Tesla

[MrH]

What email?

Nevermind.

Well, he talks about precision and tolerances as if he's just discovered the CNC mill and everyone else is handfitting machines with files and scrapers.

He seems frenetic. In any of my managers, I would say "nervous." Does he always write like that?

Lol, the part where he says all Tesla suppliers will tighten their tolerances by ten fold or they're not fit to be a supplier?  This dude is an idiot.

Tighter tolerances = more money.  You want tight tolerances where critical, but there's a lot of stuff it's not worth paying extra to maintain.  This is like basic car design stuff.

The Model 3 was sourced to the lowest bidder almost across the board too.  Pretty much the opposite of the Model S and X.  They previously sourced to a lot of established supplier, and just trusted them to handle a lot of the design, quality, and launch portions.

[Soup DeVille]

It all seems naive; like he's just figured out having contractors on T & M means you get a raw deal. I don't know. I want to like the guy; I want Tesla to kick ass. I just don't see it happening.

[CaminoRacer]

This is a train wreck.

Plant capacity is not something you can just turn up with a dial.  The entire production line should be designed around a takt time.  Everything from cell layout and labor planning to material movement and flow is designed around this takt time.

What was the original designed takt time of the line?  Is it something ridiculous, and there are really that many process failures that they're just working through each bottle neck every time?  Or are they redesigning the entire line with every capacity increase?  None of what he's saying makes sense.

Now he's going to switch from 10 shifts a week (2/day, 5 days/week), to 21 shifts a week (3/day, 7 days/week) to try and hit his production target that he was going to be able to do on just 10 shifts?  There are MASSIVE productivity issues going on then.  This is a very, very costly way to try and fix the issue.

The whole email is a mess.  Encouraging mass insubordination?  Approving every expense?  Just a massive pump before some really depressive Q1 numbers roll out.  Then Musk will play it off as "it was just an internal email to try and rile the troops".

Bro, if they haven't heard of TPS you think they know what takt time is?

[Soup DeVille]

Bro, if they haven't heard of TPS you think they know what takt time is?

Of course they do. Tesla recruited heavily from my area; which means people who are already familiar with the industry. And they were apparently told to forget everything they knew and reinvent the wheel.

[MrH]

Yeah, there are a lot of seasoned automotive guys there.  There's some good talent too.  But you can't launch a car by just winging it.  You need really, really mature processes to ensure everything launches on time and in budget.

[CaminoRacer]

Tesla's problem is that they are trying to change EVERYTHING. There's no reason they have to change production concepts or put the vent controls 4 menus deep in the center screen, but they do.

[MrH]

Tesla thinking they are smarter and better than Toyota at manufacturing meant they were doomed to fail.  LONG LIVE THE KING!

[AutobahnSHO]

Of course they do. Tesla recruited heavily from my area; which means people who are already familiar with the industry. And they were apparently told to forget everything they knew and reinvent the wheel.

[12,000 RPM]

Between this and the FB Cambridge Analytica shit maybe Silicon Valley is getting the humble spanking it needs

Auto industry just sitting back like

[AutobahnSHO]

Between this and the FB Cambridge Analytica shit

Meh, that crap was a scare but nothing will happen.

I wasted time last nigh reading the entire Zuckerberg vs. Senate transcript. So lame.

[Galaxy]

Meh, that crap was a scare but nothing will happen.

I wasted time last nigh reading the entire Zuckerberg vs. Senate transcript. So lame.

However on May 25 the new EU General Data Protection Regulation comes into force, which would punish something like the Cambridge Analytical situation with a fine of 4% of world wide turnover.

[93JC]

Elon Musk

The most fundamental difference is thinking about the factory really as a product, as a quite vertically integrated product.

Jeffrey Straubel

It's treating it as more of an engineering and a technical problem as well.

Brian Johnson

Right, which is the Toyota Production System.

Elon Musk

Yeah. We don't think so.

Jeffrey Straubel

I think that generally it's more of an optimized operational problem, being extremely lean and really managing the flows of materials and the supply chain. They're great at it, but this is I think a different approach, looking at it really from a deep technical lens in terms of automation, robotics, process.

Holy dogshit, the delusions of grandeur at work here are impressive. And the marketing-speak gobbledygook—wow...

[Tave]

It's almost like you guys don't understand the automotive manufacturing industry has the highest barriers of entry of any market on the planet, or that (other than China and India), there hasn't been a significant new player in almost 100 years.

Tesla may or may not succeed long term, but absolutely no one, Tesla included, will break into that sphere by doing the same thing everyone else is doing. It's economically impossible for a new business to beat the GMs/VWs/Toyotas of the world at their own game.

[Tave]

Between this and the FB Cambridge Analytica shit maybe Silicon Valley is getting the humble spanking it needs

Auto industry just sitting back like

Why does the biggest driver of free market wealth of the last 20 years, as well as perhaps the largest single reason why we're not in the throes of a second Great Depression, "need a spanking" as opposed to the industry oligarchs who have continuously fed off the public teat for the last century while actively lobbying against every centimeter of progress?"

[Soup DeVille]

It's almost like you guys don't understand the automotive manufacturing industry has the highest barriers of entry of any market on the planet, or that (other than China and India), there hasn't been a significant new player in almost 100 years.

Tesla may or may not succeed long term, but absolutely no one, Tesla included, will break into that sphere by doing the same thing everyone else is doing. It's economically impossible for a new business  to beat the GMs/VWs/Toyotas of the world at their own game.

Those barriers come from safety regulation, not means of production.

It would be far simpler to adopt the highly optimized methods of production already being used than to re-invent the wheel. None of that would make them make product changes, which is what the customer sees; a unique prioduct.

[Tave]

Those barriers come from safety regulation, not means of production.

It would be far simpler to adopt the highly optimized methods of production already being used than to re-invent the wheel. None of that would make them make product changes, which is what the customer sees; a unique prioduct.

The barriers are at every level, economically and politically.

It's impossible to conjure GM's manufacturing cabalities out of thin air if you plan to do everything exactly like them. You're competing against a 100-year war chest of unlimited revenue streams and capital improvements. It'd be like running a presidential campaign on $5.

[2o6]

How is following a tested procedure at all related to "beating them at their own game?"



It's like making a cake, but instead of using butter and sugar, you use Stevia and coconut milk, and don't account for any changes. Then wonder why your cake is shit.


Following a Toyota-style manufacturing process would only enhance the final product

Those barriers come from safety regulation, not means of production.

It would be far simpler to adopt the highly optimized methods of production already being used than to re-invent the wheel. None of that would make them make product changes, which is what the customer sees; a unique prioduct.

This.

[Soup DeVille]

The barriers are at every level, economically and politically.

It's impossible to conjure GM's manufacturing cabalities out of thin air if you plan to do everything exactly like them. You're competing against a 100-year war chest of unlimited revenue streams and capital improvements. It'd be like running a presidential campaign on $5.

Actually, GM would absolutely kill to be able to build a plant with no union concerns, no thousands of pages of their own standards, no commitment to component interchangeability.

I don't think you have the faintest idea what the interdepartment politics are like at GM.

Tesla wants to buy a robot; they buy the best one for the job and have it put in.

GM wants to, they buy one that meets the CRW standards they've written, make sure it conforms to the NACCL, make sure its supplied from an approved supplier, installed by an approved union contractor, have an equal number of UAW millwrights on the clock while the riggers move it in, have an equal number of UAW electricians sit around while the IBEW guys wire it up, have a low level engineer watch as its programmed by the contractor; have the program approved by his boss, have a safety and movement audit done, and so on and so forth.

[Tave]

Actually, GM would absolutely kill to be able to build a plant with no union concerns, no thousands of pages of their own standards, no commitment to component interchangeability.

I don't think you have the faintest idea what the interdepartment politics are like at GM.

Tesla wants to buy a robot; they buy the best one for the job and have it put in.

GM wants to, they buy one that meets the CRW standards they've written, make sure it conforms to the NACCL, make sure its supplied from an approved supplier, installed by an approved union contractor, have an equal number of UAW millwrights on the clock while the riggers move it in, have an equal number of UAW electricians sit around while the IBEW guys wire it up, have a low level engineer watch as its programmed by the contractor; have the program approved by his boss, have a safety and movement audit done, and so on and so forth.

GM owns $80 billion of capital just in land, property, and equipment. It probably has twice that in R&D and IP, and ten times that in human capital.

It already has thousands of robots and hundreds of factories or supplier factories at its disposal.

Yes, there is a ton of inefficient bureaucratic nonsense among the existing players, but that doesn't change the fact that volume auto startups are literally nonexistent. If it was as easy as moving to a union-free state/country (which is extremely common already) then where is the competition?

[Tave]

How is following a tested procedure at all related to "beating them at their own game?"



It's like making a cake, but instead of using butter and sugar, you use Stevia and coconut milk, and don't account for any changes. Then wonder why your cake is shit.


Following a Toyota-style manufacturing process would only enhance the final product

You're saying this as if they rewound the tapes to 1908 and started with a blank-slate Model T.

Is it more likely that no one in any decision making capacity there has knowledge of modern manufacturing processes, or do you think maybe they're trying to exploit what they perceive to be inefficiencies?

[Soup DeVille]

I think doing business with GM is hard to describe to someone who hasn't been through it themselves. The word "inefficiency" doesn't really describe it fully.

Automation has made tremendous gains in the last 20 years; yet many GM factories are still running 30 year old equipment in some places. Many factories are built around fixed Chain on Edge conveyor systems that can be traced back to WWII. Some of them have been converted from DC power and still have the power shafts from steam engines still on the ceiling.

Moving to non-union states does not solve union problems; as GM actually has to make deals with the unions to build those factories, often ensuring a supply of union contractors and skilled trades. a

In any case; how can you make the argument that paying for experimental and non-standard automation is somehow more cost effective? 

I have worked in this industry most of my life; usually in the building of plant and automation equipment. I worked on the comissioning of the Oshawa II plant, Lake Orion's retooling for the G6, the decommisioning of Shreveport, and a dozen smaller projects. I am telling you; what Tesla is doing doesn't make a lot of sense

[93JC]

It's almost like you guys don't understand the automotive manufacturing industry has the highest barriers of entry of any market on the planet, or that (other than China and India), there hasn't been a significant new player in almost 100 years.

Tesla may or may not succeed long term, but absolutely no one, Tesla included, will break into that sphere by doing the same thing everyone else is doing. It's economically impossible for a new business to beat the GMs/VWs/Toyotas of the world at their own game.

So they're going to break into that sphere by doing... what exactly? "Thinking about the factory as a vertically-integrated product"? That's gibberish. "Look at it really from a deep technical lens in terms of automation, robotics, process"? That's what the old guard does!

[Tave]

I think doing business with GM is hard to describe to someone who hasn't been through it themselves. The word "inefficiency" doesn't really describe it fully.

Automation has made tremendous gains in the last 20 years; yet many GM factories are still running 30 year old equipment in some places. Many factories are built around fixed Chain on Edge conveyor systems that can be traced back to WWII. Some of them have been converted from DC power and still have the power shafts from steam engines still on the ceiling.

Moving to non-union states does not solve union problems; as GM actually has to make deals with the unions to build those factories, often ensuring a supply of union contractors and skilled trades. a

This is all well taken and correct, but we'll have to agree to disagree if you don't think the economic barriers to volume automotive production aren't a separate and significant factor.

In any case; how can you make the argument that paying for experimental and non-standard automation is somehow more cost effective?

Didn't you just argue that? I didn't comment on the efficacy of Tesla's strategy in those posts; I said that they're not going to out-GM GM. 

I have worked in this industry most of my life; usually in the building of plant and automation equipment. I worked on the comissioning of the Oshawa II plant, Lake Orion's retooling for the G6, the decommisioning of Shreveport, and a dozen smaller projects. I am telling you; what Tesla is doing doesn't make a lot of sense

Very well perhaps it doesn't. There are plenty of negative indicators to go along with the positives. What I don't get, and I don't think you are doing this at all, is taking such abject glee in their predicted failure. I don't know whether they'll succeed, but they're trying to do the impossible and have come further than anyone else, so I'm not going to actively ruminate on their demise for the pleasure.

[Soup DeVille]

I've used non-standard in a confusing way.

What Tesla has the opportunity to do is to use the most modern and flexible automation that has already been developed, and could do it cheaper and quicker than any established automaker.

What they've supposedly been attempting to do though is making a revolutionary rather than an evolutionary step; even though they have little idea how to do that. Take Musk's statement about having their tolerances "ten times" tighter than competitors. Depending on what part of the car he's talking about, this can be changing a
tolerance from +/- 1 mm to 0.1 mm or from a ten thousandth of an inch to a hundred thousandth. Any body who knows about these things knows his means increasing cost by at least the same factor. For what purpose? Who will notice? Who will care? What will work better?

Modern manifacturers make money by making the best product at the lowest cost; not by randomly increasing their costs for increasingly minimal gains.

I think you're also unaware of some industry practices. GM and Ford do not for example, do a lot of their own factory engineering. They design the products; and although they guide, coordinate, and write standards, most of the actual engineering and the entirety of the installation and machine building is contracted out.

Tesla has access to and uses most of the same contractors for this.

[GoCougs]

It's almost like you guys don't understand the automotive manufacturing industry has the highest barriers of entry of any market on the planet, or that (other than China and India), there hasn't been a significant new player in almost 100 years.

Tesla may or may not succeed long term, but absolutely no one, Tesla included, will break into that sphere by doing the same thing everyone else is doing. It's economically impossible for a new business to beat the GMs/VWs/Toyotas of the world at their own game.

Well, doing everything objectively bad isn't a way to break into that sphere, either.

But yes, the way to break in is to copy the #1 player, and that is Toyota. Tesla didn't and as a result it is doomed.

[Soup DeVille]

So they're going to break into that sphere by doing... what exactly? "Thinking about the factory as a vertically-integrated product"? That's gibberish. "Look at it really from a deep technical lens in terms of automation, robotics, process"? That's what the old guard does!

vertical integration isn't nonsense. It's how the first textile and pottery mills in England were built in the mid 19th century after all.

[12,000 RPM]

It's almost like you guys don't understand the automotive manufacturing industry has the highest barriers of entry of any market on the planet, or that (other than China and India), there hasn't been a significant new player in almost 100 years. 

I am pretty sure nearly every auto maker started up in the last 100 years. Only Ford and Daimler come to mind as those who came before.

More importantly, there have been several startups in the last 50-60 years; many of which set up shop in the US long after GM became the big kahuna.

Very well perhaps it doesn't. There are plenty of negative indicators to go along with the positives. What I don't get, and I don't think you are doing this at all, is taking such abject glee in their predicted failure. I don't know whether they'll succeed, but they're trying to do the impossible and have come further than anyone else, so I'm not going to actively ruminate on their demise for the pleasure.

They are not trying to do the impossible. It's not so much glee as it is disappointment. Elon cannot get out of the way of his own ego, and seems to seek to make challenges where he doesn't have to.

You look at something like the Iphone, the revolution was in the product. The manufacturing was and remains secondary and conventional. Teslas are similarly game changing. There is zero need to fuck that all up trying to reinvent manufacturing from scratch. Especially at the direction of a guy with zero experience in it.

[93JC]

vertical integration isn't nonsense. It's how the first textile and pottery mills in England were built in the mid 19th century after all.

The statement itself is nonsense. It's empty, meaningless jibber-jabber.

[Soup DeVille]

The statement itself is nonsense. It's empty, meaningless jibber-jabber.

It appears to be, yes.

Musk does not however, whatever he may be, seem to me to be a flim-flam man who throws out nonsense in order to befuddle his audience. For better or worse, I think he believes his own hype, and does not seem the type to throw out nonsense for the sake of nonsense. I think he knows what he means, even if its ridiculous.

Of course, I may simply be drinking too much for a Wednesday.

[AutobahnSHO]

However on May 25 the new EU General Data Protection Regulation comes into force, which would punish something like the Cambridge Analytical situation with a fine of 4% of world wide turnover.

Good stuff, but it appears the GDPR was in the works long before any Facebook shenanigans were going on.

https://edps.europa.eu/data-protection/data-protection/legislation/history-general-data-protection-regulation_en