You originally made a broad claim with no specifics, and now you dive into the nitty gritty details of auto manufacturing and a the very specific divisions of labor.
That’s all fine, but the specifics don’t prove the general. And if the general is true then the specifics are redundant.
I do appreciate the rundown of how typical plants are run but divisions of labor are inherently artificial. They may make sense in one situation but not another.
Pissing on someone’s background is a compliment, coming from me.
The point is that Musk didn’t hire Shotwell because she’d successfully brought another rocket startup from nothing to dozens of launches a year. Nobody had done that. Musk hired her because he saw something in her that went well beyond her modest background. Likewise for Mueller.
Does a similar person exist for Tesla? Maybe, but you aren’t going to find them by shopping around for VPs or executives at other auto companies. It’s going to be someone with a solid education, with some experience but in a place where their talents aren’t being utilized. And someone with an overall vision compatible with Musk (whether in a “make everything electric” or “send humanity to space” sense) but willing to fight on the details on how to get there.
As far as micromanaging people 6 levels below him goes, that’s at least as common at SpaceX as it is at Tesla. And it’s probably a huge component of SpaceX’s success as well.
Ok, suppose it cost billions. How much is that knowledge worth? If it accelerates the production of the Model Y, was it worth it?
It’s not like Tesla hasn’t been learning. Forget Tesla vs. the rest–just look at Tesla vs. Tesla. The entire Model 3 development and production cycle has gone vastly better than their three previous products. It’s a better car, it was developed faster, it’s shipping in much higher numbers, the production has only been roughly a year vs. several, and overall just much less nonsense going on.
If the Y ramp is as much better than the 3 as the 3 was to the S, it’ll be pretty much on par with anyone else production wise.
I can barely parse that transcript at all. I don’t have time to listen to the audio recording right now to find out what he’s really trying to say. Maybe someone else has a better transcript.
For now, I’ll note that he’s talking about rates. A “half million per year run rate in 2018” does not mean producing 500,000 cars in 2018. It means that at some time during the year they will produce 9,600 cars a week, or 1,370 a day, or some other rate.
I suppose I’ll also note that he mentioned pulling in the volume target from 2020 to 2018. So, they greatly accelerated their ramp and managed to only do it 1.5 years early instead of 2 years. This is supposed to be some great disaster?
so basically no one can disagree with you, since you can hand-wave away any and all disagreements because reasons like “everything works the way I think it does.”
got it.
This ain’t worth it anymore. you’re right about everything. you know everything there is to know about building cars and rockets. I’m sorry I ever questioned you.
So if Elon says he will produce 100-200k cars in 2017, and get a “rate” of 500k in 2018, you’re telling me that he actually just meant there would be a few weeks where they produced 10k cars?
I don’t see the words “week” or “day” in there. How do you know that he didn’t mean that there could be a few five minute periods during 2018 where five cars came off the line? Such a five minute stretch could also give Tesla a “rate” of 500k/year. Maybe they have already reached that rate!! Win!!
ETA: poor show of claiming the transcript is hard to understand, and there must be transcription errors. Why not blame that saboteur/whistleblower of editing that website to make Elon look bad?
I genuinely have no clue what the second paragraph is supposed to say. It’s not in English. It may well be Elon being his usual self, or it may be because it’s a dodgy machine translation.
I know I don’t really have to explain this to you, but since the snark is at an all time high and I can no longer tell the difference between sarcasm and not, here is a link about “run rate”:
(the link spoke of revenue and not units, but the same principle applies)
Yeah, but you’re substituting the plain text of Elon’s typical garbled speech for a very narrow definition that isn’t supported anywhere else in the text.
Besides which: “Additionally, given the demand for Model 3, we have decided to advance our 500,000 total unit build plan (combined for Model S, Model X, and Model 3) to 2018, two years earlier than previously planned.” http://ir.tesla.com/static-files/041a2223-4f7b-4007-b030-b48965cc6ad9
If you want to quibble with this, fine, that’s probably like 450,000 Model 3s and the rest Ss and Xs. But that sentence clearly isn’t referring to annualized production drawn from a short period of time (month, week, day, whatever).
Quibbling would put it at 400k, since S+X is about 100k.
But sure. I’ll acknowledge that the letter puts his statement in a different light. And listening to the conference call, he does more or less say that as a rough guess, they’d do 100k-200k in the second half of '17 (the transcription is still seriously wrong, though, sometimes amusingly so–“particular date” turned into “secret base”).
So in 16Q1 we have an estimate of 100k-200k in 2017. Later, in 17Q2 we have an estimate which is much lower than that–1500 in Q3 and 5k/wk in Q4. Obviously, this can’t be anywhere close to 100k, let alone 200k for that year. And in 18Q2 they finally hit that 5k/wk rate.
How you interpret all this depends on how you look at your endpoints. If you compare again the early, early 2020 volume estimate, they’re looking good. If you compare against the 16Q1 statements, they look quite bad. And if you compare against 17Q2, well they didn’t hit those either but it’s only a 6 month delay and that’s not so horrible.
At any rate, I hope you get to drive one soon. They’re great!
I dunno what you’re talking about. I didn’t make any broad, sweeping statements about Tesla or SpaceX. I said some things about the kinds of people Musk tends to hire and I think they’re supported by facts. I’d recommend reading Vance’s biography if you want a better feel for Musk’s style.
I’m actually really curious who you think would be a good fit for a Tesla equivalent of Shotwell. I am fully on board with the idea with the idea that Tesla could use an operations officer to leave Musk with the tech stuff. I just don’t know who. I don’t see any incumbent person being a good fit. You said yourself that most industry people want to stay away. And yet there are people that are both compatible with Musk’s idiosyncratic style and yet can run a company in a sensible fashion. So who?
Why is it on me to identify someone? I don’t work for Tesla, and surely not in their recruitment/talent acquisition. I’m merely saying they need someone to instill some discipline in that operation. if you’re going to poo-pooh me because I can’t come up with someone you’ve heard of, too bad.
I didn’t pooh-pooh you; I agreed with you on that particular point. Yes, they would be absolutely be better served by having a talented operations officer, and in particular one focused on manufacturing. And on the specific comparison against Shotwell, I’ve given her credit in the past for essentially holding together SpaceX.
The only thing I dispute is that there’s just automatically someone out there that can fill the role. And really, I was more curious about types of people than specific names. Best I can think of is someone at a manufacturing automation startup of some kind, but with a reasonable amount of background in the industry. I.e., someone that knows the ropes but also with a dissatisfaction with the way things are done and a willingness to take risks (both personal and business).
When Musk was hiring for SpaceX, he couldn’t find a chief designer. He found propulsion people (Mueller), operations (Shotwell), as well as avionics, ground support and other stuff. But no good fit for, basically, chief rocket scientist. He says he took the role himself not because he was a great fit (his only knowledge in the area coming from reading a ton of books) but because there was no one else.
Well, it worked out in the end and he remains chief designer. And of course today he probably does count as one of the top rocket scientists in the world. Along the way a bunch of mistakes were made, but somehow they survived. Probably the same will be true of his role at Tesla.
Zero serious problems with mine. Phone key is, as mentioned, something like 95% reliable. It’s still a net win over having to carry a dongle, but I do wish it was >99%. Toggling the Bluetooth always fixes it so it’s probably more on the phone side than the car side.
I’ve become somewhat obsessive about washing mine. I suppose I’ll get tired of doing so when the new car aura wears off. But in the meantime I’ve found that hand-washing it is super-quick due to its seamless exterior (no door handles, no grille, simple wheel covers, etc.) and because of the lack of exhaust and almost complete lack of brake dust (due to regeneration). So that’s nice.
Anyway I would keep going, but the gist of it is that it takes several years to commission and tool an auto plant and it also takes several years to ramp up to full production. I specifically picked new plants, producing new models to have as close to an apples to apples comparison.
If any of the veteran board members want to cite other cases of new car manufacturing where the chassis and drivetrain is completely new and is not shared with other models or is a derivative of earlier models and their respective commissioning timelines, production numbers and investment amount so we may have a proper idea of what to expect from Tesla. Overall though, it seems to me that the Model 3 ramp up production seems to be in family with other manufacturers.
and by the by, that gutting also included a thorough cleaning decontamination of the stamping facility and body shop to prevent iron (or other metal) particles from getting onto the aluminum body panels and causing corrosion.
now obviously they were under immense pressure to get it done as quickly as possible given the critical nature of the F-150, but it’s what can be done with proper planning and discipline.
without knowing the actual plans, logistics, and a bunch of other details from the ones you quoted, you really don’t have enough information with which to say such a thing. Especially in the cases where they were building an entirely new plant from the ground up (including the building itself) then something as mundane as site selection can take up an incredible amount of time.
Really liking it. Only issue I’ve had is the rear license plate bolt inserts weren’t correctly installed so they just spin - Was kinda difficult to take the dealer plates off and put the new plates on, but should be an easy fix at the service center.
I also had some initial problems getting my contacts to sync via bluetooth, but this turned out to be due to unsupported characters ( “(”, “&”) in the contacts list, so spending 2 minutes renaming a few of them fixed everything.
Yeah, still not a huge fan of the phone-key - It has lots of potential- when it works it’s great, but lately my success rate is something like 50%. Occasionally the “waking” process on LTE takes ~40 seconds to respond, or the bluetooth connection doesn’t pick up right away, or I can somehow get in the car with the phone but it still tells me to swipe the keycard because it doesn’t pick up the Bluetooth. All this is easily resolved with getting the phone out of my pocket, unlocking it, and occasionally having to disable/re-enable bluetooth on my phone, so maybe ~5-10 seconds, but still annoying. I think most of the issue is phone-side (due to restrictions placed on “always running” bluetooth connections for power management reasons), but still - Tesla really should have offered a fob as a 3rd option beyond the card and phone, even if it cost extra, at least until all of the bluetooth issues could be sorted out, or that they can guarantee that the latest iphone or android can work 99.9% of the time without any issue or delay.
But yeah, once the car is put in “D” and the music is on, i tend to forget all about the little stuff that bugs me. After getting used to the Model 3 for a few weeks and then going back to my wife’s car ('15 Honda CRV), I just get annoyed at the sluggishness, vibration, shuddering, hesitation, and transmission shifts. All that stuff never bothered me before, but yikes, it’s front and center now. Granted, it’s not exactly fair to compare the two, considering one is a under-powered SUV and the other is a much more expensive sports-sedan, but still.
Thank you for contributing to the thread. While certainly super impressive feat, this is apples and oranges. The 8 week timeframe applies to only the teardown and installation of new tooling. That tooling that was installed was obviously designed and manufactured years in advance, the new aluminum stamping processes tested and trialed in small batches offsite years in advance also and the entire line was warehoused in anticipation for the switchover.
The cases I pointed earlier are more analogous to the Model 3 situation:
-New chassis
-New drivetrain
-New processes
-New people
The case of the F-150 is not applicable because there are multiple factories producing it, new models have largely unchanged processes, drivetrains are shared between multiple models, etc.