Slashdot Mirror

← Back to Stories (view on slashdot.org)

Tesla Relied On Too Many Robots To Build the Model 3, Elon Musk Says (theverge.com)

Posted by BeauHD on from the helping-hands dept.

An anonymous reader quotes a report from The Verge: Elon Musk says Tesla relied on too many robots to build the Model 3, which is partly to blame for the delays in manufacturing the crucial mass-market electric car. In an interview with CBS Good Morning, Musk agreed with Tesla's critics that there was over-reliance on automation and too few human assembly line workers building the Model 3. Earlier this month, Tesla announced that it had officially missed its goal of making 2,500 Model 3 vehicles a week by the end of the first financial quarter of this year. It will start the second quarter making just 2,000 Model 3s per week, but the company says it still believes it can get to a rate of 5,000 Model 3s per week at the midway point of 2018. Previously, Tesla has blamed bottlenecks in the production of the Model 3's batteries at the company's Gigafactory for the delays. But in a wide-ranging (and largely positive) interview with CBS's Gayle King, Musk also admits it was Tesla's over-reliance on robots in the production. Musk then said the company needs more people working in the factory and that automation slowed the Model 3 production process. He alluded to a "crazy, complex network of conveyor belts" the company had previously used and said the company eliminated it after it became clear it wasn't working.

28 of 103 comments (clear)

  1. Crazy, complex network of conveyor belts by DontBeAMoran · · Score: 2

    Everyone knows you need to use tubes, not conveyor belts.

    --
    #DeleteFacebook
    1. Re:Crazy, complex network of conveyor belts by dos1 · · Score: 2

      Welcome to the Tesla Motors computer-aided Enrichment Center.

    2. Re:Crazy, complex network of conveyor belts by torkus · · Score: 2

      Go to B&H in NYC and they'll prove you wrong.

      I get the reference, but their system is actually pretty damn impressive and worth a mention:)

      --
      You can get rich if you own a politician, but you have to be rich to buy one in the first place.
  2. What? by Luthair · · Score: 3, Insightful

    Has anyone ever said that? Everyone I've seen points out that the issue is Tesla has no experience building at scale and has had issues with QA / consistency on their existing lines.

    1. Re:What? by DontBeAMoran · · Score: 5, Interesting

      Exactly. They have no experience so they tried to use too many robots.
      And industrial robots currently suck at doing their own QA.

      --
      #DeleteFacebook
    2. Re:What? by bhcompy · · Score: 5, Informative

      And if we want to look at automaking efficiency with robotic integration, it's not like Toyota hasn't designed their production lines for maximum efficiency already.. oh, wait, that's what they specialize in and what they've built their reputation on. They've recently removed some robots from their production line because people did the jobs better and faster, with less waste.

    3. Re:What? by JourneymanMereel · · Score: 4, Interesting

      If only there was a state out there that was full of people that knew how to make cars. Perhaps if such a place existed, it may have been a better place for Tesla to HQ than silicon valley was.

      --
      Life has many choices. Eternity has two. What's yours?
    4. Re:What? by Mr+D+from+63 · · Score: 4, Insightful

      So, Musk's plan to 'revolutionize' the auto industry with his advanced production techniques isn't going as planned. Maybe other car manufactures actually did learn something through all those years of production.

    5. Re:What? by Anonymous Coward · · Score: 3, Funny

      Musk’s ego is writing checks his companies can’t cash.

    6. Re:What? by Anonymous Coward · · Score: 5, Funny

      Hey! What the fuck are you doing using facts and correlating data from one of the biggest auto makers on the planet? This is a story about Tesla where facts and data have no place - only raw emotion and horseshit count.

      Quick! Someone make a comparison to Ford, because Ford totally operates hundreds of thousands of solar panel installations across the US, and so Ford is clearly a 1:1 comparison!

    7. Re:What? by 140Mandak262Jamuna · · Score: 2
      First american car maker outside the big three to make more than 100K cars a year since Studebaker folded ubn 1970s.

      Every one who followed Detroit ended up bankrupt. Including several divisions of GM, Ford and Chrysler.

      --
      sed -e 's/Chuck Norris/Rajnikant/g' joke > fact
    8. Re:What? by steveha · · Score: 4, Interesting

      Tesla bought its factory super cheap: $42 million. That price is so cheap that essentially Toyota was investing in Tesla. That factory is in California, not Michigan.

      https://www.sfgate.com/business/article/Tesla-paid-only-42-million-for-Nummi-plant-3187254.php

      Tesla's big cars have been extremely successful (they simply took market share away from other luxury car makers). The Model 3 is selling as many as they can make, and they are selling only the most expensive options for it right now. A year from now I expect there will be almost 200 thousand Model 3 cars out on the road. So overall I'd say that things are working out pretty well for Tesla.

      --
      lf(1): it's like ls(1) but sorts filenames by extension, tersely
    9. Re:What? by yodleboy · · Score: 3, Insightful

      "A year from now I expect"
       
      Tesla is always a year away from hitting their stride with production. Kinda like how true AI is only 10 years away. Practical fusion is only 25 years around the corner you know?
       
      Tesla doesn't scare any auto maker. 100K total sales in luxury cars? Mercedes, BMW and Lexus EACH sold over 300K cars in 2017 alone. At the lower end, the established makers sell hundreds of thousands of entry level luxury cars every month. Tesla is like a Ferrari for the geek crowd. I guess it's a nice way to feel like they are part of a revolution.

    10. Re:What? by dj245 · · Score: 4, Interesting

      Depends on what QA needs to be done. Surface finish, dimensional inspections, color, electrical checks, etc are all relatively easy to have automated QA. The problem is that you may not know what parts of your process will have issues, and many of these checks must be custom-setup for each model. Setting up these processes and ironing out all the issues takes a lot of time. Much more time than Tesla had budgeted apparently.

      Not only that, but what do you do with all the products that come off the line with a defect while you are in the tuning process? In most cases, these can't be fixed by running them through the line again, so you need time and people to fix them.

      --
      Even those who arrange and design shrubberies are under considerable economic stress at this period in history.
    11. Re:What? by Joce640k · · Score: 2, Interesting

      On the flip side, Musk appears to be able to admit when he makes mistakes (and change direction accordingly).

      When was the last time you saw that?

      --
      No sig today...
  3. Oh,sure! by jddj · · Score: 4, Funny

    That's it! Blame the robots! Always blame the robots first! What has a robot ever done to you, Elon?

    1. Re:Oh,sure! by torkus · · Score: 4, Funny

      Built his car so poorly he had to send it to space to get rid of it?

      --
      You can get rich if you own a politician, but you have to be rich to buy one in the first place.
    2. Re:Oh,sure! by Mr+D+from+63 · · Score: 5, Funny

      That's it! Blame the robots! Always blame the robots first! What has a robot ever done to you, Elon?

      Its not so much the robots, its the fact that you have to watch them constantly and be ready to take over.

    3. Re:Oh,sure! by CaseyB · · Score: 4, Funny

      That one was mostly hand built in England. Which makes the poor quality entirely understandable.

  4. Automation is a must! by EETech1 · · Score: 4, Interesting

    If he really wants to build 5000 cars a month, there will HAVE to be some crazy network of conveyors, they'll be making about 250 cars a day.

    You don't push those around on carts by hand!

    Everything must come together in a continuous process, an uninterrupted flow of parts in, cars out.

    They're not there yet, and that's why they can't make it work. Running at 10 percent capacity with parts missing (or incorrect) everywhere along the line. Piles of sub-assemblies sitting everywhere that can't be completed. Ok this part is in, these 20 pieces can be completed up to the next screwed up part of the process.

    I've seen that happen. It sucks!

  5. Not the robot's fault... by cmdr_klarg · · Score: 2, Interesting

    The robots do EXACTLY what they are programmed to do.

    Either their process hasn't been ironed out completely, or there was an incompetent automation setup. Possibly the setup was done too quickly.

    --
    THE SOFTWARE, IT NO WORKY!!!
  6. wrong people by AlanBDee · · Score: 5, Funny

    He alluded to a "crazy, complex network of conveyor belts" the company had previously used and said the company eliminated it after it became clear it wasn't working.

    He didn't hire the right people. He should track down the top https://www.factorio.com/ players and have them design the conveyor belt system. We do that shit for fun.

  7. 5000 a week by rmdingler · · Score: 4, Interesting

    Henry Ford had those production numbers one hundred years ago.

    I would recommend the tried and true industrial production method: Locate and hire a guy who's been a keystone at a competitor's assembly line.

    --
    Happiness in intelligent people is the rarest thing I know.

    Ernest Hemingway

  8. Summary by raftpeople · · Score: 3, Insightful

    Automating complex systems is difficult.

  9. Not Detroit! by Anonymous Coward · · Score: 2

    First, Detroit has a reputation of being old school. Pretty hard to build excitement and hype being based in Detroit!

    Now in SILICON VALLEY Valley valley valley (hear the dramatic echo), that's where innovation occurs - and they have a monopoly on disruptive innovation and genius.

    And Elon being a disruptive innovative genius had place his disruptive innovative company where it can be noticed and appreciated.

    But don't worry, when Tesla goes belly-up, those same stodgy boring old-school automakers will buy up Tesla's assets for pennies on the dollar and actually make something of it - it IS a pretty damn good idea - it's just the implementation is being screwed up.

    All at the expense of current Tesla shareholders and bond holders.

  10. no money no car by AndyKron · · Score: 3, Insightful

    A story about a car I'll never be able to afford. Or should I say I can afford it, but I'm not paying that kind of money for a car.

  11. Re:Learning from history by Actually,+I+do+RTFA · · Score: 2

    In fairness, between 1985 and 1995 we also learned that streaming video over the internet was impossible. It's completely fair to claim that robot technology has improved in the past 25-30 years in a way that obviates those lessons. Now, hiring some people who participated back then and had institutional knowledge of failure points would probably be start.

    --
    Your ad here. Ask me how!
  12. Re:This is why I read slashdot. To point and laugh by yodleboy · · Score: 2

    X Nobody will ever make a fast electric car.
    There were fast electric vehicles long before Tesla. No one doubted it could be done, it was just a matter of the tech becoming cheap enough to use in a consumer product.

    X Well, Tesla will never make an electric sedan that people actually want
    There are ALWAYS going to be buyers for unique $100k cars. No company enters that level of the market without being pretty sure they will sell

    X It takes 12 hours to charge an electric car
    When people were saying that, it did take a long time to charge an EV, maybe not 12hrs, but not fast.

    X They'll never sell more than a few thousand of them
    Fair enough...

    X They'll never sell 100K cars
    Fair enough...

    X They'll never make a $35K electric car with >200 miles of range
    Making them? Sure... Delivering them? Well that remains to be seen. If they can't get production under control and start filling the backlog, they may find people just walking away. Especially if some other company surprises the market with a competitive EV before they can get sorted.

    X They'll sell maybe 50K of them
    All those pre-orders don't count until a car is actually made and delivered. It looks like a Kickstarter campaign at this point.

    X SpaceX will never be competitive
    X SpaceX will never reliably land boosters
    X Well, they haven't reused them yet

    These aren't cars. Considering that the established players hadn't come remotely close to landing and reuse and had plenty of studies to show how hard it would be, I don't think some skepticism about a private company pulling it off was unwarranted. It was a hell of an impressive achievement.