Cost To Build a Tesla Model 3 Is $28,000, German Engineers Say (www.wiwo.de) 141
Rei writes: An interesting report came out the other day from Germany, where an engineering firm purchased four Tesla Model 3s on the grey market to study on behalf of an anonymous major German auto manufacturer. Among their key findings: due in part to a huge reduction in cobalt in the batteries (2.8% in the cathodes versus a typical 8%) and a number of simplifications, the parts cost of a Model 3 (in units of 10,000 vehicles per week) is estimated at $18,000, along with $10,000 in production costs. Note that the teardown was for the long-range version with the premium upgrades package.
On Reddit, users with access to the full report added further details. The 75kWh battery is 40% of the components cost ($7,200); the interior is completely symmetric (facilitating RHD); there are only 4 kinds of screws used in the underbody (a typical German luxury manufacturer uses 40); many parts of the car are designed specifically so as to be easier for robots to grab; and the battery pack is harder to remove than on the S/X (e.g. not battery swap capable). After studying the individual components, they concluded that German EV manufacturers would not be capable of producing a similar vehicle at this point in time. Asked on Twitter whether Musk agreed with their price conclusions at a rate of 10,000 vehicles per week, Musk replied: "Definitely." That said, Tesla is still in the process of moving from 3,500 to 5,000-6,000 per week by the end of this quarter, and is not expected to reach 10,000 vehicles per week until next year.
On Reddit, users with access to the full report added further details. The 75kWh battery is 40% of the components cost ($7,200); the interior is completely symmetric (facilitating RHD); there are only 4 kinds of screws used in the underbody (a typical German luxury manufacturer uses 40); many parts of the car are designed specifically so as to be easier for robots to grab; and the battery pack is harder to remove than on the S/X (e.g. not battery swap capable). After studying the individual components, they concluded that German EV manufacturers would not be capable of producing a similar vehicle at this point in time. Asked on Twitter whether Musk agreed with their price conclusions at a rate of 10,000 vehicles per week, Musk replied: "Definitely." That said, Tesla is still in the process of moving from 3,500 to 5,000-6,000 per week by the end of this quarter, and is not expected to reach 10,000 vehicles per week until next year.
Opportunity Cost strikes again (Score:2)
So a German engineering firm says that it only costs 28,000 to make a Tasla Model 3 but that German EV companies are not capable of producing a similar vehicle at this time. Does that mean that their German cars cost less then $28,000 to make? Otherwise they'd be foolish not to try and compete.
Re:Opportunity Cost strikes again (Score:5, Interesting)
Learned this in my first econ class ever.
"double your price of whatever you make it for. You will make money. We have all these formulas and tricks to tell us what it should do but we have no idea. 2x just works".
The class then ran dozens of simulations. The guy was right.
Re:Opportunity Cost strikes again (Score:4, Insightful)
Sounds reasonable. Keep in mind that the $28K is just parts and labor. It does not include a share of fixed costs -- overhead, debt servicing, operating costs, taxes ... etc. My experience with that sort of accounting is tangential (IANAA) and not really applicable to mass manufacturing. But the costs of running a business tend to be pretty impressive.
Re: (Score:3)
Sounds reasonable. Keep in mind that the $28K is just parts and labor. It does not include a share of fixed costs -- overhead, debt servicing, operating costs, taxes ... etc. My experience with that sort of accounting is tangential (IANAA) and not really applicable to mass manufacturing. But the costs of running a business tend to be pretty impressive.
R&D...
Re: (Score:3)
how does 2x work if no one will buy the car?
Obviously if you can make 2x for the car you will make money, it does not take simulations to prove that, but the problem comes in with demand versus price point.
Re:Opportunity Cost strikes again (Score:4, Informative)
2x is what you need to avoid losing money on each sale. It has to include unknowns like warranty costs, manufacturing issues and so on.
It's pretty standard to use 2x when costing anything in industry. You want to add a â5 widget to the car, you assume it will add â10 to the price.
Re: (Score:3)
I worked in customer service for a large company, and I could see every cost, and price associated with every product we made, from service parts, to complete units, and you are correct.
We doubled our cost when we sold it to the dealer, and they doubled their cost when selling it at suggested retail.
There were a few things that we completely raped people for (a $3 shaft for $120) but the majority was the double double.
While I was there, I learned another valuable lesson for the double double, and that has t
Pride in my students (Score:2)
I incorporated programming and testing a logic circuit implemented on an FPGA into a lab course. To motivate including this in the lab to the students, I had them do a comparison with the complexity of wiring 7400-series logic to do the same hardware function. The lab manual was revised to ask them to time themselves doing the wiring manually using a Protoboard.
TAs taught the lab sections, but I was required to attend lab sections to supply mandatory evaluations of the TAs. Sitting in on one such lab
Re: (Score:2)
That doesn't make sense.
"How long did it take you?"
"The stopwatch said 18 minutes, so it took us 36 minutes."
Re: (Score:2)
double your price of whatever you make it for
Unless you can sell it for more, then ask whatever you can get.
Re: Opportunity Cost strikes again (Score:2)
"what the market will bear" is a very important consideration, as is targeting consumers with money to burn on premium products where reliability is secondary to image.
Case in point: Remington electric razors. The company was profitable when Victor Kamman bought it. He immediately doubled the price and marketed them as a premium product. Sales skyrocketed and it became even more profitable.
The point about Tesla (and upmarket German cars) is not that it's a Halo product, but that it's a premium and desireabl
Re: (Score:2)
Most German car manufacturer produce EVs, and they sell quite ok.
So, what is your point?
However most EVs sold in Germany are french (surprised?) and Asian.
Re:Opportunity Cost strikes again (Score:4, Informative)
However most EVs sold in Germany are french
To say "french" is a bit too generous. German can manufacturers have not put serious efforts into the EV business yet. Neither really has anyone else in Europe with the single notable exception of Renault.
That is changing. 2017 saw the Renault Zoe head and shoulders above all other EV manufacturers. German manufactures had position 3, 7, 8 and 10. With the remainder being a few Asian cars and some Teslas.
In January this year the e-Golf outsold the Zoe for the first time. Year to date paints the Germans in a very different picture:
1: Smart For2 ED 1947
2: VW e-Golf 1905
3: Renault Zoe 1719
4: Kia Soul EV 1668
5: BWM i3 1620
6: VW Golf GTE 1113
7: VW Passat GTE 1066
8: Smart For4 ED 1058
9: BMW 225xe 943
10: Mercedes E350e 813
Number 11, 12, 16 and 18 are also German manufacturers.
Expect to see many more German EVs on German roads.
Re: (Score:2)
"It could be easier if one were to partner with Tesla*, but what's in it for them?"
Aggravation? Lots and lots of aggravation.
Re: (Score:2)
> German EV companies are not capable of producing a similar vehicle at this time.
Clearly, Tesla'a Apple scientists are better than Germany's Apple scientists.
Re:Opportunity Cost strikes again (Score:4, Informative)
The german manufacturers don't have the structure to keep up with Tesla at this point.
Tesla uses vertical integration to cut the costs of the middleman(the auto suppliers and the dealer network in this case). They produce a lot of key subassemblies themselves, where other's subcontract to various levels.
Traditionnal car OEMs can't do the vertical integration today.
Re: (Score:2)
The german manufacturers don't have the structure to keep up with Tesla at this point.
Tesla uses vertical integration to cut the costs of the middleman(the auto suppliers and the dealer network in this case). They produce a lot of key subassemblies themselves, where other's subcontract to various levels.
Traditionnal car OEMs can't do the vertical integration today.
And other car manufacturers actually deliver cars. Millions of cars where Tesla can't even produce a few thousand 2 years late without have to recall or update them every two months to fix serious quality issues.
Re: (Score:2)
Name another manufacturer selling long-range electric vehicles at the current rate of Tesla. You can't because there isn't one.
Tesla now builds more Model 3s in a week than Chevy builds Bolts in a month.
Re: (Score:2)
Re: (Score:2)
Don't forget that the only Model 3's sold to date cost at least $50k ($35k base, +$9k for the long range option/model, +$5k for the premium upgrade package, +$1k for delivery, even if you live in Fremont).
"4 kinds of screws" (Score:5, Insightful)
That's huge. Last weekend I helped a friend replace the motor in his 1993 Toyota, and IIRC there were 10, 12, 13, 14, 15, 17, 19, and 21 mm bolts plus several different other Phillips screw sizes. Better than GM which my friend had to buy several sizes of Torx drivers or my other friend that has a Harley we replaced a clutch on that needed a couple of exotic Torx drivers that we couldn't find locally. T25 was too loose and T27 wouldn't fit in one of the bolts. He also replaced one of the heads on it, and the special Torx driver from Snapon was more expensive than he paid for the used head!
Re: (Score:2)
> Why did you need to replace the engine?
The engine only had a little over 480,000 miles and still ran great, but Jiffy Lube on 85th in Kirkland forgot to tighten the oil plug after his last oil change. They gave him $2k for a nearly twenty-five year-old car without much hassle so I'll have to give them credit for that. The month before that bad oil change, he drove his Corolla to Reno, NV for a poker tournament which is about 1,500 miles roundtrip so it not only ran, it ran well enough for him to depe
Re: (Score:1)
Yeah.. A hard won lesson. I had the same happen with my favorite Bronco.
It lead to an engine upgrade... I am very thankful I wasn't financially strapped at the time, they didn't pay enough to replace the standard 351 V8 that was in it at the time.
Jiffy Lube==Check all valves, caps, seals, plugs after service.
Re: (Score:2)
Re: (Score:2)
Re: (Score:1)
Snapon was more expensive than he paid for the used head!
Yep. My bother is a mechanic that works for an independent shop that mainly works on German cars, His monthly payment to Snap-on is larger than his house payment. Using only four different fasteners, if true (and I don't think it is but assume it is), is huge like you said.
Re:"4 kinds of screws" (Score:4, Interesting)
Re: (Score:2)
T25 was too loose and T27 wouldn't fit in one of the bolts. He also replaced one of the heads on it, and the special Torx driver from Snapon was more expensive than he paid for the used head!
Maybe those renegade torx were actually triple-square (m) heads? If you ever work on an audi, you'll be blown away at the odd assortment of sockets and devices you need to do actual work on them.
...and make money... (Score:2)
Re: (Score:2)
So long as they don't 'break molds' in the DeLorean sense...
Re: (Score:2)
the 28k component price was for the long range model that is 44k.
Re: (Score:2)
$49k.
Today, you have to buy the Premium Upgrade package, which is another $5k.
Re: (Score:2)
Re: (Score:3)
If by "rest of the costs" you mean R&D and SG&A, yes, but "no duh" to that. ;) (Also worth noting: they were tearing down a Model 3 LR with PUP, not a base model)
Ford's average vehicle margin is 10%. Now, Tesla's structure means they need a higher margin (as they own all of their own stores and service centres), but "greater than 20%" is more than a healthy margin. And that's with no options (which are overwhelmingly profit).
Re: (Score:2)
That's not the gross margin [ycharts.com]. Gross margin is just (revenue - COGS) / revenue. If you're including R&D, SG&A, etc, you're thinking of the profit margin, which is around 4%.
(ED: turns out the gross margin is was a bit higher than I remembered, 14,76% last quarter. But still nothing spectacular)
Re: (Score:3, Informative)
The $10k is not "labour", it's "production costs", which includes tooling depreciation.
And FYI, but a normal automotive margin is around 10%.
Re: Elon Mask is not going to sell it for 35K (Score:2)
Normal automive margins are 10% on a complete vehicle but most carmakers are in the business of selling finance and/or car parts.
Re: Elon Mask is not going to sell it for 35K (Score:2)
Hit submit too soon....
By that comment I mean that the car is simply to get you in the door. They make money on interest payments and servicing.
Something doesn't add up (Score:3)
So either the Germans are able to get much cheaper materials and labor than Tesla, or this is just a FUD piece. Since it's written by an actor, sources anonymous German engineers, and has a paywall for the full article, I'm leaning towards the latter.
[1] A $5 billion Gigafactory on a very aggressive depreciation schedule (double declining balance with a useful life of 10 years) would be $1 billion for the first year, which would add about $10,000 to the cost of each vehicle if they make 2k vehicles / week.
Re:Something doesn't add up (Score:4, Interesting)
I think you are confusing model 3 with model S. Model 3 is the 35k to 44k car. In fact the germans are saying that they cannot current meet how good the price is on the model 3, its not fud, it is the opposite.
Re: (Score:2)
Re: (Score:3)
They must be doing well out of the options they sell, which are often mostly software. Autopilot is pure software, and full self driving isn't even available for years but you can pay them thousands of dollars for it now.
Bioweapon defence mode is a cheap HIPA filter and software switch. The performance modes are software.
Re: (Score:2)
I think Performance models require the dual motor option, which is definitely hardware.
Re: (Score:2)
also they are ramping up to 10k per week.
Re: (Score:2)
Re: (Score:2)
Find a nation that wants to build jobs and ask how much their gov will pay for every worker who has a full time production line job.
Thats how the worker price can be adjusted. Then it is just the cost of parts
Re: (Score:2)
According to the last investor call, Tesla claims their depreciation of the factory (at 5k cars/week) is $2,000/car.
Re: (Score:2)
"Well under" $2k, if I recall the wording correctly.
Also, the above $10k in the German study is total production costs, not just labour. That includes depreciation.
Re: (Score:2)
True it also assumes all those costs are spread over 10k/cars/week, not 5k/cars/week. So, those should probably be doubled today to $20k/car production costs.
Re: (Score:2)
If the components are $18,000, labor is $10,000, and factory depreciation is $10,000 [1], then Tesla should have been making a profit on those $70,000 premium Model 3 sales, but somehow that's not reflected in their Q1 report.
Is that before or after you take into account constant factory modifications? Telsa is ramping up and spending money doing it. And they're not done: https://www.teslarati.com/tesl... [teslarati.com]
Also that labour estimate is final labour costs of assembly. That doesn't include overhead, rework, factory downtime, production changes, all of which seem to be happening constantly as Tesla tries to ramp up. Looking at the numbers it would seem that they imply a steady state operation.
Re: (Score:2)
I don't understand. How can a $70 k car, even at 100% profit, cover the depreciation of a $5 billion factory?
Tesla does other things with those batteries apart from cars.
Re: (Score:2)
Giga is designed to produce packs for 500k TM3 vehicles per year (Fremont is capable of ~400k, with ~100k of those being S/X, Tesla hasn't disclosed where they want to complete the other ~200k TM3, whether via local expansion or elsewhere). The average TM3 sale price is expected to be $45k. Even with only a 25% margin, that's $5,6 billion gross profit per year. Now of course, that's not money you can just take to the bank - you have R&D, SG&A, etc to pay for. But it's a lot of money.
Awesome, duh, and buh, respectively (Score:4, Insightful)
there are only 4 kinds of screws used in the underbody (a typical German luxury manufacturer uses 40); many parts of the car are designed specifically so as to be easier for robots to grab; and the battery pack is harder to remove than on the S/X (e.g. not battery swap capable).
The screw thing is awesome. As someone who has done a transmission swap on an Audi A8, I can tell you that it is a truly massive improvement for the technician.
When robots install parts, they are always designed to be installed by robots. That's how it's done.
The battery pack on the Model S was installed partially with adhesives, so....
Re: (Score:2)
When robots install parts, they are always designed to be installed by robots. That's how it's done.
Designed to be installed by robots, and specifically designed to be easier for robots to install are not the same thing.
The battery pack on the Model S was installed partially with adhesives, so....
Not sure how that was relevant. The point here was the the Model S battery pack was designed and demonstrated to be swappable in 40 seconds. The Model 3 not so.
Re: (Score:2)
When robots install parts, they are always designed to be installed by robots. That's how it's done.
Designed to be installed by robots, and specifically designed to be easier for robots to install are not the same thing.
No, you don't get it. It's designed to be installed by robots, or it's installed by a human. There's no third way.
The battery pack on the Model S was installed partially with adhesives, so....
Not sure how that was relevant. The point here was the the Model S battery pack was designed and demonstrated to be swappable in 40 seconds.
Nonsense. It was installed with adhesives. You can't swap that out in 40 seconds. There is no evidence that Tesla ever battery-swapped a car outside a demo.
Re: (Score:2)
As someone who has done a transmission swap on an Audi A8
Please tell me you converted to manual. :)
Re: (Score:2)
As someone who has done a transmission swap on an Audi A8
Please tell me you converted to manual. :)
No, I'd love to, but I actually got a "parts car" for $300 that was nicer than my car, so I swapped my working engine/trans into it.
Re: (Score:3)
The screw thing sounds like more half assing and cutting corners by Tesla. You need to use different bolts because they all have different torque requirements. You can expect many model 3 falling a part because they didnâ(TM)t properly spec the bolts.
That's not how it works. You've got a required torque to do a job, and excessive torque which will damage parts. Anything in between will hold the part on. The variety of different fasteners used is due to different assembly techniques, suppliers, etc. If you design intelligently in the first place, you can reduce the number of different fasteners used. Most manufacturers don't do this because they have long-running programs with entrenched ways of doing business, but Tesla doesn't. They can declare these t
Re: (Score:2)
If you overspec the bolt it will work it self loose over time because it hasnâ(TM)t been spaced right.
Unless you take the simple expedient of including use-once thread locking compound, which is how the big boys do it. Such bolts are all over my A8, and what's more, they are labeled as being single use in spite of not being torque-to-yield. Of course, Audi also used TTY bolts on the undercarriage. Naturally, I reuse the bolts which came with preapplied thread locking compound, after cleaning them with a wire brush and refreshing it. Unfortunately, these bolts are coated with a passivating layer to avoid gal
But how much do they cost to buy? (Score:2)
All those numbers are pretty meaningless if you don't include the cost of the final product to consumers.
New German hybrid announced (Score:2)
The electric motors are as efficient as all hell, but the charging engine burns coal.
Re: (Score:2)
Don't worry, it burns Clean Coal whenever the environmental regulations inspectors stop by.
Sometimes. (Score:2)
In different places, different amounts of the power are generated from renewable or nuclear sources, right up to 100%. And as the grid becomes cleaner, so will all electric cars.
And even if the electricity supply is mostly coal, a large coal-fired plant is much, much cleaner than any car's internal combustion engine.
Re: (Score:2)
Certainly. But Germany's problem is that its industrial baseload is becoming dirtier, not cleaner, as it gives up modern carbon-free sources of power. As a bonus, the coal it is falling back on the dirtiest, lowest energy density kind. Their anthracite is all gone.
Re: (Score:2)
And even if the electricity supply is mostly coal, a large coal-fired plant is much, much cleaner than any car's internal combustion engine.
I don't think that's true.
https://www.citylab.com/enviro... [citylab.com]
Tesla possibly hoarding cars until July... (Score:4, Insightful)
This could be the reason why "Tesla is still in the process of moving from 3,500 to 5,000-6,000 per week by the end of this quarter" [youtu.be].
Basically they might be stockpiling cars until after the next quarter starts (July) so that the maximum number of buyers can take advantage of the full tax credit.
Re: (Score:2)
Indeed, both shorts and longs are expecting this. And Tesla making a huge push on Canada supports this notion (since Canadian deliveries don't count toward the US credit).
I'm sure congress knew when they designed the credit that manufacturers would try to "time" it to maximize their benefit.
Re: (Score:2)
Sure because making a loss and constantly re-tooling the factory is exactly what you do when your price insensitive customers get a few measly dollars tax credit.
They are spending a lot of money trying to increase the production speed only to not increase the production speed and continue burning money by your theory.
Re: (Score:2)
the model 3 does not sale for anywhere near 80k, it starts at 35k and the top end long range is like 44k, note the 44k module is the one that costs 28k to make, so about 16k profit.
Re: (Score:2)
No, you can't. (Score:3)
Nissan has Bolts on their books for $31K, but if you walk into a dealer anywhere but California, you'll be given reasons why you shouldn't consider a Bolt, followed by reasons why you cannot. In California, you'll be able to get one only if Nissan hasn't yet reached some legislated environmental quotas.
If those state laws are done away with, expect the Bolt to disappear entirely.
VW might produce enough EVs to fulfill their obligations under repatriations for dieselgate, but expect them to try to convinc
Re: (Score:2)
Re: (Score:1)
Do you mean Chevy and Bolt, or Nissan and Leaf?
Curious which company is tanking its EV sales.
Re: (Score:2)
Re: (Score:1)
How is Tesla tanking its sales? Now I'm even more confused.
Are you saying both Nissan and Chevy avoid selling EVs?
Re: (Score:2)
Re: (Score:1)
the model 3 does not sale for anywhere near 80k, it starts at 35k and the top end long range is like 44k, note the 44k module is the one that costs 28k to make, so about 16k profit.
You should check those numbers again. The Model 3 price goes up way higher than whatever you just pulled out of your ass. The top end is about $86k.
Yes.
https://www.theverge.com/2018/5/21/17376136/tesla-performance-model-3-specs-price-base-model
Re: (Score:2)
Nope. You can't buy a base ($35k) model at this time and probably not for another 12 months.
The minimum price for a Model 3 is currently $49k, plus $1k for delivery (everywhere in the USA, including Fremont). You can add about $10k to that price today, if you add all the Autopilot options and the larger (and less efficient) whee
Re: (Score:3)
There's nothing at all "misleading" about starting a production-limited vehicle with the more premium versions and working down from there as production ramps up. Lots of manufacturers do this. Jaguar won't be offering their base i-Pace for a year after the heavily-optioned-out version is on sale.
It's only where you're demand limited that you have to get the cheapest versions out immediately.
As for the demand for the base battery, according to polls on the Tesla forums, only about a third of reservation h
The free future of manufacturing components (Score:2, Interesting)
Consider the "component" cost of the tesla. Well that's to purchase the components. But of course those "components" were someone elses product and they in turn had component costs and production costs for that. And so on all the way down to "ore" and water before they are even taken out of the earth.
SO if you really stated the "component" cost versus production cost I think you might find it's almost all production cost.
Now imagine what happens when we make robots that can make custom manucaturing rob
Re: (Score:3)
Ha ha ha. Good one. Exactly one good flu epidemic will set humankind back potentially thousands of years. That epidemic may not be that bad, but a bad one will happen and there will be some serious societal adjustments that more than likely will preclude your rabbit hole.
Re: The free future of manufacturing components (Score:2)
The last good flu epidemic (1918) set us back no more than a decade.
The next will be about the same.
Re: (Score:2)
Suddenly EVERYTHING is cheap. Rediculously cheap
It will still be constrained by price of energy. Mining is a very energy intensive business, especially when you get to lower grade ores and are forced to use a clean process.
Re: The free future of manufacturing components (Score:2)
In his robot utopia the price of energy would for all intents and purposes be zero. All you need is robots pumping out solar panels and windmills, and other robots installing them.
Obviously you can't power mining machinery with solar directly, but once you have "free" energy it's also cheap to make hydrogen or methane or whatever else you want to use for that purpose.
Re: (Score:2)
In his robot utopia the price of energy would for all intents and purposes be zero. All you need is robots pumping out solar panels and windmills, and other robots installing them.
Obviously you can't power mining machinery with solar directly, but once you have "free" energy it's also cheap to make hydrogen or methane or whatever else you want to use for that purpose.
Try going back 500 years and describing our technological wonders to those living in the Middle Ages. Such as how one family can farm the land that 10,000 much larger families used to (leading to empty farm towns across America). How we have machines that allow a dozen men to build what would to them be a huge building. How almost anyone can fly above the clouds to go visit family half a world away. Or how we can instantly communicate with the other side of the world, not just with "letters", but audio and
Re: The free future of manufacturing components (Score:2)
Was there a question in there somewhere? I'm honestly baffled as to how your comment relates to mine ...
Re: (Score:2)
Re: (Score:2)
Redonkulously?
Re: (Score:3)
Re: (Score:1)
Oh please... Nothing is free. Even if we have self-replicating robots, the resources to build them are finite. So are the energy sources to run them. The pollution they produce will be non-zero. Their maintenance and design will require human or genius-bot input, and both of those will be limited in number.
Most importantly, the desires of humans for labor are essentially unlimited in scope. Give a family one robot, and they'll soon "need" 2, 3, and... how many computing devices do you have in your house aga
Re: Price of the battery (Score:2)
That would mean the battery in the Model S cost $80,000. Not likely.
Re: (Score:2)
No, more like $200/kWh.
Re: (Score:2)
a) Apple is one of the most valuable companies in the world.
b) Tesla's cars simply don't wear out like mechanical cars do.
https://www.greencarreports.co... [greencarreports.com]
c) Stick to your horse and cart if you want. Elon Musk won't lose any sleep over it.
Re: (Score:2)