Here’s a company, Hadrian, which is planning to build a series of factories for manufacturing of precision metal components. Their first factory is in Hawthorne, CA, and they’re building the next one nearby in Torrance.
One of the lead investors in Hadrian is the VC firm Andreessen Horowitz. The A-H partner responsible for this deal, Katherine Boyle, writes about the company, the opportunity, and why she considers it a promising investment.
Hadrian is hiring, if anyone’s interested.
Also, a thoughtful piece from Ms Boyle on the need for America to get serious.
48 thoughts on “An Interesting Startup”
Great stuff. I suppose they are in LA for the aerospace market. Can’t think of any other sane reason for a company to try and do business there. Some great possibilities when the programming and new tech manufacturing come together. We can absolutely outcompete China if we have young people with decent basic education and work ethic. IF.
Yeah, aerospace has to be the reason for LA initial location…a little surprising that factory #2 is going to be so close to factory #1.
Hope they are going to develop an apprenticeship & training program. A quick look at some of the open-job requirements suggests a refreshing absence of credentialism.
Right next to SpaceX, which doesn’t seem likely to be a coincidence. I used to drive right past it, trying to avoid using the 105. Does seem like a stupid place to be, presumably it’s only to see if the overall company can work, and they’ll hightail it out of CA as soon as they get any traction, most likely with Elon to TX…
That is cool. Someone could make a boatload of money getting this concept developed into other industries as well.
You can buy CNC machines quite cheaply now. With the skills to program them, you can learn a great deal about how modern machining is done.
I see enthusiasm, but nothing concrete here. Can anyone explain to an old man WTF this is all about?
I thought the material presented was vague, but I’ll fill in a few gaps with speculation.
Design software has come a long ways. Computers need a fair bit of processing power to run it but that has gotten ridiculously cheap. So you can design very elaborate parts and test them virtually to determine weight, stress points, motion conflicts between moving parts.
This would be fun but not useful, except that manufacturing has jumped quite a bit forward at the same time. Can’t find old school machinists? They are still important, but instead of manually producing a part on a mill or lathe you can program the CNC machine to make it. And then make 1,000 more. Laser cutting is another great tool. Ludicrously accurate. Even 3D printers are finally approaching what has long been promised. Using some of the newer materials you can make things super light, super strong and in quantity.
Its a new world. China can still crank out a zillion bits of crap parts to go into disposable products. But if you need 1000 of something at very high standards they can’t compete on price, precision or speed.
I work with students in these areas, but I’m far from the master of any of this.
Yeah I know. I can program Arduino and of course any *nix based machine, but I see nothing here that is in any way new.
Looks like a scam actually. ;)
A fair number of cool sounding but vague notions are just that. T
Dan from Madison: “Someone could make a boatload of money getting this concept developed into other industries as well.”
A-H is always looking for business plans:
Interesting. Another immigrant like Elon Musk. From Australia. After the recent election, I expect more fleeing from their insane government, Even worse than ours.
There have to be 10,000 shops all over the country trying to do the same thing. It’ll come down to whether they can deliver. They’ll be competing for machinists and machine programmers with all the rest. You have to wonder how many of the people they’re after want to stay in Southern California.
Right now they seem focused on cheap and easy to machine aluminum, I’d want to know why I should trust they wouldn’t turn a $15,000 chunk of super alloy or titanium into a big piece of scrap. Especially when there are other shops that have been doing it for years. Producing parts to spec, not scrap that is.
Good luck to them but this isn’t an app where they can spend millions programming something and sell it millions of times for free. They can throw a lot of money at different things until the venture capital runs out but I don’t see anything unique. There are a lot of buzz words but nothing that hints at how they plan to solve actual problems. There’s already a lot of automation in the process where there can be.
A couple of additional articles on Hadrian:
The second article notes the failure off an earlier manufacturing startup, Plethora, which was funded by two of the same VCs now funding Hadrian…they attribute Plethora’s failure to an attempt to automate *too much* rather than to seek the best balance between humans and hachines.
Quit both articles as the straw men started multiplying. Do you really believe that SpaceX or anybody else is just going to drop off drawings for pieces worth tens of thousands, that their entire business depends on, and then just walk away twiddling their thumbs, waiting for them to show up on their dock? It just doesn’t work that way and never has. If you want a complicated part delivered on time and on spec, you had better start talking to whoever is going to make it when you start to design it. It’s really easy to design something that no known process can produce, or if one can, it might be horribly slow and expensive. You’d better start talking to your material supplier too, unless you want to find out that what you need is either simply not available or has months or longer lead times.
Back in the olden days, engineers used to be forced to take machine shop classes to give them insight into some of the difficulties that they would encounter converting their bright ideas into real stuff. That is no longer the case.
Both articles are just regurgitations of press handouts by people that wouldn’t know a drill press if it fell on them.
MCS, why would you assume they are just expecting SpaceX (for example) to just drop off a large stack of drawings and expect the parts to show up, rather than following a sales process in which they would develop an actual relationship with the customer, who would initially give them a few parts to try out, and then take it from there? This obviously isn’t a mass-marketing play, and I don’t see anything in the info suggesting that they think it is.
Andreessen Horowitz is a very credible VC firm, with a lot of successful exits, ditto for Founders Fund. Of course not all of their investments work, but that’s the nature of the game.
“Do you really believe that SpaceX or anybody else is just going to drop off drawings for pieces worth tens of thousands, that their entire business depends on, and then just walk away twiddling their thumbs, waiting for them to show up on their dock”
I’d bet dollars to donuts that this company already has an in with SpaceX. They’re literally across the street, and Andreessen Horowitz has a close relationship with Elon–they put a bunch of money into his twitter purchase, for example.
I’ve read all the links and looked through their web site. I simply don’t see that they offer anything out of the ordinary. They’ll be using the same tools and tooling as their competitors and dealing with the same constraints. What will they bring to the table that no one else can?
“A fair number of cool sounding but vague notions are just that. T”
From where I sit, I don’t see anything but the hundredth iteration of what gave us the dotcom bust. I’m sure there’s room for another well run shop to make money in that market, what I don’t see is the sort of unrealized margins that keep VC’s in private jets and hand made wing tips.
There’s just something about this that raises my hackles for some reason. Whatever happens probably won’t take long, that’s a very expensive neighborhood to set up shop and I’m not talking real estate. There’s lots of opportunities to burn money and lots of money to make as well but the difference is less than a hairsbreadth.
Back in the olden days, engineers used to be forced to take machine shop classes to give them insight into some of the difficulties that they would encounter converting their bright ideas into real stuff. That is no longer the case.
That reminds me of my days as an engineer in Douglas AC’s wind tunnel. One day we got a new model to test in the 4 foot tunnel. It was an engine nacelle and the engineer had gotten the math wrong. When the tunnel was started up, heading for about Mach 0.8, the center cone in the nacelle model started going UP the tunnel against the wind flow. Everybody who saw that grabbed something stationary to hold onto. Seconds later, the cone came back at Mach .8 and hit the window for Schlieren photos and broke it. The wind tunnel decompressed into the building with winds at Mach 2 or so for a second. Fortunately, the building was designed for this and the roof lifted on rails to decompress the interior. I don’t know what happened to the engineer.
Here’s a fairly short video that explains how even simple seeming details can make a part anywhere from ruinously expensive to totally impossible to make. The provider is a site that specializes in education on CNC, or as I call it, machining porn.
MCS, true that the info on the website & in the articles is pretty skimpy. I certainly wouldn’t invest in a company without a lot more detail on what they plan to do & with what resources. I’m pretty confident that the VC investors *have* seen a lot more info about the company…not sure what the reason is for such a minimalistic approach to describing the company on the website and other materials.
Took a look at LinkedIn, where they have 43 employees listed, to see what kinds of backgrounds they have. Director of Operations looks pretty solid–was director of machining and of Dragon production at SpaceX, previously quality engineer at GE Aviation (jet engines). Principal Software Engineer did software work at Google, previously was a mechanical & structural engineer at Boeing, worked with titanium. Head of Programming has been a CNC programmer and programming department manager. Factory Lead has about 7 years experience as a machine operator and machinist.
So, doesn’t look crazy from a background standpoint, they’re not consumer software people or consumer marketing people or anything like that.
Something messed up with the YouTube link, points to ‘calling Russia’s nuclear bluff.’
Interesting, long winded and off topic. Let’s try this instead:
I’m not much for HTML but you can check a link in the comment box by highlighting, right clicking and choosing open in new tab. Didn’t do that and lost track of what was on the clipboard. Could have been worse.
I suppose I was just disappointed. I really wanted to see what they thought would make that much of a difference and all I got was hand waving.
One of my first jobs was running a lathe in a shop run by an engineer that a few years before had been machining the combustion chambers for the F-1 engines on the Saturn V. You might say I’ve had an interest in this for a long time.
Very interesting video. If these design engineers had ever watched their part being made…or even if they’d gotten cost feedback–“hey, these things cost $2500 each to make, isn’t there something we can do to make them less expensive?” –I’d think the would have seen the error of their ways.
Like I said, the smart ones ask. It’s usually just that simple. It’s an even better idea to keep thing like aspect ratios in mind when you’re designing a part.
You can look at tool catalogs, If you see that what you want will require some sort of exotic tool or fixture or machine it’s a sign you need to stand back and decide if you really need that detail because you’ll pay for it in money, drastically limiting the shops that can produce it and usually time. And then, asking at this stage really does wonders because you’re machinist not only knows more about machining than you do, he probably has some ideas you haven’t thought of.
I got caught out on this a few years ago when I needed a lot of long fairly narrow holes in aluminum. I absolutely needed the holes but had to completely re think the part before I found a shop willing to try to make them. In the end, I designed the part around the tools they had.
Since this Hadrian shop looks like a SpaceX shop in all but name, I think we can assume the usual economic incentives don’t apply. If Elon’s not happy with the current options, then all bets are off about how they “should” operate…
Here’s some more:
These are the small compressors used to raise natural gas pressure to the 3,500-4,500 psi for fueling vehicles. The larger ones used on natural gas pipelines have cylinders you can crawl into. Notice that they manage to actually produce these in the U.S.
Tom Brown, who has run and been otherwise involved in a number of UK industrial companies, wrote a book about his experiences, Tragedy & Challenge. One of his first jobs was in a Scottish forging company. He said the manufacturing engineers, who designed the sequence of dies needed to make a particular part, did not like to go into the shop except when required by some crisis. Understandable, in that the presses were so loud that they could be heard up to 12 miles away!…but he found this disconnect very detrimental to productivity.
Lux Capital, one of the investors in Hadrian, talks about their reasons:
One of the links is to this post:
I am very very very skeptical about this statement from Lux Capital:
6. Perhaps most importantly, our view that space has limitless potential for economic growth has only expanded and intensified over the past years.
The primary “customer” for space “companies” is still governments, and we are entering a time of severe economic uncertainty, so I don’t think this is a particularly smart bet…
Hypersonic missile production is going to require the same kinds of precision manufacturing needed for space systems….big overlap among space, missiles, and high-performance aviation.
“He said the manufacturing engineers, who designed the sequence of dies needed to make a particular part, did not like to go into the shop except when required by some crisis.”
When I was young I worked at Alcan Kingston, where we made wing struts for American fighter aircraft. We did this because our ancient extrusion press, a real monster, could do things no other press could do. I was part of the No 1 Press crew for quite a while. We could extrude 25′ of an X section, pop the dies and extrude another 20′ of a rectangular section, which was all one piece. I got to know the tool and die guys as they came to visit all the time.
The plant is still in business under the Novelis group..
“I got to know the tool and die guys as they came to visit all the time.”
Might have something to do with the fact that they are still in business…
California? The tithe to Sacramento would be too much for most companies to open shop there nowadays. Mostly those that have been there a long time remain. Their viability in question. IMO.
Given the plethora of “Boards” of this-that-and-the-other, it is a wonder that anything ever gets built, or for that matter, ever opens their doors. Once opened, they are then under the thumb of the CA bureaucracy to tell them how to run their business, what insurance they must provide, and how much they must pay their employees, and only those that are of the approved mixture to meet the local ‘palette’. “Once hired, never fired.” I doubt Space-X will be remaining where they are today. Mr M could have them on trucks on their way to TX overnight… some day.
I’d be willing to bet that the Scottish engineer’s reluctance to go onto the shop floor had as much to do with the British concept of class as noise. from the end of the war till maybe the mid ’80’s, there was considerable antagonism between management and labor in most British companies. Anyone in a suit that ventured out of his cloister was likely to encounter some accident that ruined his suit at least. Those forty years saw the end of a lot of British businesses too for some reason.
Forging is mostly done by presses now instead of hammers, so things are a little quieter. Forges and foundries are still maybe the dirtiest places in the world.
“I’d be willing to bet that the Scottish engineer’s reluctance to go onto the shop floor had as much to do with the British concept of class as noise. from the end of the war till maybe the mid ’80’s, there was considerable antagonism between management and labor in most British companies.”
Yes–Tom Brown emphasizes the problems that class-consciousness has caused for British industry–in this book, published in 2007, he talks about it as a continuing problem.
David F: “Tom Brown emphasizes the problems that class-consciousness has caused for British industry–in this book, published in 2007, he talks about it as a continuing problem.”
Brown’s book was mainly pushing the view that the UK leaving the EU (“Brexit”) was a bad idea and would lead to an even further decline in the UK relative to Europe. Of course, it can be difficult to see whether that prediction was correct when the UK and the EU are both in a green-washing, virtue-signaling dive to the bottom.
Governors and legislatures need to create the right economic incentives to restore manufacturing. So does the national government obviously. Do it without tariffs.
Jeffrey C: “Do it without tariffs.”
Yes and No.
If a US manufacturer is simply inefficient — outdated capital equipment, bloated over-paid staff, management buying back its own stock to goose their bonuses — then tariffs are not appropriate. We should not be featherbedding the inefficient.
On the other hand, what should we do if a US manufacturer is fully efficient but cannot compete on price with an import from a country where the manufacturer uses under-paid child labor, does not have to comply with environmental regulations, has no obligations for workplace safety? The only options there would be (1) for government to roll-back excessive regulations, or (2) tariffs to impose on the foreign manufacturer a cost equal to the avoided cost of not having to comply with US regulations — thereby leveling the playing field to let the most efficient producer win.
The ideal answer would probably be a flexible mix of (1) and (2) — but the chances of the bought-and-paid-for US Political Class delivering that are approximately zero.
Yeah, I remember freshman economics–tariffs are bad!, if Country A and Country B each make sprockets and widgets, then it’s optimal for Country A to only make sprockets and Country B to only make widgets, and they trade with each other for the other one, blah blah blah.
It was totally deranged, and anyone who might feel like jumping in to say that well akshually economics is way more complicated and accurate than that isn’t going to change the fact that American smaller cities and towns as well as urban cores are rubble, and all the “elites” care to do is wring their hands and say, gosh trying to unwind the past 30 years of disastrous “free trade” policies that led us to this place just isn’t possible.
The only consolation is that there’s no money left for them to impose any more stupid and destructive ideas.
See TJ Rogers in today’s WSJ:
I’ll comment later.
Our politicians disingenuously campaign for conversion to solar energy, but their propensity for top-down economic controls is forcing American homeowners to pay $2.65 per watt on average to install a residential solar system today, according to Clean Energy Associates. The equivalent fully installed residential solar costs are $1.50 in Europe, $1.25 in Australia and $1 in India—because these places practice, and get the benefits of, free-market capitalism in their solar markets.
Is any part of the above paragraph incorrect?
FWIW, here’s a breakdown (from 2018) of costs for installing residential solar systems:
One of the unmentioned costs of residential solar is additional roof maintenance.
I suspect that Rogers is missing differences in the way they are subsidized. Here, the initial cost is subsidized mostly by tax credits so the home owner pays then collects from the government. I Europe and Australia subsidies might be at other levels. I find it unlikely that the cost in Europe of anything is that much less than here. David’s link shows that the panels themselves only account for 25% of the installed cost so It’s practically impossible for them to account for that much of a difference.
Left unaddressed is the even more massive subsidy paid by electric rate payers on an ongoing basis after installation.This is both direct by massively overpaying for power sold back into the grid and indirectly paying the cost associated with meeting demand when they are producing nothing.
Then there’s the slippery concept of trying to determine just what a “watt” of solar power installation consists of. Doing a quick look, and finding an appalling mathematical incoherence, the numbers published seem to be based on peak, fully insolated output which for most installations never happens. The actual area needed to provide a given amount of energy over a year is going to vary wildly from place to place. If anyone other than the government was paying, you can be sure the numbers and the guarantees would all be out in the open.
I get maybe 8 calls a day from solar salesmen, unsolicited and unwelcome. I actually had a nice young guy come out and give an estimate but he told me that I needed to replace my old air conditioning unit to cut electricity cost first. I did so and it cut the cost in half. Then, since I decided I had finally found an honest solar guy, I would call him for a new estimate. Now, I can’t find him. Maybe he decided it was mostly a racket and is doing something else.
You can do a couple things with solar panels. If you just want to charge batteries its quite simple and all you need are panels and a controller.
To actually power a house or generate AC is quite a different proposition, and requires an inverter/charger setup and other equipment if you need to interface with the grid.
One is cheap, fairly easy to set up and useful if you just need batteries charged. The other is expensive and may, or may not be actually useful.
TJR: “Clean Energy Associates reports that for every $0.35 per watt the American consumer paid in 2021 for Chinese solar panels, the U.S. solar industry collected another $2.30 per watt to mount them on houses, convert their erratic DC current to grid-legal AC current, and hook them up via the “internet of things” to the sophisticated Network Operation Centers that monitor the performance of millions of American solar systems.”
“Mount them on houses” is basically similar to any other home remodeling work…respectable work requiring some skill, but I don’t see what installing solar panels on houses in inherently more value-added than other home-improvement crafts.
“convert their erratic DC current to grid-legal AC current”…he’s talking about inverters. So the *revenue* for the inverter goes to a US company, but how much of it is value-added? I’d be surprised if a high % of the semiconductors in that product aren’t made somewhere other than here.
I’m very curious about what the current estimates for lifetime, efficiency, etc., for solar panels is, actually in the field and not according to boosters. I was promised decade lifetimes for LED bulbs, and except for the ones in my lamppost that never get turned off, I can’t say I’ve noticed they last longer than conventional bulbs in my house. I recall reading somewhere recently that the performance of solar panels degrades pretty severely much sooner than is usually advertised by proponents…
“the performance of solar panels degrades pretty severely much sooner than is usually advertised by proponents…”
Indeed! It is called “sundown”.
Comments are closed.