By on March 20, 2017

Tesla Model 3 Unveil, Image: Tesla Motors

After deciding that its analytic tools are top-notch and will tell it everything it needs to know, Tesla is skipping the “beta” phase of the Model 3’s development cycle.

In a conference call to investors last week, Tesla CEO Elon Musk said the automaker doesn’t need to bother with that crop of pre-production models. Early release model 3s built on the company’s production assembly line — not specialty one-offs built in a corner of the shop — should be rolling by this week or next, he said.

Last month, Tesla halted production at its Fremont, California assembly plant to prepare for the Model 3. Starting February 20th, the plant went dark to upgrade its paint shop. There’s the tightest of timelines in play for the automaker’s upcoming “affordable” electric sedan, with deliveries scheduled to begin in the fourth quarter of this year.

By skipping this step, Tesla boosts its chances of meeting the production target date. Another benefit is money saved. There’s also a chance, or so Musk says, that consumers can expect the final product to be much more glitch-free than the problem-plagued early Model S and X. Thank the enhanced computerized scrutiny for that.

Of course, there’s also the possibility that a last-minute realization of a problem with those nearly-ready models could leave the company scrambling. With well over 300,000 reservations to fill from the get go, Tesla no doubt hopes for the rosier scenario.

[Source: Gas2] [Image: Tesla]

Get the latest TTAC e-Newsletter!

Recommended

65 Comments on “Tesla Skips a Step, Goes Straight to ‘Early Release’ Model 3s...”


  • avatar
    FreedMike

    Well, nothing could ever go wrong by putting a car right into production and letting the customers beta-test them. Worked splendidly for Chrysler. Remember the Dodge Aspen?

    • 0 avatar
      Erikstrawn

      Whenever I hear of “skipping beta testing” I think of how buggy Windows 98 was and then having to pay again for Windows 98SE, which still crashed once a week.

      Tread carefully, Tesla.

      • 0 avatar
        mcs

        @Erikstrawn

        Except they aren’t skipping the beta testing, just skipping building the cars with beta tooling. They’re using the production tooling – which is a good move.

        • 0 avatar

          It’s still an issue you now have expensive tooling. If something goes wrong your either stuck with the problem or your stuck with soaring costs and delays. This leads to basically encouraging problems going out the door as they are to expensive to fix.

        • 0 avatar
          JimZ

          no it isn’t. hard tooling is expensive and time consuming to change. There’s a reason the rest of the industry does things a certain way, and it’s not because they love spending hundreds of millions of dollars doing it. It only looks like a “good move” to someone who doesn’t know how cars are designed and built.

          • 0 avatar
            stuki

            Without knowing details about Tesla production internals, things aren’t necessarily that obvious. Compared to most cars, the complexities of electric ones are to a larger extent in the components. SO there’s less to go wrong at final assembly to begin with.

            Then, even model 3 volumes are smaller than those of most cars, and in a segment that changes quicker. So, “production tooling” is less optimized. Hence less rigid. Hence less of a leap compared to “beta tooling.”

            But I’m sure another reason is just a higher level of risk tolerance at Tesla than at most legacy makes. Just not the whole reason.

  • avatar
    silentsod

    Tesla’s phenomenal track record with robust releases and few glitches means the 3 should really catch fire.

    • 0 avatar
      SCE to AUX

      That would be shocking.

    • 0 avatar
      islander800

      Yes, Elon does a fantastic job burning through investor and PUBLIC funds. I understand that he’s paid back what generous governments at many levels had loaned him, but equity investor funds are being consumed at a voracious rate. Tesla has never shown a profit, so everything to date is to be considered “investment in the future” of a promised money-maker.

      I give Elon credit. He’s as good a salesman as Preston Tucker, and as long as the funding keeps flowing in, he just may pull this off.

      • 0 avatar
        Rick Astley

        Elon Musk can do absolutely fantastic things as long as he’s burning through somebody ELSE’S cash like a well-oiled cash burning machine.

        Something that has often bothered me about him is his ability to spend other peoples money, have them shoulder all of the risk, and yet he retains the profits.

        Truly a personality type that would do well in the government sector.

      • 0 avatar
        hgrunt

        Preston Tucker sold 51 cars bearing his name, Tesla has sold 75k Model S in 3 years, a bit of difference there.

        Amazon also didn’t show a profit for a very long time because they were investing for the future, and shareholders are finally being rewarded for it.

    • 0 avatar
      SuperCarEnthusiast

      Yes, but they can handle it now! All setup and ready to blast off!

  • avatar
    SCE to AUX

    I’ll take mine in blue, glass roof, more battery, RWD, std motor.

  • avatar

    In other words, the first cars off the line won’t be sold to the general public, but instead will be used as validation cars.

    When car companies set up pilot operations to build validation samples of a new product, like GM has in Pontiac, they try very hard to replicate assembly plant equipment and conditions. It’s not the same as Metalcrafters building a one-off concept car.

    • 0 avatar
      orenwolf

      Indeed – I presume new tech helps there too, hence his comment that even without one-offs, the validation cars will likely have less issues as a result.

      You’d have to imagine analytics and robotics would get us to a point soon enough where “test cars” *not* on the assembly line would actually introduce more defects (due to differences in process) than they find.

      • 0 avatar
        indi500fan

        Another thing to consider is how many of their component suppliers have successfully mastered the super short development approach. That will be a big part of the success / failure of this rollout, both for quality and schedule.

        • 0 avatar
          JimZ

          “Another thing to consider is how many of their component suppliers have successfully mastered the super short development approach.”

          you have zero knowledge that this is in any way true.

      • 0 avatar
        JimZ

        “I presume new tech helps there too, hence his comment that even without one-offs, the validation cars will likely have less issues as a result”

        indeed, very presumptuous on your part.

        “You’d have to imagine analytics and robotics would get us to a point soon enough where “test cars” *not* on the assembly line would actually introduce more defects (due to differences in process) than they find.”

        and what knowledge/experience do you base this presumption on? Elon’s tweets don’t count.

        • 0 avatar
          orenwolf

          I base it on the simple fact that cars, thanks to robotics today, are coming off the line with far less defects than they did when they were put together entirely by humans. I seriously doubt that’s in dispute.

          • 0 avatar
            28-Cars-Later

            “far less defects than they did when they were put together entirely by humans”

            Defect 1: Cannot see out of one.
            Defect 2: Cannot see ground.
            Defect 3: Tires now cost $400 apiece whereas my old car’s were $70/wheel.
            Defect 4: Fingerprints all over the dash.
            Defect 5: V6 fuel economy without V6 power.

    • 0 avatar
      SCE to AUX

      They also need to produce a bunch (in my opinion) for test-drive purposes at their stores. I won’t be buying mine unless I can drive one first.

    • 0 avatar
      indi500fan

      I’m not at all familiar with how they’ve configured the old Fremont plant. Are they going to build the S, X, and 3 on the same line, or is the 3 getting an additional assy line?

  • avatar
    CarnotCycle

    DoD industrial complex tried similar approaches for F-35, LCS, and CVN78.

    Doh.

    • 0 avatar
      BigOldChryslers

      Worked okay for the Boeing 777 though.

      • 0 avatar
        dal20402

        777 line number 1 continues to fly for Cathay Pacific in Asian regional service today, almost 23 years after first flight.

        Boeing executives must cry a little tear every time they see it, after having to make successive decisions to never sell first 787 line 1, then 787 lines 1-3, and then finally (last year) 787 lines 1-5. The first three went to museums, the fifth is partly disassembled and will be scrapped, and Boeing is keeping the fourth around for engine testing.

      • 0 avatar
        CarnotCycle

        Didn’t work for Boeing 787. Especially so regarding a certain kind of battery employed on the aircraft.

        I actually don’t think Model 3’s are going to be flaming down the road or burning up houses charging per se; but rushed stuff is always buggy.

      • 0 avatar
        TrailerTrash

        Not sure this is true.

        The FHA requires tons of testing on commercial AIRCRAFT BEFORE ANY CAN BE USED. No way they could send out untested aircraft.

        Unless this is only for the sample cars and not consumer purchases, not sure it would even be legal.

        And in my humble opinion not likely accepted here, Tesla has always considered themselves more of a high tech co than an automobile company.
        And tech companies LOVE using customers as their beta testers…..

        • 0 avatar
          dal20402

          777 LN 1 (together with the next five 777s) went through more than a year of flight test. After the flight test program, the FAA and its international counterparts certified the 777-200 type. After the flight test program ended, the first 777s were refurbished and sold to customers. The very first 777s delivered to customers were the first ones not used in the test program.

          Boeing wanted to do the same thing with the 787, but the first one was so damaged as a result of the Potemkin “rollout” that the refurbishment couldn’t be accomplished, and it would have been so expensive for nos. 2-5 that Boeing decided it was cheaper to build new airframes for the customers that would have bought the aircraft.

  • avatar
    Asdf

    Tesla should postpone this model until it has sorted out its range and charging time issues.

    • 0 avatar
      SCE to AUX

      The ones customers aren’t screaming about?

      • 0 avatar
        Asdf

        They may not be screaming now, but a significant percentage will once they realize the severe limitations of their EV, especially considering the exorbitant price they paid for it. Not wanting to be fooled twice, they’ll ditch the very idea of an EV, buy a good ole’ ICE-powered car and never look back. And Tesla will have alienated an EV buyer for good.

        So, you see, it’s in Tesla’s interest to sort out these very serious issues now, lest they risk ruining the future market for the product they’re selling.

    • 0 avatar
      dal20402

      What range and charging time issues? Their batteries do what they say on the tin.

  • avatar
    dukeisduke

    Taking lessons from Microsoft now?

  • avatar
    indi500fan

    Tesla currently has a much different customer base than the typical car. My buddy’s daughter (who makes serious $$ repping pharma) has a Model S that’s required many significant warranty repairs. And she had a minor wreck with the oft mentioned lengthy time for body parts from the factory.

    But talking to her, it’s the greatest thing ever. Like she has to validate her “out of the mainstream” choice. My view is if she had similar problems with a Bimmer she’d be vowing never to buy another one.

    Now we have the much higher volume Model 3 with the “fast to market” development plan and perhaps a different customer base.

    • 0 avatar
      markogts

      Agree. While I do admire Tesla and Musk and I wish he succeeds in changing the car industry, there are some design choices that should be criticized, but the fan base considers indisputable: super slow and over engineered falcon-wing doors, touch screen to save on dials and buttons, un-validated, self-proclaimed “auto pilot”, the ridiculously small size of the model 3’s trunk hood… just try to comment critically on such items, and you’ll be overwhelmed by screaming fanboys.

  • avatar
    Felix Hoenikker

    Maybe Tesla will pull this off without major rework or recalls. But, this reminds me of costly fixes required on industrial gas plants where new technology was adopted, and one or more steps of the scale up were skipped because “we don’t need to to that anymore”. Problems that would have been found and fixed for a one or two hundred thousand dollars were fixed in the field for tens of millions of dollars.

  • avatar
    dal20402

    The commenter above who pointed out the example of the Boeing 777 is right on.

    The 777 was Boeing’s first all-computer-designed airliner. There was no wood mockup, no actual prototype aircraft, and a whole lot of puckered sphincters. Accordingly, other than the reliance on computers, the schedule and process were pretty conservative. The actual aircraft came in on time, but well over budget as engineers scrambled to implement tons of last-minute fixes. But the result was great. The aircraft performed beautifully, except for some engine issues that were GE’s fault. Every 777 built went into revenue service with an airline, and most of the first ones (including the very first) still fly today, 20+ years later.

    By contrast Boeing tried to get fancy with process with the 787. They set up a reality-free project schedule and shoved a lot of QC responsibility onto their suppliers. Aircraft subassemblies arrived from suppliers needing so much rework that it would have been cheaper for Boeing to build them on its own. The “rollout” of the first 787 was a Potemkin affair with an airplane that had no innards and whose fuselage was held together with Home Depot fasteners. The project ended up three years late. Boeing was unable to sell the first five 787s (~$600 million worth of aircraft) to customers at all. The next 61 aircraft required varying degrees of invasive rework, and Boeing is just getting the last two of those early 787s out the door this year, nearly a decade after they were built. Now, there is a millstone of ~$20 billion in unanticipated expenses around the 787 program’s neck, and it’s unclear if the 787 will ever turn a profit.

    Time will tell whether this process is reasonable or too ambitious. If the latter, we’ll know pretty quickly.

    • 0 avatar

      Aircraft are realativley low volume. It’s a very different game. The tooling is not designed for 1000’s of parts. Making a change late in the game is expensive but not nearly as bad as when you need to make 300,000 parts. Aircraft are more similar to boats you build one with good computer simulated design (or intuition back in the day) then build it and redo things till you get it right. But that means alot of testing in-between. As you mentioned the timeline would be the biggest issue with this theory.

      • 0 avatar
        CarnotCycle

        I can kinda see why Tesla would try something like this; especially given their financial clock-ticking situation. Electric cars, mechanically speaking, are relatively simple. I’d assume most the software is tweaked copy-paste from other Tesla vehicles. The brand has cachet enough to buy a kind of reliability-forgiveness from fans in same way Alfa Romeo or Audi have.

        The hard part for Tesla isn’t building a Model 3, it is stamping out thousands of Model 3’s that actually work most the time. A ‘beta’ program, especially for early-adopter brand fanboi, isn’t terrible initial idea and good way to find the edge scenarios no structured test regime can find. But the risk is huge if a ‘glitch’ hurts and/or kills some people, especially people who didn’t sign up for the beta. Such things are known to attract sharks. And Elon Musk is a good salesman, but a good salesman in shark-infested waters is only a good lunch.

        If the 787 had killed some paying passengers while getting debugged live in service, sharks would have killed the 787.

        • 0 avatar
          JimZ

          “I can kinda see why Tesla would try something like this; ”

          yeah, their fans will give them a pass.

          “Electric cars, mechanically speaking, are relatively simple.”

          BWAHAHAHAHAHAHA!

          You know the “poster child” for unreliable vehicles (FCA?) Go look at CR and see what makes them “unreliable.” You’ll find that it’s largely electrical/electronics and suspension.

          Guess what EVs have too? Electrical systems and suspension. Guess what breaks a lot on Tesla’s vehicles? Electrical/electronics and suspension. Oh, and weren’t there quite a few total replacements of those “relatively simple” drive units on the Model S?

          you people are incredible sometimes.

          • 0 avatar
            CarnotCycle

            “Electric cars, mechanically speaking, are relatively simple.”

            BWAHAHAHAHAHAHA!?”

            Count moving parts in latest E-class AMG – no, make that a Toyota Yaris – then count moving parts in a Tesla Model S.

            One of them is much simpler than the other.

          • 0 avatar
            mcs

            @jimz: The electrical systems and suspension were pretty much the only thing that ever broke on my BMWs. Even the dishwasher controller module that nearly burned down my house was a Bosch. Right now I’m leaning towards a Porsche Mission E, but it’s the electricals that have me worried. I would bet that I’ll have more warranty issues than a Tesla. Porsche also has their IMS failures in the past.

            Then again, at least it won’t have the number of sensors an ICE has. Currently trying to explain to a friend how the oil level sensor on her E-46 works and it might not actually be failing. She filled it with cheap walmart oil and it might be throwing off the sensor. My daughter has sensor issues with her car (evap system). It’s fun chasing those down sometimes.

            Tesla or Porsche I figure I’m asking for trouble. If I want reliability, the rumors are that Nissan will be really aggressive on pricing for the 200-mile Leaf this fall. I might pick up one of those anyway as the commuter car, but we’ll see.

          • 0 avatar
            JimZ

            “counting moving parts” is meaningless. if that’s all you have to fall back on, I’ll take that as a concession.

          • 0 avatar
            CarnotCycle

            ” ‘counting moving parts’ is meaningless. if that’s all you have to fall back on, I’ll take that as a concession..”

            Got tired of counting moving parts on the Yaris, didn’t you?

            And I didn’t say which one was mechanically simpler, I just stated that one of them was.

            Which one do you think is mechanically simpler?

        • 0 avatar
          dal20402

          “If the 787 had killed some paying passengers while getting debugged live in service, sharks would have killed the 787.”

          This is a good point and worth considering.

          Boeing, for all its disastrous management on the 787, did the right thing when it came to safety. The only one of the 787’s many problems that threatened passenger safety in any way was the battery design, by a supplier, and Boeing’s fix (a big fireproof box around the battery) was proven effective when another battery failed with no consequences for the rest of the aircraft. So far, the 787 has an exemplary safety record.

          Elon Musk has given no reason for confidence that he takes safety as seriously.

  • avatar
    OldManPants

    Good. Maybe this will act like a shot of WD to loosen the stuck Bolts.

  • avatar
    islander800

    This has the whiff of a “Hail Mary” pass, out of desperation.

    What’s the rush, Elon? You do realize that automobiles are much more than a set of code. Skipping the pre-production shake-out, a time-honored and standard step in the release of a new model in the AUTOMOBILE industry, is the height of hubris.

    What’s to say this isn’t a panicked pitch to the equity markets for more money, much more money, in order to have a chance of meeting his promises on production numbers and dates? He doesn’t want to see all those pre-orders walk out the door, and he’s chancing he doesn’t need a Beta test before unleashing his new product on public highways.

    Good luck with that, Elon. And good luck to public safety.

  • avatar
    JimZ

    ok. so… long established automakers can go through a normal, 3-4 year development cycle and still end up with launch issues and early quality problems.

    and now we’re supposed to believe Tesla has found a way to give that all a miss via “computerized scrutiny.”

    yeah. I wish them luck.

  • avatar
    Kyree S. Williams

    Oh, sure, I need my car to be as unreliable as beta software. But, hey, at least Tesla is admitting that the customers are getting live prototypes. BMW, for example, certainly didn’t do that when they released the E65 7-Series, which was nothing so much as a rolling prototype at least until its facelift in 2006.

    • 0 avatar
      JimZ

      man, you’re just a Luddite who hates anything new.

      • 0 avatar
        hgrunt

        There’s a lot of luddites when it comes to cars.

        I just point out how people didn’t trust hydraulic brakes when they first came out, because you couldn’t see the cable and bands. What if something broke? You wouldn’t be able to see or fix it until it was too late! Tragedy!

    • 0 avatar
      mcs

      @kyree: “. But, hey, at least Tesla is admitting that the customers are getting live prototypes”

      That’s not what they said. They are building the pre-production vehicles on the production lines. They’re still testing them. They aren’t going to customers.

      I took another look at the investor call notes. According to the notes, they aren’t using beta tooling to build the cars, but instead are using production tooling.

  • avatar
    7402

    Reminds me of the old software developer adage that beta testers are just willing victims.

  • avatar
    jkross22

    This is normal in an environment where publicly traded companies are graded on quarterly performance.

    • 0 avatar
      baggins

      you mean the same environment we’ve had for decades?

      what are you suggesting – that Tesla is just doing what everyone else does? That private stock ownership is somehow evil? That publically traded companies should not have to report quarterly to shareholders?

      Or just parroting something you read on Jalopnik or Huffpo?

  • avatar
    SuperCarEnthusiast

    Tesla cars can get over the air updates for it software so the “beta” is not skipped it just transferred to the buyer!

  • avatar

    This is going to end well.

  • avatar
    Big Al From 'Murica

    Meh. I’ll just buy an Alfa. At least then I can say “sure it is unreliable, but it has a Ferrari engine” and I’ll get to use terms like “good mechanical noise” instead of “reboot”


Recent Comments

New Car Research

Get a Free Dealer Quote

Staff

  • Contributors

  • Matthew Guy, Canada
  • Seth Parks, United States
  • Ronnie Schreiber, United States
  • Bozi Tatarevic, United States
  • Chris Tonn, United States
  • Corey Lewis, United States
  • Moderators

  • Adam Tonge, United States
  • Kyree Williams, United States