I didn’t talk about Open P2P Design and how to co-design open processes and systems; instead I talked about the business models behind the current Open and DIY projects. Running an Open business is part of the big theme “how to co-design open systems”, and it’s something I’m increasingly investigating more and more (and it seems there is a lot of interest in it).
Here’s my presentation; soon I will blog about a longer presentation about the same issues I gave in Berlin few days later:

The event was very nice and insightful and it was great to finally have time to do something together with Openwear.org. Furthermore, it confirms that Open Design has a really great place in the fashion industry (something I must confess I was’t aware before meeting the people from Openwear.org 2 years ago). Among the many interesting speakers, I’d like to outline the presentation of Jen Ballie, a PhD student at the Textile Futures Research Group and Consultancy whose research is about the intersection of co-design, web 2.0 and sustainability for the fashion industry.

Definitions of Open Hardware

The current Open Source Hardware Draft Definition is intended to help provide guidelines for the development and evaluation of licenses for Open Source Hardware and it says that Open Hardware is “a term for tangible artifacts — machines, devices, or other physical things — whose design has been released to the public in such a way that anyone can make, modify, distribute, and use those things“. The main difference with Open Source Software is that Open Source Software is collaborative, while Open Hardware is derivative: here a fork is the rule, not the exception.
Even if Open Hardware has become famous in the past 5 years, it has been around for years: The Apple I was built by hand by Steve Wozniak, and he and Jobs were members of the Home Brew Computer Club. The hardware hacking community has never gone away; it has just adapted to the changes in technology. Open Hardware is still in its first steps though, just like Open Source Software was in the 1980s, when the GNU project began, before all the infrastructure was created.
Interesting overviews of Open Hardware can be found on Make Magazine’s Blog, MIT Technology Review, Computerworld, O’Reilly Radar. Lists of existing Open Hardware projects can be found on the GOpen Hardware 2009 website, on the P2P Foundation website (here and here), on Make Magazine’s Blog, Open Innovation Projects and Open Knowledge Foundation. Open Hardware projects are not limited to gadget and interaction design projects, but they can also be about development aid projects.
Patrick McNamara defined 4 possible levels of Openness in Open Hardware projects, that can help us understand them better:

1. Closed: any hardware for which the creator of the hardware will not release any information.
2. Open Interface: all the documentation on how to make a piece of hardware perform the function for which it is designed is available (minimum level of openness).
3. Open Design: in which enough detailed documentation is provided that a functionally compatible device could be created by a third party.
4. Open Implementation: the complete bill of materials necessary to construct the device is available.

Arduino: a successful open hardware project

Arduino is arguably the most popular Open Hardware project: an open-source electronics prototyping platform based on flexible, easy-to-use hardware and software; many versions of the Arduino hardware have been commercially produced to date. It’s intended for artists, designers, hobbyists, and anyone interested in creating interactive objects or environments. You can read a comprehensive introduction to Arduino on Wikipedia or on Alicia Gibbs’ thesis.
Most of Arduino official boards are manufactured by SmartProjects in Italy. The Arduino Pro, Pro Mini, and LilyPad are manufactured by SparkFun Electronics (USA). The Arduino Nano is manufactured by Gravitech (USA).

By 2006, Arduino had sold 5,000 units; in 2007, it sold 30,000; in 2009, Arduino was reported being on track to sell at least 60,000 of the microcontrollers. The reasons for Arduino’s success are:

1. it is a good starting point for projects;
2. at $30,00 a piece, it’s cheap and durable enough;
3. there is already a thriving community and business ecosystem where to find resources;
4. it is a mature and simple enough project.

The designs for the Arduino board are released under the Creative Commons license Attribution-Share Alike: you can produce copies of the board, redesign it, or even sell boards that copy the design without paying a license fee or even ask permission (you just have to credit the original Arduino group and use the same CC license). The only piece of intellectual property the team reserved was the name Arduino, which is trademarked: if you want to sell boards using that name, you have to pay a small fee to Arduino (this is set in order to make sure the Arduino brand name isn’t hurt by low-quality copies).
Clive Thompson on Wired reported two different business models for Arduino (and other Open Hardware projects):

1. sharing open hardware to sell expertise, knowledge and custom services and projects around it;
2. selling the hardware but trying to keep ahead of competition with better products (users will buy your products because are better than the copies, but the copies will help your products become more famous).

Clive Thompson concluded that Open Hardware is a sign that hardware is becoming a commodity and that it still has not clear business models: it’s possible that open source hardware will not compete with the for-profit world but will instead fill niches otherwise ignored.

The Market for Open Hardware

In May 2010 Philip Torrone and Limor Fried collected 13 examples of companies that are selling open source hardware: according to them, these companies, generate a turnover of about $ 50 million and there are currently about 200 open source hardware projects of this kind. They project the open source hardware community to reach $ 1 billion by 2015. Adafruit, Arduino, Chumby and Liquidware have each one $ 1 million in revenue, and Torrone and Fried estimated them to reach a $ 5 million revenue soon (while many other companies will reach a $ 1 million revenue). Sparkfun alone has even a $ 10 million revenue.

Open source hardware $1m and beyond – foo camp east 2010 from adafruit industries on Vimeo.

In January 2010, Joseph Flaherty calculated that the Makerbot (an open hardware 3D printer produced by a 3-person firm) has a revenue of $ 1,350,000-1,710,000 (1,800 * $ 750-950). The industry leader Stratasys (which uses a FDM technology similar to MakerBot) had a total revenue of $ 124,500,000 in 2008, but with a considerably bigger firm and more R&D investments. And MarkeBot has just opened a retail store in New York called the Botcave.

Makerbot Opens Botcave Retail Store in Brooklyn from Dan Patterson on Vimeo

Business Models for Open Hardware

Salem & Khatib proposed four business models for Open Hardware companies; later Edy Ferreira and Stoyan Tanev further expanded these to seven business models. According to Edy Ferreira and Stoyan Tanev, there is little research on the types of business models specifically related to Open Hardware, just like there is no consensus on the definition of Open Hardware itself as well. The open asset is different from the ultimate market offer, the manufactured hardware device itself, and hence the problems with the adoption of existing Open Source business models. Ferrera and Tanev examined 4 companies, 88 market offers and 93 Open Hardware projects, and then identified seven business models for Open Hardware:

1. Services (customization, consulting) over owned or third party Open Hardware.
2. Manufacturing of owned or third party Open Hardware.
3. Manufacturing of proprietary hardware based on Open Hardware.
4. Dual-licensing: companies owning Open Hardware designs that are offered for use under either a GPL license or a proprietary license. The design does not contain any proprietary module.
5. Proprietary hardware designs based on Open Hardware.
6. Hardware tools for Open Hardware: companies selling the right of ownership of development boards for the testing and verification of hardware devices manufactured on the basis of the Open Hardware assets. The designs of these boards are entirely proprietary (another related example is Sparklelabs).
7. Proprietary Software tools for developing Open Hardware.

Furthermore, there are three more business models for Open Hardware already implemented:

1. Free service for building a greater user base: Adafruit created Adafruit Jobs Board as a marketplace for designers, makers, programmers, artists, engineers and companies who want to meet and work together. This is a free service, but in order to use the job boards users must be Adafruit customers.
2. Partnership between Long Tail Open and Fabbing businesses: Ponoko has teamed up with SparkFun Electronics to enable its users to build custom electronics products combining Ponoko’s laser cutting technology with a 1500+ strong electronics catalog from open source electronics supplier SparkFun.
3. Funding Open Hardware projects for getting good Open documentation: In August 2010, Bildr offered to fund original user projects in return for good documentation: in this way it would have promoted a bildr user by showcasing his/her work and paying for the parts to construct it. In return, Bildr would have got more information for its wiki, blog and community under the MIT software license.

Manufacturing Open Hardware

But business models are just half of the issue of developing a thriving Open Hardware project: we should also focus on the manufacturing of the Open Hardware projects, a step that we don’t find in Open Source Software. The value of manufacturers is in economies of scale: cheap high-quality objects or a superb shopping and support experience. But what about manufacturers of Open Hardware?
According to David A. Mellis most open-source hardware projects (including Arduino) seem not to have taken advantage of the distributed manufacturing models enabled by the open nature of their designs. Instead, we mostly see two conventional distribution models: centralized manufacturing (that makes the product available in many places, but increases the cost to the consumer) and artisanal production (this keeps the costs low because there’s only one party profiting from a product, but at the same time it limits the product’s availability).
Mellis suggests then to adopt a distributed manufacturing model: a number of smaller groups independently producing the same design for local distribution.

Significantly enough Chris Anderson, in his “In the Next Industrial Revolution, Atoms Are the New Bits“ article, suggested to manufacture Open Hardware projects in China using alibaba.com (until a complete distributed manufacturing ecosystem will be ready), the largest aggregator of the country’s manufacturers, products, and capabilities. Alibaba.com doesn’t aggregate only companies suitable for manufacturing Open Hardware projects, but it is an interesting company as well and it takes us directly to the Long Tail business models.
Anderson reports that Alibaba, founded in 1999, has become a $12 billion company with $ 45 million registered users worldwide. Over the past three years, more than 1.1 million jobs have been created in China by companies doing e-commerce across Alibaba’s platforms.

But manufacturing in China is also a phenomenon called Shanzai: Chinese imitation and pirated brands and goods, particularly electronics, but originally the term described bandits who oppose an authority to perform deeds they see as justified. According to Anderson, the same Shanzai companies are “increasingly driving the manufacturing side of the maker revolution by being fast and flexible enough to work with micro-entrepreneurs”.
Today, the Shanzai represent approximately 20% of the mobile phones sold in China annually, and represent 10% of worldwide phone sales in 2009 (especially in Third World countries). Moreover, some manufacturers have become so successful that they are leveraging their own brand now instead of producing pirated products.
What is interesting about Shanzai companies, it’s not just that we can use them for manufacturing our Open Hardware projects, but that at the same time they work in a similar way. Albeit pirateing brand products, they have established a culture of sharing information about the products through open BOMs (bills of materials) and other design materials, crediting each other with improvements. The community self-organizes and ostracizes those that violate it. Moreover, they understand and respond to local needs and tastes, establishing and maintaining local manufacturing and distribution bases: Tom Igoe calls it situated manufacturing.
Significantly, the Shanzai companies are almost universally bootstrapped on minimal capital with almost no additional financing: Mitchell Tseng reported that 10.000 € are enough to start such a company, and it may eventually scale to over 50 million € revenue per year within a couple years.

A place for Open Hardware communities: Hackerspaces

The Open Hardware movement is also enabled by Hackerspaces, community-operated physical places, where people can meet and work on their open source software and hardware projects. Hackerspaces are distributed throughout the world, and the up-to-date list and map can be found in the hackerspaces.org wiki. There is even a hackerspaces.org email list for talking about fundraising ideas, strategies, member donations, tax laws, or anything finance related.
Wikipedia reports that membership fees are usually the main income of a hackerspace, but some also accept external sponsors. Some hackerspaces in the USA (like Noisebridge) have 501(c)3 status, while others have chosen to forgo tax exempt status. For example, HacDC in Washington DC is an non-profit corporation and 501(c)(3) (pending); as of April 2010, membership stands at over 50 people and dues are $50 per month and include benefits such as 24/7 key access, voting rights, and more.

A Bank for enabling Microcredit for Open Hardware

There is another interesting business model for Open Hardware that is just blooming: microcredit and peer-to-peer lending / crowdfunding. The main idea is to gather small loans from single individuals or greater groups in order to start an Open Hardware project.
The best example of this trend comes from two hackers, Justin Huynh and Matt Stack, who calculated that for every small hardware project, there’s a potential to have to pay upwards of 40-50% of the initial cost of the project in just infrastructure fees. As a consequence, they have started the Open Source Hardware Reserve Bank in order to solve two main financial problems specific to Open Hardware: throwaway costs that result from repeated revisions to physical hardware during the design process, and the inability to take advantage of volume discounts for raw materials. The principles of the Open Source Hardware Reserve Bank are:

• Reduce margins and share costs for the community.
• Minimize the risk and opportunity cost of unsold inventory.
• Provide incentives for Open Hardware 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).
• Give rewards and profits back as close as possible to those who contributed.

The Open Source Hardware Reserve Bank (which still has to fully comply to the laws that regulate lending) allows only hackers (no VC or other companies) to make investments in specific projects, buying and funding at the same time, doubling then the number of pieces created and reducing per-unit costs by around 10 percent to 30 percent.

Moreover, they designed an infovis that visualizes the state of the funding and manufacturing of each copy of a Open Hardware project. Each cell identify one copy of the Open Hardware project, and the smaller cell on the top represents the 15% markup charged by the Bank (as opposed to the normal 30-50% or more charged on electronics). All the cells are numbered and sorted on a waiting list for receiving the manufactured copies. If a cell is white, there is a copy available in the queue, otherwise the color will be blue (for personal copy) or yellow (for copies funded with microcredit). When somebody funds the manufacturing of one more physical copy, he/she won’t pay the 15% markup; when the copies funded will be two, he/she will save the 15% markup and the shipping fees. Funding 5 copies makes you an investor in the specific Open Hardware project, getting a 15% return on investment.
For every funding, the Bank will issue a certificate like this:

A lesson from Open Hardware: expanding microcredit certificates

All the previous cases can be seen as the state of the art for business models related to Open Hardware projects. They can be taken as example as they are, or expanded further; anyway there is one aspect that can be stressed out and I think it can be important for managing such open projects: let’s consider how a community could self-fund its projects through microcredit as a license.
Even without considering radical projects like all the Open Money and Metacurrency initiatives that proposes new forms of currencies, we can think more about further joining existing currencies with microcredit certificates like the Open Source Hardware Reserve Bank ones. There is the need of accurate, portable and shareable tools of reputation ranking, able to interconnect different local contexts and attached to existing currencies. The Open Hardware still needs proper open-content licenses, since with current licenses we can protect the design but not the manufactured product or forks. And Open Hardware projects will have the need of warranties and conformance marks about the proper function of the manufactured product. Why don’t we use the microcredit certificates for these tasks as well? We could design microcredit certificates to act as a conformance mark, warranty and license certificates as well: only the community can issue them and use them for its own self-organization.

Other interesting resources about Arduino are:

• Build It. Share It. Profit. Can Open Source Hardware Work?, an article written by Clive Thompson on Wired;
• New Media Art, Design, and the Arduino Microcontroller: A Malleable Tool, a thesis written by Alicia Gibb.

Arduino The Documentary (2010) English HD from gnd on Vimeo.

We Make Things. from Ryan Varga on Vimeo.

One of the first documentaries of this kind was Handmade Nation, a DVD documentary and a book. Here’s a trailer:

And now we just have to wait Electromagnate to finish a documentary that explores the state and direction of the Maker and Hackerspace movement in America. They even created a page on Kickstarter to get funded for shooting the movie. I already backed the project, and they have already reached more money than they needed to start the project; anyway you still have 15 hours to back the project (and receive something in change).

ReMade: The Rebirth of the Maker Movement will reveal the stories of members of community-run workshops, including hackerspaces, makerspaces, and other collaborative organizations across the U.S.

ReMade will tie these inspiring stories with those of the leaders who have furthered the goals of this movement and inspired thousands to take part. It will follow the history that brought us to this revolution, such as the Mechanic’s Institutes of the 1870’s and the computer revolution of the 1970’s.

ReMade: The Rebirth of the Maker Movement (1st Trailer) from Electromagnate on Vimeo.

Companies, for their part, say an open approach can help them get to market quickly with products that give customers what they want—without the need for market research. Such advantages, they say, outweigh the drawbacks of exposing what are usually seen as corporate secrets.

In some ways, open-source hardware is a throwback to the 1970s and 1980s, when early computers were sold in kits or shipped with schematic diagrams to make it easier for users to customise them. But the open-hardware trend has been reborn in recent years, thanks to the rise of the internet and the success of open-source software. Some enthusiasts point to 2005 as a crucial year: that was when work began on devices such as the RepRap (a rapid-prototyping machine that will, its makers hope, be able to replicate itself) and the TuxPhone, an open, Linux-powered mobile-phone. It was also when Sun Microsystems, a computer-maker, decided to publish the specifications of one of its microprocessors, the UltraSPARC T1.

Open-hardware business models are difficult to understand, because by turning users into product developers, they turn tradition on its head, says Eric von Hippel, professor of innovation at the MIT Sloan School of Management in Cambridge, Massachusetts, and the author of “Democratizing Innovation”. That makes it necessary for companies to consider the users’ motivations too, he says. “The users have a built-in business model—they build to satisfy themselves,” he says. “The business model is ‘I can get stuff for myself, I can get a better design and I can benefit.’ The innovation is paid for within the activity itself.”

As well as tapping a valuable new source of ideas, an open approach can also lead to savings in market research, as users act as focus groups, indicating what new features they would like (and then helping to develop them).

Going open-source may also help to keep customers. “Once you’ve opened the guts of a machine, you’re a much more loyal customer,” says Mr Talley, who got a Chumby for Christmas. Sun says the primary advantage of open-sourcing the designs of its processor chips is an elusive marketing boost to its other products, such as server computers. “It builds a community that will buy our hardware,” says Sridhar Vajapey, who runs Sun’s OpenSPARC program.

An alternative approach is to make money from something other than the hardware. Chumby Industries, for instance, expects to make most of its revenue by piping advertising to its devices. “It’s a traditional media model, only with user control,” says Steve Tomlin, the firm’s founder and chief executive.

Some examples of Open Hardware that can be found in the article and in the comments are:

• Neuros OSD
• Chumby (wikipedia)>
• Tuxphone (wikipedia)>
• RepRap (wikipedia)>
• OpenMoko
• Gumstix (wikipedia)
• iRobot (wikipedia)
• TV-B-Gone
• BugLabs
• OpenSparc
• Free Telephony Project
• Adafruit Industries/ladyada.net
• Akvo: The open source for water and sanitation

via | Ponoko

Citing various sources, Bauwens notices how the openness of this project is actually very limited (CAD files relate solely to the plastic shell, while hardware and its related software remain closed), especially if we compare it to OpenMoko. Therefore, here are four different levels of openness for Open Hardware projects:

Closed: Closed Hardware is any hardware for which the creator of the hardware will not release information on how to make normal use of the hardware, in such a way that that information may be freely shared with others. A sure sign of closed hardware is requiring the signing of an NDA to receive documentation on how to make use of a device.

Open Interface: In the case of Open Interface hardware, all the documentation on how to make a piece of hardware perform the function for which it is designed is available. In the case of computer hardware, this means that all the information necessary to produce fully functional drivers is available. This is the minimum level of openness that makes hardware useful to the open software community. Surprisingly, large amounts of integrated circuits fall into this category. Any device for which you can get a complete data sheet from the manufacturer, with no limitations on sharing the data contained within, meets the Open Interface definition.

Open Design: Open Design hardware is hardware in which enough detailed documentation is provided that a functionally compatible device could be created by a third party. It is not at all uncommon for the programmer’s guides for a microcontroller to have complete instruction encoding formats, memory maps, block diagrams of the processor core, and other technical details that would make it possible to reproduce a compatible microcontroller. Open Design hardware allows you to see what was implemented and what it should do, but still keeps the finer details of how it was implemented closed.

Open Implementation: Hardware for which the complete bill of materials necessary to construct the device is available fall into the category of Open Implementation. In the realm of computer chips, this means the hardware definition language description of the device is available. For a circuit board, this would include the schematic. Everything needed to reproduce an exact copy of a device is available. This is the hardware parallel to the concept of open source software. The debate between ‘open’ and ‘free’ (libre) that exists in the software space exists for hardware as well. In this regard, the only hardware that can truly be claimed to be free, in the same manner that the Free Software Foundation defines free, is that which falls into the Open Implementation category. Unfortunately, unlike software, an idea and the desire to produce a hardware device that is free and open is not sufficient. Certainly in the semiconductor space, the ability to do so is beyond the individual and in most cases, beyond even a reasonably equipped development group.

Finally, here Bauwens reports an interview to members of Open Graphics Project (appeared on the Free Software Magazine), the project of a completely open source 3D card. In this interview we can find what are the specific obstacles to the Open Hardware diffusion (and which may be useful to reflect on the deployment of a more generic Open Design).