6 ms·

Open Source Micro-Factory

64 points by ph0rque 16y ago

ph0rque 16y ago
I wonder if YC would consider a startup application, and if anybody here would be interested in starting up something, based on this tech.
mangirdas 16y ago
Very good ideas, but too early.
nickpinkston 16y ago
There's an obsession to own your own factory - which certainly is sexy, but this plan is pretty naive to anyone in manufacturing. If you want to prototype something - there are hackerspaces* and even TechShop - the later investing $8M and they definitely couldn't make all the systems for a car. You can't ignore poor process fit - I.e. have fun using a MIG machine to weld your car together from sheet metal cheaper than GM can stamp it out. And you can't ignore raw materials - have fun casting your engine block from metal cans without any alloying tools / testing. In reality, local production of many things (cars) won't work with today's tools, and a real mix of general purpose local shops and standardized / accessible large ones is the real "Industry 2.0" path. * http://hackerspaces.org http://hackerspaces.org Disclosure: I run CloudFab, a cloud-based manufacturing company.
- tomjen3 16y ago
  >In reality, local production of many things (cars) won't work with today's tools, and a real mix of general purpose local shops and standardized / accessible large ones is the real "Industry 2.0" path. Assuming I won/t the same car GM is stamping out, they can beat me. But what I want is a flying car and an 10.1 inch Android tablet, both of which are things nobody currently produces. And tools only improve, so at some point we can defeat the empty leftovers of the industrialized age.
  - nickpinkston 16y ago
    I'm not arguing against building unique things industry doesn't / won't make. I'm saying that it's more realistic to better coordinate existing production than try to invent some small scale / limited way of doing so. Have fun making a capacitive screen with that equipment list... ;-)
  - bigiain 16y ago
    +1. I don't want to be able to make my own '83 Pinto, or to have the success or failure of my general purpose manufacturing tools to be judged on whether it's capable of producing an identical copy of an Aston Martin DB9. I wonder how far away from being able to produce all the custom components needed to build a vehicle that'd run me round my neighbourhood in reasonable safety running off biodeisel with reasonable efficiency?
    - tomjen3 16y ago
      You don't need special tooling for that, just a diesel car. But you should be aware that biodiesel is a scam, since it takes more gasoline to create it than it saves your car.
sbierwagen 16y ago
Microfactories, eh? See also: Kinematic Self-Replicating Machines (2004) http://www.molecularassembler.com/KSRM.htm http://www.molecularassembler.com/KSRM.htm Advanced Automation For Space Missions (1980) http://www.islandone.org/MMSG/aasm/ http://www.islandone.org/MMSG/aasm/ As for the idea, ehhhh. Obviously, companies have powerful incentives to reduce the price of manufacturing facilities, and yet they still cost millions of dollars. Poor nations are not poor because the tools aren't cheap enough, but because of official corruption, or uncontrolled nationalization, or because the state can't maintain a monopoly on violence. If they actually pull it off, it would be great, and might have some actual commercial applications, (like how the OLPC project resulted in the creation of the netbook concept, despite the OLPC itself being quite useless) but the stated use case just doesn't exist.
- kragen 16y ago
  The kids in Paraguay don't seem to think the OLPC is useless; although obviously the OLPC project has failed to achieve anything like its original goals, it's still producing things people want.
kragen 16y ago
I think Marcin and his crew are doing fantastic work, and their simple, low-cost designs could make a real difference, especially in the case of places that are poorly served by current mass manufacturing. One of my first kragen-tol posts http://lists.canonical.org/pipermail/kragen-tol/1998-November/000299.html http://lists.canonical.org/pipermail/kragen-tol/1998-Novembe... predicted mass adoption of this kind of thing within ten years. So far it's taking longer than I expected, although it's happened in a few niches: http://www.cleggind.com/specialstructures/mobilepartsmachineshop.htm http://www.cleggind.com/specialstructures/mobilepartsmachine... One thing that some may not appreciate is that a local micro-factory doesn't have to run cheaper than, say, GM's body panel stamping machines in order to be economically beneficial. To be concrete, consider the case where you need a certain screw. The screw that holds the left wheel of your push lawnmower on, which fell out in the grass somewhere. It needs to be a countersunk head to fit, and it has a much narrower thread pitch than most screws that size. None of the hardware stores in your area carry any screws in the right diameter and thread pitch, let alone a countersunk head. What do you do? Well, at the time, I lived in Dayton, Ohio, one of the machining centers of the world. So I went across town to a company that manufactured screws and spent about US$5 to buy a pound of the right kind of screw (I think it was about half an inch diameter and half an inch long, with a humongous countersunk head). Here in Buenos Aires, I imagine I'd just be screwed. Now, I don't know what kind of process you use to manufacture a big machine screw like that. Do you cold-roll the threads, or do you cut them with a die from a cold-forged blank cut from bar stock, or what? I'm sure that, whatever it is, it isn't cutting it out on a lathe, or they'd be sold by the screw, not by the pound. But you can cut one on a lathe, except for the slot in the head, of course. If my local hardware store had had a fully automatic CNC lathe and the appropriate software, they could have punched in the numbers and cut me the correct screw in about five minutes. Maybe it would cost me US$10, but it would have saved me hours of shopping and driving. (I probably could also have mail-ordered it from McMaster-Carr, although that still leaves me without a lawnmower for a few days.) Custom automated manufacturing isn't competing economically just with mass production, even for clearly mass-producible goods like machine screws. It's competing with mass production, plus the costs of its supply-chain management and its wholesale and retail channels, plus the delays implicit in the same.
- nickpinkston 16y ago
api 16y ago
Side note but: I've always thought that this kind of technology rather than rockets/spacecraft is the limiting factor preventing us from being able to realistically colonize another planet. When you got to Mars, the Moon, etc., you'd have to be able to unpack and set up a reliable self-sustaining industrial infrastructure.
- nickpinkston 16y ago
  Yea, I think you're right. You need a real version of a GECK* really. A lot of this is a material science problem I think - making the Martian soil into some type of metal / ceramic building material using just solar / nuclear energy. GECK: http://fallout.wikia.com/wiki/Garden_of_Eden_Creation_Kit http://fallout.wikia.com/wiki/Garden_of_Eden_Creation_Kit
samatman 16y ago
This is the same project earlier known as RepLab: http://www.replab.org/ http://www.replab.org/ "Open Source Micro-Factory" is perhaps more descriptive, which is good for the goal of a Kickstarter campaign. disclosure: I am one of the moderators of the RepLab list.
- kragen 16y ago
  One of the problems with this project is that it has too many names. Open Source Ecology, Openfarmtech, the Global Village Construction Set, RepLab, Open Source Micro-Factory...
billswift 16y ago
Their claim, "described best by Jane Jacobs, who claims that the highest level of evolution (like Maslow’s Pyramid) for cities - is for those cities to return to local production (import substitution)." is false. Jane Jacobs described import replacing as a stage from importing to exporting. It keeps capital in the city by not using as much of it for imports, while developing manufacturing skills suitable for producing exports. Also, as she points out, it just changes the makeup of imports, you import the raw materials instead of the finished items. Microfactories are a neat idea for several reasons, there is no need to misrepresent their benefits. I have been playing around with the idea since the late 1980s, the technology still has a good ways to go before it is practical though.
johnohara 16y ago
While I think I understand the localization goals of this project, I find it interesting that the project makes use of contributors from all over the world. Like it or not, we are a diverse global community and we all benefit from each others skills and abilities. Food, energy, replacement parts, etc., are all good localization goals. Locomotives, aircraft, sea vessels, automobiles, farm and construction equipment, and such, are a different matter.
ChuckMcM 16y ago
I've been playing around in robotics for 25 years, and one of the things that is just as true today as it was in 1985 is that for some mechanical parts, motors, gears, and other linkages, there isn't any 'there' there. And so the ability to manufacture small lots of things is appealing on a variety of scales. That being said, I suspect this particular effort, like those before it, won't come to fruition. The reason is that manufacturing, while easy to comprehend, is insanely complex at the nuts and bolts level. That being said I would love to see this be successful, I've got a simple idea for a widget and if I could get some time in a factory to make a few thousand of the structural bits to keep the costs down, I think I'd have a nice little niche product. While I'm sure I could get it done in China, I'm not interested in spending the time to learn the way to correctly bribe the necessary people to get the product I want. Another risk of this capability is weapons production. Making guns is actually pretty easy (the California prison population displays great creativity in this area). So if you create a capability to make small quantities of 'hard metal' products I would expect that you would be visited often by folks who would want to insure you're not tempted to become profitable the 'easy' way. I hope it works though, it would be great to have a field of small factories bloom.
- nitrogen 16y ago
  Those who really want to make guns are probably already doing so with existing manufacturing technology.

64 points by ph0rque 16y ago

I wonder if YC would consider a startup application, and if anybody here would be interested in starting up something, based on this tech.

mangirdas 16y ago

Very good ideas, but too early.

nickpinkston 16y ago

There's an obsession to own your own factory - which certainly is sexy, but this plan is pretty naive to anyone in manufacturing. If you want to prototype something - there are hackerspaces* and even TechShop - the later investing $8M and they definitely couldn't make all the systems for a car. You can't ignore poor process fit - I.e. have fun using a MIG machine to weld your car together from sheet metal cheaper than GM can stamp it out. And you can't ignore raw materials - have fun casting your engine block from metal cans without any alloying tools / testing. In reality, local production of many things (cars) won't work with today's tools, and a real mix of general purpose local shops and standardized / accessible large ones is the real "Industry 2.0" path. * http://hackerspaces.org http://hackerspaces.org Disclosure: I run CloudFab, a cloud-based manufacturing company.

tomjen3 16y ago
>In reality, local production of many things (cars) won't work with today's tools, and a real mix of general purpose local shops and standardized / accessible large ones is the real "Industry 2.0" path. Assuming I won/t the same car GM is stamping out, they can beat me. But what I want is a flying car and an 10.1 inch Android tablet, both of which are things nobody currently produces. And tools only improve, so at some point we can defeat the empty leftovers of the industrialized age.
- nickpinkston 16y ago
  I'm not arguing against building unique things industry doesn't / won't make. I'm saying that it's more realistic to better coordinate existing production than try to invent some small scale / limited way of doing so. Have fun making a capacitive screen with that equipment list... ;-)
- bigiain 16y ago
  +1. I don't want to be able to make my own '83 Pinto, or to have the success or failure of my general purpose manufacturing tools to be judged on whether it's capable of producing an identical copy of an Aston Martin DB9. I wonder how far away from being able to produce all the custom components needed to build a vehicle that'd run me round my neighbourhood in reasonable safety running off biodeisel with reasonable efficiency?
  - tomjen3 16y ago
    You don't need special tooling for that, just a diesel car. But you should be aware that biodiesel is a scam, since it takes more gasoline to create it than it saves your car.

sbierwagen 16y ago

Microfactories, eh? See also: Kinematic Self-Replicating Machines (2004) http://www.molecularassembler.com/KSRM.htm http://www.molecularassembler.com/KSRM.htm Advanced Automation For Space Missions (1980) http://www.islandone.org/MMSG/aasm/ http://www.islandone.org/MMSG/aasm/ As for the idea, ehhhh. Obviously, companies have powerful incentives to reduce the price of manufacturing facilities, and yet they still cost millions of dollars. Poor nations are not poor because the tools aren't cheap enough, but because of official corruption, or uncontrolled nationalization, or because the state can't maintain a monopoly on violence. If they actually pull it off, it would be great, and might have some actual commercial applications, (like how the OLPC project resulted in the creation of the netbook concept, despite the OLPC itself being quite useless) but the stated use case just doesn't exist.

kragen 16y ago
The kids in Paraguay don't seem to think the OLPC is useless; although obviously the OLPC project has failed to achieve anything like its original goals, it's still producing things people want.

kragen 16y ago

I think Marcin and his crew are doing fantastic work, and their simple, low-cost designs could make a real difference, especially in the case of places that are poorly served by current mass manufacturing. One of my first kragen-tol posts http://lists.canonical.org/pipermail/kragen-tol/1998-November/000299.html http://lists.canonical.org/pipermail/kragen-tol/1998-Novembe... predicted mass adoption of this kind of thing within ten years. So far it's taking longer than I expected, although it's happened in a few niches: http://www.cleggind.com/specialstructures/mobilepartsmachineshop.htm http://www.cleggind.com/specialstructures/mobilepartsmachine... One thing that some may not appreciate is that a local micro-factory doesn't have to run cheaper than, say, GM's body panel stamping machines in order to be economically beneficial. To be concrete, consider the case where you need a certain screw. The screw that holds the left wheel of your push lawnmower on, which fell out in the grass somewhere. It needs to be a countersunk head to fit, and it has a much narrower thread pitch than most screws that size. None of the hardware stores in your area carry any screws in the right diameter and thread pitch, let alone a countersunk head. What do you do? Well, at the time, I lived in Dayton, Ohio, one of the machining centers of the world. So I went across town to a company that manufactured screws and spent about US$5 to buy a pound of the right kind of screw (I think it was about half an inch diameter and half an inch long, with a humongous countersunk head). Here in Buenos Aires, I imagine I'd just be screwed. Now, I don't know what kind of process you use to manufacture a big machine screw like that. Do you cold-roll the threads, or do you cut them with a die from a cold-forged blank cut from bar stock, or what? I'm sure that, whatever it is, it isn't cutting it out on a lathe, or they'd be sold by the screw, not by the pound. But you can cut one on a lathe, except for the slot in the head, of course. If my local hardware store had had a fully automatic CNC lathe and the appropriate software, they could have punched in the numbers and cut me the correct screw in about five minutes. Maybe it would cost me US$10, but it would have saved me hours of shopping and driving. (I probably could also have mail-ordered it from McMaster-Carr, although that still leaves me without a lawnmower for a few days.) Custom automated manufacturing isn't competing economically just with mass production, even for clearly mass-producible goods like machine screws. It's competing with mass production, plus the costs of its supply-chain management and its wholesale and retail channels, plus the delays implicit in the same.

nickpinkston 16y ago

api 16y ago

Side note but: I've always thought that this kind of technology rather than rockets/spacecraft is the limiting factor preventing us from being able to realistically colonize another planet. When you got to Mars, the Moon, etc., you'd have to be able to unpack and set up a reliable self-sustaining industrial infrastructure.

nickpinkston 16y ago
Yea, I think you're right. You need a real version of a GECK* really. A lot of this is a material science problem I think - making the Martian soil into some type of metal / ceramic building material using just solar / nuclear energy. GECK: http://fallout.wikia.com/wiki/Garden_of_Eden_Creation_Kit http://fallout.wikia.com/wiki/Garden_of_Eden_Creation_Kit

samatman 16y ago

This is the same project earlier known as RepLab: http://www.replab.org/ http://www.replab.org/ "Open Source Micro-Factory" is perhaps more descriptive, which is good for the goal of a Kickstarter campaign. disclosure: I am one of the moderators of the RepLab list.

kragen 16y ago
One of the problems with this project is that it has too many names. Open Source Ecology, Openfarmtech, the Global Village Construction Set, RepLab, Open Source Micro-Factory...

billswift 16y ago

Their claim, "described best by Jane Jacobs, who claims that the highest level of evolution (like Maslow’s Pyramid) for cities - is for those cities to return to local production (import substitution)." is false. Jane Jacobs described import replacing as a stage from importing to exporting. It keeps capital in the city by not using as much of it for imports, while developing manufacturing skills suitable for producing exports. Also, as she points out, it just changes the makeup of imports, you import the raw materials instead of the finished items. Microfactories are a neat idea for several reasons, there is no need to misrepresent their benefits. I have been playing around with the idea since the late 1980s, the technology still has a good ways to go before it is practical though.

johnohara 16y ago

While I think I understand the localization goals of this project, I find it interesting that the project makes use of contributors from all over the world. Like it or not, we are a diverse global community and we all benefit from each others skills and abilities. Food, energy, replacement parts, etc., are all good localization goals. Locomotives, aircraft, sea vessels, automobiles, farm and construction equipment, and such, are a different matter.

ChuckMcM 16y ago

I've been playing around in robotics for 25 years, and one of the things that is just as true today as it was in 1985 is that for some mechanical parts, motors, gears, and other linkages, there isn't any 'there' there. And so the ability to manufacture small lots of things is appealing on a variety of scales. That being said, I suspect this particular effort, like those before it, won't come to fruition. The reason is that manufacturing, while easy to comprehend, is insanely complex at the nuts and bolts level. That being said I would love to see this be successful, I've got a simple idea for a widget and if I could get some time in a factory to make a few thousand of the structural bits to keep the costs down, I think I'd have a nice little niche product. While I'm sure I could get it done in China, I'm not interested in spending the time to learn the way to correctly bribe the necessary people to get the product I want. Another risk of this capability is weapons production. Making guns is actually pretty easy (the California prison population displays great creativity in this area). So if you create a capability to make small quantities of 'hard metal' products I would expect that you would be visited often by folks who would want to insure you're not tempted to become profitable the 'easy' way. I hope it works though, it would be great to have a field of small factories bloom.

nitrogen 16y ago
Those who really want to make guns are probably already doing so with existing manufacturing technology.