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

[jerry33]

When you pick the FSD option on the order page of any Tesla you see the below. Any reasonable consumer would then expect the car to ‘respond’ to traffic light and stop sign and get ‘automatic driving on city’. I think this is what @Singuy was alluding to. If this deadline is not met it will provide amunitions to Tesla haters, even if we at TMC understand the fine details on how the company works and we don’t mind the occasional delays given the massive progress the company is making.

Coming this year:

• Recognize and respond to traffic lights and stop signs.
• Automatic driving on city streets

There are already FB posts with pictures showing that stop signs work.

 

[KarenRei]

It's not just the headline, the whole article is carefully written to make it appear that the SEC may be investiging Tesla. Every single time they mention investigation it is of "the firm", never UBS nor Tesla. Totally reads like a FUD piece.

Ironically, it's on a website run by a Tesla bull (Vincent).

 

[jerry33]

On CNBC today, I believe the "DH" who interrupted Cathie Wood also asked questions about Toyota's market cap in order to try and demonstrate that Wood doesn't know the auto industry well enough. Cathie answered the question satisfactorily, and then the "DH" clearly was trying to set up a trap regarding Tesla's future market cap vs. some number that "DH" was going to spout off. Cathie saw the trap, didn't take the bait, and shut him down. To me, this was more fun than watching any football game.

That's setting the bar pretty low.

 

[Fact Checking]

But even this layout isn't ideal: it would mean that the Model Y bodies would have to travel the whole length of the building to get into the paint shop. So either they build a second paint shop for the Model Y, or the current Model 3 BIW line has enough capacity to make the Model Y as well:

Note the very purposeful topology of the assembly lines: stamp shop feeds into the body shop on one side of the building (which has the saw-tooth pattern loading docks), which feeds into the paint shop, which feeds into the start of the very long general assembly line, which spews out finished cars into the parking near the stamp shop. The car, as it is assembled, takes an almost perfect circle within the factory and always stays close to loading bays.

Note how the phase 3/4 building is a mirror image of the phase 1/2 building: the highest throughput loading docks are at the outside boundary of the factory, and the dense, heavy shipments of raw metal to the BIW ships are loaded via the saw-tooth loading bays in the middle of the factory.

Ok, to better illustrate the layout of GF3, I've corrected the 3D perspective distortions of the photo of the screen and created an overview/map view of the GF3 layout, with arrows added to illustrate in what direction the workflow is probably going:

​

The counter-clock rotation factory flow is easier to see in this image:

• Tesla cars start their GF3 life probably somewhere on the boundary between the square stamp shop and the rectangle shaped BIW shop - which is a mostly automated section with hundreds of welding robots.
• As they move through the BIW they get into the paint shop, and probably exit the paint shop somewhere in the south-western corner.
• The general assembly line is hundreds of meters long and mostly linear.
• Finished Model 3's exit the factor on the right side end of the GA line, right next to the stamp shop. This is consistent with where we saw new cars emerge on drone videos:
  • 

Those new Model 3's emerge at the right side end of the GA line, and get charged immediately, before being parked at the larger parking lot. This charging flow will have to chance once they make a new car every 3 minutes - hence the much larger row of Superchargers at the parking lot.

Here's a side view of the new construction I've found in the latest video (shout-out to @KarenRei):

No way are those new foundations large and long enough for Model Y assembly line, nor for a BIW lines - but note how the new foundations are approximately twice the size of the stamp shop. Maybe they are adding more hydraulic presses to double the stamping capacity? That would be an ideal location for such an expansion, and might explain why Elon said that they need new construction for Model Y at GF3.

 

[Artful Dodger]

It's not just the headline, the whole article is carefully written to make it appear that the SEC may be investiging Tesla. Every single time they mention investigation it is of "the firm", never UBS nor Tesla. Totally reads like a FUD piece.

Nawh, its not FUD its an ESL piece... but they're learning and tryin' hard.

Cheers!

 

[Sean Wagner]

Elon drives over a traffic cone, SP goes up. Can we get him to do this every day?

Dear Deities, please no. Or did you want to start cheering Elon the cone man?

It would be nice - but even if Tesla could make 2TW of batteries in 2024, I think they might have a little bit of a difficult time producing 26.6 million vehicles (unless they have another 10 or 20 unannounced Gigifactories they plan to build in the next 48 months.)

Agree this is a fantastical notion.

However, let's forget that sky-high number and consider what kinds of vehicles might be cheap and sell in volume sometime down the road: say a fully autonomous, geo-fenced metropolitan rover limited to 35 mph and built to last a million miles [just throwing ideas out there, please don't nitpick].

Cheap and long-lasting. Boxy with a small footprint. The low-cost battery pack and robust motor* bundled with autonomy would be the core product. In small van guise great for businesses, great for high-frequency "mass" transit where you seat six to eight people per pod... how many classic cars would that obsolete?

*maybe one per wheel, with built-in gearing.

I'll admit the idea of Tesla building a vehicle family that's pared-down and constrained by design does not ooze s3xyness, but would it help the mission while being profitable? The core platform and software could be modified by third parties.

The takeaway being to not only extrapolate but keep in mind that the solution space is changing too. What remains is the requirement for massive amounts of cells and a smarter grid.

 

[KarenRei]

That's setting the bar pretty low.

They clearly meant to say, "More fun than watching a SpaceX rocket launch"

 

[mongo]

There are loading docks on both sides of the building - but I think the angled/sawtooth loading docks on the other side are for the BIW lines.

But my guess is that there is so much overlap in final assembly of the Model Y and Model 3, that I'd expect the two lines to run in parallel and next to each other, and workstations to be 'mirrored' in a fashion - and maybe even shared and crewed by the same people, and the parts prepared and conveyed in unison. Given that throughput is about 1 car every 3-4 minutes, it's not a problem for supplies to cross the assembly lines - as long as the second line is still close to the loading dock and the unpacking area.

The Model Y shares 76% of the parts with the Model 3 - and I guess a large part of that percentage is the chassis/panels, which are assembled by robots. At the general assembly level, which is about ~70% of the factory footprint of total car assembly, I'd expect the part sharing factor between the Model 3 and Model Y final assembly lines to be as high as 90%, maybe more.

If that is true then it would be natural to line up the two assembly lines and build them in parallel, and only bifurcate the workflow where there's real differences.

The Model Y might still require a new building: a BIW body line next to the stamping shop. The new pile-drivers are close to such a location. This new Model Y BIW building would be relatively small, compared to the length of the Model Y final assembly line.

But even this layout isn't ideal: it would mean that the Model Y bodies would have to travel the whole length of the building to get into the paint shop. So either they build a second paint shop for the Model Y, or the current Model 3 BIW line has enough capacity to make the Model Y as well:

View attachment 486708​

Note the very purposeful topology of the assembly lines: stamp shop feeds into the body shop on one side of the building (which has the saw-tooth pattern loading docks), which feeds into the paint shop, which feeds into the start of the very long general assembly line, which spews out finished cars into the parking near the stamp shop. The car, as it is assembled, takes an almost perfect circle within the factory and always stays close to loading bays.

Note how the phase 3/4 building is a mirror image of the phase 1/2 building: the highest throughput loading docks are at the outside boundary of the factory, and the dense, heavy shipments of raw metal to the BIW ships are loaded via the saw-tooth loading bays in the middle of the factory.

Just ... slapping a Model Y building next to the stamp shop breaks all this current and future symmetry, disturbs the smooth flow of cars within the factory. This makes no sense to me. I think both the current BIW and the general assembly line has enough free factory space to host Model Y assembly.

Could be wrong though. Maybe @Krugerrand has some ideas about how they might approach this?

I claim that BIW and GA are both single line and will remain that way for one main reason: material flow.
Body (possibly handles 3 and Y) uses a raised transport system, it would require up and over arms for loading bay provided materials (stampings could flow easily) to get the GA or second line. Plus GA and body are separated by a wall.

One GA line is bad enough, you need to get parts across the line to the opposite side of the car from the loading doors. If you put two lines in parallel, then all the second line's material cross the first line. This requires gap in the line of cars and storage in the inner loop or between lines. A single long line only needs storage between the door and the line. Look at the pics (also attached) in https://ir.tesla.com/static-files/47313d21-3cac-4f69-9497-d161bce15da4
The final inspection is a single lane offset in the bay to the building center with dual dynos next to it. GA is a single line, body is a single line, only paint has parallel lines of painted cars, one blue, one white, one empty (multiple booths upstream or just staging?) You can also see all the prep areas that take up the remaining space.

 

[KarenRei]

It would be nice - but even if Tesla could make 2TW of batteries in 2024, I think they might have a little bit of a difficult time producing 26.6 million vehicles (unless they have another 10 or 20 unannounced Gigifactories they plan to build in the next 48 months.)

Halve that. The 2TW is supposed to be half for automotive, half for energy.

Still a crazy-big number, of course!

 

[Tslynk67]

Pure gold - $TSLAQ adding Lin Wood to their block list (where's the ROFLOLSMS) icon when you need it?

 

[KarenRei]

Ok, to better illustrate the layout of GF3, I've corrected the 3D perspective distortions of the photo of the screen and created an overview/map view of the GF3 layout, with arrows added to illustrate in what direction the workflow is probably going:

View attachment 486745
​

The counter-clock rotation factory flow is easier to see in this image:

• Tesla cars start their GF3 life probably somewhere on the boundary between the square stamp shop and the rectangle shaped BIW shop - which is a mostly automated section with hundreds of welding robots.
• As they move through the BIW they get into the paint shop, and probably exit the paint shop somewhere in the south-western corner.
• The general assembly line is hundreds of meters long and mostly linear.
• Finished Model 3's exit the factor on the right side end of the GA line, right next to the stamp shop. This is consistent with where we saw new cars emerge on drone videos:
  • View attachment 486742

Those new Model 3's emerge at the right side end of the GA line, and get charged immediately, before being parked at the larger parking lot. This charging flow will have to chance once they make a new car every 3 minutes - hence the much larger row of Superchargers at the parking lot.

Here's a side view of the new construction I've found in the latest video (shout-out to @KarenRei):

View attachment 486737

No way are those new foundations large and long enough for Model Y assembly line, nor for a BIW lines - but note how the new foundations are approximately twice the size of the stamp shop. Maybe they are adding more hydraulic presses to double the stamping capacity? That would be an ideal location for such an expansion, and might explain why Elon said that they need new construction for Model Y at GF3.

If they add more presses there, they'll need some serious foundations The foundations for the current press line looked like they were digging for the Lost Ark when they were building them

 

[dc_h]

Ironically, it's on a website run by a Tesla bull (Vincent).

I don't think it is intended to point to Tesla, just not worded very clearly. When he says the firm, he means UBS. If he was talking about Tesla he would say the company. Technically this is correct, a financial firm, the manufacturing company. Best practices is to avoid pronouns and just say the name to avoid ambiguity.

 

[jerry33]

Lately it feels like the only violations that Twitter cares about are copyvio.

Copyright violations are the only ones that Twitter can get in trouble for.

 

[Tslynk67]

I don't think it is intended to point to Tesla, just not worded very clearly. When he says the firm, he means UBS. If he was talking about Tesla he would say the company. Technically this is correct, a financial firm, the manufacturing company. Best practices is to avoid pronouns and just say the name to avoid ambiguity.

I did DM on Twitter to advise some adaptation, but no response. He's registered here too, but I don't know the handle...

 

[Boomer19]

I did DM on Twitter to advise some adaptation, but no response. He's registered here too, but I don't know the handle...

appears to be him opening these threads - example below

Tesla's Made-in-China Model 3 May Qualify For Government Incentives in China [Rumors]

 

[jerry33]

However, let's forget that sky-high number and consider what kinds of vehicles might be cheap and sell in volume sometime down the road: say a fully autonomous, geo-fenced metropolitan rover limited to 35 mph and built to last a million miles [just throwing ideas out there, please don't nitpick].

Not nitpicking, but a 35 mph speed limit will only work in a few places such as Tokyo. I don't think it would need to be geo-fenced either. It should be just as cheap to make one with a max speed of 65 mph which would be viable in most cities.

 

[Hock1]

Sales for Taycan, eTron, etc. would have been off the charts if they had prices and specs even remotely comparable to Tesla (and the production capacity to match). But they were obviously inferior in multiple ways for the price. That just leaves a bigger hole for Tesla to fill. As the EV price curve continues to decline, it's really all about how quickly Tesla can expand its production capacity (which further drives cost declines). Which is why so many of us are looking forward to the upcoming Investor's Battery Day.

This question of 37M cars in 2024 has been a question in my mind for quite a while. Is it possible Kathie is talking about EVs, not BEVs? If so, that is doable IMO. (i.e., including PHEVs). The only way 37M BEVs is going to happen is 1) a massive increase in battery pack supply and 2) the licensing of Tesla's operating and battery management systems to others.

 

[jerry33]

This question of 37M cars in 2024 has been a question in my mind for quite a while. Is it possible Kathie is talking about EVs, not BEVs? If so, that is doable IMO. (i.e., including PHEVs). The only way 37M BEVs is going to happen is 1) a massive increase in battery pack supply and 2) the licensing of Tesla's operating and battery management systems to others.

The way I understood it is the 37M is BEV although Kathy only used the term EV. In either case this makes Tesla's 17% share 6.3M (and like Apple, they are likely to take the lion's share of the profits). Most likely, a number of those BEVs are going to be the very small ones from China, which will use far less material (once the Chinese manufacturers figure out how to make higher quality batteries. The low quality batteries are hurting them currently). I believe she's relying on VW to make up a substantial quantity.

 

[jbcarioca]

...
The Model Y shares 76% of the parts with the Model 3 - and I guess a large part of that percentage is the chassis/panels, which are assembled by robots. At the general assembly level, which is about ~70% of the factory footprint of total car assembly, I'd expect the part sharing factor between the Model 3 and Model Y final assembly lines to be as high as 90%, maybe more.

If that is true then it would be natural to line up the two assembly lines and build them in parallel, and only bifurcate the workflow where there's real differences.

....

​

Could be wrong though. Maybe @Krugerrand has some ideas about how they might approach this?

Two specific areas for which Y should probably be materially different than 3 are the wiring (IIRC Model Y expected to have much less wiring than does Model 3), and stamping processes which might have the huge complex stamping that has been patented but not yet appeared. In both of those cases the factory floor would have similar structure with two exceptions. One: the presses would probably be physically larger/stronger than previous ones:
Tesla Goes Patent Crazy With Huge Casting Machine For Model Y
Two: The wiring changes seem specifically intended to reduce weight and complexity while being easily adaptable to robotic assembly.

I am not an expert, but it does seem that these two items might reduce costs quite materially, plus increasing durability and quality, thus reducing warranty costs too. Is there any clear notion whether these are being implemented for Model Y production in both Fremont and GF-3?

If so, what direct impact will they have on manufacturing and warranty costs? These two have accounted for quite a few labor hours as well as testing/servicing expenses. Perhaps some manufacturing cost accountant among us might make educated guesses.

I have some ancient ~1993 data for what was then considered to be a very good assembly plant. Wiring alone was ~3% of direct assembly costs. BIW assembly was ~6.5%. I know those bear minimal relationship to modern automated processes, but that complex stamping and wiring pre-assembly/automation have been two of the more intractable challenges.

If I am incorrect in these assumptions please advise me..

 

[ZsoZso]

When you pick the FSD option on the order page of any Tesla you see the below. Any reasonable consumer would then expect the car to ‘respond’ to traffic light and stop sign and get ‘automatic driving on city’. I think this is what @Singuy was alluding to. If this deadline is not met it will provide amunitions to Tesla haters, even if we at TMC understand the fine details on how the company works and we don’t mind the occasional delays given the massive progress the company is making.

Coming this year:

• Recognize and respond to traffic lights and stop signs.
• Automatic driving on city streets

It is called "marketing message"
Same as when Audi told the world in 2011 they would release their eTron EV "next year", they were only late 7 years, i.e. 700% time slippage...