Can Elon Musk and Tesla Reinvent the Way Cars Are Made? | Hacker News Reader

Can Elon Musk and Tesla Reinvent the Way Cars Are Made?

nytimes.com

38 points_fizz_buzz_77 comments

Orthodox automakers have reasons behind their methods that aren't legacy bullshit: they're the product of years of practical experience in production line design: a finely-honed way of cranking out lots of an identical product at a cost that should be recoverable from a normal rate of sales. There are still duds every year, when a particular make's particular model is clearly the worst-received out of a large field of dozens, but by and large this process works very well.

Musk has thrown out the book entirely, and has taken the opposite step at almost every stage. This is a mistake, because some of those decisions have real consequences that will hurt later.

For example, most carmakers make lots of an identical product, from the same batch of parts, for a particular run. If a part (or a manufacturing process) turns out to be defective, they know that all of that run could be compromised, so they recall a batch of VINs. While a portion of these cars is probably fine, recalling a broader set allows some vehicles to be checked and serviced before problems happen or people get injured. If Tesla tweaks the process very often, the cars instead become practically hand-made, to where there are barely some that are alike. This makes it much harder to proactively identify risks about production siblings.

As someone who spent almost half his career as a software engineer in automotive companies and the other half in a couple of large tech companies, I could completely understand why there is a high chance of this happening.

Most automotive companies are run like factories, including the hardware, software, and R&D departments, in a factory, the focus is more on the process than on achieving the “correct” results, which although might work great in a factory could be extremely hampering for innovation in development and technology teams.

If you ever wonder why infotainment systems in most cars are stuck in the 90’s UI it’s because it would take weeks to change a single line of code where it would need to go through multiple code reviews, get approved by people who’ve been using the same development tools for the last 20 years and consider IBM RTC as cutting edge, document every single thought that went through the developers head and mindlessly test the crap out of every possible combination.

The way tech companies operate is extremely data-driven, if it works and you have the data to back it up that is usually the end of the discussion. However, in process-oriented companies conformance to the established processes is what is usually valued, a statement that is usually echoed is that it’s more important to follow the process and produce incorrect results than not follow it and produce correct results.

For the last decades, automotive OEM’s have been bringing manufacturing processes to software development, my hope is that any tech company (and Tesla being one of them) would be able to bring agile software development practices to manufacturing.

This is offtopic insofar as it’s meta but: the HN attitude of telling me why everything won’t work is really depressing. If I took many of the top comments on here seriously: pretty much every new technology or breakthrough is nonsense. Surely doing things that are new and challenging mean some hyperbole, but it doesn’t make you sound smarter when you’re constantly telling me why things don’t work.

>why everything won't work is depressing

That's funny, because I find the Silicon Valley attitude of "no one knows how to do anything but us" equally depressing.

Tesla is a great example of this hubris.

As far as I can tell, that's what the people running the site want it to be. It's been a problem for years and years now and no action has been taken to stop it.

Why should it be stopped?

I prefer critical analysis to unthinking yes-men. To the extent that HN administration agrees with me, I appreciate them.

I'm surprised how many people on HN has a problem with this ide...Engineers are trained to be pessimistic opportunists (for good reasons IMHO). The thought process is always This doesn't work > ask myself why it doesn't > prove my wrong and show me how it works. There are many foundational rules and laws that are easy to disprove especially when it comes to something mechanical. That's one of the beautiful thing of mechanical and manufacturing engineering. Knowledge from the past tend to be relevant for a good reason.

That's reality. Most pie in the sky ideas won't work because of fundamental constraints of reality. The trick is to find the leveraged areas where you actually can do something groundbreaking. E.g. solar roadways was a ridiculously stupid idea but solar shingles was brilliant.

I'm unsure what it is exactly Tesla's doing that's supposed to be such an earth-shattering heart-breaking staggering-ten-quintillion-Einsteins-level genius unprecedented changes-the-course-of-history-forever thing.

As far as I can tell, the thing they are struggling to do is... build and run a car factory.

And this runs through everything Musk does. His more successful ventures like SpaceX have basically consisted of taking something where someone else already did all the expensive fundamental R&D, doing some iterative improvement on it, and promising to use economies of scale to bring down the price. Except, because it's Musk, it's buried under a Mount Everest of breathless superlatives praising him for doing it.

Dude's trying to build a car factory to churn out cars that are incremental improvements on the prior state of the art. A rolled-up combination of Galileo, Copernicus and Newton he ain't. Maybe he'll succeed at it and make some money. Maybe he'll get his lunch eaten by people who know more about (and don't disdainfully sneer at existing knowledge of) building and running car factories. I have a hard time caring, but HN seems to worship the ground he walks on, and I simply don't understand why.

Find reasons in the present to appreciate life. Anticipation of new things shouldn't be your sole source of happiness. Criticism of flashy and ambitious but unworkable ideas shouldn't make you depressed because there should be plenty that already exists that keeps you happy. It's a big world out there.

Elon’s just turned 47. He slept at the factory last night. Hope he gets 700 Model 3’s off the line today.

>Hope he gets 700 Model 3’s off the line today.

Personally, I don't see the significance of putting all-hands-on-deck to reach an arbitrary production number that they themselves set. Quality is already a question, and I can't see that improving under such circumstances.

I dont know if you have been following Tesla closely, but they choosed to addreess quality in the beginnng, thats why the started with very very low production numbers.

I don't know if you have been following Tesla closely, but: https://twitter.com/bgluckman/status/965664986298306561

That was actually my point! The tweet is from febuary, and there have been varius quality issues, acknowledged by Tesla. They haven taken it to heart, and waited to ramp production till now. (Not only because of quality issues)

At least part of the point is building/repairing credibility with the capital markets. Elon Musk has made numerous black-and-white promises to the capital markets and has come up (well) short on most of them. Having previously promised that they would become cash-flow positive and not need to raise further capital, it would be unfortunate if they came up short on that promise AND short on significant production number promises leading up to the next capital or debt raise.

I suspect another significant motivation is more local leadership: set the tone of aggressive milestones for the company and then lead the company to achieve a goal that few thought was possible.

Disclaimer: I generally admire Elon Musk and his achievements, most specifically Space-X, though I am short $TSLA right now as I don't see any way they're going to navigate positively through what's ahead of them in the next 8-12 months, and don't think that Tesla as a car company is worth more than GM or 1.4x Ford.

"At least part of the point is building/repairing credibility with the capital markets"

In order to do that it would probably help to not cut off investors asking legitimate questions with snide comments to then devote significant parts of that call to answer questions by a fawning Youtube "personality" fan during an investor conference call.

I think we will see record recalls for Tesla after frantic build up. These type of burst production rarely ever goes well when you are shipping large scale hardware with thousands of components that require proper QA.

Musk thinks its like shipping software and wants to exit with a golden parachute. He doesn't care if ppl get killed by Model 3s or that they are putting down money for a junk and a lengthy recall process. He's a giant egomaniac with little regard for the safety of it's occupants.

IMO Musk last bullet is to crank hard through this bottleneck enough to get enough air to have another shot at redesigning car production.

Tesla cut jobs recently, removed some options (internet), opened the model 3 payment period. With the truck thing and other side money intakes I think he's just trying to cross the survival line investor agreed upon.

Maybe he was "changing production processes on the fly"... as the article mentions. Really funny how being rich and powerful can make your failures look 'revolutionary'.

There needs to be a corollary to Betteridge's law, to the effect of "It's a stupid question."

Tesla is not trying to reinvent the way cars are made. Car manufacturing lines already rely heavily on automation. What's notable about Tesla is that they have been trying to automate production to a higher degree than is typical for the industry.

Don't get me wrong here, this is a big deal, and it's caused Tesla some problems that they could have avoided. But "more automation" is not revolutionary and it's not reinvention.

Betteridge's law would suggest that the answer is "no". In fact, they have already failed to reinvent it, and are reverting to standard practices in some sense.

This. too many Tesla cults shilling online. The fact of matter is experienced and established brands are coming out with superior quality EV.

In fact Tesla is ill equipped to handle the flood of EV competition. Once Elon is forced to leave Tesla, it will be the end of the brand.

There's just no way I would put down 70k for a Tesla when there are so many superior options coming down in the next 2 years.

I'd even reckon Tesla won't be around within 5 years.

The automotive market is filled with energetic visionaries who ultimately crushed by established automotive giants. Tesla is no exception in fact it will be far more crushing in that they basically paved a way forward for EV and other brands will fully reap the profit, not Tesla.

Such is the cold reality of a saturated race to zero market. Tesla is not even a luxury brand like BMW or Lexus. It's steaming pile of shit that kills drivers on autopilot.

And yet without them we probably wouldn't have EVs today.

Just like how, without Apple, we wouldn't have had rounded corners.

Pretty much the only reason any car manufacturer has ANY EV program is because of the threat of Tesla.

Bob Lutz said w/o Tesla my Volt (or now Bolt) would not exist.

The regulatory pressure federally and by California to decrease emissions plus incentives to buy EVs matter too.

The car companies just made compliance cars (and created astroturf org) to battle that. Until Tesla shamed them, they never really had to make serious effort.

Shaming is very different than presenting an actual threat which some people claim Tesla was or is.

When I say "shame" I mean the consensus from the auto industry was/is:

1) EVs are slow 2) EVs can go far on a single charge 3) EVs are ugly 4) Batteries degrade a lot over time 5) EVs are expensive.

etc, etc, etc

Tesla has knocked out most of those bullet points.

> The regulatory pressure federally and by California to decrease emissions

That pressure is dictated by the administration in power and the current administration has already started the process of rolling back the prior administration's regulations by changing CAFE requirements. In addition to that they've also questioned that Waiver that gives CARB it's authority. If successful they would strip CA of it's regulatory power.

> plus incentives to buy EVs matter too.

And those expire. Once Nissan hit's 300k units of the Leaf sold in the US they'll lose their credit and there will be very little motivation for people to purchase it. Unless they're able to significantly reduce the MSRP or dramatically increase the batteryt capacity, I wouldn't be surprised if the ceased production shortly after hitting 300k units.

That seems unlikely, considering that the first "Golden Age" for EVs was in the late 1890s and early 1900s.[1] Musk was the first to use lithium-ion batteries for a highway-legal vehicle, but is that really a unique innovation that wouldn't have occurred to anyone else?

[1] https://en.wikipedia.org/wiki/History_of_the_electric_vehicl...

The 'innovation' is that car companies did not (and still do not) want to make EV cars. There's so much inertia behind ICE vehicles the incentive to make EVs is small.

1) Oil companies don't want them. 2) Dealerships don't want them. 3) Automakers don't want them.

Tesla's energy has forced the issue.

> The 'innovation' is that car companies did not (and still do not) want to make EV cars

Car companies want to sell cars that can make a profit. They're willing to wait for the technology to improve to the point that that is feasible.

Unfortunately our planet isn't.

Ah, so we’ve crossed from the “it’s impossible” to the “it was inevitable” stage of disruption :p

I "reckon" we'll revisit this comment in 2 and then 5 years.

This may be true, but this is also what people have been saying for years. All the other auto manufactures do is release cool concept cars at car shows and then go back to producing the same watered down garbage.

Right now there isn’t an EV that even comes close to a Tesla. And even the traditional luxury brands like BMW and Mercedes are losing to Tesla on the tech side which is increasingly the part of luxury that people care about.

Everyone else keeps talking about these imaginary cars that other manufactures are going to make that are going to put Tesla out of business. It’s been FIVE YEARS of that talk… and nothing.

The iPace from Jaguar is a really good start however it still suffers from:

1) Dealerships: they still hate selling EVs. (My personal experience - my GM dealer spent hours trying to convince me to not by my Volt).

2) Charging: Tesla has the only serious world-wide charging network.

3) Software: Telsa (for good or bad) updates their software all the time. People get 'goodies' every few months. My Volt, only a handful of VERY minor software updates in five years. People are coming to expect that.

And the biggest problem: Jaguar plans to make only 20k cars in the first production year. I hope they have plans to increase production considerably, or it will be another token electric car.

That's funny, because last year Tesla produced only 3000 or so Model 3s.

And in the first quarter of 2018 they produced:

Model S: 11,730 Model X: 10,070 Model 3: 8,180

The argument was that Jaguar wasn't serious because they're only making 20k of their electric car in the first year of production. I pointed out that Tesla hasn't produced 20k of any of their cars during the first year of production. That's noteworthy hypocrisy.

We know that established car manufacturers can ramp up quickly if they want to. Can Tesla? So far, the evidence indicates they cannot.

Tesla also didn't have the decades of experience and existing production lines to work on.

>We know that established car manufacturers can ramp up quickly if they want to.

The evidence so far indicates that they don't.

I don’t think bmw or Audi are losing to Tesla in technology. They both have pretty good autopilot now and each has its own cool tech features, like the 360 camera in bmw or the digital dashboard in Audi.

I am sure bmw or Audi can release a fully electric sedan to compete with Tesla, but most people I know who own bmw or Audi - they don’t like the idea of a fully electric car. They want a sense of security in having a gas engine. I guess the market is just too small for bmw and Audi to even try and go compete with Tesla at the moment. Maybe later...

Yes, I'm expecting that it'll take a while for the market as a whole to get used to the idea that they don't actually need a gas engine.

So I expect that EVs with range-extenders to launch first, but there still doesn't seem to be any major commitments along that front.

This isn't 2010. It is settled, range-extenders have a very limited self-life. I've had a Volt for five years, love the car. For me and most of my fellow Volt owners, our next (and then on) car will be a full EV.

> our next (and then on) car will be a full EV.

Yep, but I think many others will have to go through the same journey before they believe in having a pure-BEV.

I haven’t had an Audi, but I have had many BMWs and a VW (which, UI wise, is the same as an Audi). It’s not really close to a Tesla. The digital cockpit is OKAY, but the car still has 50 buttons and it’s buggy. The infotainment system resets itself all the time. Audio glitches with car play. It’s half assed and never updated like everything else with the auto industry these days.

this

I'll admit Tesla has a few quality issues to iron out but jeebuz, every modern car has HORRIBLE software implementation(s). Horrible UI, buggy infotainment, resets, etc.

At least with Tesla you'll get constant upgrades and updates. Tesla's choice of almost no hardware interface, again, mimics the iPhone. Before that, everyone crammed a crappy keyboard on the phone that it turns out you don't need. "But how will I compose emails!?"

> But how will I compose emails!?

Slowly, and not-at-all in the rain being the answer.

I'll never buy a car that doesn't allow me to turn on the air conditioning without taking my eyes off the road. In other words, physical buttons and dials that are always in the same place. Something I can feel with my fingers and always rely on being there.

Touchscreen controls sure are flashy, but as far as I'm concerned when employed in a multi-ton piece of machinery traveling at 60+mph they're borderline homicidal. I know I'm not the only one who thinks this way.

I'm open to voice controls. I don't like speaking to machines, but having machines speak to me is something I appreciate (for instance I always have my GPS talk to me, I never look at the screen) and I could see myself coming around on talking to machines. But a big old glorified ipad stuck in the dash? No way.

Aren't there physical controls for simple things like A/C on the steering wheel, and the tablet is just for more involved workflows? That was the impression I got.

If the controls for everything that I need are physical, then I don't need the digital controls.

The climate controls aren’t ideal on the Model 3, but “Auto” does the job 90% of the time. As a complete system the Tesla UI is everything that BMW and Mercedes has tried to be for the last decade with its stupid wheels and touch pads and other gimmicks. Audi/VW is probably best in class in terms of a streamlined and feature complete set of cockpit controls but it’s jus so damn buggy. Doubly so in the US where you have to go to a dealer just to update the software.

I much preferred the fully manual controls on the late 90s early 2000 BMWs without nav to ANYTHING they have put out since. Mercedes is even worse with multiple different input gimmicks, all of which are hard to use. It’s like every year they make it worse. Somehow.

Tesla can improve with voice commands and better mapping of steering wheel controls but everyone else really just needs to scrap their entire UI and start over because their current software is just 15 years of monkey patches from the first navigation control models in the early 2000s.

This.

So much talk about fantastic new options coming to market "real soon now", and I'm all for that. What seems to be missing is the orders for batteries (cells or packs) for any large scale manufacturing of an EV line.

That'll be the actual indicator that someone is finally taking EVs seriously, and not just using it for a marketing opportunity.

That hydrogen future Shell has been pimping for 20+ years.

If I had a dollar for every fuel cell car I see in a year I'd have... a dollar.

I too saw a single one this year.

Tesla has been at this for fifteen years. The competition is coming, but it takes time.

Cool, so I’ll look for BMWs answer to today’s Tesla in 2028. I’ll just mark it down on my calendar.

>This. too many Tesla cults shilling online. The fact of matter is experienced and established brands are coming out with superior quality EV.

I'm afraid to say that today even Chinese do.

For the under $30k niche, Chinese automakers completely aced the process.

On other hand, all Chinese "Tesla killers" to be, came out with really awful cars which, according to what first owners say, are prone to rapid disassembly after few minutes of driving in bad road conditions.

Chinese cars, not built to the safety standards demanded by other nations or built to the bare minimum at best, tend to be man-killers not Tesla-killers.

In regards to performance, it is not a problem at all to make a car with Tesla like horsepowers to weight ratios. But after doing that, you still somehow need to make a driveable car.

Few people who drove Nio ES8 for example can't hold expletives for the suspension setup, and extreme body flexing that makes doors to not to close after a single rough ride. Were not for their NDAs, Nio's creditors would've been calling back their loans now.

Tesla has one durable advantage: supercharger network. There is nothing remotely comparable and it limits the utility of their competitors. Until that changes they have a pretty good moat.

BP just bought a large charging company, while Tesla has suspended their program.

Honestly,how hard do you think it is for a gas station to provide chargers? It isn't rocket science.

I googled for “Tesla suspend supercharger” and got no relevant looking hits.

Are you saying Tesla is putting supercharger rollouts on hold? I just listened to the latest shareholder meeting and I don’t believe I heard anything like that. Usually they’re always expanding that network.

Most gas stations don't have the real estate. It takes 20 minutes just to charge to 50%. Charging needs to happen where people are already going to spend time while their car is busy charging; home, work, commuter parking lot/garage, mall parking lot, etc. Highway rest stops with gas stations and fast food could add chargers in the parking lot and adapt to people eating there instead of only stopping long enough to pee and get take-out/drive thru; it will make long road trips take longer but not acceptably so.

Most gas stations will realize they have the real estate once they calculate how many additional hotdogs those bored EV drivers will buy while they're waiting for their charge.

They bought the network to retard it's growth at best.

https://en.wikipedia.org/wiki/Patent_encumbrance_of_large_au...

The problem is gas stations are not exactly high profit businesses, so I doubt they’ll want to pay to install them.

Unless BP wants to eat the cost the way Tesla did it might take a long time to get a good network of chargers available.

There are chargers here and there but the fastest of them are half the speed of the superchargers and that makes a big difference. And the superchargers are methodically placed to enable long trips. It’s not about the free energy — Which is what has suspended — It’s about freedom of movement. Teslas can replace a gas car for almost everything; without that infrastructure no other EV can (yet).

I keep seeing this touted in every Tesla thread. But then every time I go look it up, the closest supercharger network to me in a metro area of ~4m people is like a 45-60 minute drive away. There's like 2-4 total chargers in the entire metro area, and they're all way out there.

The total capacity of superchargers for this metro area of 4m people is like a couple dozen charging spots.

I don't see why this is a relevant part of the discussion.

Supercharges are for people traveling long distances and need to charge up quickly after using up their battery that was fully charged over night. People in a city need to have a charger at home or at work for daily use.

They are not for routine use, they enable long distance travel, which is pretty impractical without them.

They don't reinvent as much as keep the hype machine running well, and they are doing excellent job at that.

This world runs on hype. Tesla's and Apples are good at it.

After that, whatever they do is 'reinvention'.

Reply on news.ycombinator.com