Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

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

This site may earn commission on affiliate links.
Speaking on the EU ruling about the sound requirement:

One, stupid. That’s a benefit to an electric car, not detriment.

Two, does this also affect Plug in Hybrids? They’re just as quiet a lot of the times at lower speeds. Got a ride today from a buddy and his Prius, couldn’t hear it roll off into the distance after he dropped me off.

And three, ugh! I was hoping for the Jetson’s car noise. Just how ‘traditional’ does the sound have to be?

Can we choose? Can it sound like my Harley-Davidson with the TAB pipes? My ‘70 Firebird with the Flow-master set up? It’d be interesting for it to go from 11 to 0 after crossing from 12mph to 13mph.
 
Trade-war cease-fire - Nasdaq futures and pre-market are way up across the board.

It's not just a cease-fire, it appears Trump blinked first and basically capitulated in the trade war against China: they gave China concessions, such as removing some Huawei restrictions.

While I don't have high expectations of what Trump is aware of, he seems to be aware of the fact that a trade war escalation has a fair chance to trigger a recession in the U.S., which would further reduce his already bleak re-election chances.

I.e. chances are that the China trade war will stay at the current levels until November 2020.
 
Good summary of the known unknowns but beware of the unkown unkowns. Supply chains are tricky enough.

Yeah, but unknown unknowns happen all the time, and supply chains are always tricky.

I was thinking of major risks that stem from strategic (design) decisions: the original Model 3 expansion at Fremont was effectively a large, crazy swarm of such factors (most of them self-inflicted), and none of those or similar risk factors appear to be present in GF3. Tesla probably performed dozens of capacity upgrades of the seat factory, powertrain factory as well - but those didn't get in the news because they were incremental and low risk.

GF3 is, to a large degree, basically an incremental upgrade, but at a different geographical location.

Note how they don't start Model Y at GF3 - although they could certainly have tried that. Totally conservative Model SR+ production expansion - like like what traditional automotive and their tooling vendors have been doing for the past 50 years. So apart of a force majeure such as a key ship with a large press machine sinking or an earthquake or a hurricane, this should be a pretty "boring" expansion.

I like boring.
 
What kind of problems do you expect at GF3, which would delay volume production by 6 months?

So yes, the initial Model 3 ramp-up was a cluster-sugar of Dreadnought hubris, lack of organization of a hyper-growth company and a doomed effort to deliver a baby in half the time by throwing money at the problem.

I was thinking of major risks that stem from strategic (design) decisions - none of those appear to be present in GF3.

Last time was hubris but this time it is different and you can't see how things could go wrong? I think the lesson with Tesla is that things going as expected is unexpected.

Given that last time there were strategic risks and you don't see them now, did you point them out last time? My point is, are these strategic decision and risks only apparent to you in hindsight?

The factory build has been fast but it is July and they are still building the structure. At best they will be making a handful of cars this year in the new factory. With Model 3 ramp-up they were already starting production at this point of the year.
 
Last time was hubris but this time it is different and you can't see how things could go wrong? I think the lesson with Tesla is that things going as expected is unexpected.

Given that last time there were strategic risks and you don't see them now, did you point them out last time? My point is, are these strategic decision and risks only apparent to you in hindsight?

The factory build has been fast but it is July and they are still building the structure. At best they will be making a handful of cars this year in the new factory. With Model 3 ramp-up they were already starting production at this point of the year.
Give me a break!

Your last statement is completely ingenuous. They already had the factory up and complete. Come on dude, at least put your comments in context. Geeze.

Dan
 
Last time was hubris but this time it is different and you can't see how things could go wrong? I think the lesson with Tesla is that things going as expected is unexpected.

Given that last time there were strategic risks and you don't see them now, did you point them out last time? My point is, are these strategic decision and risks only apparent to you in hindsight?

The factory build has been fast but it is July and they are still building the structure. At best they will be making a handful of cars this year in the new factory. With Model 3 ramp-up they were already starting production at this point of the year.

That’s a tad pessimistic. With M3 here, the factory and everything was already well established. Workers, equipment, batteries. And I don’t think they started true productions until like, August or September?

Now that building is going up *quickly.* I don’t doubt they’ll be able to get enough of the interior built up to start making cars by the end of the year, and they know the car manufacturing like well enough that they can replicate what is already used.

I don’t think they’ll produce 100k/month by December, but I think it’ll be in the 10,000 level by end of the year at least.
 
Given that last time there were strategic risks and you don't see them now, did you point them out last time? My point is, are these strategic decision and risks only apparent to you in hindsight?

I do think the 600% expansion going from ~100k/year to ~600k/year at Fremont within a 12-24 months time frame was super risky and crazy - and if they attempted it today I'd say so.

Last time was hubris but this time it is different and you can't see how things could go wrong? I think the lesson with Tesla is that things going as expected is unexpected.

This time it's different because they scaled back their ambitions by an order of magnitude: they are going from the current ~500k/year at Fremont with an incremental increase of +150k, which is a +30% increase only.

It's a factor of 20 slower expansion, and it is done in an entirely new building that isn't burdened by the constraints of existing production and historic layout of the NUMMI factory. They also stated that they are going to use the production methodologies of Fremont, they won't prototype crazy new production equipment at GF3.

Basically with GF3 Tesla went from a crazy everything-at-once expansion cycle to something similar to the more conservative "Tick-Tock" model that Intel is using to reduce expansion risks:


Furthermore, note what the primary 'ramp-up troubleshooting tool' at Fremont was: more manual labor. In China "more manual labor" happens to be both broadly available, and much less expensive as in the U.S.

While I expect Tesla to make all new mistakes in their quest for a better planet, I don't expect them to repeat past mistakes very often: and the primary mistake with the Model 3 was simply the speed of expansion. Most of the problems they encountered (chaos, lack of testing, lack of oversight, out-of-phase Panasonic expansion) was a consequence of too fast expansion.
 
Last edited:
Agree with you on GF3, it's harder to do something the first time than to do it again.

They struggled with the paint shop achieving enough throughput and good enough paint quality - which they solved by color batching, color option removal, paint option price increases and (I assume) software optimizations.

And allegedly sabotage.
Elon Musk Claims Tesla's Manufacturing Operating System Has Been Sabotaged By Employee | CleanTechnica

(As I recall there were mods to the paint shop code, but I'm not finding that specific link)
 
(As I recall there were mods to the paint shop code, but I'm not finding that specific link)

Yes, IIRC the allegation was/is that the "Manufacturing OS" - the dedicated, Linux-based system that is running on all of Tesla's factory equipment, was hacked/backdoored by Martin Tripp, with an unknown number of other changes.

Just a convenient and random seeming ... hick-up in a key piece of machinery in an end of quarter peak rate production run would be enough to halt production and cause a significant drop in the stock price, so I think that kind of paranoia is justified, regardless of whether Tripp did it or not.

Especially since Tesla is using the "Manufacturing OS" on "almost all" of their equipment, this is what gives them their ramp-up flexibility:

A Look Inside Tesla's Fremont Automotive Factory — #CleanTechnica Field Trip | CleanTechnica

"Tesla’s Manufacturing Operating System was built completely in-house and has evolved over time as the company grew. It currently supports nearly all of the company’s manufacturing equipment. The custom-built operating system has allowed Tesla to fine-tune its equipment and processes. It clearly comes at a cost, as every change must be vetted and developed internally, but the upshot of the additional internal complexity is flexibility. Tesla can quickly come up with a new improvement or change to its products, equipment, or centerlines and implement it before another automaker would even be able to get a formal proposal together to send to a vendor."

...

"The Schuler press is so large that it is hard to take in at a single glance. We walked around the Schuler press and were able to see a changeover of the dies it uses to stamp out parts. Matched sets of dies are used in the Schuler to stamp out uniform aluminum body parts even faster than the rate claimed by the manufacturer, thanks to numerous improvements made by Tesla after commissioning the press."

...

"Tesla’s team installed and commissioned the Model 3 servo press in ~25% less time than the fastest installation Schuler had ever done before."

...

"Changes like this and more allowed Tesla to increase the rate of the press from 12 strokes per minute to 14 strokes per minute, a 16% improvement."​

And note the very fast ramp-up speed mentioned above - skills like that are invaluable to bringing a new factory online, and Tesla has an experienced team that has gone through 3-4 major expansion and ramp-up cycles already.

AFAIK even the German premium carmakers are nowhere close to this level of vertical integration of their industrial automation platform, they just use whatever facilities come from the tooling, and that's rather diverse (each vendor has their own platform).

Their vertically integrated factory OS is another ~5 years of hidden advantage and moat Tesla has over traditional ICE OEMs, which moat is in addition to the fact that they are not selling obsolete engines and are already 100% electrified, IMHO.
 
Agree with you on GF3, it's harder to do something the first time than to do it again.

And allegedly sabotage.
Elon Musk Claims Tesla's Manufacturing Operating System Has Been Sabotaged By Employee | CleanTechnica

(As I recall there were mods to the paint shop code, but I'm not finding that specific link)

Yes, IIRC the allegation was/is that the "Manufacturing OS" - the dedicated, Linux-based system that is running on all of Tesla's factory equipment, was hacked/backdoored by Martin Tripp, with an unknown number of other changes.

Here's the allegations in the Tesla vs. Martin Tripp court case (in Document #1):

Docket for Tesla, Inc. v. Tripp, 3:18-cv-00296-LRH-CBC - CourtListener.com

"Tesla has only begun to understand the full scope of Tripp’s illegal activity, but he has thus far admitted to writing software that hacked Tesla’s manufacturing operating system (“MOS”) and to transferring several gigabytes of Tesla data to outside entities. This includes dozens of confidential photographs and a video of Tesla’s manufacturing systems. "

"The improper means used by Tripp to acquire and disclose Tesla’s trade secrets include:

a. Breaching specific provisions of the Proprietary Information Agreement;
b. Writing software to hack Tesla’s MOS;
c. Exfiltrating confidential and proprietary data from Tesla’s MOS for the purpose of sharing the data with persons outside the company;
d. Sending third parties a confidential code or “query”;
e. Taking and sharing with third parties dozens of photographs of Tesla’s manufacturing systems;
f. Taking and sharing with third parties a video of Tesla’s manufacturing systems; and
g. Attempting to conceal electronic evidence of his misappropriation and disclosure of trade secrets."

[...]
If those allegations are true (they might not be), the pattern is at minimum industrial espionage, and sabotage for profits isn't an idea far removed, especially considering his TSLAQ affiliations.

I didn't find the paint shop connection though, and since Tripp was employed at GF1, it would be difficult but not impossible to imagine him understanding the paint shop code and procedures to launch successful sabotage. Note that there was a very interesting increase in paint quality last summer, and paint sabotage would be a particularly nasty variant as it requires a very expensive recall. So I don't exclude the possibility.

Edit: found the paint shop sabotage allegation, it's in a NYT article quoted by @EinSV:

TSLA Market Action: 2018 Investor Roundtable

"At 6:30 a.m. on Aug. 18, three robots in the paint shop at the Tesla factory in Fremont, Calif., started malfunctioning. The incident forced a production halt on the Model 3, the key to the company’s future.

Made aware of the stoppage, Mr. Musk went to the factory and worked into the night. The problem was resolved, but Tesla reached a troubling conclusion: The robots had been infected with malware in an act of industrial sabotage. And though they could not prove it, executives suspected they knew the culprit: a rogue employee, working at the behest of short-sellers."​

So I suspect the sabotage (or botched malware attack) did happen, and they suspected Tripp but couldn't link it to him.

Anyway, I think the sabotage angle was, fortunately, just a minor factor - most of the big delays in the Model 3 ramp-up were self-inflicted by Tesla.
 
Last edited:
All this talk of "hubris" at GF3 is at best wilfully ignorant, at worst deliberately deceptive.

To add to @FC's comments, one should consider that with the initial ramp of M3 in Fremont, they initiated a load of new techniques, over-automated, many of the ideas just didn't work well and were subsequently either downgraded to manual or totally revised. During the journey they setup the famous tent and simplified things even further - with ideas as obvious, in hindsight, as using gravity to roll the cars along the line.

Furthermore, output was initially stymied by the pack-production rate, which has subsequently been resolved by Grohmann's lines being installed. AFAWK right now, manufacturing constraints centre on Panasonic's battery output.

So there was a very steep learning-curve went on at Fremont and much discovered along the way that we are probably unaware of - all sort of tips and tweaks that will all come to bear at GF3, so these mistakes, at least won't happen again. Of course this assume that nothing "revolutionary" will come into force immediately, I think not, Tesla will want to de-risk this particular venture.

Another know problem for Tesla in the US, is getting enough skilled workers, at least we know about this on the battery lines. I don't think this will be an issue in China at all - lots of smart people that will be highly motivated to work for a company with the prestige of Tesla.
 
Reporting from Bangkok Thailand.
The car traffic and the accompanying air pollution and scooter noise make this city unlivable.
Transferring to EVs cannot happen here soon enough.
Yup same experience in Beijing. People in the US don't understand that it feels like committing suicide in a closed garage just by walking on the streets.
 
A couple of random observations regarding Q2 deliveries:
  • Tesla China was pulling demand levers like crazy in the first two months of Q2, but none in June AFAIK.
  • While China is much of a black box, the bullish interpretation is that they probably had customers matched up for all the Q2 inventory.
  • No significant demand levers were pulled in Europe in Q2 AFAIK, other than discounts on pre-Raven inventory S/X units.
  • In the Netherlands the final week of deliveries in June wasn't nearly as crazy as the final week of March:
  • upload_2019-7-1_13-46-9.png
    upload_2019-7-1_13-45-7.png

  • The crazy wave of Q1 was spread out over 4 weeks, with an actual reduction in the final week of June.
  • In Norway there's a similar picture for the final weeks of Q2:
  • upload_2019-7-1_13-48-6.png
  • In Norway too the final week was lower than the second-to-last week - which suggests deliberately timed delivery peak to well before the end of the quarter.
  • The bullish interpretation would be that Netherlands didn't have much Model 3 inventory unspoken of, or at least that all intended deliveries were completed in Q2.
  • All the leaked Tesla internal "motivational messaging" were related to North America deliveries, with today's leak suggesting that they were 'very close' to the quarterly record with 1 day left from the quarter:
  • (Fair article from Fred.)
To me this suggests that there were no unexpected logistics hick-ups in Q2 in any of the three major regions, and that I'm cautiously optimistic that we could be looking at a quarter with record deliveries higher than the 90,700 of Q4'18.
 
Yes, IIRC the allegation was/is that the "Manufacturing OS" - the dedicated, Linux-based system that is running on all of Tesla's factory equipment, was hacked/backdoored by Martin Tripp, with an unknown number of other changes.

Small correction -- I believe Tesla's allegations that the paint shop code at Fremont was sabotaged were regarding a *separate* sabotage incident attributed to another, unknown saboteur, and not necessarily to Tripp, who was in Nevada. (Yes. There were probably two saboteurs.)

Fact Checking found the citations. The paint shop saboteur was much better at covering his/her tracks than Tripp was, and presumably was a different person.
 
So, fair warning for people watching market action this week. We've had honest-to-goodness bear raids over July 4th week before. It's a preferred time for bear raids because the markets close at 1 PM on Wednesday and are closed all of Thursday. Lots of people take long holidays. This makes Wednesday and Friday low-volume days, prime targets for bear raids. Mon-Tue, today and tomorrow, might see a bear raid too, but it's more likely to be timed for Wed & Fri.

If you see a big bear raid this week, it'll probably stop (well, go back to the normal level of manipulation) next Monday when normal trading resumes. This is not the week to have margin loans.
 
Gali on Hyperchange mention recently that it took Toyota 10 years to scale the Prius to similar production numbers that Tesla is already achieving with the Model 3.

Prius numbers. For the first 7 years Toyota only managed to scale to a yearly production of 43k. The Prius was different to most cars produced at its release but it's nothing like the complexity of bringing the Model 3 to market. Tesla not only produced the product at scale but they also built the needed charging, dealer and service networks in less time than it took Toyota to scale the Prius. Tesla should be lauded for what they have achieved.