Mid March it was my turn to provide the session in the Escuela de los Commons (School of Commons) in Barcelona. I had promised to present (see slides) the perspectives of commons and free knowledge applied to “industrial production”. In this post I will try to summarise the session.
Wikipedia has won much “marketshare” (or: “mindshare”) while the Encyclopedia Britannica has stopped the presses. Their production model – so called “commons-based peer production” – is proving hegemonic at least in the area of encyclopedias. This logically draws resources from the market into the commons.
Something similar has happened in the domain of software. Over the last 3 decades, Free Software (a.k.a. Libre Software or Open Source) has shown the success of commons-based software projects. Studies have shown that Free Software has reduced the size of the software market by an average factor of 10. Of course people can earn a living working with Free Software, mainly by offering their services, but the Free Software commons tends to keep those market exchanges in balance and avoids monopoly formation or market domination.
Now what could the Commons mean in the domain of physical products? In particular for industrial products? That’s where many people have their jobs, where most of the population spends a large part of its money. I’d like to explore with you what impact commons-based models can have in this important domain.
Let us start by reviewing the main characteristics of the industrial production model and detect some of its shortcomings (Part I). Then we explore the commons, first of all the free knowledge commons, to see what perspectives this offers for alternative production models (Part II). In the last part we detect some of the obstacles and challenges that lie ahead to go from the current embryonic state of Free Hardware to transition to an industrial production model based in Free Knowledge and Commons-based organisation.
Part I. Mass Production
Let me confess here that, even though in the 1990′s I studied Mechanical Engineering and ended with a MSc in Production Operations Management, my in-depth knowledge about manufacturing is limited. I’ve worked just a few years in multinational corporations and I soon had enough of that organisational model. Then I started to work for myself, launching a small company offering services based on Free Software. In 2007 I co-founded the Free Knowledge Institute to expand the idea of Free Software into other domains of knowledge. My background as engineer might allow me to grasp the main processes in production organisations, but not of the details that go into designing them. That gotten out of the way, let us start with a quick overview of some characteristics of industrial production organisations.
Mass production refers to the production in large scale, that became popular during the industrial revolution. The steam engine enabled an industrial class to produce large quantities of a particular product and become powerful. Over the years more efficient machinery has been developed, especially with the advent of electricity. Highly specialised factories were set up to produce a particular product or line of products. Large concentrations of capital were necessary to make this happen. Because of large quantities and uniform outputs, the costs per unit could be rather low. It is still this low unit costs – or economy of scale – what makes large parts of industrial manufacturing turn. However, what happens when not all units are sold? To make sure most of the products are indeed sold, expensive marketing campaigns and sophisticated so called “supply chain management” solutions have been developed. To make sure shareholders get back their investments and profit for the risk they’ve taken, the whole organisation is centrally organised under control of strict hierarchies. The research and development of new products had to be strongly protected, through secrecy, copyright and preferably patents. Conventional management theory holds that for reasons of profit maximisation, companies should seek to obtain a temporary monopoly. When holding the monopoly, most profits can be made. Therefore (speedy) time-to-market and control over that market is crucial. Apart from patents, cartel formation and government licensing have shown to be strong tools to hold back newcomers to compete effectively in the same market.
Now this may seem a caricature of actual manufacturing and business processes, but these principles can be detected when studying most industrial manufacturing companies – and abound in management literature. Of course work has been done to improve and optimise their workings.
- The push marketing model has been made more adaptive to and dependent on actual demand (cf. pull/demand-driven, JIT, … leading to less stocks).
- Production lines have been made more flexible so they can be used to make other products and therefore adapt to changing demand.
- Open Innovation and crowdsourcing has helped to shift part of the costs from R&D to key users. E.g. wind surf boards based on user generated innovation may perform better, and at a fraction of R&D costs for the company. Note that “open innovation” has little to do with “open source” or “open design”. Instead open innovation refers to the trend of lowering the walls between internal and external participants, where firms come to embrace good ideas that come from outside the firm itself. In the right direction, this is however no commons yet.
- Participatory and User Centred Design methodologies (“design thinking”) have helped companies align their product development with the interests and needs of users, and have shortened the time-to-market.
Eric Hunting has observed a trend of demassification: “by the year 2000 consumer goods are now more produced in “job shop” facilities – mostly in Asia – than in traditional Industrial Age factories.” In other words, large scale manufacturing is making way for small and medium sized production, in more flexible and smaller factories. This relates to manufacturing networks, where agile small manufacturers work together to quickly adapt to new products and demands. The Shanzai networks in China have been cited to show the potential of these networks. Shanzhai are collaborative manufacturing groups that suggest a new approach to economic recovery based on small companies well-networked with each other, sharing innovations and manufacturing information.
Globalisation in the twentieth century has led to move factories towards low cost labour countries. This in turn has led to a boom in global transportation of a) primary resources, b) unfinished goods and components, and c) of finished goods towards the markets of mass consumption. However, there has been observed a decline since 2008. Reasons for that may include: 1) the resource crisis and increase of fossil fuel costs (PeakOil), 2) a decline in consumption in the Western world and 3) the mentioned demassification and start of relocalisation.
The advocates of mass production have identified the superior efficiency of the large corporation with its control over the external environment. But before we continue, let us draw a quick summary of the shortcomings of the large scale industrial model.
- Planned obsolescence: products of too high quality make their consumption decline. Nowadays most consumer goods are designed for a limited number of years. GM designer Harley Earl in the 1950s confessed: “My Job is to hasten obsolescence. I’ve got it down to 2 years; now when I get it down to one year, I’ll have a perfect score.”
- Increasing (process) innovations lead to increased productivity, which would be excellent as it means less work we have to do for producing the same output. But no, this causes a serious problem in our socio-economic model, as it leads to less jobs;
- concentrations of capital and power tend to lead to distortion of political and social processes;
- profit maximisation leads to neglect negative externalities (violence, pollution, …)
- non-alignment of interests between producers and customers (agency costs): car makers nowadays make most profit in after-sales: they have an incentive to limit reparation to their licensed car repair centres. This results in strategies for non-standard products that are hard to repair by others. Legislators can soften this tendency. As an example we can take the EC’s recent move to enforce standard power plugs for mobile phones.
Imagine manufacturers would make products that were compatible with those of their competitors. That could lead to increased competition, easier repair, reuse, the emergence of industrial networks.
Before jumping to the commons, let us take conscience of the importance of knowledge and information, which have become the dominant factor in the economy and in society. Studies have shown that the value of industrial companies is made up mostly by intangible assets, in many cases up to 90% of the value is constituted by brands, copyright, patents and other forms of knowledge embedded in the organisation and their relation within the network. More concretely Tom Peters concluded that about 90% of the costs of a Minolta camera comes from “intellect” and not from labour or materials, in other words, 90% of the costs are the tolls to corporate gatekeepers in control of that “intellectual property”. Kevin Carson says about just that: “So in a genuine free market, where “intellect” cannot be owned, the pressure of competition would — by his estimate — drive down the price of his camera by 90%. The other nine-tenths of the time we work to earn the money to buy it, is the equivalent of what the feudal peasant paid to his overlord for the “privilege” of eating the bread that he produced with his own hand.”
The trend that our economies would rely more and more on knowledge and information, and services, has long been predicted by people like Daniel Bell (“Post Industrial Society”), Robert E. Lane (“Knowledge Society”), Peter Drucker (“Knowledge Economy”), Frank Webster (“Information Society”), Manuel Castells (“Network Society”, …) Only there is a tension between knowledge and the profit-maximising, exclusive property model. In an article I wrote in 2010 with David Jacovkis and Hinde ten Berge, we made a brief analysis from historical and social perspectives about the Knowledge Society – from a freedom centred perspective. There we explore also the emerging free knowledge commons as a mode of production and organisation. And it seems that while the privatised, enclosed knowledge model is showing more and more inefficient, with shortcomings like the ones presented above, the free knowledge commons is presenting a rather more attractive alternative. At least that is what can be inferred from the many emerging positive experiences.
Part II. The Free Knowledge Commons
I’d like to compare the Commons-based model with the proprietary model and see what alternatives it can provide. Before that, we should start with some definitions.
First the classification of property models by Elenor Ostrom (Nobel Prize). She and her husband classified goods in the four categories as seen in the table.
|Subtractability / Rivalry|
|Excludability||hard||Public goods||Common Pool Resources|
|easy||Club/toll goods||Private goods|
Ostrom, V. Ostrom, E. 1977. Public Goods and Public Choices.
Common Pool Resource (CPRs) are hard to be excluded from, but they can be depleted: it can be used up. To make sure freeriding and overuse don’t become a problem, CPRs are managed. We know now that that was one of the mistakes Garret Hardin made in 1968, when he wrote the Tragedy of the Commons: he supposed unmanaged common resources.
James Quilligan (Global Commons Trust) prefers to define Commons as “inherited or created gifts that we organise, use and store in our lifetime through informal practices and rules which we pass on to future generations”.
Ostrom’s above classification is useful especially for natural resources, but less so for knowledge or intangible resources, as they are non-rival. With knowledge counts the principle of “the more, the merrier.” More users tend to enrich a knowledge resource. Following this line, the so called network effect is at work when people continue using an inferior product, such as MS Office, when a similar product developed as a commons like LibreOffice is available at no cost and provides similar or even better features.
Let us define Free Knowledge. The first definition was made for Free Software – a concrete and highly complex form of knowledge, by Richard Stallman in the early 1980′s. Based on that, similar definitions have been developed to cover a much wider area of knowledge. A very reputed one is the Free Cultural Works definition:
- The freedom to use and perform the work: The licensee must allow to make any use, private or public, of the work. For kinds of works where it is relevant, this freedom should include all derived uses (“related rights”) such as performing or interpreting the work. There must be no exception regarding, for example, political or religious considerations.
- The freedom to study the work and apply the information: The licensee must be allowed to examine the work and to use the knowledge gained from the work in any way. The license may not, for example, restrict “reverse engineering”.
- The freedom to redistribute copies: Copies may be sold, swapped or given away for free, as part of a larger work, a collection, or independently. There must be no limit on the amount of information that can be copied. There must also not be any limit on who can copy the information or on where the information can be copied.
- The freedom to distribute derivative works: In order to give everyone the ability to improve upon a work, the license must not limit the freedom to distribute a modified version (or, for physical works, a work somehow derived from the original), regardless of the intent and purpose of such modifications. However, some restrictions may be applied to protect these essential freedoms or the attribution of authors (see below).
In the field of physical products we can apply these four freedoms to the information to produce the product. This is what we call Free Hardware. And when covering really everything needed to manufacture a certain product – from design files (CAD), documentation, Bill-of-Materials, manufacturing information (CAM), the software needed, etc – we talk about Free Manufacturing.
In another post, introducing the various terms used in the Free Hardware Movement, you’ll encounter various similar terms, with often only small – but relevant – differences:
- “Open Design” refers to having access to design files via Internet: this is a minimum requirement and some use it as an all encompassing term, it however doesn’t guarantee the right to modify and distribute work based on that of others.
- “Open Hardware” is hardware that results of open designs, similar to open source code. Note that Open Hardware is also used as a trademark of the Open Hardware Specification Program. It doesn’t guarantee all uses people can make of it (e.g. to make derivative, adapted versions, and sell these).
- “Free hardware design” refers to a design which can be freely copied, distributed, modified, and manufactured. It does not imply that the design cannot also be sold, or that any hardware implementation of the design will be free of cost.
- “Open Source Hardware” originally referred to computer hardware for which all the design information is made available to the general public. It has been redefined to equate to Free Hardware, “hardware whose design is made publicly available so that anyone can study, modify, distribute, make, and sell the design or hardware based on that design”: http://freedomdefined.org/OSHW
It becomes clear that many people use any of these terms to refer to the hardware knowledge protected by the four freedoms, and while Open Source Hardware is now largely (re)defined along those lines, I tend to use “Free” when referring to anything that comes with those freedoms, while I understand “Open” is associated to “open access”, which is a precondition to the four freedoms. Hopefully the definitions help us to keep things clear
How do we guarantee the freedoms? Such hardware is generally protected by licenses, much in the same fashion as free software and other free licenses. This means their protection is based on copyright, not on patents. Note that the license thus only governs the plans (computer files), but not the resulting device. Using patents has proven extremely cumbersome for putting a hardware design in the commons. But patents are thé tool to “protect” ideas or product designs. Is this a serious problem?, we could ask. But once a design is published on the Internet (with a clear time stamp), we have proof of “prior art” in case some profiteer would like to patent our idea a posteriori. That should avoid the enclosure of the product design in particular, – or give us a tool to make the patent claim void. Of course, someone might make use of our free hardware design, make a modified version and keep the modified design for himself, while selling the resulting products. Apparently, that’s what happens quite frequently. This seems hard to avoid, and it is questionable if it constitutes a serious threat to the hardware commons at all. Most licenses at least come with the copyleft clause, which means that they require all derivative works also to be published under the same license. So if the person in this example would like to publish her modified version, she would need to do so under the same free license, and thus contribute to the commons.
Licenses used include the free licenses within the realm of Creative Commons, the GNU FDL and several specific free hardware licenses: the TAPR Open Hardware License (OHL) and the CERN Open Hardware License (CERN-OHL). A special case is the OHANDA trademark: this community uses their (non-registered) trademark to label a free hardware project with the four freedoms. You can compare it with other quality certificates that consumer products often carry. Only products that have their designs published under a free license and registered on the OHANDA website, can attach the OHANDA label to it. The label combines copyright and trademark laws, and is expected to be stronger to protect the hardware commons.
Now it’s time to review a few cases. Cases that you will most likely already know, but that I’ll use to compare with the conventional industrial production model. For each case we’ll see how they organise the core functions of a manufacturing organisation: R&D, Marketing, Production, Transport and Finance.
R&D: by the Arduino team + key users in the global community, in a distributed commons under free licenses.
Production: through an ecosystem with official and unofficial boards and shields. The first production order in 2005 was of 200 units produced by SparkFun. Later came AdaFruit and other companies who started to manufacture boards. In 2008 the Arduino founders set up their own robotised flexible manufacturing cell in Ivrea to manufacture their boards themselves. Until 2008 50.000 units have been produced.
Marketing: by the community; no large marketing budgets are needed.
Transport: from manufacturer to end user mainly through cheap and quick international courier services.
Finance: little capital is needed and small batch production is viable with product prices of 20-30 €/unit
R&D: the machine is developed by the maker community under free licenses, and thus constitutes a commons; the product designs are – at least partially – shared as a commons (through Thingiverse, etc)
Production: the machine can be DIY/self made, or with others in workshops, or can be bought in self-assembly kits from the founders’ company Ultimaker Ltd. With the machines, products can be printed very low-cost at home, in hackerspaces, FabLabs, in any place where such machine is available; it basically consumes some plastic and electricity.
Marketing: done by the community; no capital-intensive campaigns
Finance: there is little capital needed to get one machine, about 1200 € for a kit to self-assemble the machine. The founders won’t need much capital either to run their machine production shop: the manufacturing of the machines is outsourced to a third party and most functions are performed by the community.
Organisation: the main functions are performed in a distributed network, so no centralised hierarchies are needed. It has been said that the right to fork provides the ultimate protection against hierarchy. And this leads to a strong advantage of the knowledge commons above the proprietary model: one project builds on the results of others and participants innovate by modifying and improving existing ideas, while no contract negotiations are needed to obtain any permissions.
Open Source Ecology
To get to a more complete vision of the commons in industrial manufacturing, the Open Source Ecology movement presents a radical proposal. “Imagine if you could build cars, industrial robots, engines, and other things in your own back yard. The only problem is, these require billions of dollars of infrastructure in the current industrial system. Not for long – if we succeed with the Open Source Micro-Factory.” Marcin Jacubowski and a growing community around the world is contributing to this vision. Their first objective is to produce the 50 tools for producing most products for modern comfort – the “Global Village Construction Set”. See my other post with an in-depth case study on Open Source Ecology (OSE).
In short, OSE bases on freedom, open development and self-sufficiency. Local communities are started, working together through global networks. Their principles include ecology and design for durability and reuse (modular, standardise, repairable, …). Also on the level of exchanging products and services with the market, their business plans and knowledge is shared with other groups who would like to follow the example. They call it the “distributed enterprise”. The resulting machines are around 8-10 times cheaper than off-the-shelf industrial versions – but may require self-assembly. The factory model they are working on is one of maximum flexibility, while aiming for efficiency and high performance. We could call it a job shop, or microfactory.
If we study the main organisational functions as discussed in the industrial factory model in the beginning of the post, we can come to the following model.
- R&D: global community
- Marketing: global community
- Production: local community
- Financing: crowd funding + market exchange
- Transport: little, mostly local
- Environmental Impact: low
- Social: less work, more autonomy, creativity, solidarity
Open Source Ecology has shown initial prototypes for various machines, but needs still a long way to reach maturity. The initial results, further blueprints, crowdfunding, plans and models show however the path to reach that. See for more details a case study in another post at the Microfactory Blog.
Summarising, it seems to be increasingly possible to produce industrial goods through the Commons. The main function of industrial production can be organised as a commons by applying straightforward rules to share knowledge. The Commons offers – at least potentially – the following advantages for industrial manufacturing over the proprietary model.
- No need for profit
- Low overhead
- No “Intellectual Property” costs, no monopoly rent
- Takes intangible assets out of the RoI equation (*)
- The result of all this is that much less financial resources are needed.
Regarding the functions of industrial production organisation, we have seen the following changes:
- Production: from mass scale towards flexible, small-scale, production and in distributed networks
- This can lead to relocalisation
- R&D: knowledge can be shared as a knowledge commons for design files, manufacturing information etc
- Marketing: community members interested in the project and its results help spread the word through social networks, no (big) marketing budgets needed.
- Customers turn into users, co-owners, participants, True Fans (Clay Shirky: a free culture project just needs 1000 TrueFans to sustain themselves)
- Finance: Crowdfunding, donations, market exchange
- Organisation: distributed networks of individuals, companies, cooperatives, associations, foundations
(*) While intangible assets can be taken out of the business equation, this does not mean that the costs of producing the needed knowledge becomes zero. As observed by David Jacovkis and Javi Creus during the session, researchers, developers etc are very much needed. But instead of accounting their value as “intangible” assets of the conventional company, their results become available to all at low cost and in such way that no such assets can be accounted for in conventional book keeping. There’s simply no proprietary hook that allows for exploiting the product and obtain serious margins: if there was, anyone with the right interest and capacities could do the same and drive the margins to zero. So at least in theory there is no place for these intangible assets in the books.
We may ask if there is a threat of not providing the necessary incentives to generate the necessary R&D for society’s needs. While most areas of knowledge are taken care of by people with direct needs to solve a personal or community need, there maybe areas where additional incentives are needed. Think of certain medicines that aren’t developed if society doesn’t designates special funds or privileges to develop them. As has been shown by several researchers, the current situation doesn’t resolve these needs very adequately either, where medicines for non-western diseases in low-income countries are hardly invested in by the pharmaceutical industry. Several solutions have been proposed. One could be to designate public or collective funds for particular developments in areas of high social needs, and assure that the results are equally available to all companies and individuals.
Arguably the biggest threat to developing a vision along these lines comes from Patents. Patents are a system of state-granted private monopoly that may have served its purpose in the past, but is becoming a growing obstacle to the commons (and society at large). Incumbents use them to threaten new entrants and to protect their dominance in a particular market. Patents provide at least two obstacles for the Commons. First, a person or group with an interesting idea may be tempted to register a patent to get at least the illusion of one day becoming rich. Fortunately it’s very costly and time-consuming to acquire and effectively use a patent for a new entrant. Second, many ideas have already been patented, which can be an obstacle for designing and building practical solutions for everyday needs (a legal mine field). While patents generally expire after 20 years in most cases, this can be a problem in the case of new technologies. And even more so when we consider that patent descriptions are often rather vague to be as widely applicable as possible. One way to go around this is to build for yourself and avoid market exchange. Possible patent infringement can hardly be effectively checked for personal use or within a community. Better would be to get rid off the patent system at all…
Demands to the State
A libertarian would say: we just need the state to take her hands off of the commons. Current reality has however many state interventions, like patents. We should work for 1) breaking down this kind of corporate privilege, and 2) positive policies towards the commons.
- Stop relying on (or eliminate!) patents
- Open Standards, compatibility
- Require published specifications
- Favour non-market sharing
- Public funded research into OA repository
How to get started?
We need basically three things: Spaces, People and Resources. Spaces where we can get together, learn, experiment and produce. People specialised in various areas and with different interests: some might want to take the lead, others follow, some may want to experiment, test and produce and make a living of these activities, while others may just want to consume its results. Resources can be the tools and raw materials or the financial resources to obtain them.
Additionally we need a general vision, rules and governance frameworks. Find some in general aspects. I set up a library with some links to web sites in these categories. You’re welcome to enjoy it and help me build that library.
We have seen how the paradigm of free knowledge and the commons are demonstrating successful in various domains. The oceans, air and woods are long time commons, often threatened by privatisation. In more recent years, the Internet, Wikipedia and Free Software have emerged as the most visible ones of the knowledge commons. Recently physical products and electronic hardware are increasingly designed according to the same principles and rules as found in the free knowledge commons.
The industrial sector – despite its decline in favour of the service sector – is still dominant in economic terms and arguably the hardest one to organise as a commons. In this post we reviewed the main shortcomings of the industrial production model as we know it and discussed the way in which we can imagine a commons-based industrial production taking shape. In order to do so, we reviewed the ways in which some well known free hardware projects are organised in comparison with the conventional industrial model.
In short, in a commons-based industrial production model we see how the knowledge is shared as a commons: R&D and manufacturing knowledge is as much as possible shared through Internet platforms. Given the distributed nature of participants and projects, one builds on top of the works of others, develops and publishes improved iterations which in turn are also available to next generations of users. Communication goes by word of mouth, assisted by ever smarter social networks, while the community may run campaigns to get a sufficient number of users or investors in a crowd-funding campaign. The production doesn’t need large scale factories any longer for most types of products. Small and mid-sized batch production is done in flexible job shop facilities, which can rapidly adjust to new demands and don’t need much capital to start off. Though a fully commons-based, cooperative structure might make the production facility itself more a commons, this is not needed to shift towards the envisioned commons-based industrial sector. After all, some may produce to satisfy their own needs – as we saw in the Open Source Ecology vision – while others may produce for the market. When the knowledge itself becomes truly a commons, market exchange can allow people to produce and consume what they need according to their interests and specialisations.
We saw how commons-based production tends to reduce the market-size. Indeed, when a commons-based project becomes successful, there is an exodus from the market where those services/goods had been produced previously, to the commons. See Wikipedia vs. Brittanica or Free Software for that matter. When the commons makes its inroads in the industrial production, we can expect the same to happen. This promises to make socially attractive products and solutions available for much lower costs than they cost today. Think for example renewable energy, electric vehicles and food production technology. Once a sufficient number of people directs their attention to these areas, we can see better and cheaper solutions. In fact the increasing number of renewable energy projects in the free hardware community, of permaculture and “open source cars” suggests exactly that.