notice other car manufacturers talking about LIDAR and high number of radars. AP2.0 has only 1 radar... so what happens, if the front requires 3 radar... and the back has to add 2 more? At that stage - it will be AP3.0 - and the 'no retrofit' strategy becomes worse, as now there are more owners stuck with AP2.0...
Dont get me wrong - while I am bummed about the AP2.0 retrofit issue, I am rooting for AP2.0... before long, will need to change cars... and so, will eventually be an AP2.0 (or beyond) user.
The irony is that retrofits are Service center based activities - and should have no bearing on the innovation curve... not to mention - there are folks that want to pay.
So you go to Costco and prepay for a $10 combo pizza (sausage, veggies, pepperoni and cheese). You go to pick it up and is told that the pizza has no cheese because cheese costs too much. If you want cheese in it, too bad, you have to pay $15 for a new pizza with the expensive cheese. What are you going to do?
So you go to Tesla and prepay $100K for a car (electric, AP2, FSD). You go to pick it up and is told that the car has no FSD because FSD costs too much (requiring new HW). If you want FSD in it, too bad, you have to pay $150K for a new car with the new HW3. What are you going to do?
Now please tell me what are you going to do?
I'm going to Burger King!
Yes. They have to give me my $10 back and I go to BK.
As long as Tesla gives me back my money..I'm fine.
All I want is cheese :-(
Once L5 is up and running and approved by the regulators in Colorado, I will upgrade to AP5.0. Until then yawn.
The upgrade price will be to buy a new Tesla and sell my current car. My personal opinion is that will happen 10 years or more from now.
The problem with your example is pizza is already made with those items - they are not inventing something :) This is way more complex then that ;)
If you wanted the cheese that wasn't made yet, but possibly would with disclaimers, you should probably wait until they actually have the cheese made and working :)
(If that was the ONLY thing you wanted :)
They have a new and better tasting cheese that might be available in 3 weeks. When available it will cost $5 but you are given an option to pre-order it at a discount now for $3 with the caveat that they might never be able to make it. Basically you are taking a risk and compensated by a $2 discount. If the probability that it can be produced is 60% then you break even since $5 * 60% = $3. The prices of $5 when available or $3 now with a risk are both fair (if we neglect time value of money which is not too bad given the short time frame and current low interest rates).
Applying the same logic to $4k to activate when available or $3k now we can imply a probability of successful FSD at 75% :-)
(Of course this is just for fun. I don't believe it is 75%. I am convinced it is 100% and the discount is to reward the true believers ;-p )
I think if FSD required an ugly roof mounted LIDAR monstrosity retrofit, I'd skip it and forget FSD. Be careful what you wish for - just because you imagine something perfect, doesn't mean there isn't a way for regulators to force automakers into a really crappy solution.
The 'rhino horn' of LIDAR is probably going to get addressed sooner than later... search for Hyundai Ioniq autonomous tech - they demoed this at the latest CES and it had a concealed LIDAR.
I honestly don't think there will be a credible case to force Tesla to return the money (or cheese... ;-)...) - because as burdogg mentioned, it is safe to say that AP1.0 addresses autonomous highway driving (more or less)... but to larger point of innovation, if u go by Elon's latest tweets...
- AP1.0... First HW September 2014... first software load by end of 2015....reliable SW or 8.0 by mid 2016....
- AP2.0... First HW October 2016... first software load by end of 2016... reliable SW by end of 2017/early 2018??
- AP3.0... first HW mid/late 2018?? More radars?
- AP4.0... 2020...??
The problem with the latter stages is that more users are affected... and there is reasonable consensus that most of autodriving tech is going to have a significantly larger retrofit market - than new cars. In one of the previous interviews, I remember Elon mentioning global car annual production was 100million... while there are 2 billion vehicles on the road...
Will definitely be interesting to see how Tesla's view on autodriving tech evolves...
I think we are starting to see what the retrofit group is doing: HEPA filter now available for retrofit on X ordered without PUP. But more involved spinal chord replacement won't probably be offered.
iconel - where did you get the HEPA filter available for non PUP? Curious, as I don't have PUP, not sure I would do it or not, but had not heard you could.
Thanks inconel for the link :)
You are welcome. Hopefully we will have more of those retrofits coming. Also if it's now available for $750 after delivery Tesla might as well offer it as a standalone option when ordering.
Why is anyone surprised there will be no retrofit to go from AP1 to AP2? Tesla has never retrofit any sensors and it is not in their best interest to do so. They get valued on deliveries. Want a new feature? Trade in or sell your car and buy a new one. That's what Tesla is incentivized to do as they "pull demand levers". They have zero incentive to retrofit cars, as I've said in previous posts. People who buy Tesla's must learn to understand their call will be outdated in 6-12 months in some way.
@elguapo - I have generally struggled with this concept of "Tesla wants to just sell cars" ... not sure, what's the data behind this hypothesis... if you take a look at their upgrades (for a moment, ignore Autopilot) - you will notice a pretty healthy set of upgrades...
Elon's tweets are also referring to "dont want to stop the innovation engine" - which is a more plausible explanation.
While the end result is the same i.e. AP retrofits looking bleak... though, the motivations are different.
Tesla is more like "the geeks that want to finish the task" ... as opposed to "wall street guys looking to maximize profits at all costs"...
@lhanspal I respectfully disagree. Like it or not, Tesla burns cash faster than a forest fire. The only way to raise it is to show they're getting closer to making money which comes with significant YoY deliveries and the Model 3 (that's a whole separate thread).
Showing investors that they've retrofitted 10,000 cars, for example, doesn't excite anyone. Showing continued growth in deliveries helps prove the total addressable market is big and growing and Tesla's a leader in that market. That's what gets investors excited that the company will continue growing. Focusing time and effort on retrofits just doesn't do anything for them. I agree, there's little real evidence either way, except the only headline every quarter is how many cars they delivered and how much cash they burned.
As it relates to software updates - those need to continue coming to spur new buying demand. As long as the tech supports it, one of the selling points of the car is the OTA updates. It's important to note, however, that Elon has said the Tegra 3 processor is pretty much tapped out at this point. So the AP1 and older cars will likely start to see fewer SW improvements other than UI.
Again, it's all conjecture, I just believe, after having an S and two Xs and following them for years, that they have absolutely zero interest in providing meaningful hardware retrofits, even when they clearly can. It started with the $10,000 parking sensors for those of us who had an original S...