Why AP 2.0 Won't Be Here Soon, and It Won't Be What You Think It Is

[number12]

I hear this a lot but don't understand the reasoning behind it. If new cars start to roll off the lin
I get e tomorrow with AP2.0 hardware, how does that affect Q3 numbers?

I for example would cancel my 9/27 delivery immediately.

He risks all cars built and not delivered. Granted most would gladly take $2500 hit for a few more weeks without car, but would really effect q3 numbers prior to capital raise.

 

[number12]

What I don't understand is how tesla gets end of q4 deliveries.. Say what everyone will about born on dates, people outside of tesla think of cars in years. Dec 2016 car not worth more than a Jan 2017. And also a Dec 17 not worth THAT much more than Jan 17 car... And so on.


But what would really boost Q4 sales this year would be 2.0 autopilot announcement. 3 year old trade ins alone with boost numbers. Q3 incentives and faster processing time of cars. Q4 2.0 hardware. Not a bad plan. But just a long shot guess.

 

[Canuck]

I for example would cancel my 9/27 delivery immediately.

Why? You don't want AP2.0 hardware? Once someone takes delivery of a car with AP2.0 hardware, all future deliveries will have it. No one will be cancelling to try to get old hardware.

So I still don't understand the argument and I don't think I'm missing anything unless you can explain to me what I am missing.

But I don't think people understand how hardware additions work, like parking sensors, folding mirrors, AP, and even the recent rear cup holders and USB's, and everything else, EXCEPT for the "D" reveal. AP2.0 will not be like the D reveal. That was completely different than these types of hardware additions.

He risks all cars built and not delivered. Granted most would gladly take $2500 hit for a few more weeks without car, but would really effect q3 numbers prior to capital raise.

Even if your argument does make sense, and perhaps you can better explain it to me, the vast majority of people buying a Tesla do not read here, and have no idea of the difference between AP1.0 and 2.0 hardware, so to suggest "He risks all cars built and not delivered" is more than a bit of a stretch. A few would be risked, but not all that many to affect numbers, if you can even explain to me how adding hardware will affect numbers.

 

[number12]

Why? You don't want AP2.0 hardware? Once someone takes delivery of a car with AP2.0 hardware, all future deliveries will have it. No one will be cancelling to try to get old hardware.

So I still don't understand the argument and I don't think I'm missing anything unless you can explain to me what I am missing.

But I don't think people understand how hardware additions work, like parking sensors, folding mirrors, AP, and even the recent rear cup holders and USB's, and everything else, EXCEPT for the "D" reveal. AP2.0 will not be like the D reveal. That was completely different than these types of hardware additions.



Even if your argument does make sense, and perhaps you can better explain it to me, the vast majority of people buying a Tesla do not read here, and have no idea between AP1.0 and 2.0 hardware, so to suggest "He risks all cars built and not delivered" is more than a bit of a stretch. A few would be risked, but not all that many to affect numbers. Again, however, I am all ears.

You said new cars roll off the line. I assumed you meant production line. Which mine is off of said line. Meaning my car would not have 2.0 hardware. There has to be a cut off somewhere. I could look at my car. Determine no new hardware and walk away right?

risks not loses. Not sure how to better explain that. He adds some risk to all pending q3 numbers... Not all pending q3 numbers/deliveries will cancel, but all pending will have some new added risk if story gains enough traction. Not everyone knows 1.0 vs 2.0 but everyone reads the news. Especially an article about tesla days before picking up a $100,000 car after waiting a month.

The question is why risk any? Wait 20 days and not risk any (like my order and several others on here) and instead have a new surge of orders in q4. As of today no new orders have any chance of getting delivered in q3. So he focuses on keeping orders live and moving inventory cars at a discount that he can deliver in 1-2 weeks.

 

[Stoneymonster]

I can't square AP 2.0 coming soon with the realities of HW/SW development and the timing of the Mobileye breakup. The only way I see this as possible are:

1. Tesla found a new HW partner sometime last year, before Mobileye knew they were being ditched, and no one has leaked this fact.
2. Tesla has been developing their own hardware since late 2014, around the time they signed up with Mobileye for production.
3. Elon really is an alien, and he got some of his buddies from Proxima Centari to drop off some cast off tech they weren't using.
4. It's not possible and everyone should calm down and expect AP 2.0 sometime in the latter half of next year.

 

[Canuck]

You said new cars roll off the line. I assumed you meant production line. Which mine is off of said line. Meaning my car would not have 2.0 hardware. There has to be a cut off somewhere. I could look at my car. Determine no new hardware and walk away right?

Roll of the line, take delivery, whatever. There's always a cut-off line. If you think new owners walking around cars looking for AP2.0 hardware will affect numbers, you are dreaming. You're the rare exception, practically everyone else who will take delivery around the cut-off line are the rule. This has happened before, with AP itself (started showing on cars in Sept/14), and it did not affect numbers. There is no risk at all.

4. It's not possible and everyone should calm down and expect AP 2.0 sometime in the latter half of next year.

Perhaps, but Fred Lambert says "we were told by sources familiar with the program that the Bosch Model S prototype features very similar level 3 autonomous features as the upcoming Autopilot 2.0." Fred seems like a straight-up person to me and I doubt he's making up information or his sources. He could have innocently obtained wrong information but that seems unlikely to me. I think we will see AP2.0 hardware long before the end of next year, since that's when the Model 3 comes out and they need to get it on the S and the X before then.

 

[RogerHScott]

Elon really is an alien, and he got some of his buddies from Proxima Centari to drop off some cast off tech they weren't using.

I thought you were only listing possibilities here, not established facts.

 

[Stoneymonster]

Roll of the line, take delivery, whatever. There's always a cut-off line. If you think new owners walking around cars looking for AP2.0 hardware will affect numbers, you are dreaming. You're the rare exception, practically everyone else who will take delivery around the cut-off line are the rule. This has happened before, with AP itself (started showing on cars in Sept/14), and it did not affect numbers. There is no risk at all.



Perhaps, but Fred Lambert says "we were told by sources familiar with the program that the Bosch Model S prototype features very similar level 3 autonomous features as the upcoming Autopilot 2.0." Fred seems like a straight-up person to me and I doubt he's making up information or his sources. He could have innocently obtained wrong information but that seems unlikely to me. I think we will see AP2.0 hardware long before the end of next year, since that's when the Model 3 comes out and they need to get it on the S and the X before then.

Well then Bosch is doing most of the HW and they must have started working with them a while ago.

 

[number12]

While I would not be the norm I believe the type of buyer tesla attracts is fairly educated and still an "early adopter".

They probably do more research than most, and a 6 week wait time gives you plenty of time to "Google" your upcoming car and read all about 2.0.

I know you've mentioned several times you missed AP and are completely at peace with that. Even before I take delivery, I am also at peace with missing 2.0. I want the car regardless, and realize the deprecation hit I will take. I more disagree with teslas business model.. Or more just enjoy discussing it.

I would: Announce it early. Let people get discounts right before if it's not important or let them wait. Seems like in the long run people would be more content and more likely to repurchase. And this to pertain to major hardware changes like facelift, AP, AP2.0... Not cupholders, etc.

Not my company and not even sure if it's best strategy.. It's just what I would do.

 

[Canuck]

Even before I take delivery, I am also at peace with missing 2.0.

Given the articles, and your delivery date, you may actually be in for a surprise. But even if you miss out, if you're like me, we're not getting any younger, and if you wait for AP2.0 you could be waiting a long time, and one day without driving a Tesla is too long for me. I had my car the summer before AP was released in Sept 2014. That was a great summer with my family and my car, that I will never get back.

But I agree with your position that Tesla's business model leaves a lot to be desired. I missed AP by months, not days. Missing it by days would have been much more difficult. But then again, I still would have got what I ordered and paid for. It's tough to figure out how to do additions without waiting for model years like traditional auto makers. I don't think there's an easy answer.

 

[wdolson]

Yes... our 2+ year old P85 is far from obsolete--it's still great. Sure it is no longer the fastest model by a lot, and yes it doesn't have AP, but it's wonderful to drive. Obsolete is such a misused word.

By many measures the only car in the world better than a 2014 Model S is a newer Tesla. Other car companies are that far behind.

 

[krnshh]

I think one of the biggest pre-requisites for autonomous driving will also be vehicle to vehicle communication. That is a key aspect that cannot be overlooked. An autonomous car cannot exist as an independent entity in the real world, imho. It requires all the sensors to decode the environment presented to it yes, but the environment (cars, roads etc.) needs to give it significant amount of localized data for it to maneuver through safely.

And V2V will become a leveller because cars from other brands will have to communicate using the same protocols. That said, smaller improvements (as mentioned by OP) will definitely arrive beforehand but don't expect to sleep when your Tesla drives you around and wake up, soon.

 

[bob_p]

Is it possible that when Mobileye found out Tesla had already decided to use a different vendor for AP 2.0 hardware, that's when Mobileye made the demands (claimed by Tesla), which really appeared to be their attempt to pressure Tesla to continue as a Mobileye customer - and when Tesla refused to shift back to Mobileye for AP 2.0, that's when Mobileye made the situation public, pre-emptively striking to make it appear they had ended the relationship, when instead, it may have been Tesla that had decided to end the relationship, but hadn't been planning to announce anything until they were ready to announce AP 2.0 was in production?

If Tesla announced AP 2.0 was imminent, but not yet in production, they risked impacting Q3 sales - since some customers might cancel or postpone their orders and wait for AP 2.0. So that put Tesla in a difficult situation, because they aren't ready to announce AP 2.0, so they won't discuss their plans on what hardware they plan to have in the new AP, so the best they can do right now is talk about the 8.0 upgrade - and how it will improve AP 1.0.

Still believe the AP 2.0 announcement will be sensors and increased processing - and once they make that announcement, it will be clearer what happened with Mobileye.

 

[wdolson]

I think one of the biggest pre-requisites for autonomous driving will also be vehicle to vehicle communication. That is a key aspect that cannot be overlooked. An autonomous car cannot exist as an independent entity in the real world, imho. It requires all the sensors to decode the environment presented to it yes, but the environment (cars, roads etc.) needs to give it significant amount of localized data for it to maneuver through safely.

And V2V will become a leveller because cars from other brands will have to communicate using the same protocols. That said, smaller improvements (as mentioned by OP) will definitely arrive beforehand but don't expect to sleep when your Tesla drives you around and wake up, soon.

Converting from one technology to another is always a bigger problem than when the entire world is on one tech. The big headache that has to be dealt with when designing an autonomous car is it being the only autonomous car on the road. An autonomous car in a fully connected driving environment is a simple problem by comparison. It is 100% vital that problem is solved because it's going to be decades before it will be possible for every car to be autonomous and accommodations still need to be made for classic cars owned by hobbyists.

The average car in the United States is 11 years old and the fleet is getting older every year. That means 1/2 of the cars on the road were built before 2005. I've been driving a 24 year old car until recently. Requiring people upgrade their car would be a massive burden on the poor. If someone lives in the core of a major city, public transportation is usually fair to excellent in a few places, but get outside the city core and public transportation quality drops off a cliff.

When I lived in the Seattle area, I was a 1/2 hour drive from downtown Seattle and a 1/2 hour drive from Redmond (where Microsoft is) if I left at 5:30 in the morning. It was an hour's drive if I left any later. I looked at public transport options. There was bus service that picked up on my street, but it only ran a couple of times a day and only went to downtown Seattle. If I drove to the nearest park and ride (15-20 minute drive) and got a bus there, it would still take me 3 bus transfers and 2 hours to get to work in Redmond.

By Seattle standards, I lived "close in", but public transport was so poor, driving was my only realistic option.

Getting any further out from a major city and it becomes even less viable. There are people who work in the San Francisco Bay area and Los Angeles who live 60 miles or more from work. For most of them, there are zero public transport options. Most of these people live so far from work because they can't afford to live any closer.

The US also has a lot of rural areas where buses would burn more fuel than have people drive their own cars because the population density is way too low to make it worthwhile.

Autonomous electric cars in some kind of taxi service will help to some degree, but it's going to be a long time before there are enough cheap Ubers in rural areas to where the rural poor can give up their cars. Many will still refuse to because the rural poor also tend to be more conservative about adopting new technology than urban dwellers (most of time, there are always exceptions).

The United States alone has over 250 million cars on the road and worldwide production is about 80 million cars a year. Even every car built starting Jan 1 was fully autonomous and there was some sort of program to give poor people autonomous cars, it would take the US consuming every car built in the world for three years just to replace all US cars.

The autonomous engineers have had to deal with the assumption their car is the only autonomous one out there because that will be the reality at first and even many years down the road, there will still be "dumb" cars autonomous cars will have to deal with.

 

[RogerHScott]

The US also has a lot of rural areas where buses would burn more fuel than have people drive their own cars because the population density is way too low to make it worthwhile.

Until the Tesla Model B(us) comes along.

 

[MarkS22]

Roll of the line, take delivery, whatever. There's always a cut-off line. If you think new owners walking around cars looking for AP2.0 hardware will affect numbers, you are dreaming. You're the rare exception, practically everyone else who will take delivery around the cut-off line are the rule. This has happened before, with AP itself (started showing on cars in Sept/14), and it did not affect numbers. There is no risk at all.



Perhaps, but Fred Lambert says "we were told by sources familiar with the program that the Bosch Model S prototype features very similar level 3 autonomous features as the upcoming Autopilot 2.0." Fred seems like a straight-up person to me and I doubt he's making up information or his sources. He could have innocently obtained wrong information but that seems unlikely to me. I think we will see AP2.0 hardware long before the end of next year, since that's when the Model 3 comes out and they need to get it on the S and the X before then.

He's an awesome source for the community. No doubt.

But last November, there was an article stating supplier (Mobileye) hinted at AP 2.0 before the end of the year (2015). Interestingly, the title of that article has since changed to "within a year."

So, nobody except Tesla really knows and I'd take all "reports" with a huge grain of salt, even some of the best sources, like Fred.

 

[Stoneymonster]

He's an awesome source for the community. No doubt.

But last November, there was an article stating supplier (Mobileye) hinted at AP 2.0 before the end of the year (2015). Interestingly, the title of that article has since changed to "within a year."

So, nobody except Tesla really knows and I'd take all "reports" with a huge grain of salt, even some of the best sources, like Fred.

The thing is, both articles could have been right at the time (or right-ish). It wouldn't be the first time a SW/HW company had ambitious plans that met with reality.

Consider too that discovering what they could do with the existing radar might have given Tesla some breathing room to spend more time on AP 2.0, so they changed plans and extended the schedule.

 

[Canuck]

The thing is, both articles could have been right at the time (or right-ish). It wouldn't be the first time a SW/HW company had ambitious plans that met with reality.

Consider too that discovering what they could do with the existing radar might have given Tesla some breathing room to spend more time on AP 2.0, so they changed plans and extended the schedule.

Right and add the fatality accident into the mix and there's perhaps further reason for the delay.

So, nobody except Tesla really knows and I'd take all "reports" with a huge grain of salt, even some of the best sources, like Fred.

Right too. But it's still fun to speculate.

 

[MarkS22]

The thing is, both articles could have been right at the time (or right-ish). It wouldn't be the first time a SW/HW company had ambitious plans that met with reality.

Consider too that discovering what they could do with the existing radar might have given Tesla some breathing room to spend more time on AP 2.0, so they changed plans and extended the schedule.

Well, maybe sometimes but they quoted Mobileye noting: “Today we are already preparing with one of the OEM, a first vehicle based on 8 cameras, one radar and ultrasonic around the vehicle."

The implication was that this could be AP 2.0, but (as I argued at the time), it was a Nissan test car that was announced around that time. It literally had 8 cameras running 5 Mobileye EyeQ3.

My point being, I don't believe there was ever any plan for AP 2.0 with 8 cameras and one radar from Tesla by the end of 2015. But yes, I do believe good intentions can slam into reality.

 

[wdolson]

Until the Tesla Model B(us) comes along.

And available in large numbers.