Slashdot Mirror
DARPA Pays $3.5 Million For New TechShops and Secret Reconfigurable Factories
pacopico writes "Businessweek reports that DARPA will pay for the creation of two new TechShops in Washington D.C. and Pittsburgh. The $3.5 million deal includes 2,000 TechShop memberships for military veterans and will have DARPA employees performing top secret work at night. 'The project is called iFab. For a month, a given factory might use dozens of machines to make parts for helicopters. Then you reboot the software controlling the machines, and out come the parts for the drive train system in a tank. The Darpa workers at TechShop will try to figure out which tools and methods can be used to rewire factories in this fashion.' Maker mayhem."
Is this a hedge in case China decides to stop making shit for the US? Or plain ol' pork?
Confucius say, "Find worm in apple - bad. Find half a worm - worse."
Haven't got to the article yet, but in the summary I keep reading...
Then you reboot the software controlling the machines, and out come the parts for the drive train system in a tank.
I still don't get what a reboot has to do with this. Is it running Windows?
Flexible bare-metal recovery for Linux/UNIX
Always link to the printable version in the future please! http://www.businessweek.com/printer/articles/26828-techshop-paradise-for-tinkerers (still have the splash, but then it's one page not 5 or whatever).
"Factories like this" being quickly reconfigurable to manufacture a variety of products? You don't see any potential commercial uses for that? Do you really not understand the push toward manufacturing on demand?
Nowhere in the article is any mention that the DARPA employees would be doing TS work.
Periods processing of the sort required to do TS work at night in a facility used by civilians during the day is basically impossible...
retrorocket.o not found, launch anyway?
Whether we want a mechanical army which is within reach or maybe a clone army that could fight better.
The project is called iFab.
Cue Apple trademark lawsuit in 3... 2... 1... :-)
It must have been something you assimilated. . . .
"The only time you need factories like this is n preparation for a global conflict."
If there were a prize for "Absurd Asserted Conclusion", your post would win it.
Foxconn reconfigures by issuing orders to hordes of expensive workers.
US manufacturing could reconfigure taking advantage of technology, and negate the Chinese labor advantage.
"This post is an artistic work of fiction and falsehood. Only a fool would take anything posted here as fact."
The only time you need factories like this is n preparation for a global conflict.
Is the DoD expecting WWIII?
That's kind of their job. Expect the worst, hope for the best.
These are the guys who make contingency plans for just about everything, so on some level, yes, they probably are expecting WWIII.
"My God...it's full of trolls!"
"Don't reconfigure at the touch of a button."
Yet.
"This post is an artistic work of fiction and falsehood. Only a fool would take anything posted here as fact."
So...a factory that can more quickly and efficiently adapt to changes in demand? That can, instead of needing mass layoffs or closing up shop entirely, reconfigure their processes and retrain employees (increasing their skill sets if they ever need different future employment) to produce different things? Moving suppliers one level closer to being able to swiftly and effectively respond to the economic climate?
And all this research is only going to cost $3.5 million or so?
If they can make this work, and can be spread to other US suppliers, that $3.5 million investment will be paid back in no time in economic development. Hell, if it's a significant enough improvement, it could eventually help revitalize the US manufacturing industry by significantly upping our competitive advantage.
Foxconn reconfigures by issuing orders to hordes of expensive workers.
Expensive workers? When did Foxconn leave China?
Have gnu, will travel.
It's more than a bit concerning that the most flexible, agile, and innovative part of the economy is the military.
Any one else think we need CARPA - the Civilian Advanced Research Project Agency? Preferably one that has nothing to do with the government.
Yep, nothing DARPA has ever put money into has ever been licensed for commercial use. Everything they put money into is immediately used for war. Seriously? I'm could come up with a nice long list of links for products that are used commercially today that were originally funded by DARPA and a much longer list of DARPA sponsored projects that never came to fruition at all, let alone ever got used by the military.
Sounds like standard CNC capabilities.
When I was working for Boeing, a decade ago, they were transitioning from fixed jig assembly to laser coordinate measurement driving floor mounted hydraulic positioning equipment.
The benefits were:
1) No more huge jigs. Need to adjust a setting? No need to mod the jig, just tweak the s/w.
2) Eventually, each assembly line could handle any model. Just punch a button and the jacks position themselves to hold any body section.
3) Everything was modular, floor mounted and relatively compact. Union problems? Just load your production equipment into a couple of shipping containers and move it to a more hospitable environment. Any large building with a flat floor will do.
Have gnu, will travel.
The human COST is far more then the sum of wages.
Good-bye
I always had this idea that somehow the DIY concept in the hands of frontline troops would dramatically reduce the cost of our weapons systems and result in more effective and practical equipment.
The current procurement process is:
1. DOD compiles a bunch of specs (the people compiling them are usually bureaucrats or desk general, who are always refighting the last war.)
2. Give specs to a bunch of greedy civilian contractors (who over promise vaporware).
3. Then designed by civilian engineers and scientists whose lives don't directly depend on the product.
So it is in a way the blind leading the blind, and the process takes years or decades. We've known just as long that this process is broken beyond repair.
Now if frontline soldiers who are in harms way, have an effective means to "evolve" their own equipment and weapons on-the-fly in direct response to the changing tactical situations. Soldiers are always the best hackers of their equipment, and they have been hacking since war was invented. But it has always been impromptu and crude improvisations.
Supposed there are fully equipped hackerspaces with CNC machines and raw materials close to the frontlines that soldiers have access to. Some of the more complex gear may require dedicated design and manufacturing facilities, but these can probably be handled directly by more sophisticated hackerspaces at the homefront.
This approach may not result in fancy gold-plated bad-ass looking stuff, but most of the effective gear isn't glamorous looking at all. It just needs to be effective, practical, and custom fitted to the tactical situation and the enemy at hand. It will save the taxpayers a heck of a lot of dough.
Why Pittsburgh and DC? DC is where all the bureaucrats are and as far as I know, Pittsburgh has no major military research laboratory. Why not put the lab near a major military research organization that does actual hands on research and would actually be interested in using these services?
> reboot the software
*eye twitch*
It's a hackerspace for feds.
Is the DoD expecting WWIII?
Yes.
But the technologies involved do have civilian applications. I'm thinking just-in-time manufacturing, small-lot manufacturing, boutique manufacturing, that sort of thing.
Understanding the scope of the problem is the first step on the path to true panic.
If they could pull a FANUC and go to lights-out manufacturing, they would.
"This post is an artistic work of fiction and falsehood. Only a fool would take anything posted here as fact."
I have a TechShop membership, and have spent a lot of time there. What goes on there is mostly not all that high-tech. Most of it is hobby artwork. Some people are repairing cars. Others are making furniture. The electronics facilities are basic and little used. Much of the machine shop usage is by pros from companies nearby that need some machining done.
At times, it's rather pathetic. iPhones and iPads are made in China. Here in Silicon Valley, we have people making bamboo cases for them, and cheap plastic things to hold them on dashboads with suction cups.
Quite a bit of work was done on this back in the mid 1980s. The versatile factory capable of quick redirection came along with the concept of inventory taxes. The notion being that if a factory could convert from making fishing reels to brake assemblies or whatever in a few hours then many product lines could keep going with almost no inventory in storage. It was going fairly well back then but there was an issue with the price of the help needed to keep everything in order back then. That was mostly due to very inadequate computer systems. The machinery involved was also costly but it did work.
War is coming...
Of course it is. Not much of an observation since there's always a bunch of wars going at any point in time.
My take is that someone in the US government is exploring ways to deal with the foreign supplier problem and other supply disruptions. For example, in any military dispute (not necessarily including outright war) between China and the US, it's going to be a no-brainer for China and allies to withhold goods that are important to the US military. Then such a factory can compensate for that supply disruption. Or as in the case of the Iraqi invasion, a monopoly supplier of small arms ammunition simply couldn't keep up. Such factories might be quickly retooled to cover supply problems (until traditional manufacturers can enter).
Yes, there are obvious military applications here. I don't see it as an important piece of some US plan of war however, unlike say the widening of the Kiel Canal prior to the First World War was for the German military (which was instrumental in German naval planning since the largest German ships couldn't otherwise use the canal, a serious vulnerability in case of war).
War is either always coming, happening, or just ended.
In the land of the blind, the one-eyed man is usually crucified.
This does come up against something I've been trying to work out the numbers on:
- what's the $ per hour figure for running such a machine?
- how does that balance against the efficiencies of ganging up elements (when possible) for production?
I've begun making wooden cases for my archery gear, and have the tools to do all the cuts efficiently (save for routing out the stopped dado / groove in the end pieces):
http://www.3riversarchery.com/images/Contest2010/WilliamAdamsTakeDownCase.jpg
It doesn't take long to cut a set of dovetails once one is practiced at it (and one can clamp multiple boards together to cut several sets of tails at a time --- pins need to be referenced off the matching tails and cut individually), and drilling some holes in the right place is just a matter of a template/jig which can be flipped over --- if I get an electric router then each stopped groove is a quick pass w/ the router (once I build a jig to place it in).
Will I be able to make a machine like a Shapeoko pay for itself for straight-forward work like this?
- how long does it take to mount a piece for cutting?
- how long does it take the machine to make the cuts?
- how much clean-up will said cuts require?
I'd love to have a CNC machine to try out, but am still a bit dis-heartened that Shapeoko's envisioned $300 price point comes to $649 for a full kit at www.inventables.com
William
Sphinx of black quartz, judge my vow.
The amount of capital outlay required to do something like that is immense. It's simply not cost effective except for certain product lines.
As manufactured products use more and more specialized materials (because of weight savings, usually), the manufacturing becomes more and more complicated.
Back in the day, blacksmiths could manufacture a wide variety of metal parts, each one unique by modern standards. Then, with the industrial age, interchangeable parts with tight tolerances became the norm, but an auto or tractor factory could still be retooled to make tanks or planes by following the prints and using the same machinery. Not so much anymore.
Yeah, sure, CNC machining has gotten much better, for a host of reasons. But it's not just material removal that's relevant. Stuff like heat treating is today even more of an art which is not easily replicateable for high-precision materials and parts. Or, to pick a nonmetallic composites example, there's no way that you could quickly ramp up massive new production of, to take a civilian example, something like Boeing's Dreamliner. Military composites even less so.
I remember when maintaining heavy manufacturing capability which could be quickly switched to wartime use was a matter of national priority and national pride. The emperor has no clothes today.
Dad?
Cheap storage VM.
Dream big.
Cheap storage VM.
You can't compete with private industry, but it's fine for private industry to hire prisoners at $.30 / hour.
Makes perfect sense...
Cheap storage VM.