Elon Musk Apparently Wants The Cybertruck As Precise As ‘LEGO And Soda Cans’

Ev Cybertruck
ADVERTISEMENT

The Tesla Cybertruck is sort of in production and a leaked email purportedly from Elon Musk shows how important it is that the edgy, bright metallic vehicle look perfect. The email gives some nice insight into Musk’s leadership strategy. Tesla, though, won’t be alone in the electric car revolution. Volvo previewed an electric MPV, which is a vehicle we’re sorely lacking. The EV revolution won’t be easy, however, as a new report from J.D. Power shows that tech adoption in EVs is going a little rougher than expected. Plus, VW figures out chips.

LEGOs, Soda Cans, And Tesla’s Cybertruck

Tesla Lego

Inside all of us are two Elons. There’s the Triumphant Elon who manages to clearly see the problem and look past the obstacles that seem to restrict other people and also manages to achieve a seemingly impossible solution. Then there’s the Misguided Elon who sees a solution so clearly that he’s blind to all the problems and obstacles he creates.

The Model S/Y and Supercharger network are all stunning examples of the Triumphant Elon who took on every major OEM, government regulators, dealership lobbyists, and reluctant consumers to deliver world-changing EVs. It was, and is, difficult to make electric cars, but it would be more difficult to imagine this happening without Musk’s steadfast commitment and admirable reluctance to settle. The same can be said for SpaceX which, unless you’re a homeowner or a marine bird in South Texas, has fundamentally altered the trajectory of space flight for all of humanity.

Then there’s the Misguided Elon of Twitter (thus far), of the Boring Company, and of the Model X Falcon Wing Doors. These are solutions to real problems, but I’m not sure any one of these has actually been successful. Twitter was a train wreck, but it’s leaning towards full-blown nuclear meltdown. The Boring Company is faltering in most places and, where it is in existence, is mostly a silly sideshow. While the Model X’s Falcon Wing Doors are neat, they were riddled with problems and Musk admitted himself that it was “hubris” for trying to do too much at once with the model.

I think the biggest challenge of writing about Elon Musk is that the same hubris that was necessary to take on NASA and Volkswagen can also clearly backfire, but it’s not always easy to know when he’s seeing the future. If you’re curious, go listen to the “Land of the Giants” podcast that our own Patrick George helped make for Vox and you can be your own judge of Musk’s motivations.

Where does the Cybertruck fit into all of this? Which Elon is at the helm? I think Musk is totally correct that people want electric trucks, and I respect the boldness of the product. One can imagine a way more conventional vehicle that looks more like a Model Y Ute. The Cybertruck is a big swing and I have trouble faulting anyone for a big swing.

Still, the vehicle has not been without its challenges. There’s an internal email, leaked to Cyber Truck Owners Club and picked up by Electrek that gets into one of the biggest challenges:

Due to the nature of Cybertruck, which is made of bright metal with mostly straight edges, any dimensional variation shows up like a sore thumb.​
​
All parts for this vehicle, whether internal or from suppliers, need to be designed and built to sub 10 micron accuracy.​
​
That means all part dimensions need to be to the third decimal place in millimeters and tolerances need be specified in single digit microns. If LEGO and soda cans, which are very low cost, can do this, so can we.​
​
Precision predicates perfectionism.​
​
Elon​

I can’t verify the email, but the fact that Electrek picked it up makes me think it’s likely real. Plus “precision predicates perfectionism” is an extremely weird and extremely Musk way to talk. While Tesla has ramped up production across the globe with impressive speed, I’ve personally seen numerous Tesla vehicles with sub-par build quality. This has improved a lot in the subsequent years, though I’ve got friends with a new Model Y who just had to take their vehicle in to fix an issue.

The Cybertruck, with its bright metal edges, cannot hide its own defects. If two panels are not perfectly aligned there are no complex curves to mask the issue. Musk, correctly, points out that soda cans and LEGO bricks are perfect. The difference is that soda cans and Lego bricks are small and produced in the hundreds of billions. Trying to apply that kind of detail to something that’s initially been hand made is quite the reach.

Still, people have made more money betting with Musk than against him, so it’s a different kind of hubris to assume he’ll allow himself to fail.

EV Owners Have More Problems With Tech Than Non-EV Owners

Here’s an interesting state from a soon-to-be-released J.D. Power study on new technology, via Automotive News:

Battery-electric vehicle owners reported 4.2 more problems per 100 vehicles than owners of internal combustion vehicles that — besides powertrain — contain the same technology, according to J.D. Power’s 2023 U.S. Tech Experience Index Study.

Huh. More from the article:

Interior gesture control proved problematic. EV owners reported 18.4 more problems per 100 vehicles, said Kathleen Rizk, senior director of user experience benchmarking and technology at J.D. Power. The function lets drivers use hand gestures to control vehicle features. For example, drivers can increase radio volume by turning their fingers clockwise or lower it by turning them counterclockwise.

Advanced technologies are more difficult for consumers to use, no matter the vehicle type, Rizk told Automotive News. EV owners, many of whom are first-timers, already grapple with a different ownership experience as they learn to charge and consider battery range.

This appears to be a perception issue as much as an actual technology issue, at least at first glance. EV owners seem to have expectations for how a vehicle should operate and also, likely, are bigger users of the technology that’s already there.

The Volvo EM90 Is Volvo’s First EV MPV

Volvo Mpv

The MPV is a vehicle that doesn’t get much love in the United States. Offering a mix of minivan flexibility with a more compact size, America never really caught on to the Kia Rondo or Mazda5. You could argue the new Kia Carnival is an MPV, as Kia calls it, but it’s a little more minivan-ish than the Rondo was.

Don’t tell that to Volvo, which is jumping right into the MPV space with the EM90. Here’s what the company is saying:

The Volvo EM90 is our first ever fully electric premium MPV, and is designed for you to make the most of the time spent in the car, like a Scandinavian living room on the move. The new EM90 doesn’t just allow you to travel from A to B – it creates room for your life.

The Volvo EM90 will make its global debut on 12 November 2023. Pre-orders will start for customers in China on the same date.

MPVs are more popular in China and Europe (if you’re in Europe get yourself a Dacia Jogger). It’s not clear yet if the EM90 is going to come to the United States, but I sure hope it does. MPVs are good.

VW Makes A Direct Deal To Fend Off Chip Issues

Volkswagen Id. 2all Concept Car

I’ve already covered how the chip shortage dramatically impacted new car sales, so I’m pleased to see a major automaker like Volkswagen planning for the future. Instead of buying chips through suppliers, the company is skipping the middle man and going straight to chipmakers.

Per Reuters:

The German carmaker, which previously relied on its component suppliers to purchase chips, began striking direct deals with chipmakers last October to ensure its supply was secure, according to Karsten Schnake, head of a Volkswagen-wide taskforce for component supply founded in 2022.

“Global market capacity is not sufficient. We must get active,” said Dirk Grosse-Loheide, purchasing chief for Volkswagen’s passenger car brand.

This seems smart.

The Big Question

Which Elon is making the Cybertruck? How many will there be on the road by the end of 2024?

Popular Stories

About the Author

View All My Posts

184 thoughts on “Elon Musk Apparently Wants The Cybertruck As Precise As ‘LEGO And Soda Cans’”

  1. Some reference for those not in engineering…

    “Standard Tolerances” are +/-127 microns (0.005″)
    “Tight Tolerances” are +/-50 microns (0.002″)
    “Tolerances which require further processing than the “standard machine” that made it: +/-13 microns (0.0005″)

    He wants better than that, on pieces of sheet metal many feet long. Simply not possible without increasing costs WAY beyond anything rational.

    To the engineers here, he wants to jig-bore each hole, and centerless grind any OD turned part. A BRAND NEW HAAS VF3 machining center can hit +/-13 microns, but probably not consistently. A used/aged HAAS VF3 is more like +/-25 microns, while a heavily used one is likely +/-50 or more.

    “Standard Tolerances” is what “any given machinists” could/should hit with a Bridgeport mill or Hardinge lathe. This is what I expect as an engineer, and design to. If I want better, it costs MORE, and/or parts are sorted (and bad ones scrapped), also bringing costs up MORE.

    Hope this helps.

    1. To add to this, stainless steel expands 10-15 microns per meter of length per degree C. Lets say you have a panel that is about a foot (or just under 1/3 of a meter) long. Heating it up by 5C (9F) will exceed this 10 micron tolerance. So… what happens as the shop making these panels operates throughout the year with varying temperature? What happens when a panel made on a cool day is measured on a hot day, or vice versa?

    3. Elon is calling for what me and my cohorts call ” CYA ” tolerances . Engineer says ” Plus or minus .015 on welded 18 Ga. 316L stainless sheet , tube etc on something that could easily be + / – .062. I hit .018 and guess who’s the bad guy ?

      1. Well, sort of, he employed a staff of highly skilled engineers and technicians and coached and managed them toward his overall vision, he had a technical mind, but he wasn’t an engineer himself. His inventing was more “we need to make an x”, then assigning that to a team as a project and micromanaging it along

      1. I have made that comparison a lot. See also Thomas Edison, huge prick to his people, unforgiving and unrelenting, brilliant, but not a pleasant person to work with.

    3. About the only commonality is money and pursuing their ideas whatever someone else might say. Elon is a stoner who thinks he’s brilliant for the ideas he has when he’s high and is too full of himself to reevaluate them when he’s sober and realize they’re stupid and nobody around him will tell him otherwise because he pays their salary (not that he would listen). Hughes had the balls to fly his own aircraft designs in an era where death from such things was commonplace while Muskrat doesn’t even have the balls to fight little Zoidberg in a refereed fight.

  3. Outside of the unforgiving topography of the body panels, my next concern with stainless was surface finish. Just forming those panels marks them up pretty well at the edges. I used to work in fabrication, and the enclosures we made out of stainless were a pain in the rear with all the extra processing. People always assumed it was just the material cost which made them so much more expensive than painted steel, but a decent chunk of that premium was due to the extra processing.

  5. I don’t want to hear about thermal expansion/contraction, I want sub 10 micron precision! I made a truck out of LEGOS when I was a child, so there’s no reason we can’t make a truck to LEGO precision!

    1. Cybertruck 2: ultra lightweight plastic frame wrapped in thin aluminum (or maybe thin aluminum frame with plastic on top, who knows). Don’t worry, it’ll have advanced AI that will make it avoid accidents.*

      *AI will show disengagement before any impacts to avoid liability. Please be sure you, the driver, don’t engage in dangerous driving activities such as driving on public roads or off-road.

  7. VOLVO EV VAN! VOLVO EV VAN! VOLVO EV VAN! VOLVO EV VAN!

    …anyway, this JD Power study as well as the current lousy sales of BEVs SHOULD tell manufacturers what we’ve been saying here in Autopia all along….no one wants these goddamn Apple Store tech hell world interiors, and the people who want full BEVs have already switched to them.

    The market will continue to say no thanks to the $60,000+ EV techno wizardry crossover nonsense. Find a way to make an EV in the 20s/30s that more or less feels and operates the exact same as a regular car and you’ll absolutely print money. Or make more PHEVs. They’re amazing and are way better for most people. I’ve been helping my dad figure out how his operates and it’s cool as hell.

    1. Hybrids suck because you have to maintain an ICE Engine, Emissions Control System, Automatic Transmission, and Drive Batteries. While an EV does have the Batteries, it has replaced the Automatic Transmission with a Gear Reduction Box, which should last the life of the car. The Prius is popular here in Portland, though most are being scrapped at eight to ten years of age because they suck to keep serviced…

      1. The ice part of this seems pretty well understood at this point. My phev has an e-cvt not a conventional transmission. 100kwh+ batteries are a terrible use of resources in the current environment when most people would be entirely served by 15-20kwh (plus Rex or supplemental gas power on occasion), which is still small enough to basically L1 charge in your driveway overnight for the many people who have driveways.

      2. Eh, in practice using beefier motor/generators as starter and alternator reduces common categories of failure and the transmission is generally way simplified compared to ICE. Add in the reduced stress on the ICE engine from the electric part taking up the slack and in practice they’re probably on par with ICE.

        Not sure where the “scrapped at 8-10 years” idea comes from, though. Even the 3rd-gen prius, considered the least reliable of the bunch, is notorious for starting to need a lot of maintenance at a mere 200,000 miles.

  8. Elon aiming for LEGO-like precision is great, but better-than-LEGO durability is needed as well. If the Cybertruck breaks as easily as today’s LEGO bricks, he’s in for a world of frustration.

  9. People keep saying Twitter is a train wreck, but is it? A train wreck implies something that should be doing X is doing Y instead. What is objectively WRONG with what Elon is doing with Twitter? Not subjective, but objective. It’s his, he can do as he pleases, he’s got money to burn and people to influence, we are the product, not the software. Thus, maybe this is exactly what Elon wanted to do, which makes it not a train wreck as the intention of the tool is not off the rails. The real train wreck is people still using things where they are the product.

    2. Objectively wrong, his announcement of turning off the block function would put the app at odds with both Apple and Google app store policies, which require apps relying on user interaction to have a block function.

      The choices he has made have lead to a significant decrease in ad revenue – some places reporting as high as 50%.

      Now, use of “objective” in this case is a way for you to deflect any criticism with a “well, that’s like, just your opinion, man” but it’s clear that his decision making is leading to problems for the company. The X rebrand – with the loss of brand capital with it – is a poorly thought out disaster. The lax moderation policy scares off many users and advertisers. The verified system decreases user engagement. And so on, there were problems at Twitter before, but the problems it faces now are the result of his leadership.

      1. All true but only if you are evaluating the success of X by traditional company measures (return on equity, subscriber growth, revenue, customer satisfaction, employee retention, etc.) X is now Elon’s plaything. He has enough money to wreck it so he is going to remake it in his image. I am not sure if he cares if it is successful in the conventional sense.

        His ideas for a communications platform are libertarian (small l). Absolute free speech. Lie, call people names, call for insurrection, say anything you want and let the users fight it out and decide for themselves. He might believe that an unmoderated platform will be a commercial success but I don’t think he cares. His definition of success is free speech, period. Money be damned.

        1. Redefining success for a company as “we don’t care if we make money as long as fascist bigots have a platform to spew hate” is an…interesting…take.

        3. “His definition of success is free speech, period. Money be damned.”
          If that’s the case why has he removed users who he doesn’t agree with? His definition of success is free speech as long as he agrees with it. Morals be damned.

        4. He doesn’t have unlimited money to piss away on the platform and he has investors /loan guarantors to please. Also nothing what he does is small or big L libertarianism. The speech is not free on the platform it’s entirely at his whim

      2. Those are all assumptions as to Elon’s objective with Twitter. We don’t know his objective. Only he does. That’s not moving the goal posts, that the reality of privately owned businesses. Elon is trolling Apple and Google as he has trolled others for the last 10-15 years.

        You keep saying those are problems for Twitter but Twitter doesn’t exist in the form it did when those were problems addresses by a publicly traded company. Privately run businesses existene and definition of success is up to them and the confines of the law not apple, google, you, or me. The fact you are so annoyed and condescending about it speaks volumes. I’m not saying you are wrong, just that you cannot be certain you are right.

  11. Surely accepting any tolerance at all (however ridiculous) excludes perfection.

    Perfection isn’t necessary, the trick is appropriate quality.

    I just stuck a +0.5/-0 tolerance on a hole because half a millimetre makes no functional difference to the assembly. That’s money we can spend on something that matters.

    1. Perfection in manufacturing is hitting the tolerances you said you were going to hit. One could call that appropriate quality, but, that’s a little different. If the appropriate quality is tighter than the tolerance, then, they’re manufacturing it perfectly to the (wrong) spec, but they don’t have the appropriate quality metrics, spec, for the design.

      1. “ Design for 6 Sigma” and/or “design for manufacturability” cannot be emphasized enough. All too often, I observe a lot of waste on the factory floor due to over-processing trying to achieve specs that the customer gives 2 shits about. Truly understanding CTQ features separates the experienced engineers from the newbs.

    2. I’ve done similar. I released a part that had a fairly tight flatness spec on one face. That was due to bolting a cast aluminum housing to it and the spec came from directly from that components manufacturer. I added a flange to the part to attach a completely unrelated thing. I just needed to angle that flange away from the AL housing enough to get a zip tie around it. 15° +/- 5° it was!

      Someone else revised that part later and changed the angle tolerance to 0.5°. Sure enough, I got a call from the supplier saying they can’t meet it. I said “Don’t, and no one will ever notice”. Sure enough, no one ever noticed!

  13. “Interior gesture control proved problematic.”

    I didn’t even know this was a thing. WTF. Why. WHY. Motion controls have NEVER worked that well. Why the fuck would you want people trying to do it while driving? Smaller screens and physical controls are so obviously superior in a car to anything else and yet every company is rushing to touchscreens and, apparently, an even worse control scheme than touch screens in these gesture controls.

    1. I think gesture controls were an attempt to mitigate the issues of putting more controls on a touchscreen, as they knew putting lots of things on a touchscreen was going to be a distraction.

      Touchscreens are mandated by directives from on high to reign in costs. I prefer good buttons, switches, toggles and knobs, but good ones aren’t cheap.

      When gesture controls proceeded to not pan out, yeah, criticizing them is totally fair. But if they actually worked very well, I think we might be less critical of trying to stuff as much as possible in a touchscreen.

      Meanwhile, you can pry my buttons and knobs from my cold, dead hands. Though I still prefer a touchscreen for navigation when the vehicle is stopped and I’ve not yet begun my journey.

    2. My mom has this in her X5, and it’s actually hilarious watching her use it. Moving her finger in a circle trying to get it to register when the volume knob to do the same thing is like 2 inches in front of her finger makes for some good comedy.

      My father and I make fun of her relentlessly for it, but she seems to like them.

      1. I have the gesture controls in a BMW. It’s really good at picking up intended gestures.

        The problem it does have is false positives. It’s occasionally triggered by shadows when driving though the mottled sunlight under a tree. The big failure is when the front seat passenger leans a knee on the center console and then rests their hand on the knee. The car picks up an unending stream of pause / unpause commands.

        So yeah I turned it off.

    3. It’s all too real. The thinking behind it is so dumb. The example of turning an imaginary radio dial versus a physical knob is among the best examples of this stupidity.

      1. Was just about to post this. I’m imagining some of my Italian relatives having a conversation while the console overloads and throws sparks like a Star Trek bridge computer.

  14. That means all part dimensions need to be to the third decimal place in millimeters and tolerances need be specified in single digit microns. If LEGO and soda cans, which are very low cost, can’t do this, neither can we.​

    Perfections a preposterous proposition.

    (Fixed it)

    1. I wouldn’t call Lego a low cost item. Lego bricks average 9 cents a piece, which seems awfully high considering the age of the tech and the volume involved. I think that’s what you pay for precision.

      1. Lego bricks used to be $0.10/piece 30 years ago. Inflation adjusted, they’re 53% cheaper than 30 years ago, and today’s LEGO plastic quality is actually superior, cleaner to manufacture, and of similar or higher tolerances.

        Long time LEGO fan. Absolutely remember what I paid as a kid, and am baffled that they actually managed to effectively reduce cost to buyers as they have.

        Haven’t bought any in at least 25 years though. 25-35 years later, they all still work perfect!

  16. “ any dimensional variation shows up like a sore thumb.”

    Whoa…so a variation from the 4th dimension would look like a battered Musk thumb? Can’t be his left one because he favors the right.

    Totally Far out, man!!!

    1. I mentioned this in the Discord, but the colloquial saying he tried to use is “[it] sticks out like a sore thumb”.

      His version doesn’t make any sense.

  18. EV Owners Have More Problems With Tech Than Non-EV Owners

    If we built EVs with the level of complexity to which cars were built in the late 1990s, but with modern technology for the batteries/powertrain/chassis/ect, the damned things would last forever and be repairable with basic tools. Why can’t we have this? Why does everything have to be tech-laden and locked behind proprietary software? Give me an EV that is built with simplicity in mind, with no touchscreens, and actual buttons in place for all the functions, a durable LiFePO4 traction battery that will last for decades and mid 6-figure mileage or more without hassle, and nothing but a basic-bitch OBD-II reader needed to diagnose any issues that crop up(most of which would inevitably be from the 12V electronics).
    EVs are not rocket science, yet most of the industry can’t get them right, and the depreciation curve of many of the offerings tells a story all its own.

    1. This. It’s as if auto makers just can’t get out of their own way and have an uncontrollable impulse to make vehicles worse than possible.

      1. It’s the marketing guys, they insist of adding features that impress potential buyers in the showroom.

        90’s features with ultra reliability doesn’t wow customers as much as colour changing charge lights.

        3. I think a big part of the problem is focus groups. Most people have better things to do than attend focus groups so all that data comes from a particular subset of people who likely won’t even have the funds to buy new cars.

    2. locked behind proprietary software?

      Because this ensures future service transactions that the OEM’s get money from. It’s the Apple model and a complete cash grab.

    3. For that, I think we’re largely going to be stuck with fitting modern BEV hardware to old cars via the aftermarket, but not only are old cars already expensive enough, but they have decades of potential damage and corrosion underneath them as well as having to contend with the greater weight (unless battery tech improves). As for the OEMs, if they make a soulless modern car that lasts forever, what’s the incentive to buy a new one? If they’re all ugly and uninspired, might as well keep driving the old one, but that doesn’t fit with the constant growth model the board needs to see and the dealers demanding work for their service departments.

      1. There’s no reason it has to be soulless. To improve the range per dollar ratio, it would by necessity be lightweight and aerodynamically streamlined. And it would be dirt cheap to give it more horsepower. A sub-3,000 lb streamliner sedan with a 25-30 kWh battery and 300 horsepower drive system shoved in it would be one way to make an affordable sub-$25k EV with 200 miles range, and such a thing designed correctly would be a lot of fun to drive. I think if such a thing was marketed, there would be a demand for this, but then it would cannibalize the sales of higher-margined products and dealerships would be highly displeased.

        As for that constant growth model, it is not at all sustainable from a financial or environmental standpoint, and is eventually going to reap a banquet of negative consequences to deal with from which society may not ever recover from.

        1. Yeah, you say that every day and I’ve designed things similar to that myself that I would love to build . . . for myself because nobody else wants that. They aren’t interested in how good something is to drive because they’re too stressed to live in the present and have never been exposed such a thing to known what they’re missing in the first place. Even if they did, they wouldn’t compromise on what they want—range (which, in lieu of intelligent design—the human kind, not the BS religious one that’s readily disputed by a basic understanding of anatomy and meeting more than 3 people—means weight, the blunter of responsiveness and feel), they want comfort (even if these overweight tall vehicles tend to ride like ass and they put seats that turn to stone after a few hours in them), they want flexible space for people and cargo (even though too much of that paper space is useless height instead of internal footprint and giant consoles and bunker doors intrude on passenger width), they want to stand out without standing out too much, they want the comfort of the familiar, at least the idea of safety and resale value—they don’t even want colors or are at least more than ready to settle for not having them. They don’t appreciate how special it is to be a trichromat mammal! Maybe you can sell a few expensive oddities to affluent weirdos who have multiple other vehicles for when they don’t want to wear that particular technicolor dreamcoat out on the town, but that isn’t even a broad market solution for the upper end. It’s also why we keep building stupid houses out of junk and I’m sure almost any other industry is plagued by the same lack of imagination that frustrates innovators and thinkers. It’s cheaper and “better for the economy” to keep people spending on overpriced, substandard stuff and as long as they’re too dumb, distracted, or complacent to care, that’s all we’ll get unless we’re wealthy enough to afford to go our own way with bespoke stuff or have the ability to DIY, like you’re doing. I went to school for Industrial Design, specifically for transportation. Now I’m in telecommunications largely because I got disillusioned and realized that I don’t want to design mainstream products to marketing-specs for boring people.

          The current crop of bland cars in the present form factor could certainly be improved in terms of serviceability and reliability, but until the market demands those things, they’re going to keep building junk that keeps them coming back for repairs and replacement.

        1. So, car companies will never make what you say you want, because you won’t buy it. They’ll go after the market of people who *will* buy their cars, and they apparently want fuller featured units. The profit margin of “maybe” is why the stick shift is dying out.

          1. If a car was made that appealed to me and was within my price range, I certainly would buy it. Had I had the money I do today back when the Lotus Elise first came to the U.S., I’d have snatched one up. That car ticked most of my boxes. Same with the 1st gen Tesla Roadster, issues it had notwithstanding. A Honda Insight 1st gen with a manual also kind of came close. The GM EV1 did as well, but they refused to sell them. These days, a used Lotus Elise or Alfa Romeo 4C is on my short list once I get some land purchased to live on, but I’m not going to pay “bring-a-trailer” prices for them either.

            1. You might want to see if you can get into some focus groups for manufactures that might get you what you want. Marketing pays some level of attention to those because they have to pay you. Weird, but true.

    5. If we built EVs with the level of complexity to which cars were built in the late 1990s, but with modern technology for the batteries/powertrain/chassis/ect, the damned things would last forever and be repairable with basic tools. Why can’t we have this?

      What shareholder-owned, durable goods company wants to produce durable goods? We’ll buy whatever they produce that we can afford, so where is the incentive to make something that isn’t broken or obsolete in five years, and in need of replacement with the new generation? So with phones, now with cars.

    6. Because actually getting from A to B is incidental for most folks anymore. They may as well start putting badges on the fender that advertise the #/size of the screens the vehicle wraps you in to distract you from its true function as a transportation device. The funny thing is that at the end of the day, all of the goofy crap/functionality they try to develop is not as good as your phone and really pointless if people are being honest with themselves. Like yourself, I wish carmakers would just give up on that stuff and stay in their wheelhouse.

    7. Imagine if we could build decent cars like we used to…
      My 1984 Audi 4000q, 1978 Ford E150 Cargo Van were examples of a well made product.
      And now, not only is it impossible to buy a new Quattro or Full Sized Van with a Manual Transmission in the US, it is also impossible to buy a new car that was made so it could be serviced…

  19. Don’t worry Elon! I’m sure the Cybertruck, if it’s ever produced, will be as attractive as Legos and as durable as a soda can. Even IF Tesla was able to produce parts designed to these tolerances, the first time a door or hood was opened and shut would expand those tolerances by 10x, easily, if not 100x. The real world is not likely to conform to Elon’s fantasies.

    The Volvo EM90 is an exciting idea that will work in the real world, however. I’ll definitely be looking at the electric Volvos when my Polestar 2 lease is up.

    1. I hope you don’t live in a climate where the temperature changes ever, or park outside in the sun!

      This will be perfect for Vancouver where it’s always 20°C and cloudy

  23. I truly believe that success from Tesla/spaceX are successful in spite of Elon, not because of him, also, it’s hilarious that he thinks Lego is that accurate while also being cheap and easy to produce.

    1. I did read somewhere that people at SpaceX have a program that they turn on whenever ol’ Musky comes into the office so it looks like they’re doing what he thinks that doing space stuff looks like.

    2. Word on the streets is that SpaceX engineers, being rocket scientist brainiacs, are quite skilled at diverting Musk’s attention towards trivial things and keeping him out of the meaningful areas/problems/etc. whenever he tries to role play as an engineer.

    3. Uhhhh, Musk is a lunatic and I do agree some of his companies have succeeded in spite of him; but he is also right about Lego. Lego are estimated to be that accurate. While Lego don’t publish their numbers, it’s generally guesstimated that Lego probably has something like 50% the tolerance window of similar precision injection molded parts. That would put them in the neighborhood of something like +/-0.02mm or less even. Considering other cheap injection molded toys probably have a tolerance window 5x-10x bigger than that, I’ll give Elon the nod here. A broken clock is right 2x a day after all.

      1. Yeah. Hi. I actually measured a shitload of Lego with micron accurate calibrated equipment, because at one point, I figured “that can’t be right for ABS.”

        That’s because it’s not. Legos are small parts, where tolerances are easier to hit, which my sample set came in about +0.1mm -0.25mm, but I don’t remember wall thickness. It was out there. Because guess what? The secret’s in the plastic and the mold. All the excess flash is in the ‘void’ area, and when the brick is undersized, the plastic stretches.

        Hitting 0.001 on giant stainless steel panels that are going through press forming? No. Just fucking no. Period. You don’t get to argue it, because it’s a statement of metallurgical fact and simple physics.

        1. a mantra from 6sigma is “designers don’t define tolerances, processes do”…
          and processes drive cost.
          unnecessarily tight tolerancing is fiscally irresponsible and probably unattainable on large metal stampings like body panels.

            1. Yeah; to be clear, +0.1-0.25 is ridiculously tight for an injection molded part like that. The molds themselves can go way past that just from injecting molten plastic. So don’t take it as me pooh-poohing on Lego as much as providing a reality check on the hard numbers.

              The way they’ve designed their molds is legitimate genius, because they’ve effectively compensated for natural variances and inherent problems. Flip some Lego bricks over, and check the void sometime. That’s where you’ll find all the ‘slop’ where tolerances are basically “does it work or not?”

        2. I’m not saying that level of tolerance is achievable by Tesla because I know it’s not for large parts. I’m just saying that that the 10 micron window Elon referred to is actually within the the ballpark of Lego. Elon’s reference to Lego’s tolerancing was correct; his comparison is wrong because very small injection molded plastic is nothing like large stamped/bent metal, and worse yet his expectation is wildly impossible.

        3. I assumed the Lego tolerances are in the molds, because they probably go through a lot of them and they can’t be all wonky relative to other molds.

          The molds are probably made with allowances for the behavior of the ABS, so that even if the ABS comes out of the mold a tiny bit wonky, the tight tolerances of the molds allow the blocks to stay in spec.

      2. I wasn’t saying Lego doesn’t achieve an extremely high tolerance at all.

        I was saying it’s hilarious that he thinks it’s done cheaply and easily and could just be translated into sheet metal as easy as saying the words. Lego isn’t cheap at all.

  24. All parts for this vehicle, whether internal or from suppliers, need to be designed and built to sub 10 micron accuracy

    If this is really the standard for a production vehicle, we will literally never see one. 10 microns is 0.0004″.

    Just for context, when I worked at automotive suppliers, many of our fabricated parts were toleranced to +/- 0.015″ or even 0.030″ on most dimensions without any issues. The proposed Cybertruck part tolerance is literally 1/100th the size.

    Not only is this type of tolerance virtually impossible for many standard manufacturing processes, it’s not even remotely necessary for many parts to work correctly. Something as large and with as many components as a truck, with *every* part built to +/-0.0002″ would be so expensive that only Elon Musk could afford one.

    1. I’ve designed a few parts with +/- .0002″ tolerances. The tolerance applied to a few diameters only, the rest of the part was +/- .010″ mostly.

      Average price: $75 000 each.

      And forget sheetmetal at those tolerances.

      1. “Precision costs” is a truism. I have to do micron-level calibration of laser source and receiver often enough that I had to buy an alignment cube.

        That 10mm piece of fused silica cost as much as a decent condition used car.

    2. In fairness even a tiny relaxing in those metrics yields huge reductions in net cost. +/- .0008 for example is still hard, expensive, but quite doable for many parts. Here though… I think tolerancing like this is completely the wrong attitude. You don’t need tolerances that tight, you mainly want low variance on the tolerancing surface. A single part being consistently + 0.001 is much less noticeable and damaging than a part that is +0.0005 and -0.0004 on the same surface.

      1. I spent a day in a controlled temperature room with a micrometer, a sheet of think stock, and a VW feeler gauge. I was tasked with checking the dimensional consistency of the thin sheet stock over the entire sheet looking for variations. I had to check the micrometer every few minutes to determine whether the heat of my hand changed the micrometer enough to be not accurate. I took my feeler gauge to the metrology lab to calibrate it for this use. Turns out that a basic feeler gauge set is actually very accurate.

    4. I honestly had no idea what 10 microns is before your post, but I just had a feeling that Musk’s mouth was spewing more bullsh*t when I read it. Thanks for confirming my suspicions.

    5. Echoing everyone else in this thread. It’s not impossible for some parts, but yeah good luck getting a car built to those tolerances. I’ve dealt with parts that can be held to those tolerances but you pay a lot to specialty shops for that, and these are small parts (6” in length or less).

      1. Yeah it’s typical for high precision aircraft-grade bearings and the like, but even that stuff is rare in automotive.

        And it’s laughable for bigger stuff or non-critical parts.

    6. Holding these micron-level tolerances to surface profiles (especially sheet metal) is, effectively, a no-win.

      Suppliers won’t be capable, and the OEM won’t pay for it. What usually results in that scenario is that the quality is compromised to a level that the OEM would pay.

    7. Guys, come on. He only said “microns” because he heard that once and thought it sounded cool. He honestly has no idea how big a micron is, other than “super-duper small”.

    8. I can only assume that if Musk notices a picture on the wall is a slightly askew, his fix is to plane the frame down to perfect flatness with a tolerance of 0.0004 millimeter.

Leave a Reply