Thursday, March 5, 2009

Introducing the Open Source Hardware Central Bank

This blog post started with a cross-country driving trip, Zen and the Art of Open Source Hardware, and culminated with discussions during and especially after Justin's Open Source Economic Council (OSEC?). The list of people I interviewed and talked to about this is too long to list individually, so instead I'm going to throw credits to everyone who's been kind enough to help on a wiki soon. This coming weekend, Justin, Andrew, and I will be "launching" the Open Source Central Bank. This is the first in a series of blog posts describing it, and next week I'm headed up to MIT and Harvard to talk to some economists and business professors about how it will work. I'm excited!

Why does Open Source Hardware need a bank?

Because Open Source Hardware is different from Open Source Software. Software can be made with time, but hardware needs time and money. The same kind of "openness" principles from the Open Source Software "time economy" transition nicely to the Open Source Hardware-based "time economy," but they seem to get muddled in the OSHW "money economy." Need proof? Just try to answer any of these questions: who makes money from it, who funds it, why do they fund it, and who's helping to make it sustainable for the community? Open Source Hardware lacks a way for individuals to come together, make a cool project, and get something out of it - without taking a second and third mortgage on their houses!

Right now, the status quo, emerging trend for OSHW DIY'ers has been: build something, put up a bunch of money to build a few of them, if people like it, scale it up, raise money, realize you might lose all that money, charge a margin on top of it to cover your potential losses, start a small company to resell more, cross your fingers, maybe get lucky or maybe not. Setting up each little company takes an infrastructure investment like incorporation legal fees, Paypal transaction costs, and website hosting fees to name a few. For every small hardware project, there's a potential to have to pay upwards of 40-50% of the initial cost of the project again in just infrastructure fees - that's prohibitive and ridiculous for little guys like me.

I initially built the Illuminato with financial help from some friends, but mostly from a former mentor of mine who sponsored the project by helping to get scaling costs for the inventory. That worked the first time, but I've been stuck with a decision of how to fund it. If I only build 25 at a time, the cost will be around $50 apiece, which is just wrong. So I've been sitting here trying to figure out whether to take out a loan, pass the hat amongst friends, try to pitch it to a VC, or try something else? This is the OSHW "money problem" - how do you fund Open Source Hardware?

A vision for Open Source Hardware

Looking at Open Source Software, it's a thriving ecosystems of communities, projects, and contributors. There are a few companies, but they mostly offer "paid-for" services like consulting, tech support, or custom code/build-to-order functionality. I'd like the same for Open Source Hardware. I'd like the money problem to go away for small contributors like me and others. And I'd like to help guys like Chris and Mike and Mark and David and Jake build more cool stuff because it's fun.

I happen to believe that success for Open Source Hardware is not a distributed, highly-fragmented ecosystem with hundreds or thousands of individual companies, each structured around a single project. That seems wrong, and the transaction and infrastructure costs alone make that hard to stomach, let alone the time it takes to set all that up. I also don't believe that Open Source Hardware should ever be venture-backed. This is a controversial topic to some people. But speaking for myself (and quite a few others, apparently!), if I'm contributing my hard-earned time and money to projects and giving them away for the community benefit, I want to know, like Mark and Justin have taught me, that the community is reaping as close to 100% of the benefits. I don't believe in middle-men or intermediates just for the sake of it, or in speculators profiting off of my spare time. I get enough of that during my day job, so I want to eliminate that from my "spare time!"

Principles for the Open Source Hardware Bank

Justin, Andrew, and I have put together what we'd consider a beginner's set of principles for the Open Source Hardware, which the bank will operate under. Naturally, these are also on the wiki. These principles are described in terms of what we think Open Source Hardware needs to succeed:

A mechanism to:
  • Reduce margins and share costs for the community
  • Minimize the risk and opportunity cost of unsold inventory
  • Provide incentives for Open Source projects to move to production without risks
  • Allow the building and distribution of low-quantity, non-scalable products (e.g. niche applications that are potentially non-VC fundable, since "bad business idea" isn't the same as "bad hardware idea")
  • Give rewards and profits back as close as possible to those who contributed

A platform that:
  • Minimizes economic transaction costs to high-paid non-laborer economic types
  • Reduces barriers to contribution
  • Rewards innovation and encourages new ideas
  • Encourages project-level (not necessarily company-level) competition

People who:
  • Participate because they are getting as much or more out as they put in
  • Do it not just to make money and profit off of others for free
  • Have rare, valuable skills who volunteer their talents for recognition or fun
  • Are willing to build a more sustainable hardware innovation system
  • Are willing to teach others for the gratification of helping others learn new skills

What's the main issue the Bank is trying to solve?

Open Source Hardware has two main financial problems that the Open Source Bank will try to alleviate (in addition to a number of other tool-based problems, but others in the community are working on those thank goodness): "Throwaway Costs" and the "Quantity Monopoly." As if the current economy weren't bad enough already, both of these problems seriously hurt DIY'ers and potential Open Source builders who want to participate in the growing Open Source Hardware community.

Throwaway Costs - building physical hardware takes revisions. Early revisions have things wrong, like misplaced traces, wrongly sized solder pads, or just bad luck. In the OSS software world, when things go wrong, you just fix the code, hit compile again, and the only thing it really "costs" is your time finding and fixing the error. But in the OSHW hardware world, errors mean broken, non-functioning junk PCB's that cost money to make. And that means lost money. Who pays for this? Guys in college, or guys who just lost all their money in their houses can't afford to build 2, 4, or 6 revisions of hardware before it works!

The Quantity Monopoly - this is a term I'm giving to the fact that large companies, especially PCB houses and component suppliers, offer volume pricing discounts. Normally this is a good thing, but only if you're building 10,000's of finished products. In the DIY OSHW world, we're talking about building 1's to 10's to 25's at a time, and so the community gets burned every time by paying "quantity tax" to large suppliers. The has the side effect of pricing individual DIY builders out of many potential hardware developments, simply because they don't get cheap enough until you make 1,000's. It's a quantity monopoly, because there's only 1 quantity number that anyone wants to build: 100,000 of anything. This is a difficult topic, and in my interviews and conversations, I've found many people on both sides of the fence about this - some for, some against. The bottom line is: if stuff were cheaper, Chris, David, Mark, Mike, Omar, Justin, and I'd all personally be able to build and share more, so anything prohibiting this is what I'll call "bad."

The Solution: how the Bank will work

The Open Source Hardware Bank will work to eliminate the scaling and quantity pricing problem for OSHW projects by funding the build of 2x the quantity of any Open Source Hardware product. That means, if a project has found a way to find 10 potential buyers, the bank will put down the money needed to fund 10 more, for a total of 20 products. If a project has found 25 community members to buy in, the bank will fund another 25, to bring the total quantity down to 50. This should reduce the unit costs by around 10-30% of any hardware project, and in the case of the Illuminato, it'll reduce costs by almost 40%!

In return, anyone who pitches in money to the bank will get a modest and sustainable return on their investment, somewhere between 5-10%. Normally, this wouldn't be a huge amount, but given what I've learned about the "real" economy recently, 30-50% return on investment may never have really existed in the first place, let alone represented "sustainable growth." This money gets paid back and cashed out when the rest of the inventory is bought as a check that Justin, Andrew, or I write and sign personally.

So Andrew, Justin, and I will see to it that the Open Source Hardware Bank does not default, and each of us will guarantee every investment. Maybe you could call it AJMIC (instead of FDIC insured)! No one is trying to become a millionaire (without lots of hard work), a high paid investment banker (ugh), or Alan Greenspan (was he ever right about anything?). We're just trying to build a sustainable little financial institution to help Open Source Hardware DIY'ers. Consequently, we're also human and realize the limits of spare time, so no one's rushing out to build 50 projects, just 1 or 2 or 3 at a time will be perfectly fine, thank you!

The Open Source Hardware Bank is "Open Source"

The bank is funding Open Source Hardware, but it is also trying to be a step in the direction of Open Source Finance. As a result, the bank is also going to be "Open Source." It will run on a wiki, everything will be transparent, and it is open to anyone who'd like to join in any of the following roles:

Open Source Banker - these will be rotating positions, and Andrew, Justin, and I will do it first until it gets unsustainable and we need help (hint hint Mark and John)

Open Source Hardware Investor - by buying anywhere between $1,000 and $5,000 Open Source hardware T-bills

Open Source Economic Council - attending bi-monthly Open Source meetings (OSEC) in NYC and Boston to vote on what the bank will fund

If you're interested, or just think Justin, Andrew, and I are totally nuts, just send me an email! :) inthebitz at gmail... and in the meantime, every ridiculously crazy project needs a respectable logo, so here's one for the Open Source Bank:

Naturally, the text around the logo reads "Open Source Hardware" in ASCII, there's wreath of resistors representing overcoming resistance (buh dump chhhhh), and a course a fancy set of circuitry in the middle, and a hexadecimal base 16 set of stars around the center...

Here goes nothing!!!


Chris said...

That logo is simple, yet very inspirational

Great work!

Matt said...

thanks :)

Justin said...

I can't wait to see where this goes!! Nice job Matt

UnwiredBen said...

I really like this... the HTINK group in Brooklyn has been thinking about a similar idea, but more grant-oriented. I think you've got something really nice fleshed out here, and I can't wait to see more details on your plan.

jeff.kantarek said...

What about designing an investor structure similar to Micro-payments and large groups of investors

Unknown said...

Very interesting... But can it be somehow international ? I think it will appeal makers from all over the world (I myself am french).

Marc said...

this is great! Definitely worth spreading the word.

Matt said...

@UnwiredBen - thanks! Justin and I are like chickens with their heads chopped off coding, writing, drawing... by the way, i checked out the HTINK, and couldn't help but notice someone linked Odyssey of the Mind at the bottom. I actually participated in that as a kid in elementary school up in CT - talk about memories!

@jeff - I hadn't seen kiva before, but I really like their wording "in need of funding - not marketing material". That sums up pretty much my experience with real banks as of late... is there some way lenders pool together?

@Tom - hmmm, want to start the OSHW Bank European division? That was one of the more difficult parts I haven't quite figured out - how to alleviate international distribution costs for hardware. I was kind of thinking about a distributed model with an "office" aka "basement" that someone could help ship from in Japan, Europe, and maybe Brazil? If you knew 2 people were in the same country across the world, there has to be a way to save on shipping costs, right?

@toaster_oven - thanks! p.s. i used you this morning to make a set of pop tarts, since i was too lazy to pour milk in a cereal bowl :)

Anonymous said...

Wouldn't it be much easier to pitch a hardware project on a website, and have people pledge to buy a certain number of items when they become available at (or under) a certain price?
That way, it's the users who determine what to invest in, instead of a central entity.

Also, as for cost of revisions etc: make your own PCBs to try stuff out! Much cheaper and much shorter turn-around times than having everything fabbed.

Anonymous said...

Fub's idea is also pretty good. Might be a way to be transparent about how you are giving loans out as well.

The problem you are trying to solve is the high cost of prototyping. One solution is the bank providing leverage. Another might be to act as a clearinghouse for investment offers. Offer to hold money in an interest bearing account for general investors, then loan it out to qualifying projects, which would be responsible for selling out the resulting overstock prototypes and paying back. You wouldn't get fdic insurance, so you'd have to realize decent profits to make it worth it to invest.

Specific investors could lower costs just by looking at the list of ideas and applicants and offer to either invest in a specific project or commit to buy a prototype. They'd drop in escrow to the bank and then the bank would handle the middle for a small fee.

jeff.kantarek said...

@Matt - Kiva has a huge list of people looking for small loans ($500-$3000), the users just decide who they personally to loan to. Money comes directly out of a paypal account and is deposited back when the loan is repaid. Most of the loans have 0% interest but that is by design. It also seems like there are actual back end banks vouching for and verifying the people who are applying for loans

DU said...

Where do I send my money?

Unknown said...

Interesting idea! You might be interested in my story - I derive a full time income from selling open hardware products like the IP04 IP-PBX. Many other companies and individuals are also deriving significant income from related IP-PBX technology. Here is a video of a talk from 2009 on Open hardware Business models. Lots of other material on my blog on Open Hardware.

Matt said...

@fub - a few years ago i met up with some of the guys that run threadless, and that was one of the complaints and caveats they mentioned. pledge or pre-commit works sometimes, but every once in a while you left with a ton of t-shirts that were made that you thought would sell, but then actually don't :( i think you're right though, maybe there's a pre-stage where externally proposed products get submitted, voted on to set the initial build quantity targets, and then moved into the order queuing system. the guess-work is definitely in "how big should the initial order queue be?" so perhaps either a tally system, vote-system, ordering, priotitization first, funding second?

@mattk - do you have an email address? i'm building some pictures right now trying to explain just that... i think there are two entry-points for money: specific investors that fund a project, and general investors that fund the system. either way, i'm going to finish them up this weekend and try to post them next week

@jeff - justin has a list that he showed me a long time ago of all these different distributed financing systems he had collected, (i can only remember zopa and prosper)

@DU - wait 3 more days, maybe? then send me an email at inthebitz at gmail... i'm really excited about how many people are actually interested in making this work. one guy even said, if the open source bank could return even 1% interest, it'd be more than he was getting on his current investments, so he'd be willing to switch banks - ha!

@davidrowe - wow, it's really nice to meet you, i really liked the video presentation you made, and btw whatever software that was that showed the topics on the side while you spoke was pretty cool. i also really like the way the red daughter boards look... makes me want to play it like a piano :)

Michael said...

When you start representing an organization as a bank you start getting into the realm of lawyers.

You may want to make it some sort of corporation instead. Heck, I consider throwing in if it were even a not-for-profit!

Jean Vincent said...

Here's the business model I propose for OSHW:

1 - Make OSHW as OS software - i.e. don't build anything, just provide the design with no physical experimentation.
2 - Test extensively using comprehensive open source testing software.
3 - Provide support services to OEMs.

The key to this is therefore the availability of a comprehensive open source testing software. If there is nothing satisfactory then this should be the focus of OSHW proponents, the key building block as the GNU project was instrumental to the success of OSS.

Anonymous said...

I absolutely love the ideas expressed. Please, please, please, make sure you have an entity to perform due diligence. People can and will, by human nature, abuse this idea. I wish you all the best.


Unknown said...


I am not quite sure what to make of this. You address issues that I have not yet figured out. I am building a small scale hardware/software project. It is a hand held chord key-board.
The software-only version is very easy to write and distribute, the hardware version has been a nightmare. CNC machining is so difficult, largely because the CAD/CAM software does not know about lots of things that are issues. I did learn from your pages that I should get someone to do a design review of the circuit board. Much easier than building lots of versions! I am less worried about the money to build a bunch of boards. I am more worried about just bringing all the pieces together, the marketing, the cnc machining, the circuit board, the software.

Having said all that, I think that you are slightly missing the boat. What is needed is a hardware incubator. Finance is just part of a larger problem. Mechanical design, ergonomic design, electrical design, software design, a store, and marketing all need to be brought together.

Christopher Lozinski

Matt said...

@christopher - ps, i'm mathematically inclined, and i like the idea of a 5 key one :) anyway, yeah i totally agree on the incubator thing. actually, on the road trip, chris and i realized that there are lots of problems with open source hardware - expensive proprietary pcb tools, design reviews, ridiculously expensive trace layout software, lack of physical machine shops, and poor financing. i looked around, and realized that there were some people fixing some of those other problems, but no one was dealing with the finance part... so that seemed like a fun one to tackle :)

Jhony said...

Very very nice blog.

Kaspersky keys
Kaspersky Activation key
Mobile Softwares

enjrolas said...

I think there's also an opportunity for volume purchasing in components. Every circuit board someone designs will be unique, but often the parts overlap between projects. A great example is the Atmega168 processor used in many Arduinos. You can get decent volume pricing off digikey and mouser in quantities of 25, 100 and 500, but if you get into quantities >=1000, you can order from china and get prices (depending, of course, on the part) 10-25 times cheaper than you'd find on digikey, even at mid-sized volumes. In this case, it's kind of like DIY speculating, except it's self-reinforcing. If you predict that a particular chip is going to be really popular (at90usb, adxl330, etc), buy a lot of 1,000 from China and offer it at a slight margin to someone whose product is supported by the bank. You don't want to force or limit people to using certain parts, obviously, but there's some basic parts that will always be popular (or, at least, for the next ~2-5 years) in DIY projects, and it could be a huge step in bringing down the cost of open-source hardware projects.

SocialCritic said...

As the creator of the "CubeSpawn" project I have an interest in the thinking related to the open source licensing model, although I think the more its adopted, the less relevant it becomes, in an ironic inversion.

The assumption that you can "control" something as ethereal as an idea is some pretty weird thinking in and of itself, since nearly any new creation builds on the work of thousands of previous innovations.

the territoriality of the 18th and 19th centuries is fading from our worldviews (at least in developed countries) due to our being inundated by an embarrassment of riches in innovations everywhere we turn,

In the past, they had less, and so grasped what they did have, harder. But its time to realize than innovation should be open to anyone who cares to try their hand at it and the age of exclusive "ownership" of innovation will begin to close.

A recent example is the fallacy demonstrated by pharma companies fighting against distribution of generic AIDS medication in Africa.

Seems perfectly reasonably in western courts to "protect their investments" but millions of sick people being denied access to protect a private enterprises license makes no actual sense, its simply a distortion of any reasonable value system.

My project will be an attempt to circumvent the issues by building an “Open Source” equivalent to whatever tool is required for small scale mass production.

See it here:

Matt said...

@socialcritic - interesting post, and i enjoyed diggin around your site too... interesting, are you offsetting the cost of scaling by donations? that's a curious model, for sure...

Sepp said...

Matt, has this project gone anywhere?

My question is really, whether there have been further developments, whether the proposed method of financing is working and to what degree, and finally, if there is a more recent article updating us on further developments.

I'm asking for the p2p foundation, which blogged this article, and the person who did it is wondering what the current status is...

Unknown said...

I also looked for updates as to the status of the OS Hardware Central Bank. How is it coming?

I would also like to point out another point of attack of providing efficiency of scale to small producers. I am working on an Open Source Electronic Manufacturing Line. The idea is to lower the cost of production on a small scale until the scale advantage is eliminated.