Anyone who has had to do with the manufacturing side of car makers (German ones anyway) knows how highly optimized from end-to-end their processes are. It seems doubtful that anyone could leapfrog this amount of accumulated process knowledge and optimization simply by way of being new.
Automation is not a black-and-white thing, either, as it is often portrayed. Sure, some things are fully automated (e.g. assembling the car body), but many things are partially automated. For example, there is a tool which is essentially 5/6/7 torque wrenches coupled to a lifting arm. This tool allows a line worker to mount a tire with little physical exertion in seconds. You see this sort of thing all the time at car assembly lines; offloading mechanical power to the machine, while having a human line up and steer things. (This also avoids a whole bunch of safety concerns you normally have when operating autonomous equipment in the same physical space occupied by humans)
Another example: The body of virtually any car is self-supporting (there is no frame), so getting it right is rather critical. To achieve both the high strength and low weight of a modern car body the steel is deformed very closely to the maximum deformation it can withstand. This needs very good matching of the process to the steel. So, the manufacturer of the steel coil tests each coil individually for its exact properties. For mass production lines only coils meeting very tight tolerances are used. For low volume lines the process is adjusted individually for each coil based on the manufacturers testing and some in-house testing.
It's also shocking to anybody who had studied Toyota's history. Toyota spent decades honing their processes, but are notoriously automation-averse [1], as automation locks in a (hoped-for) short-term productivity gain but prevents the human-driven continuous improvement that has made them highly productive. [2]
In some ways, I think the long software-is-eating-the-world boom has been bad for us as an industry. We have been so successful with "throw tech at it" solutions that we don't know when that isn't a good idea. E.g., the Silicon-Valley-reinvents-grilled-cheese flop The Melt [3], or the $120m clown show that was Juicero [4].
American car companies spent decades trying and failing to learn Toyota's approach. They even had enthusiastic help from Toyota, who even went so far as to take GM's worst plant and redo it as a joint venture. (Interestingly, it's the same plant that Tesla now uses. [5]) For those unfamiliar with the story, I strongly recommend This American Life's episode on it. [6] But my takeaway was that executive arrogance kept GM from learning that there was a much better way to make cars. That sounds more and more familiar these days.
Yeah I also read about how American car factories used super sophisticated and expensive computer system back in the 80-90s only to realize that the Japanese accomplished the same with just people and some cleverly placed sticker symbols.
I think underlying a lot of this is a disdain in America for blue collar workers. People at the top thing of them as dumb monkeys and that there is nothing to what they do. Hence the falsely assume a computer or robot can do the job just as well.
The Japanese showed that by utilizing what humans are actually got at you get better productivity. Using humans the same way as robots is a very bad way of utilizing the flexibility inherent in a human.
> I think underlying a lot of this is a disdain in America for blue collar workers.
It could simply be that everybody knows that blue collar job are eventually going to disappear. You don't want to build your latest factory using technology that will make you obsolete in a few years.
Obviously we were not quite there yet in the 80's/90's. Otherwise you would be reading the opposite story on how Japanese Car Factories failed because they used human instead of machine like in the US. Actually you can replace "Japanese Car Factories" in the previous sentence with a lot of thing that are actually true today.
Now, about Tesla, there is some relevance. They are still building car that are very much like regular cars and other manufacturer are building their own factories regularly for new platform/model, not held back by legacy issues. So it does indeed seem that Tesla is at an experience disadvantage.
I guess what happens is that this is Tesla trying to leverage a manufacturing bet on top of its successful "electric car" bet. You can see that on how harshly they lashed at journalists talking about their manufacturing safety, and draw some parallel with the initial review of the model s ( compared to today where you can make a 100 million view video bad mouthing about their car and Musk would probably not even be notified )
Tesla real goal is to find a manufacturing process that can get them producing car 1 order or 2 order of magnitude faster than the competition (and you know such a process cannot include human). They want to be the market leader by volume on every single class of electric vehicle in 10 years.
> everybody knows that blue collar job are eventually going to disappear
Everybody does not know that. Indeed, a lot of people know the opposite. Including Toyota, who went from a small-time player in post-war Japan to the world's largest car company based on the belief those jobs weren't going anywhere.
How is just assuming that without evidence not disdain for blue collar workers?
> Tesla real goal is to find a manufacturing process that can get them producing car 1 order or 2 order of magnitude faster than the competition (and you know such a process cannot include human).
I'm not sure this is true, and if it is, I don't think it's a particularly smart goal. Toyota spends ~30 worker-hours on building a car. Getting that down to 3 doesn't do much to reduce costs, and getting them down to 0.3 does way less.
To make that happen, Tesla will have to invest a lot of money in less flexible, less repurposable hardware and software. This a) doesn't save money in the short term, b) can introduce significant lags (because they're spending time up front to save time later, and also because they've increased delay risk), and c) may lock them in to suboptimal techniques in conditions where human workers could keep improving.
> How is just assuming that without evidence not disdain for blue collar workers?
In order for that to be disdain, you need to associate people identity with their job.
I'm not a millennial, but I'm still young enough that the concept of job as identity does not fit the social reality of the people I know. They have moved between position enough that they cannot just be summarised by their present occupations.
So this is not disdain for the people, but for the jobs. And not only those jobs, but all the jobs, including mine. I want Star Trek for my great-great-grand-children, not grinding at work till they drop.
> I'm not sure this is true, and if it is, I don't think it's a particularly smart goal. Toyota spends ~30 worker-hours on building a car. Getting that down to 3 doesn't do much to reduce costs, and getting them down to 0.3 does way less.
I don't know what Tesla is doing, but what you say is correct, there is no gain to make by replacing a human by a robot if the robot is not providing massive economies somewhere. I personally don't think that hate of blue collar worker is really that valuable to Tesla, so I'm thinking more of massive throughput increase. Since it appears that Tesla want to spread in every market segment, being able to build a "10x" production line would allow them to do it at a fraction of the time of their competitor.
Note that I'm not suggesting that Tesla is going to make it, or that the technology is ready. Musk and Tesla investors do.
Another point which one always has to keep in mind with Musk. Maybe that's also related to SpaceX. If humanity has a future in space, it needs to be able to build stuff with as little human as possible.
edit: I just noticed that I missed the proof. Like the majority of the people on HN, I'm in the business of automating other people job away. I have friends working in the engineering field and I have yet to hear a story about some new tech requiring more human being than before. It is the same as white collar work: next release will require less people or less qualified people.
I was enlightened by the link above to NUMMI. Here is a transcript of that video: https://www.thisamericanlife.org/561/transcript
You have seen the same advice from Paul Graham: "Do things that don't scale", http://paulgraham.com/ds.html.
In the case of NUMMI, the emphasis was changed from quantity to quality. It didn't mean automation, it meant listening to suggestions from the workers. The worker adopted the change because it meant a change from industrial mayhem to a fulfilling work experience. The changes at that plant have (too) slowly trickled through the US car industry, as the managers involved moved around the GM organisation, finally giving the US car industry the high quality we see today. Tesla is similarly learning that, rather than automation being the magic answer, it's the one-off things that humans can do by coordinating that will bring costs down and quality and production up.
> It could simply be that everybody knows that blue collar job are eventually going to disappear. You don't want to build your latest factory using technology that will make you obsolete in a few years.
Plumbers are not going to disappear. There is heaps of repair work that won't disappear.
But beyond repair, look at construction. There are all sorts of claims about how 3D printing and full automation of house construction is coming, but go and look at almost any construction site you see.
(edit: Innovation is hard with construction in part because people want things that will definitely last 50 years. Fail fast is great on a web app, but what about failure in 10 years
because 3D printed concrete doesn't work correctly in your location? How about brick laying robots cement failing in 15 years when the company is insolvent? Are you up for the risk? )
Power tools introduced in the 1950s or earlier have made a huge difference, but these jobs do not appear to be disappearing.
In Australia I've heard of teachers and people who can write mean python code quitting to become builders.
People with power tools are productive and very flexible. We're a long way off from a machine that can attach a deck to a house and make a hole in a wall for a window.
> It could simply be that everybody knows that blue collar job are eventually going to disappear. You don't want to build your latest factory using technology that will make you obsolete in a few years.
This is circular reasoning.
Blue collar jobs will only disappear if no-one builds factories using them, because everyone knows these jobs will disappear.
> you know such a process cannot include human
Sounds like you're just assuming this is true, because "more tech is always better/faster/etc".
I'm not so sure this is correct.
Knowing what will happen eventually isn't always as useful as most people think, even if that belief turns out to be true. The transition can take decades or centuries. And in the long term, as Keynes used to say, we're all dead anyway.
The Melt was one of the most moronic restaurants to ever grace this Earth. I have no idea who thought it was a good idea to charge people fifteen bucks for half a slice of a grilled cheese and a small cup of soup.
Interesting article. I never heard of them, but the original CEO blames real estate as being the cause of failure. He overlooks the fact that their sandwiches were MICROWAVED. Who in heck microwaves a GRILLED cheese sandwich?
$6 for a tiny, soggy microwaved cheese sandwich.
$3 for a tiny bowl of soup.
That's a weird thing for the CEO blame. Commercial real estate for restaurants is a very well understood marketplace. Why would you go into that business without understanding one of the most important components to both your budget and your availability to customers?
> while yes, the headaches of real estate in the Bay Area—where The Melt first set roots—are well established, fingering real estate is a bit like pointing out that it gets cold every winter
To each their own maybe? I thoroughly enjoy grilled cheeses at The Melt and still visit it regularly. I don't really care how they run their businesses, what tech they use, whether the CEO came from tech or food. I like the grilled cheese there better than most places. There are only a couple of other places in SF/bay area with better/comparable grilled cheeses to me. (American Grilled Cheese Kitchen, Boudin SF)
Also, you said you haven't heard of them, but you made the claim that it's a microwaved soggy grilled cheese sandwich. I've never had a soggy grilled cheese at The Melt.
Grilled cheese is so low maintenance that I don't understand why you would pay a premium for something you can make way better yourself for practically pennies.
No matter how good they make the grill cheese, I feel confident that I will be able to make a grill cheese that tastes better for cheaper every single time and I believe this sentiment is shared by most people.
A lot of sandwiches that I would want on a whim I do not have the groceries for. I almost always have the ingredients for a grilled cheese. Not only that, sandwiches you could buy from delis all over Palo Alto are substantially cheaper that sandwiches from the melt.
I know this is a joke but just for safety's sake, I want to mention that this would be a bad idea in a modern toaster. The grates close, covering them in melted cheese, and now you have either a ruined toaster or a fire hazard on your hands.
I remember when this image was making the internet rounds a couple years ago and I was surprised how many people hadn't thought about the issues with it. Luckily, I know no one that had tried it for real.
It's definitely inspired some folks. I recently went to a local event and there was a stand there selling "cheese toast" at $8 a pop. To be clear, it was not a sandwich, they weren't using nice bread (just storebought Texas toast), and the cheese selection was just the standard fare (cheddar, parm, jack). No clue where they got that exorbitant price from.
I have a friend who worked at Juicero, so unlike 99% of the people here on HN, I have a lot of info on it, and didn't just read the headlines. In fact, I had more than a few glasses of Juicero drinks whenever I visited her place (generally hated it, but I'm not their target demo).
Overall, I never liked the idea, and I told her that. I said it was way too expensive and they would have to decrease the price by 80% for it to be viable. But for the people who got it, mainly celebrities and millionaires, they loved it. I think I remember that Beyonce or Katy Perry or Britney got it for all of her dancers, etc.
The entire "green juice" industry is large, and the founder made his first millions by selling chains of these types of juices. The target was to make a glass of green juice less expensive than what you would find at a high-end gym or juice shop. The plan was to build it out with the people who were willing to pay for them, like celebrities and influencers, and then make them mass-market. The biggest expense was getting the food chain up and running, trying to get enough local farmers, etc, and getting the freshest ingredients possible. The internet connection, while ridiculous on the surface, was necessary to ensure that the ingredients were still fresh, because they literally had no preservatives. They wanted to avoid having people drinking a glass of mold or worse especially if they weren't mindful of expiration dates. This could also help if they sold these presses to juice shops (a later business strategy) and the person selling the juice might have old expired stock.
But once it was set up, the plan was to scale out and lower prices through higher volume. Unfortunately, the bad press hit just as they were doing a round of funding and went bankrupt soon after because funding dried up.
Do I think it would have been a hit? Probably not. But the idea wasn't absolutely stupid. It's about the same as any other Silicon Valley idea I've heard. Cater to a very select market of influencers first, and then scale out and create a new market based on green juices. The company absolutely believed that green juices would help the world, because it's a lot healthier, etc, so if you buy into that mindset that they could create a new industry, then it made sense.
Did they make pure business strategy mistake? Yes, just like other companies. Did the founder have too much of a Steve-Jobs-Deity complex by trying to over-design the press? Yes. But it's not nearly as bad as people believe from the headlines. The goal was to create a new market, like how Uber created a market by lowering prices for ridesharing. If you could lower the price of green juice to the price of a Starbucks latte from $15, then they really felt like this company would be a tremendous success.
Thanks for posting your insider's view. I agree their mistakes are not unique, but the scale of the failure and the level of dumbness makes them a much bigger failure than most.
They started in 2013, long after books like "Four Steps to the Epiphany" and "The Lean Startup" explained exactly why scaling up before you have traction is a huge mistake. They could have learned the "this isn't working" lesson for an order of magnitude less money. If they had done that, they a) would have had capital to recover from their early mistakes, and b) would have been much better placed to find that low-cost solution.
When I look at a crater like that, I think about the 50+ seed rounds (or 20 A rounds or whatever) that could have happened but didn't because they got high on their own slide decks. They blew up not because they couldn't have discovered these problems, but because they didn't want to.
Juicero was ridiculous even if only because the machine was ridiculously overengineered. They should have used two steel rollers, I bet the machine could be made for $20. Instead they made an absurd contraption that crushes the bag between two plates.
I have plenty of food in my refrigerator more perishable than green juice, and no QR-code reading iOS app, and have still managed to avoid killing myself with food poisoning.
If you could lower the price of a Ferrari to the price of a Camry it would be a tremendous success too. You can make a business selling to Katy Perrys and Beyoncés, but not at a $700 price point, and not by raising $100m in VC.
> The internet connection, while ridiculous on the surface, was necessary to ensure that the ingredients were still fresh, because they literally had no preservatives. They wanted to avoid having people drinking a glass of mold or worse especially if they weren't mindful of expiration dates.
It was ridiculous on the surface, but also ridiculous all the way through to the core. We eat a lot of perishable items, and we have a well understood mechanism for dealing with it: Expiration dates. Making the expiration date into a QR code that gets scanned and then sent to a server so it can be checked by a computer is a thing you can do, but you can also just put it on the packaging, the way every other item in your pantry and fridge do.
If you really wanted to be crazy, you could encode the expiration date in the QR code and check it that way - again, without a net connection.
> But once it was set up, the plan was to scale out and lower prices through higher volume.
Not a bad plan. But as has been pointed out ad nauseam at this point:
1) One of the reasons the price was high is that the machine was poorly designed and hideously over-engineered. Relying on prices to drop as your volume increases is great, but if you do an even mediocre job at product design, you need to rely on a lot less volume.
2) The fundamental idea of the machine wasn't very good. A supply chain to ship bags of high quality juice to people? Great idea. A fancy machine to (effectively) open the bag? Not a great idea.
> But the idea wasn't absolutely stupid.
There were elements of the idea that weren't absolutely stupid. But taken as a whole, as implemented by Juicero? ...yeah, it was absolutely stupid.
Let's be clear: The business model was selling a hideously expensive gadget for very high prices, so that you could then sell the buyer a bunch of juice packs. Selling a cheap gadget for high prices and abandoning the juice packs might have been a good strategy (albeit scummy); the mainstay of stupid infomercial gadgets. Selling an expensive gadget for a low price so you could make it up on the juice packs might have been a good idea (it's a tried and proven strategy). But an expensive gadget at an expensive price, so you can sell juice? All you're doing is crippling the sales of the juice packs.
> Did they make pure business strategy mistake? Yes, just like other companies.
Just like other failed companies, yes. The only companies to succeed after that level of strategic failure were the ones lucky enough to pivot. This wasn't a viable idea!
> If you could lower the price of green juice to the price of a Starbucks latte from $15
I dunno about where you live, but around here a green juice from a juice bar is already a lot closer to a Starbucks latte than to $15, and the Juicero packs were actually more expensive than just going to a juice bar.
You're totally right that cheap, high-quality juice for the masses is a viable idea. And you're right that initial high prices, falling due to volume is a good idea. What I think you're missing is how little these things have to do with Juicero.
> Making the expiration date into a QR code that gets scanned and then sent to a server so it can be checked by a computer is a thing you can do, but you can also just put it on the packaging, the way every other item in your pantry and fridge do.
The expiration of these packets were measured in days or maybe a week, not weeks. Also, as I mentioned above in another comment, there could be particular product recalls like the current romaine lettuce recall. I'm not arguing that it wasn't overengineered, I'm just pointing out why they did it. I wouldn't have done it, but they chose to.
> One of the reasons the price was high is that the machine was poorly designed and hideously over-engineered.
It was totally overengineered. When my friend read the product breakdown from one of the blogs about how it was as if completely unexperienced mechanical engineers had designed the press, she said this wasn't too far off the mark. But the main reason for the failure wasn't the press, it was the food supply chain. That was way too intensive and expensive and wouldn't scale the way they wanted it to. They spent money on food scientists to make sure that the combination of vegetables they used wouldn't interact during the time that they were packaged together, etc. The people they were marketing to were pretty price-independent. $600 for a dumb press wasn't a big deal if they got a great green juice, which they did. Their phase to expand and lower the price was in the future, before they died because funding dried up.
> Just like other failed companies, yes. The only companies to succeed after that level of strategic failure were the ones lucky enough to pivot. This wasn't a viable idea!
Firstly, 90% of all startups fail.
I never said it was a viable idea. I thought the idea wasn't good. But the investors took a chance. That's how it works. Even YC admits that the idea is less important than the people who are attempting the startup. The founder had fundamental experience with launching juice bars on the East Coast, and the idea to create a new market wasn't horrible. Investors took a chance, and then the management fucked it up.
And yes, you can laugh at how dumb the idea is now, but the same goes for Soylent, or even the infamous Dropbox comment in HN. But at least their idea wasn't just "sell $600 juicers ad infinitum and make $$$$". That's my point. A lot of the article miss this.
Juicero is just another failure like all other failures we see all the time in Silicon Valley. There are far worse examples of bad investments out there. I'm sure it's definitely in the bottom quartile of bad investments, but in my opinion it's not like it's the worst ever and doesn't deserve the level of humiliation associated with it. It was just a mismanaged idea that got bad press and then the Internet Lynch mob jumped on it.
> there could be particular product recalls like the current romaine lettuce recall.
I don't know if you can give this as an example in support of internet connectedness - we seem to be doing OK with the romaine lettuce issue without overpriced IoT salad bowls.
>The expiration of these packets were measured in days or maybe a week, not weeks.
That isn't a particular hurdle. There are plenty of products in my local supermarket that have an expiration date measured in hours after purchase (usually about 3 or 4 days in total, the sticker is printed on-location, this especially concerns some raw meat products).
>Also, as I mentioned above in another comment, there could be particular product recalls like the current romaine lettuce recall.
That might require an internet connection but not permanent either. The juicer could simply have a blacklist of product IDs. Supermarkets also already implement this at the checkout. The cashier will be unable to add the item to the bill, even if a supervisor gets out their admin keys and tries it manually (the computer will even block same-price manual entries after a blocked item is scanned during checkout).
In the lean startup world, people talk about early products in "painkiller vs vitamin" terms. If you're going to build a new kind of business, you want a product that people are desperate to have. If it's a nice-to-have, it's much harder to find enough early adopters to build a sustainable business. (Not impossible, but it takes a lot more capital and is much riskier.)
Juicero was literally a vitamin play. Nobody needs it, but some people like it. Coffee, on the other hand contains an addictive stimulant that 60+% of Americans consume every day. People need their morning coffee, and will literally develop headaches if they don't have it often enough.
Juicero=Nespresso is one of those comparisons that only makes sense if you don't think about the details. Sure, it's got some surface plausibility, but only if you don't think much about the details.
You insist that the plan wasn't stupid, but the plan was to initially market to ultra-wealthy celebrities who want to pinch pennies on juice sold at gyms or juice shops?
Yep that was the person but memory is rubbish for names - my production eng lecturer at college mention him in the context of the collapse of the UK Motor Bike industry
My Corolla manufactured in Mississippi is even better made (in terms of visible gap tolerances, at least - they both experienced pretty amazing reliability) than the previous one, made in Fremont. I think it's Toyota's manufacturing process that's key, not the geography of where the car is manufactured.
Same as people bashing "made in china" when the highest quality laptops (Macbooks) are made there. It is the region, low quality things made in china are low quality on purpose.
These stories of SV excesses (especially Juicero) remind me of the anecdote telling how the NASA spend a lot of money to develop a ball point pen than could write in zero gravity. The Russians? They took pencils.
It's a funny story, but it's also not true. You don't want pencils in space (not because they don't work, but because you don't want graphite dust everywhere).
I like the part where you skipped right over "a short and amusing or interesting story about a real incident or person" so you could quote "an account regarded as unreliable or hearsay".
It's okay to be wrong. I distinctly remember believing the (false) pencil story once, before reading more about it and learning the (far more interesting) actual story. And also realizing that the false pencil story pushes certain goals and agendas, too, so probably isn't accidental.
I love that story because it reminds me to always look for the deeper reasoning behind shallow first impressions - the only reason a pen seems like an expensive waste is because I'm ignorant of the full complexity behind the decision to use pens over pencils. As a programmer, it's easy to walk into a room full of seemingly simple manual processes and tell everyone they need to be automated out of a job, but it's highly disrespectful to their years of specialized experience.
Always good to be humble and remember that I'm not smarter than a rocket scientist in their own field.
Both NASA and the Russians used pencils in the beginning. Once a pen the worked in zero gravity was developed (by a private US company, and not by NASA or with NASA money), both NASA and the Russians started to buy and use those pens.
> Anyone who has had to do with the manufacturing side of car makers (German ones anyway) knows how highly optimized from end-to-end their processes are. It seems doubtful that anyone could leapfrog this amount of accumulated process knowledge and optimization simply by way of being new.
I think with this way of thinking there will never be progress. It's thanks to this try-and-fail that this man was able to build things that nobody thought to be feasible. We are talking about mistakes that were done in the '80s, meaning that this man is trying to leverage what we have nowadays and he still can't, which is pretty sad, considering how advanced we think of ourselves compared to 40 years ago.
Sometimes it's this way of thinking that really prevents innovation: we tried, it didn't work. Sure, you tried under different circumstances, I am quite sure you gave up on that for whatever reason and you are fine with it. Good lesson. However, there will be a new generation that will make it, no matter all your "I told you so".
I want to see if in the future, when we'll have highly developed AIs, we will think "oh, 200 years ago we tried already and it didn't work".
Well, trying to build AI using failed papers from the 1980s have near zero value today.
What Tesla is doing is like building a modern app but refusing to do A/B tests or code review or version control, generally universally accepted methods in the current industry, and going back to re-try software development patterns of the 1980s that failed.
>>Well, trying to build AI using failed papers from the 1980s have near zero value today.
I think if you had picked any other example you'd have been correct, but many current-day AI advances are based on decades old techniques. One caption from an article about Geoff Hinton states: "This publication from the mid-1980s showed how to train a neural network with many layers. It set the stage for this decade’s progress in AI."[1]
The issue is not one of carefully acknowledging the past and noting that some of its lessons don't apply.
The issue is how Musk typifies a certain sort of arrogance that runs through much of Silicon Valley tech culture. This is why the n-gate summaries love to describe Hacker News reinventing everything from first principles -- there's this ingrained assumption that it's literally impossible for anyone outside of SV to have ever had any kind of relevant experience or possessed any kind of relevant knowledge, and so there's no point in studying the history at all.
Tesla is now learning, in a very painful way, why hubris is traditionally considered a major flaw, and that maybe this stuff is harder than they thought and they could have benefited from standing on the shoulders of some giants.
> The issue is how Musk typifies a certain sort of arrogance that runs through much of Silicon Valley tech culture.
Is that the same arrogance of the opposite party saying "I told you so, son"?
I have never been in the Silicon Valley, so I won't talk about that. Yet, I think that this mentality creates at least 1 good thing after trying 99 times.
I have to be honest, I may be a bit biased here, because I am bit frustrated by this mentality that innovation needs to follow a specific path, that you need to always follow what your "parents" have been telling you, etc - and this is due to the environment I live in, unfortunately.
If Musk had been able to achieve what he had envisioned, now we would be talking about something totally different - OMG, what a genius, how come we didn't come up with that earlier? How come this young man made something we thought impossible, despite all the studies, etc.? And so forth. That's the same cowardice that doesn't allow us to think further, to think bigger. He does, he fails, he tries, he fails again, and who knows, maybe he makes it. If he doesn't, it doesn't mean anything. Without that way of thinking, we wouldn't have put our foot on the Moon.
I have to be honest, I may be a bit biased here, because I am bit frustrated by this mentality that innovation needs to follow a specific path
Who said it has to? I didn't.
What I said was that Silicon-Valley-type people often display a certain amount of hubris in believing that nobody else has ever done what they're trying to do, and therefore disregard free knowledge that others obtained at great expense.
Musk is attempting the revolutionary, never-before-heard-of, herculean task of... building a factory to make cars. Strip away all the PR-speak from Musk's and Tesla's public announcements and statements, and that's what you find underneath. "Build a car factory" is not a new problem. We have over a century of prior art on things people have tried, what succeeded and what failed. It's the willful ignoring of this prior art that I'm complaining about; other companies have learned about the advantages and disadvantages of automation, for example, but the whole enterprise of Tesla is wrapped up in so much mythology around breaking new ground and being completely new that they were constitutionally incapable of admitting "yeah, this is a thing other people have done, too, maybe we should study how it went for them and learn from that".
Most things Tesla seem to come down to SpaceX and Mars. When I have a conversation about a car company and it ends up veering off to Mars, I realize that it was never really a conversation about cars to begin with. Having said that, I don’t know how to cut through the mythology, some of which seems deserved. I was one of the people who thought that SpaceX was probably going to fail years ago, and I was quite wrong. Some of what you’re saying was part of my reasoning too, insofar as building and launching rockets wasn’t a new thing, but really quite an old one.
Even still, I find it hard to see the future for Tesla that real fans seem to perceive. SpaceX has literally and figuratively taken off, but Tesla just seems stuck. I’m especially leery of how willing they seem to be to beta test software that can cause loss of life and limb. I also don’t see how existing Titans in the automotive space can’t just eat Tesla’s lunch, given that they can crank out a couple orders of magnitude more cars per year than Tesla. What is Tesla doing that they can’t copy, given that Tesla is still trying to figure out the assembly line?
SpaceX I think has a better chance of succeeding simply because it's Musk doing the thing that's historically worked for him: moving into a space where somebody else already shelled out for the fundamental R&D, then delivering iterative improvements slathered in tons of marketing sheen.
Agreed. This is not the tech industry where you can change direction anytime and do something totally new. Manufacturing physical goods is very hard especially something as complex as a car. I have worked at a German manufacturer for a while (chainsaws) and it's amazing how many people are working on optimizing processes. This is not something you can pick up in a few years.
While I agree with the sentiment, I think we can be empathetic about where Musk comes from. The same arguments have been raised about making rockets, how non-states cannot make them, or how they cannot be reused. They probably even had good reasons for them. But still Musk and his team upended the status quo and showed it’s possible. His teams push the envelope forward all the time. Sure they are bound to have some failures - but hey we need people like him to understand what’s not possible(now)
The same arguments have been raised about making rockets, how non-states cannot make them, or how they cannot be reused.
Non-state entities have made all of the USA's rockets.
They weren't reusable because at the time they weren't cost effective. It's only recently that material sciences has reached the point where reusable rockets are potentially cheaper than single-use rockets (and even then, only after 5-6 launches).
Really depends on what you mean, doesn't it? I'm pretty sure all, or nearly all funding for every orbital+ rocket launched in the world came from governments before SpaceX. Yes, the work was often done by private companies. But all of the money came from governments. SpaceX was the first company to successfully launch an orbital rocket without government backing. [1]
That is belittling SpaceX accomplishments. They were told by all the "big boys" in the rocket industry that what they were doing could not be done.
Nobody had landed a rocket that way before SpaceX and Blue Origin did it.
There is nothing fancy about Falcon 9 rockets materially wise as far as I know. The key innovation was in the control software and using software industry style rapid iterations to try again and again until they nailed it.
SpaceX developed rockets at much faster pace and lower cost than the competition thanks to this iterative approach vs the water fall method used by the competition which cost 10x as much and is significantly slower.
Your talk of material science would be more relevant to the space shuttle I think. That was an attempt at reuse using sophisticated technology and materials. The spaceX innovation is reusing a cheap rocket. NASA developed a reuse system based on a very expensive and advance design.
> Nobody had landed a rocket that way before SpaceX and Blue Origin did it.
Let's not give SpaceX accomplishments that weren't actually theirs - it makes it easier to dismiss everything else they're doing. The DC-X did this (as a sub-scale SSTO demonstrator) in the 90s.
A crucial difference, IMHO, is that rockets were a largely non-competitive industry, whereas the car business has been a cut-throat competition for near a century.
Musk had to have his arm twisted to make them reusable, sorry, re-landable. All the innovative ideas are already out there, they just have to be found and pursued.
I mean, in the case of Uber, Lyft, and AirBnB a lot of their advantages came from breaking laws as well, something that can be tolerated a lot more in those industries.
It is still worthwhile to try though. Yes, many attempts fail, but they might still give some useful nuggets. And they are often cheap: a failure of established process may 10000 times more expensive than a disruption attempt (e.g. 1.5T F-35 vs 100M for cheaper drones).
During their most recent earnings call, there was a very interesting back and forth between Elon, JB Straubel, and a stock analyst. The stock analyst pressed them to distinguish their proposed vision of a fully integrated factory from the industry leading Toyota Production System (TPS). You got the distinct impression that Elon and JB didn't even know what TPS was. Very eye-opening that the CEO and CTO of an automotive company lack basic industry knowledge.
I appreciate the value of thinking things through from first principles, but there comes a time when you have to stand on the shoulders of giants and save a few billion dollars of wasted cap-ex...
Ironic given that their Fremont plant was used by Toyota. You'd think they would have gotten some knowledge transfer on car assembly 101 when buying the plant.
Some things are surprisingly hard to automate. When your parts are hard to grip and have high tolerances and some are broken, good luck in getting a robot working reliably.
Humans are in many conditions surprisingly fast, accurate, flexible, space saving and cheap when compared to a special robot that does the same thing.
The process used by the car industry is state of the art and highly competitive. But the product itself has areas that have seen very little innovation for decades - just incremental optimization.
Tesla started with a clean slate, of course they will repeat some mistakes but they have the prospect of building a revolutionary product. In a way, a bit of chaos comes with the territory.
What do you think the Model 3 shows when compared to the Bolt?
The Tesla obviously has better mind share at the moment, but various aspects of the vehicles are pretty comparable (even though they aren't really aimed at the same market segments):
I don't see a generational advantage for Tesla when I compare the 2. I see Tesla having a very short amount of time to match quality levels and production volume.
It paints a hopeful picture if you're interested in the future of electric cars, I think. The Bolt is something GM threw together with the minimal amount of effort necessary, understanding that this generation would have pretty low volume, and it's honestly really good for what it is.
> What do you think the Model 3 shows when compared to the Bolt?
An additional 72 mile range for the Model 3. And a standard compact car form factor to Chevy’s sub sub-compact. I’m sure that yields some shoulder room and other comfort benefits.
The greater range costs $15,000, it is mostly coming from a bigger battery (Tesla does have a modest edge on range/kW-h).
The form factor is also not a technical thing, it's the sort of thing I was trying to head off with they aren't really aimed at the same market segments. I think it is fair to say that the Model 3 is the more desired car, but that's not the same as demonstrating a big lead in electric platforms.
> No one else is trying to make a sexy electric car at a low price.
Whatever Tesla is trying, they aren't actually delivering that, either. The model 3 may have curb appeal (the interior is another story) but it certainly isn't at a low price (the base price is above the median total—options and taxes included—transaction price of a US new car purchase.)
Name one thing that Tesla has done that is revolutionary rather than evolutionary? (Working features only, otherwise you're comparing it to a very long list of revolutionary features being researched by mainstream auto companies)
Before Tesla, EV were seen as clunky, not very convenient and half baked conversion from already existing gas vehicles chassis. A few people were buying them because "hey, I'm saving the planet" (not really, this sweet electricity is still coming from fossil fuel).
But Tesla made EV kind of sexy, partly due to Musk other adventures that made people dream, mainly SpaceX, but also because of the "sporty" feel of the Model S and the "high tech" image of it.
Frankly, personally, I'm a bit tired of all the "praise Musk like a god" thing. The guy is refreshing in a world where everything must summarized in Excel and companies seems to be managed more by accountants rather than technology.
The guy is interesting as it has made the public aware of new possibilities regarding cars. He has unlocked something. But should we deify him for that? no. EV are still under conventional cars in term of autonomy, cost and reloading time. And personally, I think that cars are kind of a waist of resources to begin with: it costs a ton, it consumes a lot of energy per passenger and it's a mostly underused complex piece of machinery (few hours of commute every day at most).
Nowadays, I kind of wish for Tesla to crash and burn. I know it's irrational (and I will probably be down voted for it), but at least, it will shut-up the other irrational part of the audience praising Musk constantly and without reserve.
We might be able to sit in our seats and say that anything they’re doing as a company is a technology that has existed for many years, but for many years people were saying that the tech wasn’t feasible on a large scale for xyz reasons. GM could’ve made an electric vehicle (and they since have) but they weren’t going to make a fast charging network (and they aren’t going to). It was never going to be realistic to have one without the other, so Tesla did both.
Now the market exists where there are companies doing the infrastructure and companies doing the cars. I’d argue that jumpstarting the market was revolutionary. I’d also contend that we’re seeing the beginnings of the another market jumpstart from Tesla in the pairing of battery storage with renewables.
Battery operated cars, battery operation ecosystem, over the air updates, over the update that enabled auto-pilot, radically redesigned internal controls, direct sales, no haggle pricing.
I wouldn't even give Tesla the credit of revolutionary for connected cars. On Star has been in GM vehicles for a very long time (1996 says Wikipedia), offering various levels of connectivity.
There were smartphones before the iPhone, but they all sucked. Then Apple made a good one (with a touchscreen), and immediately the next year, every smartphone looked exactly like the iPhone. They changed the market basically overnight. But did they invent the smartphone? No.
Same deal with Tesla, if they can get Model 3 production up.
People don’t like to acknowledge this as an accomplishment. But the idea, and successful implementation, of selling an electric car that isn’t what a soy patty on gluten-free bread is to a burger was quite a revolution.
Being cool is a matter of marketing, not engineering. The Prius design is beloved in Japan, where it was designed. The Bolt is considered extremely stylish for a compact sedan/hatchback. And both are considered cool within their target markets.
Tesla doesn't advertise (yet). They have a huge word of mouth / Elon Musk cult following. He must be doing something right for getting all that loyalty and hype. A few hundred thousand preorders with 1000 bux down is something.
Tesla has had billboards all over LA and SF for many years now, so your definition of advertising must be very different from the normal definition. Any company following GAAP (standard accounting practices) would also consider stunts like the "flame thrower" and the Roadster in space to be marketing activities.
The only marketing activity Tesla hasn't actually used is paid TV commercials, unless you include Hulu or online streaming, in which case... Tesla has actually used every mode of advertising commonly used by its competitors.
The fact that Tesla doesn't actually categorize its marketing activities as marketing in its financial statements just means that Tesla isn't following GAAP.
I've seen the articles where you got that from... But that's BS. Just read their 10K (this is from Q4 2017):
Marketing, promotional and advertising costs are expensed as incurred and are included as an element of selling, general and administrative expense in the consolidated statement of operations. We incurred marketing, promotional and advertising costs of $66.5 million, $48.0 million and $58.3 million in the years ended December 31, 2017, 2016 and 2015, respectively.
Fair enough, so they put some marketing stuff as expenses and might have some visibility in the real world, but it also says in the same document:
Historically, we have been able to generate significant media coverage of our company and our vehicles, and we believe we will continue to do so. To date, for vehicle sales, media coverage and word of mouth have been the primary drivers of our sales leads and have helped us achieve sales without traditional advertising and at relatively low marketing costs.
Nothing in the car industry was EVER revolutionary. Cars have naturally progressed from earlier related technologies.
In fact I would claim most technologies are like that. Hence to call something revolutionary, it makes most sense to talk about who has made a radical improvement.
It is kind of like how there were smart phones before iPhone, but iPhone was such a radical improvement that it represented a revolution.
In much the same manner, the Model S was a revolution, similar to iPhone. It pushed electric cars significantly forward and set the bar for what an electric car could be.
The entire car is revolutionary. It’s like the iPhone - there wasn’t a new scientific discovery behind the iPhone but the sum of the product was clearly revolutionary. That much was obvious to anyone who picked it up in 2007 and had even the slightest ability to imagine a mass-market user.
On range, Tesla Model 3 only has a slight advantage over the Chevy Bolt, 20 miles based on third party test. Tesla isn't remotely close to 2x on range unless you're comparing the Model 3 to a first generation Nissan Leaf.
Speed? Definitely not 2x, unless you're actually claiming Model 3 can do >180 mph (Bolt is rated a 91 mph max speed, and the first generation Nissan Leaf had a max speed of 93 mph. Model 3 is rated at 140mph max speed).
Acceleration? Chevy Bolt is 7 seconds 0-60, while the Tesla Model 3 is 4.6 seconds. Not 2x.
Comfort? Almost tied. Advantage? Bolt, considering that the most expensive Bolt (with all options) is still several thousand $$$ cheaper than the cheapest Model 3 available for the foreseeable future.
Brand recognition? Depends on where you are. Everyone has heard of Chevrolet. Not everyone has heard of Tesla.
Design? Depends on what you're looking at. If you mean just the lines of the vehicle as they were intended, sure, the Tesla wins. If you mean the lines of the vehicle as put into production, the Bolt wins, hands down. The Model 3 is already infamous for misaligned panels, large gaps in the exterior between panels, misaligned window panes, misaligned handles, etc, and these are all signs of poor design failing to account for manufacturing tolerances. You won't find a Bolt with any of those problems.
So...basically what it boils down to is that Tesla is evolutionary at best in some categories and a step backwards in others.
I agree with you the Bolt is a good value. I am concerned with your comparisons. A car that accelerates 0-60 in 4.6 seconds is in a different league than a car that does it in 7 seconds. Sure it is not a 2x increase, but these aren't computer specs. Using 2x as a baseline of comparison is unfair given these machines are used for transportation. Perhaps only in range would it be fair.
Tesla's branding tactics are no different than the other auto manufacturers. I don't think you can ding them on that. They all play to an audience or niche, it can be heavy handed. It can be GM talking about trucks and ruggedness or BMW talking about performance machines. Tesla focuses on "innovation." It can all be a bit cheesy.
I am waiting for the rest of the autos to catch up and deliver their flagship electric cars. I think consumers will wise up and choose the car that provides the right value for their $. Porsche will price their upcoming electric sports car cheaper than the Tesla roadster...I would take the Porsche.
I read recently that (intended) sudden acceleration is a risk factor in automotive crashes. Can't find reference right now, but if that's true then faster acceleration isn't necessarily a feature.
But anyway, how often does the average driver drive around with their foot flat on the accelerator? I don't understand the obsession with getting to 100k/hr right now.
In these parts, most people seem to be quite content with a car that will get to 60m/hr in 15 seconds, as evidenced by all the Toyota Yaris and vehicles optimised for off-road four wheel driving. I imagine a Yaris owner would think it'd be nice if their tiny little car could do that, but it ain't happening without a set of tyres the defeats the purpose of owning a subcompact.
Try again. Frunks were around before Tesla...as far back as the 1960s. This is just an evolutionary iteration of modern cars, which are design to absorb crash energy in a crash by crumpling. And even Tesla admits that its higher rates are almost entirely due to the frunk providing a larger crumple zone.
Also, Tesla is known to misrepresent in it's PR releases. See https://www.cnbc.com/2017/07/06/insurance-institute-for-high..., in which the IIHS notes that Tesla only received an "acceptable" for the type of crash that actually results in the most fatalities.
The takeaway here is not that Elon/Tesla did anything “wrong”. I’m glad the article acknowledges that if Elon has listened to the experts he would not have started in the first place. And that by over-shooting Tesla may end up ahead even after the blips. Many journalists who point out the staff size disparity fail to note, unlike ARS, that Tesla outsources way less.
As someone who works in the car industry in a manufacturing plant, what surprises me is the lack of complexity in their product (especially in the powertrain i.e. battery and motors vs fuel tanks and engines) and yet their inability to ramp up their volume.
If you look at most production facilities (at least in Europe) there is usually one line which produces several models, in various body styles each with various different engine/transmission combinations and additional parts for specific emissions markets and climates. (That's just some of the complexity in the Powertrain system, there's more related to the interior, exterior, electrical systems, suspension, HVAC)
Tesla's model 3 line has virtually none of this complexity but still has issues in productions. Needless to say, if this was an established automaker producing this line, they would easily be up to full production volumes within months. (For me, this is part of the reason why people in the car industry are bearish on Tesla, they understand the challenge in setting up a whole new production line by a novice to the industry. By the time, Tesla have ramped up to a significant volume at good quality, the VAG, PSA/Opels, FCA, GM, Fords of the world would have their relatively simple (complexity wise) EVs designed and ready to ship (or almost ready to ship))
To me it seems like their current woes are a production quality issue or logistics and planning issue.
The Model S was their first attempt at volume production so should have been a good chance to resolve the basic issues, learn from their mistakes and build up good practices. The Model S was notorious for its poor panel alignment, squeaks and rattles and other minor quality issues but things had improved as the production went on. Dishearteningly though, it seems that this wasn't enough. The challenge of producing a low volume product like the Model S really doesn't compare to the Model 3's 6000/week ambitious target. Previously, they could ship a Model S and have their service centers fix all the production issues. At significant volume, you need first time through quality to be almost perfect.
For continuous production systems to work, all the processes need to keep to their 'takt time' and also be done to the correct standard (sufficient 'process capability'). Making sure your processes are capable is the hardest part and where your Six Sigma (Master) black belts are worth their weight in gold. The other thing they need is suppliers delivering on time and to spec. Logistics planning also is huge, with delays in supplier shipments potentially stopping the whole line. Supplier quality can also be an issue so having your own knowledgeable engineers working with them is key.
They have a huge task ahead and everyday they build below target, they are burning millions of dollars of revenue potential.
People made the same argument about Apple when the iPhone first happened. "Apple is not going to just come into this product space and get it. It's too complicated."
What we know now is that Apple had been investing in the supply chain and the manufacturing side of things for a very long time before they actually got into this business.
Tesla seems to me to be a company based on the idea of, "Yeah, Apple just went off and pwned the smartphone market, just like they pwned the mp3 player market. So, yeah! We're going to do the same thing. But with cars."
I don't really think that Tesla really understands the situation, the planning, and the investment it took for Apple to really do what it did.
Given that Tesla hired a bunch of car people to scale their car production, it's safe to say that understanding what Apple did isn't their forte. I suspect it wouldn't have helped; you can't cost-effectively send final assembly of a car to a contract manufacturer, but you can for a phone.
"Anyone who has had to do with the manufacturing side of car makers (German ones anyway) knows how highly optimized from end-to-end their processes are."
> For example, there is a tool which is essentially 5/6/7 torque wrenches coupled to a lifting arm. This tool allows a line worker to mount a tire with little physical exertion in seconds. You see this sort of thing all the time at car assembly lines; offloading mechanical power to the machine, while having a human line up and steer things.
I think it comes from a basic lack of respect or understanding of the actual skill requires to work an assembly line properly. Highly educated creative class are trained their whole life to look down with disgust on the people who do repetitive labor. The name shows it.. They call it unskilled. And yet how many of them could do it?
It takes an enormous amount of focus and concentration, the ability to learn complex routines quickly, to create efficiencies without breaking rythyms or rules. All while being payed less and treated like a second class employee.
Why is toyota different? Because they dont think like this. Line workers are thought of highly. Their ideas are integrated into the system. Most of all they are not disposable cogs
Fun fact: This seems like a very American thing to me (that is being imported into my culture). Where I live apprenticeships (2/3/4 years work and school) are frequently required for assembly work and the like, and has usually been seen as a solid way to make a good living. In the last 10-15 years there has been a pronounced rise of arrogance against journeyman ("Gesellen" i.e. those who completed an apprenticeship) and state-certified technicians ("Techniker" i.e. those who completed an extra two years of school on top of an apprenticeship; this usually entails a bunch of other additional qualifications and certifications); parents act like their children are failures if they don't manage to study at an university and put immense pressure on them.
Rationally speaking that doesn't make any sense; Gesellen and Techniker work full-time years earlier than anyone who goes to university and often climbed the pay ladder quite a few steps in those years as well. Depending on the profession they make good money. In some professions even a lot (e.g. industrial field service can pay six figures all things considered).
University and Abitur (the highest level of secondary school) aren't for everyone, but parents act as if it were so, often with detrimental effects to their children.
50% of all German school students graduating this year will be awarded with an Abitur. The changed view of learning a trade isn't just in the peoples minds, it's actually encouraged by the government. Thanks to the Bologna reform, the old way of going to a trade school is being looked down upon and graduating from college is the new default. Even when you are completing trade school, you are nowadays often awarded with a "Fachabitur" which again encourages people to go to college.
In my city for example the number of college students doubled within the last 10 years. Rents for small apartments got raised accordingly, obviously, and this put more pressure on blue collar workers to put their children into college.
Nothing wrong with exposing as many children as possible to coding. Forcing them isn't good, but making sure that in grades 1-4 they get to play good logic/coding games in computer lab time is a plus.
Even for basic assembly jobs? I doubt a time served mech apprentice would want to take that job :-) let alone a Techniker (which is effectively what I did albeit in a hyper specialised area Thermofuids which had been set up to cater for the local industry)
The tech world seems particularly self reinforcing of this mindset. If your immediate job title doesn’t sound appropriately “tough” then you’re a fool and ignored.
Here’s hoping I’ve just ran into a round of bad eggs.
Otherwise I’m glad to hear about jobs that cannot be automated away.
We have an "Enterprise Architect" where I work, with several minions under him. His primary role is to hold a wet finger up in the air, sense the air currents, then plunk said finger down into Gartner's 'magic quadrant' and select whichever company/technology it landed upon.
I think for Tesla, it is less about the workers and more about the entire design process at other companies that they're skeptical of. There are alot of items in the product design and manufactoring for GM, etc. that are there because "they've always been done", and it's difficult to tell from the outside what's important and what isn't.
Toyota's approach is different in that it takes in more input from people on the manufactoring line who are actually doing the work. It also helps that when they were getting up and running, their geography meant that pretty much everything they need is in a relatively short distance and they can make changes without needing to worry about long shipping times.
The "Big Three" which cares more about central control of the product and using their size to get everything done more cheaply. Their advantage is having the world's largest economy available to them so they have many suppliers to choose from.
Tesla's approach is that they're better at the technology and if you can throw more computers/robots at the problem eventually you can conquer anything. I think they're learning that this isn't always true, but it's still one of those cases where they're going to reach for the toolset they know best first.
The pen versus the hand. What you do manually isn't expanded on paper, procedures and books. Yet it does require finesse, patience, precision, attention.
There is a whole range of trades that go into making a car I doubt that any one working on the line is "unskilled" in the formal definition of the term, semiskilled is probably the term for a lot of line workers
But there is a difference between a time served machine tool setter and some one working on the line attaching wheels.
I agree entirely with the sentiment of the article, and I think Elon and Tesla are hopelessly deluding themselves with belief in their superior engineering and product design. But...
Just because a few companies failed at factory automation in 1980 doesn't mean it couldn't succeed in 2018. It's patently absurd to knock Tesla for trying it again in the era of ultra-high resolution cameras and advanced machine learning algorithms.
This is true. But a key point: Tesla doesn't actually have to beat state of the art in automation to be profitable. Their cars are already great, electric cars are simpler, they've streamlined the sales process in a way that already gives them at least 10% greater cost advantage if everything else is equivalent (while also eliminating a major pain point for consumers).
But it goes beyond just the fact that technology has advanced: Tesla is willing to pursue alternative approaches where others have long stopped (and this is a point the author of this article makes later on). The fact that Tesla doesn't have the conservative culture of other car makers means that they'll make mistakes like this, but also that they'll find new solutions (or recognize existing solutions in similar fields that haven't become industry standard) whereas others would not.
You cannot automate EXISTING car designs much. And Model 3 does make improvements, but in many ways is built similar. It's like trying to automate soldering components by hand. Automation of electronics required a change in the fundamental way electronics were built. Through-hole is hard to automate but relatively easy to build by hand. Surface mount is simple to automate and ultimately better in several ways, but is super annoying to build by hand. Tesla needs to find the surface mount of car manufacturing. And I think they're trying.
Something particularly hard to automate for automobiles is the wire harness. There are many degrees of freedom; it's kind of like tying your shoe: easy for humans, hard for machines. The Model 3 uses less wire than the Model S/X, but only by a factor of 2 or so. Model Y, on the other hand, is supposed to use like an order of magnitude less. This is like optimizing a circuit board for automation by only using a few through-hole components, relying mostly on surface mount.
So Tesla is going to have to continually redesign their vehicles to be more and more amenable to automation. They can't do this as a step function, and Musk has realized that now. (This is another good point the author makes.)
Tesla has yet to sell a car at a profit (using standard accounting practices, known as GAAP, for the auto industry or even in general). They're still losing money on a per-car basis, so the more they sell...the more money they lose.
And don't give me that crap about R&D blah blah blah. Other car makers include R&D and other related capex in their per-car profit accounting.
This doesn't sound right. As far as I heard they make something like 20% or more profit on sold cars but due to constant heavy investment they burn through cash reserves.
Tesla makes a 20% profit per car if you ignore standard auto industry accounting practices and use magic voodoo accounting and ignore the capex that goes into developing and building those cars. Notably, the only sites claiming that Tesla realizes per-car profits are tech companies which defend the use of non-standard accounting practices.
Note: Most other car companies make hundreds of thousands of cars at each of their factories, and sell hundreds of thousands to millions of cars each year, which substantially reduces the per-car capex. Tesla sells comparatively very few cars, so its per-car capex dwarfs the rest of the industry.
Except that Tesla is now producing at a weekly rate equivalent to hundreds of thousands of cars per year at their Fremont factory, comparable to other car companies.
But they're still ramping up. The factory will be outputting something like 500,000 to 700,000 cars per year after they're done ramping. They'll need Model Y and/or Semi and at least another factory before they're going to be comparable to other mid-range manufacturers. They'll need at least half a dozen factories to be a major manufacturer. Or they'll need to change how cars are made entirely.
If they can get millions of cars from a single factory, then the high automation that they're shooting for (and which Model 3 had to back off of) might make sense (and is required for getting that kind of production from a single line). But even that will likely need multiple factories to justify the R&D into the factory line itself.
Tesla is now producing at a weekly rate equivalent to hundreds of thousands of cars per year
That's only true if you're a time traveler from the future. But it's only April 2018 here, and Tesla is producing about 2600 cars a week. They produced 12000 cars during the first 3 months of the year.
It's an annualized rate of hundreds of thousands if you include all the models. For the model 3, 2600/wk makes for well over 100k/yr. Model S and X combined are going at an annualized rate of about 100k/yr, extrapolating from last quarter's 24,728 number. Altogether it's over 200k annualized.
It's not "magic accounting" it's looking at different aspects of the business. You can look at the cost to make each vehicle (in terms of components + labor) vs. the price you're selling it at, or you can look at the cost to run the entire business divided by the total number of sales.
Both are important, but if you're looking to invest in a business that's growing, you care more about the first number because you're assuming that if the business grows enough then the second number will take care of itself. If Tesla has been in business for 100 years and is unlikely to grow its base, like more established automakers, then you care alot more about that second number. And you also care about a lot of other numbers like the dividends paid out, how likely that something will change to move the stock price, etc.
I don't know accounting stuff so I can't say, but I wonder if they stopped every and all activities except building and delivering cars and fired all personnel not involved in those, would they stay afloat?
It doesn't make sense to compare Tesla to other car companies that way. The others have much higher volumes to spread costs on and they have been doing this stuff for over 100 years. That means the basics of car manufacturing has already been developed.
Tesla R&D isn't just about creating yet another model, but about creating the very first car model in that category that they have ever done.
I think it would be very odd if they kept having equally high R&D when upgrading their models.
Good analogy. With gas cars it seems that the engine bay has become so crowded, all the degrees of freedom are used up by making everything fit. Electric drivetrains give an opportunity to totally rethink the topology of the car with automated assembly in mind.
The failure isn’t in trying to automate. The failure is in expecting automating everything to go perfectly and therefore making the line troublesome to carry out when specific automated sub tasks fail.
I've known a few people who accidentally stopped an automotive production line. The plant manager will be there in minutes to see why the F$%* production is falling short to the tune of 5 figures per minute.
When I was younger I had a less severe but equally enlightening experience. A transmission case line with 24 CNC stations had our UI software running on PCs to make them easier to use. They could run fine without the fancy UI but it was how the operators had come to run things. I went in one afternoon to load a software update from floppy disk. The plant manager came out. "What are you doing?" "I'm updating this UI software." "You work for $company?" "Yes." "Is there gonna be a $company guy on site for 3rd shift tonight if something goes wrong with that software?" "No." "You aren't gonna update it then." I went home.
Go fast and break things doesn't work in that environment, and yes the critical processes work far more than 99% of the time because as you say there wouldn't be much output. It sounds like that's the exact lesson Tesla is learning now.
[edit] the reason all the processes on the line have to work is because it's basically impossible to store a bunch of cars (or doors or engines) anywhere until things get fixed. Ad-hoc fixes are just not feasible at that scale.
Agreed. It's been long enough to revisit the limits. Let's review SpaceX mistakes and compare them to the 80s, they probably did a lot of them. Nobody thought it would be possible to make reliable reusable low cost launchers. And here we are.
ps: not a Musk fanboy, just a middleground spectator.
The space industry isn't really comparable to the automotive industry. The former has been stagnating on government contracts while the latter has been heavily innovating for many decades due to massive competition and markets.
I don't think it's absurd either, but I think Tesla was the wrong company to try to do it in 2016-2018. Musk and everyone else agrees that Model 3 is a do-or-die time for Tesla, and the financial picture for Tesla looks dire as it is sitting on a sea of red ink from an enormous burn rate financed by bond debt. That is not a situation in which to take a likely profitable but unnecessary systemic risk which imperils the entire Model 3 program by its nature. If GM or Toyota tried an experimental lights-out factory, it's not going to potentially kill them. For Tesla, that's a possibility.
> "Far from leapfrogging the techniques of conventional automakers, Tesla is now struggling just to match the efficiency of its more established rivals."
I don't think this is true any more. Tesla switched Model 3 from high automation approach at the end of 2017 to a more conventional approach using human labor where appropriate, but still is ramping up quickly and seems to find using human labor very effectively. That doesn't sound like they're struggling with the more traditional approach, but rather embracing it successfully (for Model 3).
But I thought this line was good:
> "Instead of easing robots onto the line a few at a time, providing for inevitable debugging problems with redundant equipment, GM bet the entire Hamtramck production system on the proposition that leading-edge automation would work instantaneously."
Tesla does seem to be learning this lesson. And this is the right takeaway: learn respect for the way things are done, without losing the realization that something far better is possible. That reminds me of the ramp up of Falcon 9.
> "Musk likely could have spared himself a lot of short-term headaches if he had relied more heavily on auto industry veterans to warn him against repeating mistakes made by other car companies in previous decades. But if he had done that, he would also be less likely to discover ways to optimize the manufacturing process—particularly optimizations that work particularly well for a company specializing entirely in electric vehicles."
1) Consider an assembly line where all tasks and material movements are done by humans.
Over years, line operations are tuned to improve faster & better production.
Even the vehicle designs are influenced by the capabilities of the line.
2) Machinery is introduced incrementally, supplementing or replacing each
human task or material movement. Operations and designs continue to be tuned.
3) Eventually, a line has no humans - but the line design was created
around humans, and still inherits constraints for human interaction.
4) Now, imagine a line where the entire line is designed for robots from
the start. No humans doing production tasks at all, ever. Even the
vehicles being produced are designed to facilitate robotic production.
Henry Ford was level 1. Most modern US production is level 2, hoping to
reach level 3. Tesla hoped to reach level 4, settled for available tech
at level 3, and is having to fall back to level 2 in some areas. (IMHO).
Is anyone manufacturing entire cars at level 4? How about major components
(engines, transmssions, drivetrains)?
> Musk said he wants all parts of the company ready to prepare 6,000 Model 3 cars per week by the end of June, triple the rate Tesla has achieved in the recent weeks.
This sounds like madness, and I wonder if anyone believes them.
Has anyone made a timeline or visualisation of Tesla's promised production rates vs reality? They appear to be constantly off, and now they're apparently doubling down and promising a production target that seems even less realistic than the ones they already failed to achieve.
Auto final assembly lines make about a car a minute. Some run a little faster, some a little slower. About 500 cars per shift. 2 shifts a day, 6 days a week will get you to 6,000. Beyond that, you need another assembly line. Going to 3 production shifts is rare; you need some maintenance downtime.
This doesn't work like a growth curve. Once you get it working, an assembly line produces at a constant rate. Tesla's low output indicates serious downtime problems. They've shut down for some line rework, which is probably a good move.
There's still no discussion of the Fremont plant's problems from someone who has both access and the knowledge to understand what they're seeing. So everybody is still guessing. There's some CBS video, but it doesn't tell you much.[1] You need to see a video of a full cycle at the stations with problems, and they're careful not to show that. CBS shows 2-second clips of assembly stations but has detailed coverage of Musk's sleeping bag.
Robotic precision alignment by force feedback tends to be slow. Here are some DARPA videos of recent work they funded.[2] It's embarrassingly slow. (I look at that and think "ROS", which runs on Linux and is not hard real time, so you have to go slow.) Most of the stuff you see robots doing fast is preprogrammed motion. Sometimes, there's a vision system finding a target followed by preprogrammed motion to reach the target.
Apple used to have a highly automated factory in Fremont. Steve Jobs was very proud of it. Productivity wasn't very good.[3] They centered the factory around an automated storage and retrieval system, big racks with a mechanical system to put things in storage and take them out. This gets you a tidy plant, but not necessarily a productive one.
Back-of-napkin math suggests that it'd take about 2-3mo, assuming that they'll need to triple capacity to make their target and that they can triple capacity twice again before exponential growth must slow to logistic growth.
It's a peak rate, not a sustained rate. I think their peak rate has been about 2500-3000/week over a day (so about 350-425/day), so 6000 peak rate isn't totally out of the question. It wasn't long ago that they were only doing 1000 per week for the Model 3, so why (fundamentally) couldn't they increase it to near 6000 in a few months, too?
Remember this is also an internal target from an internal email, not supposed to be a public statement. You don't reach your high external goals by setting the bar low for internal goals.
> It wasn't long ago that they were only doing 1000 per week for the Model 3, so why (fundamentally) couldn't they increase it to near 6000 in a few months, too?
It was almost 4 months ago that they hit a peak rate of 1000 per week if this article is correct:
Right, and they hit over 2000 a week about 2 and a half months after that (end of March). Setting an internal goal of 6000 per week (burst) in another 3 months from then seems appropriate if they want to actually achieve 4000-5000 per week in that time frame.
EDIT: If their production ramp doubling time really is 2.5 months, then based on the Bloomberg Model 3 tracker's estimate of 2600 Model 3s per week right now[0], they'll have about 4900/week by the end of June: https://www.google.com/search?&q=2600*e^(2.26*ln(2)/2.5)
I just don't understand this implicit idea that setting lower internal goals would actually help them achieve a higher production level sooner. That goes against common sense and experience.
Or they could set a realistic and achievable goal, and not stress out their entire workforce (and not expose their workforce to excess health and safety risks) trying to reach a goal that they know is impossible based on their current process constraints.
Tesla’s existence itself was never realistic nor achievable, from the very start, a point that the author makes. Also, the goal isn’t impossible; in fact, the goal is close to their current ramp rate. Improbable is not a synonym for impossible, no matter how common that misunderstanding is.
As far as safety and health, I have seen no objective evidence that Tesla is worse than the industry standard range, although certainly they can improve. Media sensationalist coverage of anecdotes doesn’t count as objective evidence of being worse than the industry overall. Admittedly, this is a low bar.
They are only measuring physical injuries, not psychological burnout, then they are missing the point. You cant treat automotive manufacture like a AAA video game. Nobody dies if Arkham Asylum 5 is janky on playstation. But if you are missing bolts on your brake lines....
> As far as safety and health, I have seen no objective evidence that Tesla is worse than the industry standard range, although certainly they can improve.
You mean, apart from the investigative reporting at Reason that uncovered systematic under reporting of injuries.
This is a management technique perfected by Elon. No matter what date you set, it’s going to slip. There’s little reason to set conservative milestones.
This technique has its limits. If you squeeze enough people will be in a constant panic state that allows no forward thinking because everybody is in a rush to deliver now which then will predictably fail. I see that in my company. Every "6 month" project turns into 2 years but with the depth of thinking of a 6 month project.
I haven't been following the particulars of this storyline, but as I understand it Tesla is trying to metaprogram the automation of auto manufacturing. If it works out, everyone's estimates will be off like considering acceleration to be constant when improvements are being made at a higher level.
> At the same time, it's rarely a good idea to underestimate Musk. Musk has a long history of setting optimistic deadlines for his companies and then failing to meet them.
What? Don't underestimate him because he can't do what he says?
Maybe you should quote the entire paragraph instead of truncating it in a way that makes it look ridiculous. As I said in the next sentence, "Musk is persistent and a quick learner."
And then as I write later in the piece:
"Musk ignored the conventional wisdom, and he has gotten much further than anyone expected. He has sold hundreds of thousands of cars and has hundreds of thousands more people eager to buy the Model 3 as soon as it's available. Moreover, Tesla has had a huge influence on the broader car industry, forcing every major carmaker to take battery electric vehicles seriously."
I didn't think the rest of the paragraph helped your cause in any way. You appear to admit he fails at his objectives, but threw in an unsubstantiated claim that he's a quick learner. If he was a quick learner, he would have learned he makes outrageous claims, and tempered his statements.
The only question is if his milestones are actually any better than the milestones for the most of the market. Blitzkrieg only works if the other parties do not take it into account.
China, sure. I believe that. But the car standards are different. These cars are not available in the west because they do not meet standards of developed countries. (Not saying they won't, eventually! I expect BYD and other Chinese firms to eventually be competing in the West as well.)
My comment intended no racism whatsoever, just acknowledging the current state of the Chinese automobile market, and I acknowledged this was only temporary.
You are correct about buses, though. BYD is dominating everyone else in electric buses.
Although the article is focused on automation, I'd say almost all of the software industry is in a 20 year period of failing to learn much from TPS/lean (with no end in sight).
A friend of mine who worked at one of Elon's companies for a number of years was often flabbergasted by the hubris involved in a culture that seemed to insist on repeating almost every single mistake of the past in search of solutions known to those skilled in the art. Valuable time and very large amounts of money are being wasted on engaging in a constant game of reinventing the wheel.
I was told many people brought this up to Elon internally, often with materially negative effect for them. My friend never did, BTW, because he saw what happened and didn't want any part of it. He kept his mouth shut and moved-on on to a competitor where engineers don't waste time and money reinventing every last thing.
He said that Elon is a very smart and truly remarkable guy, but one who lives in an echo chamber within which nobody has the balls to oppose him. Elon seems to be known for unleashing verbal hell in meetings at anyone who dares, insulting them and generally chopping them up in front of their colleagues. Very soon nobody wants to say anything more than necessary, if that. Everyone is in fear of going into a meeting with Elon and getting shredded.
The above is an account I heard from a third party. I have no clue as to the veracity of these statements. The reader is advised to assume them to be false and conduct their own research.
>Musk is discovering that large-scale car manufacturing is really hard, and it's not easy to improve on the methods of conventional automakers. And while automation obviously plays an important role in car manufacturing, it's not the magic bullet Musk imagined a couple of years ago. Far from leapfrogging the techniques of conventional automakers, Tesla is now struggling just to match the efficiency of its more established rivals.
This was the money quote for me. Maybe this dose of humble pie will be good. Not for Musk but for the legion of hero-worshippers who believe that some guy who founded PayPal can just "think from first principles" and outwit nearly a century of manufacturing and industrial innovation.
who believe that some guy who founded PayPal can just "think from first principles" and outwit nearly a century of manufacturing and industrial innovation
I believe someone surely could, maybe Musk or maybe someone else. It does happen. The right smarts at the right time with the right money, etc. But were I to worship that hero I'd certainly keep my mouth shut until after he has delivered first.
"... automation works best when it's added incrementally to a production process that's already working smoothly. And Musk seems to have made the same mistake Smith did: bringing in way too many robots, way too quickly, leaving little time for testing and refining the process."
There's a certain naivete to the Tesla story. It has both strengths and weaknesses. I think Tesla will eventually be sold to an existing major car maker or just shut down.
"I think Tesla will eventually be sold to an existing major car maker or just shut down."
I think Musk will lose interest once EVs are fully commoditized, progress slows down and Tesla needs to be optimized to every detail. He seems to be more interested in "heroic" achievements and big steps. Maybe they'll find a manager who is willing to run a regular company and then Tesla will just be another manufacturer.
That fits with Tesla's stated plan, which is "to accelerate the advent of sustainable transport by bringing compelling mass market electric cars to market as soon as possible." Tesla doesn't have to be the #1 auto manufacturer, nor does it even need to survive. As long as EVs are "fully commoditized," the company achieved its mission.
It's hard to miss because they say it at least once during every earnings call, and it's been on their blog for around a decade.
Moreover, wouldn't investors be happy to read it, according to your original point? If Musk becomes bored with a company that isn't doing what he cares about anymore because it's merely producing in-demand EV cars under a great brand, and he hands it to someone with skills like Tim Cook who can actually make the company run smoothly and profitability, isn't that the best of all worlds?
and, beyond merely selling off their shares, investors who could prove that management pursued a goal which significantly obstructed a reasonable return on investment probably have a viable lawsuit against management.
the managers of publicly traded companies are supposed to pursue providing a return on investment. they aren't entitled to exclusively pursue other obstructive goals, even goals generally beneficial to society, company workers, the poor, etc. that's not what investors are buying into.
of course, that doesn't mean investors would sue and it doesn't mean the government would necessarily investigate such a company. but there would be reasonable grounds for a lawsuit...
i could easily be wrong, but, IMHO the hard challenge to being just another car manufacturer is that car manufacturing is already so competitive and dominated by large, well-funded global players.
how can small upstart Tesla compete with that? up until now, the answer has been: by having this unique, kick-ass, smart, rare, high-priced and prestigious electric vehicle.
the trouble is that the global majors are catching up and starting to offer their own kick-ass smart EVs.
my humble guess is that no small upstart can survive as just another car manufacturer. they need a special advantage, some secret sauce. i fear Tesla won't have that in the future.
And the only one I've seen that might be a Tesla rival is BMW, but they're not much larger than Tesla and by the time they catch up to Tesla's electric car tech fully (some time in the mid 2020s), Tesla may actually have exceeded BMW's production capacity (2.5 million per year).
And it's too bad. Toyota had a good thing going with Prius, but squandered a 2 decade head start by under-investing in pure electric and investing billions in the hydrogen dead-end. And GM has been under-investing in both the Volt and the Bolt, being content to just try to peel off a few other EV-centric customers from Nissan and Tesla instead of actually making a car with the marketing campaign and appeal of a mass market car.
Charging networks is another place where the other car makers have been falling on their face. They often talk up charging networks, but it's still not possible to buy a car with something like Supercharger capability (only 50kW is broadly available on the car side, but only 25kW is typical for actual DC fast charge stations in the real world, at least in the US). Meanwhile, Tesla has continued to expand the Supercharger network.
BMW knows how to produce cars profitably after having a much higher production and a company infrastructure paid for over decades. Tesla will also be profitable when they're producing a couple million cars per year. And, given the streamlined buying process, will likely have higher margins.
I think Elon enjoys challenges, he's a holistic thinker with control and ownership developing more pieces of the puzzle than just electric vehicles, and he's incentivized with up to $55B if Tesla reaches $650B value. I believe he'll get there.
“that if Musk had listened to the experts, he probably wouldn't have started Tesla in the first place”
Except he didn’t start Tesla, Martin Eberhard and Marc Tarpenning did. Musk soon joined, but that’s not what was implied.
I honestly expect better from Ars Technica. To me this is just more glorification of personality to god-like status. Yes he’s an impressive guy, but he’s never the only one singlehandedly making all the decisions, innovations, or executing the work.
Technically those other two filed the incorporation papers and started noodling around. It’s easy to imagine Tesla with Elon but impossible without. The “founders” didn’t get very far and the future was bleak.
I’m not saying he wasn’t important. He’s considered a founder by Tesla. I’m just taking issue with the statement that he wouldn’t have started it if... because he didn’t.
A common story in manufacturing is an anti-union story: labour pricing itself above what capital wants to spend. And a common sub-story in this story, is trying to use automation as a big cudgel to beat labour, especially organized labour with.
And, Elon Musk is said (by unionists) to be pretty anti union. Silicon Valley is said to be pretty anti-union. New capital, the Ayn Randeans, they're pretty anti-union.
So, I suspect there is another side to this story: If you want to try and use robots to avoid confronting your labour issues, you need to understand the value labour brings to an endevour. There is no dumb labour, there is only expensive and cheap labour. Cheap labour is rarely worth it. Expensive labour is often necessary.
Cutting corners includes trying to avoid having to employ people.
Jeff Bizos is also very anti unions. But, he automated very slowly. I think his work floor practices in the packing mills he runs is very bad, but if we discount that, it is perhaps far more sensible: learn how to run supply chain logistics with the minimum sensible automation, then slowly increase.
I suspect had Musk swallowed his pride, and his anti union and anti labour views, and employed people, and organised labour, he might have avoided this problem. Of course, he would have walked into another problem: organised labour is not cheap.
But then, a tesla isn't cheap either.
How does he make rockets? Does he use robots, or people, or a blend of both?
I have devoted my professional live to manufacturing automation.
From my point of view, what Musk tries to do is very hard, but if someone could do it it is him.
There are differences between the 1880 and the 2018, in particular AI(artificial intelligence) today is "out of this world" tech compared to 1880s. Today you could solve things just by brute force AI that was simply impossible in the past. And Musk will probably do just that. If I had hundreds of millions of dollars I will do too.
The great thing about automation is that once you solve something, it works forever. It takes way more work, of course getting it to work in the first place.
Musk has as his best gift his total lack of doubt like young people that believe that nothing is impossible. It is very typical for this young people to crash over walls as they repeat the mistakes others have done. At the same time, it is young people who change the world when conditions change(and making the same mistakes do not yield the same output).
“The reasonable man adapts himself to the world: the unreasonable one persists in trying to adapt the world to himself. Therefore all progress depends on the unreasonable man.”
I'm confused that Tesla did not hire these experts. I see Munro & Associates dropping criticism, and am baffled, given so much expert knowledge available that no one hired these guys or their competition.
Not dissing Munro & Associates, they seem highly competent - but i was amused that they literally have the conjoined triangle of success as their logo.
Why would you hire a bunch of critics? “Experts” are stuck in the past. It’s hard to find experts willing or able to let go of their baggage and apply their expertise to a current situation.
If you're trying to diagnose and fix problems, aren't critics exactly the kinds of people you want to to hire? Having a bunch of uncritical yes-men won't help you improve. When I watched the Munroe video, there's only one or two of their criticisms that I think a reasonable person would disagree with. Janky weatherstripping and doors that rub against the body of the car are bad regardless of any personal baggage.
In an alternative universe Elon-5000 is slowly figuring out that his plan to fully autoprimate his vehicle plants using the latest in primate tech- humans; slow and weak swarm entities that are extremely cognitively flexible, cheap, energy efficient (fueld by almost anything green), with built in error checking, is not working out. A mix of traditional robotics from the 80s and advanced primate tech will be most efficient.
My guess, based on the history of thousands of bankrupt companies past, is that a legacy auto manufacturer will buy Tesla's assets at an enormous discount and will serve Tesla's existing customer base as long as it's profitable to do so.
Large companies rarely die, even in "failure" -- they just become part of a different company or are resurrected by hedge fund investors.
I pulled behind a Model C a couple of days ago. Coukdn’t believe how poor the body panel alignment is, visible from 10’ away. There are no cars with these issues, not even economy cars.
TL;DR the time and investment it takes to perfect fully automated car production is still prohibitive and has a poor return on investment. GM learned this the hard way in the 80s and Tesla is getting the same hard lesson now.
You didn't read until the end of the article. For instance:
> "Musk likely could have spared himself a lot of short-term headaches if he had relied more heavily on auto industry veterans to warn him against repeating mistakes made by other car companies in previous decades. But if he had done that, he would also be less likely to discover ways to optimize the manufacturing process—particularly optimizations that work particularly well for a company specializing entirely in electric vehicles."
Seriously. I sometimes thing if these articles are doing for Elon Musk.
So Mr Musk is attempting something, let us call this X. If he simply do it, no one else notices, and the future potential investors might fail to notice...Same if he fails.
So he has to make sure that they notice when he succeeds. What better way than to make a bunch of articles that says what he is trying to do is impossible. Then if he succeeds, then the (artificial) voices that were saying what he is doing is impossible, ends up amplifying his success 100 fold. If he fails, then not much of his current reputation is changed.
So if you are handling PR for Elon Musk. Why wouldn't you make these articles happen?
Automation is not a black-and-white thing, either, as it is often portrayed. Sure, some things are fully automated (e.g. assembling the car body), but many things are partially automated. For example, there is a tool which is essentially 5/6/7 torque wrenches coupled to a lifting arm. This tool allows a line worker to mount a tire with little physical exertion in seconds. You see this sort of thing all the time at car assembly lines; offloading mechanical power to the machine, while having a human line up and steer things. (This also avoids a whole bunch of safety concerns you normally have when operating autonomous equipment in the same physical space occupied by humans)
Another example: The body of virtually any car is self-supporting (there is no frame), so getting it right is rather critical. To achieve both the high strength and low weight of a modern car body the steel is deformed very closely to the maximum deformation it can withstand. This needs very good matching of the process to the steel. So, the manufacturer of the steel coil tests each coil individually for its exact properties. For mass production lines only coils meeting very tight tolerances are used. For low volume lines the process is adjusted individually for each coil based on the manufacturers testing and some in-house testing.