Tesla, TSLA & the Investment World: the Perpetual Investors' Roundtable

[betstarship]

Came back from the road trip and had to drive the 101 today because the grapevine was closed due to blizzard conditions. It was really scary and it's going to get worse from what I'm getting fed and reading about weather articles. Re: my Model 3, some thoughts:

- It re-routed me through the Chumash Highway instead of keeping me on Hwy 101. I don't know why (maybe it was a battery charge optimization), but Chumash was a terrible highway with frozen roads and fog where you can only see 2-cars in front of you (also the rain) at high elevations that I drop 25 miles through to get back onto the 101. From a safety perspective, it should have kept me on the 101 which is all mostly flat land at that point.

- I don't know if Tesla navigation already does this, but it needs to be respectful of weather conditions somehow in the routing process. I don't know if it's something like synthesizing windshield wiper signals (Lvl 1-4) from the fleet and incorporating that into routing navigation, but after today's experience I think it's really necessary as weather continues to get worse.

- I don't know if Tesla navigation already does this, but it needs to incorporate closures somehow...for example re-routing to 101 instead of 5 when there's road closures. Santa Cruz had something similar happen at the 17 was closed today too...which might completely close off Santa Cruz from Silicon Valley during a workday.

 

[Ogre]

Thinking a bit about what Tesla’s idea of “Gen 3” is and here are my thoughts.

Generation 1:
This is the Roadster, Model S, and Model X. Cars In this generation are essential relics of the previous world. While they are electric, they are manufactured largely the way vehicles have for the past 20 years+ or so. Most of the components that make up the car are familiar, as is the form factor. It’s just the drive train which is significantly different.

Generation 2:
The Model 3 and the Model Y were designed to be very similar to the previous generation vehicles. They look the same, they have many of the same major systems. The big difference is they are designed to be manufactured more efficiently. You can see the evolution of Generation 2 vehicles over time. The Model 3 was Tesla’s first swing at trying to mechanize all the processes. But fundamentally the components were the limiting factor. Model Y was much better and the front and rear Giga Castings took things up a notch. I suspect the Highland version of the Model 3 is going to push this forward even more…. But it’s not Gen 3.

Generation 3:
I’ve thought about this a bunch and for a while I was thinking the Gen 3 vehicle they were going to talk up was the Robotaxi, but I’m pretty sure it’s actually the Cybertruck. Generation 3 vehicles toss everything out the window and are designed from the start to be easy and inexpensive to manufacture. Their “Gen 3” platform isn’t so much a sled like what GM presented with the Ultium, but a philosophy.

The Gen 3 Philosophy is toss everything we did before. Their 3rd Generation vehicles will be scratch designed to fit a purpose and to be cheap to manufacture. The Cybertruck is the first of them. This doesn’t mean all Gen 3 vehicles are going to look like the Cybertruck, in fact the Cybertruck may be fairly unique. But its very likely new vehicles won’t necessarily conform to previous norms.

The form factor isn’t the big thing though. The important bit is the fact that they are going through the way they manufacture it with a fine toothed comb. Why is the feature the way it is? Can it be made more affordable or eliminated entirely? Just as a simple example, I think Tesla is very likely to replace brake lines, steering linkage, lighting and electrical harnesses, door releases, window controls, camera wiring plus more all with a common bus/ power system. Instead of having super complex wiring harnesses, the cars will have a single strip going front to back.

This strip would decrease manufacturing costs multiple ways. First, wiring harnesses on vehicles are expensive, difficult to change quickly, and only manufactured by a few companies. This wiring bus can be printed out on an extruder by the foot for a few bucks (not exactly but close). They can use the same system on any vehicle, and they can add or remove components from the strip instantly. And… most important, it can be installed by robots instead of humans.

Plus much more…
Lots of other systems which can be removed or integrated into that central system and they all play well with Tesla’s ultimate goal which is eliminating the driver. Steer by wire, brake by wire, all can be dropped onto that bus. While these systems might be more expensive individually than their old world counterparts, the system as a whole will be comparable in price to their old world counterparts, and most important, the system as a whole will be easier and more affordable to manufacture.

This is Gen 3
I think this is what Tesla refers to as “Gen 3”. Tesla is eliminating the last vestiges of the Old Republic world way of producing cars and rethinking things as much as possible in order to scale production to insane levels. One big thing people have left out of this equation so far is affordability. If Tesla is going to sell 20m vehicles a year, they need to make these vehicles cheap… I don’t think the $25k number was ambitious enough and I’m pretty sure Tesla is shooting for even lower than that, even now.

The Cybertruck is the first. Lots of people think it’s going to be some crazy supertruck, but I’m pretty sure it’s just going to be laser focused on bang for the buck. How much truck can they cram into that chassis and still end up with an affordable vehicle?

“Gen 3 platform” Is the Cybertruck.
The more I think on it, the more I think Tesla is going to finally pull the sheets off the Cybertruck. It is their Gen 3 platform. I’ll leave this quote here to remind people what the Cybertruck mission is.

https://twitter.com/x/status/1498856948321722370

 

[MC3OZ]

Thinking a bit about what Tesla’s idea of “Gen 3” is and here are my thoughts.

Generation 1:
This is the Roadster, Model S, and Model X. Cars In this generation are essential relics of the previous world. While they are electric, they are manufactured largely the way vehicles have for the past 20 years+ or so. Most of the components that make up the car are familiar, as is the form factor. It’s just the drive train which is significantly different.

Generation 2:
The Model 3 and the Model Y were designed to be very similar to the previous generation vehicles. They look the same, they have many of the same major systems. The big difference is they are designed to be manufactured more efficiently. You can see the evolution of Generation 2 vehicles over time. The Model 3 was Tesla’s first swing at trying to mechanize all the processes. But fundamentally the components were the limiting factor. Model Y was much better and the front and rear Giga Castings took things up a notch. I suspect the Highland version of the Model 3 is going to push this forward even more…. But it’s not Gen 3.

Generation 3:
I’ve thought about this a bunch and for a while I was thinking the Gen 3 vehicle they were going to talk up was the Robotaxi, but I’m pretty sure it’s actually the Cybertruck. Generation 3 vehicles toss everything out the window and are designed from the start to be easy and inexpensive to manufacture. Their “Gen 3” platform isn’t so much a sled like what GM presented with the Ultium, but a philosophy.

The Gen 3 Philosophy is toss everything we did before. Their 3rd Generation vehicles will be scratch designed to fit a purpose and to be cheap to manufacture. The Cybertruck is the first of them. This doesn’t mean all Gen 3 vehicles are going to look like the Cybertruck, in fact the Cybertruck may be fairly unique. But its very likely new vehicles won’t necessarily conform to previous norms.

The form factor isn’t the big thing though. The important bit is the fact that they are going through the way they manufacture it with a fine toothed comb. Why is the feature the way it is? Can it be made more affordable or eliminated entirely? Just as a simple example, I think Tesla is very likely to replace brake lines, steering linkage, lighting and electrical harnesses, door releases, window controls, camera wiring plus more all with a common bus/ power system. Instead of having super complex wiring harnesses, the cars will have a single strip going front to back.

View attachment 910960

This strip would decrease manufacturing costs multiple ways. First, wiring harnesses on vehicles are expensive, difficult to change quickly, and only manufactured by a few companies. This wiring bus can be printed out on an extruder by the foot for a few bucks (not exactly but close). They can use the same system on any vehicle, and they can add or remove components from the strip instantly. And… most important, it can be installed by robots instead of humans.

Plus much more…
Lots of other systems which can be removed or integrated into that central system and they all play well with Tesla’s ultimate goal which is eliminating the driver. Steer by wire, brake by wire, all can be dropped onto that bus. While these systems might be more expensive individually than their old world counterparts, the system as a whole will be comparable in price to their old world counterparts, and most important, the system as a whole will be easier and more affordable to manufacture.

This is Gen 3
I think this is what Tesla refers to as “Gen 3”. Tesla is eliminating the last vestiges of the Old Republic world way of producing cars and rethinking things as much as possible in order to scale production to insane levels. One big thing people have left out of this equation so far is affordability. If Tesla is going to sell 20m vehicles a year, they need to make these vehicles cheap… I don’t think the $25k number was ambitious enough and I’m pretty sure Tesla is shooting for even lower than that, even now.

The Cybertruck is the first. Lots of people think it’s going to be some crazy supertruck, but I’m pretty sure it’s just going to be laser focused on bang for the buck. How much truck can they cram into that chassis and still end up with an affordable vehicle?

“Gen 3 platform” Is the Cybertruck.
The more I think on it, the more I think Tesla is going to finally pull the sheets off the Cybertruck. It is their Gen 3 platform. I’ll leave this quote here to remind people what the Cybertruck mission is.

https://twitter.com/x/status/1498856948321722370

IMO both the Cybertruck and Model 3 Highland have elements of Gen3.

A Gen3 vehicle body needs to be built in a low cost way, but that isn't enough, The battery pack, and the rest of the car build via the GA process, need to be done as cheaply as is possible without sacrificing safety or quality.

The most labour intensive part of building cars in GA, fewer GA steps and fewer parts, fewer workers needed to build a car, or the same amount of workers can build more cars. I'll be surprised if the wiring harness isn't part of Gen3, getting rid of the stalks on the steering wheel is another obvious candidate.

To eliminate parts and processes, things need to be done differently, brake-by-wire is a candidate here, if it isn't in Gen3 now, it is a likely future inclusion.

Eliminating paint is an obvious candidate, but not a certain inclusion. There are multiple ways of doing that. But ultimately the model panels need to be a material which doesn't rust, aluminium, stainless steel, zinc? Aluminium seems like the cheapest option. IMO aluminium could also support a mix of painted and unpainted, unpainted is the cheapest option, if a customer wants paint, then they pay more according to the colour they want,. No need to paint Robotaxis, just attach signage.

For the bodies, casting some parts of the sides or structural rails might result in a cheaper car, for a compact model a single price bottom casting might be cheaper.

I'm unsure if a Cybertuck style folded body is cheaper than attaching panels to a frame. It is for Cybertruck, but towing and hauling were considerations, a more conventional compact vehicle doesn't have those loads.

Model S/X can't be Gen3, and don't need to be.

 

[jkirkwood001]

Seems likely that the magic adapter can be sensed when it's docked or undocked/used. If the last customer completes their charging session and the magic adapter isn't sensed to be docked, you know who to bill...

I don't understand the risk. The whole idea is that the adapter is locked to the end of the standard cable when used by a non-Tesla. It's the same mechanism that keeps the cable locked to a Tesla when charging. Only if a v.3-compatible Tesla comes to charge (all of us) does the adapter stay in the charge unit. So it's either locked to the charge unit or the end of the cable at all times. And the adapter has power for locking because it's always connected to the charge unit one way or the other.

Sure, with brute force by breaking the locking mechanism you could rip it off, but this isn't a likely or frequent scenario.

A little like the time a flying municipal sidewalk plow ripped off my 120v cable charging overnight at a friend's in Montreal in my first year!

 

[Baumisch]

On a lighter note:

Tesla's NEW HW4 Car Computer | Munro Live (1 hr ago)

Jusk skipped through this .... Seams unfortunately like representing what green already tweeted - or was there anything new im there ?

 

[Electroman]

Nobody knows but $400 to $700 will not shock me

Waiter ! I will have what he is having.

All above is simply my wild speculation that nobody should take seriously

well .. err.. no just water please.

 

[Baumisch]

This is excellent speculation regarding Investor Day. Well reasoned.

Sections:
00:00 Intro
02:28 What's in a name?
03:51 Assumptions and Goals
05:07 HOLY TERACASTINGS!
05:15 - 1. The bottom line
06:47 - 2. Why Gigacastings won't do
09:15 - 3. Desire moves mountains
12:36 - 4. Ask Moore and Wright
13:55 - 5. Is it pos
14:04 - 6. Is it smart?
14:49 - 7. Seeing is believing
15:41 - 8. The invitation
17:31 - 9. Tesla insider spills the truth!
20:19 - 10. One piece or more?
23:07 - 11. Aluminum or Aluminium?
26:11 - 12. Bullish!
27:47 - 13. Even more bullish!
28:44 - 14. You can help too!
29:40 You won't believe this!

It's well worth watching his earlier video on Project Highland too.

Wow !

 

[jerry33]

I don't understand the risk. The whole idea is that the adapter is locked to the end of the standard cable when used by a non-Tesla. It's the same mechanism that keeps the cable locked to a Tesla when charging. Only if a v.3-compatible Tesla comes to charge (all of us) does the adapter stay in the charge unit. So it's either locked to the charge unit or the end of the cable at all times. And the adapter has power for locking because it's always connected to the charge unit one way or the other.

Sure, with brute force by breaking the locking mechanism you could rip it off, but this isn't a likely or frequent scenario.

A little like the time a flying municipal sidewalk plow ripped off my 120v cable charging overnight at a friend's in Montreal in my first year!

View attachment 911049

There is perceived risk mainly because we haven't seen one in person so there's only guesswork of how it actually works. The animation doesn't really show any details.

 

[TesslaBull]

It doesn't seem very difficult to have the adapter secured at all times. Just a locking mechanism from both sides. Locks to the charge station. Unlocks once the Tesla cable is attached and locks to the adapter. Stays locked to the Tesla cable until it is returned to the charger. Seems pretty simple.

 

[ThisStockGood]

This is excellent speculation regarding Investor Day. Well reasoned.

Sections:
00:00 Intro
02:28 What's in a name?
03:51 Assumptions and Goals
05:07 HOLY TERACASTINGS!
05:15 - 1. The bottom line
06:47 - 2. Why Gigacastings won't do
09:15 - 3. Desire moves mountains
12:36 - 4. Ask Moore and Wright
13:55 - 5. Is it pos
14:04 - 6. Is it smart?
14:49 - 7. Seeing is believing
15:41 - 8. The invitation
17:31 - 9. Tesla insider spills the truth!
20:19 - 10. One piece or more?
23:07 - 11. Aluminum or Aluminium?
26:11 - 12. Bullish!
27:47 - 13. Even more bullish!
28:44 - 14. You can help too!
29:40 You won't believe this!

It's well worth watching his earlier video on Project Highland too.

 

[CyberShy]

That Connecting the Dots channel is more Guess the Crazy Extreme Possibility that will Never Happen than actual fact-based or reasonably likely conjecture. He may get some things correct, but usually his pie-in-the-sky connections are more like not our wildest dreams.

 

[wtlloyd]

Came back from the road trip and had to drive the 101 today because the grapevine was closed due to blizzard conditions. It was really scary and it's going to get worse from what I'm getting fed and reading about weather articles. Re: my Model 3, some thoughts:

- It re-routed me through the Chumash Highway instead of keeping me on Hwy 101. I don't know why (maybe it was a battery charge optimization), but Chumash was a terrible highway with frozen roads and fog where you can only see 2-cars in front of you (also the rain) at high elevations that I drop 25 miles through to get back onto the 101. From a safety perspective, it should have kept me on the 101 which is all mostly flat land at that point.

- I don't know if Tesla navigation already does this, but it needs to be respectful of weather conditions somehow in the routing process. I don't know if it's something like synthesizing windshield wiper signals (Lvl 1-4) from the fleet and incorporating that into routing navigation, but after today's experience I think it's really necessary as weather continues to get worse.

- I don't know if Tesla navigation already does this, but it needs to incorporate closures somehow...for example re-routing to 101 instead of 5 when there's road closures. Santa Cruz had something similar happen at the 17 was closed today too...which might completely close off Santa Cruz from Silicon Valley during a workday.

That 154 highway shaves 5 minutes off the route. I've driven it many, many times in all seasons and weather, as I lived in Ventura for 20 years. Still, that section of coast from Gaviota down is lovely and worth the time lost, if you're not in a hurry.

 

[betstarship]

That 154 highway shaves 5 minutes off the route. I've driven it many, many times in all seasons and weather, as I lived in Ventura for 20 years. Still, that section of coast from Gaviota down is lovely and worth the time lost, if you're not in a hurry.

Thanks! I guess they're focusing on distance optimization. It felt dangerous with the weather conditions the way they are, so I'll take your word for it when the weather is better.

I was driving 45-50 mph.

 

[GOVA]

Jusk skipped through this .... Seams unfortunately like representing what green already tweeted - or was there anything new im there ?

I did start watching this.
Munro's understanding of new tech has always been zilch. Now they just yak about Greens findings.

Quite a low point for them. Speculation about slides.

I hope one day Ingineerix will go over this.

 

[hitchhiker]

Thanks! I guess they're focusing on distance optimization. It felt dangerous with the weather conditions the way they are, so I'll take your word for it when the weather is better.

I live in Ventura right now and occasionally head up to Santa Ynez. it always wants to route me via the 154 because its a few minutes faster. its a great road to drive during the day and in good weather. I prefer the 101 when its dark or rough weather. I doubt you usually get ice on the 154 usually -- just this crazy storm.
But in general, I find all of the mapping software (apple, google, tesla) doesn't take into account current road conditions. They try to take into account current traffic, but that's about it. And that sometimes burns them (ie: routing people into a fire area durring brush fires because that's where the traffic is lightest -- for a reason!) I'd really like to see all the software improve on that for sure. It should take into account road and weather conditions and give you the option of choosing a route that goes around bad stuff.

 

[Singuy]

That Connecting the Dots channel is more Guess the Crazy Extreme Possibility that will Never Happen than actual fact-based or reasonably likely conjecture. He may get some things correct, but usually his pie-in-the-sky connections are more like not our wildest dreams.

And people wonder why gamblers buy the rumor and sell the news.

 

[AndrewZ]

I live in Ventura right now and occasionally head up to Santa Ynez. it always wants to route me via the 154 because its a few minutes faster. its a great road to drive during the day and in good weather. I prefer the 101 when its dark or rough weather. I doubt you usually get ice on the 154 usually -- just this crazy storm.
But in general, I find all of the mapping software (apple, google, tesla) doesn't take into account current road conditions. They try to take into account current traffic, but that's about it. And that sometimes burns them (ie: routing people into a fire area durring brush fires because that's where the traffic is lightest -- for a reason!) I'd really like to see all the software improve on that for sure. It should take into account road and weather conditions and give you the option of choosing a route that goes around bad stuff.

I haven't seen Apple consider road conditions yet, it's probably coming because there are already highway hotlines that track it that can be merged in, but I have received weather alerts from Apple maps when driving into an area Environment Canada has issued storm reports for. You get a beeping alert and a pop-up that says you're entering an area with storm activity so drive carefully.

That's useful, the speed trap and accident reporting pin drops are super useful as well. I'm doing a lot of long highway hauls and last week received an alert about an upcoming speed trap, went around the bend in the highway and bam there was RCMP in waiting. Almost surprising that's even legal tbh.


If someone can pull all these data together into a super comprehensive single source, that would provide so much value to drivers and likely greatly improve safety. If vehicles could share data in real time and update those maps, just imagine that -- I remember when 5G was supposed to be the next step in enabling something like this.

 

[Artful Dodger]

“Gen 3 platform” Is the Cybertruck.
The more I think on it, the more I think Tesla is going to finally pull the sheets off the Cybertruck. It is their Gen 3 platform. I’ll leave this quote here to remind people what the Cybertruck mission is.

Yeah, this is probably overstating the intentions for Cybertruck. Elon said on the Decode podcast (Nov 2, 2018) if Cybertruck fails to catch on, Tesla will make a more conventional truck:

Elon: “I’m personally super-excited by this pickup truck. It’s something I’ve been wanting to make for a long time. And I’ve been iterating sort of designs with Franz. If there’s only a small number of people that like that truck, I guess we’ll make a more conventional truck in the future. I think this is the kinda thing the consumer would want to buy, even if they don’t normally buy a pickup truck.

So, anyway, that’s personally I’m most excited about. But like I said, it could be just like, okay, I weirdly like it and other people don’t. That’s possible. But we’re gonna make it anyway, and then we will just have a niche audience, I don’t know. But if it does, then we’ll make a more conventional pickup truck,” Musk said.

Cybertruck's inception was not another "bet the company" moment. That's why it's taken over 5 years to bring to market. It's a passion project, and a visercial attempt to show the mainstream American pickup truck buyer that 'dang, those eeVee's are some hot *sugar*!'

Gen-3 is the $25K Compact Car we first heard about during the 2018 AGM. It will likely be expanded to include a sedan and CUV (much like the S/X and 3/Y fraternal twins), and it will cost half as much to produce as the Gen-2 vehicles. As such, sales for Gen-3 could be up to 10x that of Gen-2 (or roughly ~10M vehicles per year). That's a big chunk of how Tesla gets to 20M/yr by 2030.

Cybertruck? 500K/yr was the stated goal, whereas the TAM for U.S. full-size trucks is roughly only about 3M/yr. That's not the droid you're looking for...

Cheers!

P.S. Master Plan 3 is about tonnage of batteries, not the mass of it pickup trucks.

P.P.S. Battery tonnage produced needs to be equivalent to the weight of one Great Pyramid of Giza per month, every month, until the world is converted to sustainable energy. M.P.3 will likely include partnerships with industry to get to this production level. #Predict

 

[cliff harris]

I've been thinking a bit about HW3/HW4 FSD and potential liability.

If there are 400k FSD users right now, lets say a total of 600k global owners of FSD, who have HW3. Is that accurate?

I'm trying to model the potential nightmare downside liability of HW3 being inadequate for FSD, and the law ruling that HW3 cars with FSD were mis-sold. In this case, the liability is basically the lesser of:
Refunding all those 600k their $10k
Cost of a HW upgrade for 600k users

Actually even this WORST case scenario is not *too* bad. Elon says you cant upgrade HW3 to HW4, but I suspect what he really means is that it would be an expensive annoyance that most users would not be able to justify. But the WORST case is refunding a customer $10k (and not everyone would demand that, or live in a jurisdiction ruling that). If Tesla can retrofit HW4 to a 3/Y for 9k, it still makes sense to do so.

Realistically, how tricky could it be? If the new cameras are in the bumper, then thats just a new bumper for each car plus new sensors and then some extra wiring and then obviously the swap-out for the HW4. They say its bigger than HW3 but I'm pretty sure there is room in there somewhere. Its not the size of a fridge!

Say the HW4 unit is $1500 max. New cameras... maybe $800, and the new bumper $200? (Cost to Tesla). I cant se a scenario where its more than $4,000 to do that upgrade. IF there is demand for it. So thats $4k x 600k $2.4 Billion worst case. Easily affordable.

(And arguably the refund cost could be less given that many of the FSD features already work with HW3, so any complaint would have to pro-rata for only missing functionality).

Hopefully FSD is deliverable with HW3, and HW4 is just 'FSD but even more reliable and smooth'.

 

[AndrewZ]

I've been thinking a bit about HW3/HW4 FSD and potential liability.

If there are 400k FSD users right now, lets say a total of 600k global owners of FSD, who have HW3. Is that accurate?

I'm trying to model the potential nightmare downside liability of HW3 being inadequate for FSD, and the law ruling that HW3 cars with FSD were mis-sold. In this case, the liability is basically the lesser of:
Refunding all those 600k their $10k
Cost of a HW upgrade for 600k users

Actually even this WORST case scenario is not *too* bad. Elon says you cant upgrade HW3 to HW4, but I suspect what he really means is that it would be an expensive annoyance that most users would not be able to justify. But the WORST case is refunding a customer $10k (and not everyone would demand that, or live in a jurisdiction ruling that). If Tesla can retrofit HW4 to a 3/Y for 9k, it still makes sense to do so.

Realistically, how tricky could it be? If the new cameras are in the bumper, then thats just a new bumper for each car plus new sensors and then some extra wiring and then obviously the swap-out for the HW4. They say its bigger than HW3 but I'm pretty sure there is room in there somewhere. Its not the size of a fridge!

Say the HW4 unit is $1500 max. New cameras... maybe $800, and the new bumper $200? (Cost to Tesla). I cant se a scenario where its more than $4,000 to do that upgrade. IF there is demand for it. So thats $4k x 600k $2.4 Billion worst case. Easily affordable.

(And arguably the refund cost could be less given that many of the FSD features already work with HW3, so any complaint would have to pro-rata for only missing functionality).

Hopefully FSD is deliverable with HW3, and HW4 is just 'FSD but even more reliable and smooth'.

Not sure if you're already accounting for this but the cost of FSD specifically is actually the difference between the EAP upgrade and FSD, right now that's $9k, and this is how Tesla recognizes the revenue as clarified in the most recent earnings call. It sounds like EAP revenue has been recognized immediately for a long time now, and I think the $ delta between EAP and FSD has fluctuated over time.

Revenue must have also been recognized in the past for people who were given FSD Beta access along the way, because $324million recognized in 4Q22 across ~362,000 vehicles (from the recent recall) is not a lot of revenue per vehicle with FSD.

The major cost in HW4 probably wouldn't be the camera locations but stuff like wiring harnesses and all the internals, I don't think the HW3 computer is even capable of taking more camera feeds nor does the vehicle have the wiring for it. Don't think it's even feasible, we're talking basically pulling the vehicle apart and reassembling it. Or at least that's how I understand it.