Is It Impossible To Start a New Car Manufacturing Company Today?

That’s even more true in this particular case because the Model X doesn’t have a traditional gull-wing door. There’s an extra hinge at the corner. Which means that the door lifts almost straight up, and requires almost no lateral space.

Tesla is a “tech” company because of the way it’s run, not because they have some special breed of engineer. The same is true everywhere. Engineers are engineers, but their work is frequently crippled by poor management practices.

There’s a story here about Tesla’s VP of vehicle engineering, Chris Porritt. He came from Aston Martin and was an experienced automotive engineer. The difference is in the work culture.

Other automakers have been around for a century, and during that time they’ve learned some good lessons, some bad lessons, and some obsolete lessons. It’s worth pressing the reset button from time to time, even if it means you have to relearn some obvious things.

Other manufacturers can already use Tesla technology. Tesla has already opened their patent portfolio under a “we won’t sue you if you don’t sue us” license. As far as I know, no one has yet taken them up on the offer.

SAAB fans seem to have far too generous memories of what it was pre-GM. They had two models, both of which were riding a creaky old Fiat/Alfa Romeo platform. There wasn’t much “tech” they had which made them stand out, their appeal was selling unique, “quirky” (meaning “ugly and goofily proportioned”) cars to people who cared about that. The problem is that the automotive market has become so capital-intensive that just “being quirky” doesn’t cut it anymore. You either need volume and variety (so weakness in one segment is propped up by other strong segments) or you need to be doing something so new and compelling you can pretty much name your asking price like Tesla. SAAB had neither. GM transitioning their cars from the Alfa 164 architecture to their Epsilon platform didnt make them less of a SAAB, except to obsessive fanboys who care about unimportant details.

There’s an old saying, “It’s easy to make a small fortune making cars. Just start with a large fortune.”

Called “smeerps.”

Tesla is like every other new-gen company: It likes to believe it’s Really Different, and more importantly, fans and investors want to believe it’s Really Different.

It’s not. At least, it won’t be by the time it becomes more than a niche player. Except that it never will, and when it’s a GM marque it will be even less different. Just like Saturn. Remember Saturn? Saturn didn’t do a single thing the traditional way, 'cuz management and efficiency and newthink. Saturn slowly sank into the GM morass and vanished.

Yeah, I remember how Apple was saying that the iPhone was going to be totally different and change everything, and the CEO of Blackberry said there’s no way those idiots in Silicon Valley are going to walk in and eat their lunch, because they have no idea how the phone industry works or what the customers want. And they were right, because Blackberry is still the reigning king and Apple is a tiny subsidiary of Nokia.

Ya think there might be a smidge of difference between a marque that from the very beginning was a subsidiary of GM vs. one that actually was started by a Silicon Valley entrepreneur? That, for all their talk, it might actually be hard to get away from the old GM way when you are joined at the hip?

The correct way to determine if a company is doing something different is to actually look at their processes, corporate culture, and so on. Not to listen to marketing drivel. It’s not like any company is going to claim to be old and stodgy.

to add to what I said just above, SAAB had long been too reliant on other automakers. When their 2-stroke triple became unviable in the '60s, they had to buy the Taunus V4 from Ford of Europe. they had to use the Alfa 164 platform for the 9000, I was wrong about the 900- that was still trundling along on an architecture dating back to 1968.

SAAB Automobile was doomed in the '90s. GM just gave them a 10-year stay of execution.

cars are not goddamned iPhones. the two situations are not even remotely comparable.

Well, if you’ve got some stunning insights into Tesla’s corporate culture, please do share. If not, at least quit with the “they’re different because Silicon Valley” horseshit.

By my reading of the Saturn story, that has more to do with GM’s management flaws than Saturn’s. I’m hesitant to be too rosy about Saturn’s prospects if it had been independent. Realistically, most car companies don’t make it.

But being a subsidiary of GM gave it no chance at all to succeed. Early on, 40% of their sales were to current GM owners. For an independent Saturn, that’s good news; it is only bad insofar as GM intended Saturn to help the overall GM brand compete with foreign imports. If GM had half a brain, they’d have given up the independence fiction and simply pulled the plug in '95. Instead, they increasingly stifled what was there in Saturn. By the 2000’s, Saturn’s whole line of cars were essentially GM brands with a different label.

GM is currently doing the same thing by making the Cad-bui-let models virtually indistinguishable from each other. It’s right there in their commercials: “That’s not a Buick!” No… no, it’s not.

You realize that we’re talking about a car whose major claim to fame–aside from being electric–is that it eliminates most non-critical controls in favor of a giant touchscreen interface?

You could start by reading the article I linked to. It’s still pretty thin on details, but it’s a start.

Among other things, Tesla has a very tight feedback loop between engineering and manufacturing. This is not exclusive to Tesla, even in the auto industry (the Germans are supposedly pretty good about this), but nevertheless they do embrace it at a very deep level.

Vertical integration is another area that Tesla has embraced much more aggressively than others. Virtually the entire story of auto manufacturing over the past several decades has been one of outsourcing. That has some advantages but you lose out on system integration benefits. It’s why no one, as of yet, has matched Tesla’s battery technology: not because it’s so superior technologically compared to everything else, but because it’s designed as a system. If you buy your battery packs, cooling system, and so on separately, and just do the final assembly, you end up with an inferior result.

The big automakers know this about ICE powertrains, which is why those are largely not outsourced. They haven’t figured out yet that in an electric car, the battery is part of the powertrain and also needs to be engineered from the ground up.

and what has that to do with it actually being- you know- a car? because there’s a lot (a hell of a lot) more to designing, building, and selling a car than the fucking center stack.

BS. they damn well know that. GM figured that out back when Tesla was still hacking their stuff into those pieces of junk they bought from Lotus. That’s why the Volt is the only PHEV with as much EV range as it has.

except “vertical integration” was how Ford Motor Co. got started over 100 years ago, and you’re acting like Tesla is “innovative” for re-discovering that. Here’s a hint- a lot of important stuff happened before you were born. It’d behoove you to understand that a lot of what people and companies are doing now has been done before.

And I’ll just say that a car is a hell of a lot more than a drivetrain with body panels. Software is a huge part of any car today, and not just the control systems. Consumers want a car that works like their smartphone. We’ve had years of famously shitty front-end interfaces from the likes of BMW and others. And then Tesla just went in and got everything right on basically the first try, because they actually consider software to be a first class citizen instead of something slapped on at the end or outsourced.

All 38 miles of range. Yay.

I’m well aware of this. And yes, I think Tesla is innovative for rediscovering this after a century. Quite a lot of innovation–almost all of it, really–is just recycling old ideas in a new context (or even a not-so-new context). Nothing wrong with that.

I think it can be done, but it would require radical new technology.

The new technology you would need is that you would need a form of ultra-advanced manufacturing equipment so the new car plant could fit into the space of a high school gymnasium and only cost a mere 100 million bucks or so. Junkyard scrap goes in one side, new cars comes out the other.

This is basically science fiction now but it is probably feasible.

You would also need automated design software that could replace the efforts of a thousand engineers and let you get the same work done with 10 or 20.

In such a world, where almost anybody can make cars, it would mean that scrappy upstarts could compete. Sort of how today, a couple college grads can use 3d printers and off the shelf parts and make a product that would have taken a large company to make 20 years ago. They go on kickstarter and get enough money to move forward. There are a lot of failures but there have been some notable successes.

And that as because SAAB as too small-they could never realize economies of scale. Just developing a new engine costs billions-hich tiny SAAB never had,Even Volvo had problems, despite being 10X the size of SAAB.

That’s it; I’m sure someone is going to tell us how 3D printers will be able to do that in the not too distant future!!!

The machine wouldn’t look anything like a modern 3d printer, and would be stupendously more complicated.

There’s a lot of books and articles talking about the theory behind something like this, but in essense, it’s a machine that you feed it pure gasses that are chemically bonded in a way such that you can bond the element the gas is carrying with something else. There’s a gas for 10 or so different elements at a minimum, a fancy nanoforge would have every element on the periodic table that can be obtained in significant quantities and can be bonded this way (hydrogen can’t). Isotopes are tricky and most “nanoforges” won’t be able to separate isotopes.

The gasses get made via a stage that first separates the input junk to individual elements and then chemistry makes the pure gasses.

The actual factory then, in a very long series of steps, combined the feedstock into larger and larger bits of machinery, built from the ground up. A “practical” nanoforge might work by creating human cell sized cube like robots. They might even appear silver to the naked eye, through probably not. Each robot has little wheels or legs or something and can drive over other robots. You pour the robots into the assembly bay and give a squirt of data to each one that tells it where to go. Each one asks neighbor robots that are already in place which number they are and finds it’s way to it’s spot. It then fuses in place, and others follow. Some of the cubes are specialized and have dedicated manipulation surfaces on a single face of the cube, these form into the actual functional systems of the car.

Nope.

It is not about the technology used to build the cars - River Rouge, Saturn, Viper, Toyota, Tesla or Bugatti. It’s most especially not about applying tech-industry thinking to the problem.

Cars are a specialty product with a nearly unique market/application combination. How they are made is the very least part of that complex equation. That’s why nearly every startup company and a good number of established-elsewhere sellers have failed miserably in the US market. Only those appealing to a particular niche that will put up with the shortcomings of, say, a Bugatti, Lamborghini or Tesla will survive. Not because of much of anything to do with the cars themselves - there’s a proven market for nearly every fine category of personal transportation. But because people want things that only an established, trustworthy and reliable sales, support and service network can provide.

And you can’t run one of those off with a 3D printer on VC with a tech guru at the helm.

I agree, Barbarian. I was thinking aloud what I felt would be the minimum before the big manufacturers don’t eat the lunch of the small ones simply because they can make a comparable vehicle for a fraction of the cost. If Toyota wanted to make a high performance vehicle comparable to a Bugatti and they charged customers manufacturing cost + a modest margin, the car would probably only cost around $100-200k. (Toyota doesn’t because even a mere 6 figure price tag is more than almost all car buyers want to pay, and nobody needs Bugatti performance any more than people need a personal fighter jet)

But yeah. I could see some far future world where the whole auto industry is dead because there’s a better way to do things (underground vacuum personal rapid transit pods or something, and the surface is a nature preserve where people walk) and the only new autos being made are made by tiny shops using advanced 3d printers to make equivalent’s to today’s cars.

This is true right now - you can still get parts for early steam cars, and the parts are apparently made by a single craftsman.

They did, and the price was almost double what you stated.