Open Source Hardware and Industry
Here are some of my thoughts on the open source hardware movement and industry. This is not to say that we are only concerned with industry here, but more that we saw a lot of struggling over the years on how open source software could be beneficial to business and industry.
Open Source Hardware and Industry
Dan Goldwater, Squid Labs LLC
I believe that the most effective way to propagate socially progressive change is to show that it is beneficial to as many parties as possible – individuals, governments, small and large industry. As we have seen from the open source software movement, tremendous and clear benefits exist for individuals, small businesses and developing nations [1] [2]. On the other hand the benefits for large established industry and for building new industry around the movement were less clear and took some time to establish. This article seeks to shed some light on the beginning of an open source hardware movement and how it can benefit industry. Many unanswered questions still exist, but we should keep in mind that fundamentally hardware must be manufactured and this inherently provides more commercial potential than software.
To avoid confusion, we should note that the term “open source hardware” has been used by others to describe specific types of electrical hardware. We use the term in an inclusive way which refers any and all physical objects, and includes what is sometimes called “open design”.
Open source software has had a major impact on industry worldwide. Over the past 10 years it has grown from a fringe idea to a major component of the software industry. Organizations of all kinds are adopting open source software for critical business needs and using it to power mission-critical projects [3]. In a survey of 50 $1B+ North American companies, at least 70% reported use of open source software [4]. 8 of the 25 largest software manufacturers engage in major open source software activities [5].
As the impact of open source software has increased, there has been considerable analysis of the open source software phenomenon. How can ad-hoc, distributed groups of unpaid developers produce software which competes successfully with that from the largest software manufacturers in the world? There are many well thought opinions on this matter, yet considerable controversy remains. Much has still been learned: the development of open source software has produced not only the software itself, but has also improved the means of its own development with both innovative new tools and innovative new methodologies. While surveyed corporations have given a wide variety of reasons why they selected open source software for any given application [6], it is clear that open source software is capable of outperforming proprietary software on many common selection metrics [7]. In addition, many analysts believe the benefits of open source software to industry go well beyond the selection criteria for a single software product. Employees who participate in open source projects bring back best-of-breed approaches to the corporation at large [8]. The innovative development tools and methodologies that began in the open source community are now commonly used in many industries. Most importantly, it is often observed that open source is increasing the rate of innovation where it is used [9] [10] [11] [12] [13].
“Open source brings about the reorganization of millions of software developers into global collaborative communities, amassing a strength orders of magnitude greater than what is possible in the proprietary software realm” [14]
“Open source is the ultimate global reuse library” [15]
At Squid Labs, we have seen the increased rate of innovation and the benefits to industry first hand; as software engineers we have successfully used and contributed to open source software. As electrical and mechanical engineers we believe that proven open source methods can and should be brought to the realm of electrical, mechanical and other hardware and that the benefits to manufacturing and industry in innovation, performance, cost, reliability, as well as other less tangible areas will be significant. Furthermore, hardware unlike software must be manufactured. By taking the initiative in open source hardware development, manufacturers will be in the leadership position to produce new hardware developed under an open source model. We also believe it is likely that significant open source hardware development will support new business models in rapid prototyping and manufacturing services.
Why has this revolution in hardware development not happened already the way it did with open source software? We believe that the enabling technologies are a significant factor. The Internet is commonly identified as the enabling technology for open source [16] [17]. However, other key technologies such as a common open source development platform and collaboration tools have also played an important role, and we cannot underestimate the need for a supporting community. Open source is not a viable strategy until a sufficient developer community exists around it such that any individual developer can take advantage of a vastly greater resource than they could themselves produce. Because all these enablers are built with and upon the others, the incubation period for open source software spanned 10-20 years. Our goal is to dramatically reduce the incubation period for open source hardware by identifying and constructing the key enabling technologies now. We will accomplish this goal by combining proven methods from the successful example of open source software with our significant expertise in electrical and mechanical device development. At Squid Labs we have often struggled with the difficulties of sharing and collaborating on hardware projects with remote colleagues using different proprietary development software, different proprietary manufacturing software, different proprietary collaboration software. As has been famously said – “Every good work of software starts by scratching a developer’s personal itch.” [18]
We have already begun: Our website Instructables is the beginning of a system and set of tools which will enable and support the growth of open source hardware and its development.
References cited
[1] BBC News, “Brazil adopts open-source software”, June 2 2005, Steve Kingstone. http://news.bbc.co.uk/1/hi/business/4602325.stm
[2] United Nations University, “Free software in developing countries vital to future prosperity, good governance: UNU”, March 8 2006. http://www.eurekalert.org/pub_releases/2006-03/unu-fsi030806.php
[3] Computer Sciences Corporation, “Open Source: Open for Business”, 2004. pp. 2, 23. 2004. http://www.csc.com/features/2004/uploads/LEF_OPENSOURCE.pdf
[4] Forester Research Inc. “Your Open Source Strategy”, September 2003
[5] Berlecon Research GmbH, “FLOSS Final Report – Part 2”, July 2002, pp 11. http://www.berlecon.de/studien/downloads/200207FLOSS_Activities.pdf
[6] Berlecon Research GmbH, “FLOSS Final Report – Part 1”, July 2002, pp 33-54. http://www.berlecon.de/studien/downloads/200207FLOSS_Use.pdf
[7] David A. Wheeler, “Why Open Source Software / Free Software (OSS/FS)? Look at the Numbers!”, 2004. http://www.dwheeler.com/oss_fs_why.html
[8] Computer Sciences Corporation, “Open Source: Open for Business”, 2004. pp. 10. 2004
[9] David A. Wheeler, “Why Open Source Software / Free Software (OSS/FS)? Look at the Numbers!”, 2004. section 8.5.
[10] John D. Wolpert, “Breaking Out of the Innovation Box”, Harvard Business Review, August 2002, p.4 (reprint).
[11] Computer Sciences Corporation, “Open Source: Open for Business”, 2004. pp. 22, 44, 49. 2004
[12] Michael A. Olson, “A business case for open source”, May 2002. http://techupdate.zdnet.com/techupdate/stories/main/0,14179,2865356,00.html
[13] Tim O’Reilly, O’Reilly Media. “Ask Tim”, August 2003. http://www.oreilly.com/pub/a/oreilly/ask_tim/2003/compadv_0803.html
[14] Computer Sciences Corporation, “Open Source: Open for Business”, 2004. pp. 2. 2004.
[15] Computer Sciences Corporation, “Open Source: Open for Business”, 2004. pp. 18. 2004.
[16] Minoru Development Corporation, “Open source goes mainstream”. http://www.minoru-development.com/en/opensource.html
[17] Computer Sciences Corporation, “Open Source: Open for Business”, 2004. pp. 3. 2004.
[18] Eric S. Raymond, “The Cathedral and the Bazaar”, v3.0 section 2. 2000. http://www.catb.org/~esr/writings/cathedral-bazaar/cathedral-bazaar/ar01s02.html
March 23rd, 2006 at 3:45 pm
Very interesting subject. Forgive me if this is really obvious, but it did occur to me a while ago that, in effect, patents (on hardware) go some way towards ‘opening up’ complex technology almost in an open-source way - except that you’re prohibited from doing anything commercial with the information!
Nevertheless, there’s nothing to stop prospective innovators learning from reading patents: understanding how devices work, and spotting improvements that could be made, and either waiting until the patents expire, or applying the knowledge gained in another field entirely.
As you say, “Furthermore, hardware unlike software must be manufactured,” and I guess this means that the barriers to entry for open source hardware are much higher than for software. Anyone can copy and tweak and deploy open source software with nothing more than a computer, but until everyone has multi-functional desktop RP machines and CNC mills it’s a lot more difficult to do that with actual designs. You can provide the instructions (and that’s why Instructables is so great) but it still requires more equipment/time/capability to follow those instructions than it would do for software.
OK, so how can a business model be developed? Based on that ‘barrier to entry’ point, I think it comes down to this: even if every consumer product you buy had a full circuit schematic on the box, and a manual detailing exactly how to build it, most people would keep buying the products. They can’t make them themselves, or they can’t be bothered. They certainly are no threat to the manufactruers.
In the cases where vast numbers of people do enjoy tinkering with the products they buy (e.g. cars), it was certainly true until a few years ago that all the secrets of the car could be revealed with a bit of dismantling & rebuilding (and the aid of a helpful manual). And plenty of people have built businesses based on this: tuning, customising, restoring, rebuilding vehicles, all based on knowledge gained from examination and dismantling of existing hardware. Want collaborative development? Look at any classic car club or autojumble.
In this sense, we’ve had open source hardware for years staring us in the face, just as I believe the ‘free software’ idea only occurred to Richard Stallman when he saw that software was suddenly being locked up. Up to that point it had been assumed that code would be human-readable.
Pretty much all innovation relies on prior art, hence the ability to examine and understand that prior art. In the sense, all business that is based on innovation must be based on prior art. Therefore most ‘traditional’ hardware businesses are to a large extent ‘open source’, even if there is an artificial restriction in terms of patents.
March 23rd, 2006 at 5:44 pm
nicely done. i think i should start a listing/feature on make about open source hardware projects.
March 23rd, 2006 at 6:12 pm
pt - you already are! anything where you’ve explained how to do it is open source. probably you are thinking “something with significant intellectual property”, and even then you’ve had lots of stuff where the CAD files, firmware, etc. are published by the project author. the point where we’re not seeing much yet on the open source hardware side is:
(1) collaborative projects
(2) projects that are at the forefront of technology
March 24th, 2006 at 9:30 pm
What a great article, a coherent argument for open design.
There have been examples in the past, obviously and like you say Dan L, its really a reaction to open source software that means that we have to define the same for hardware/products.
For me the fundamental characteristic of any open source movement is the sharing of instructions to make a product, which in the hardware case is almost always restricted by patents. There are numerous examples of open source collaborative projects already existing on the web (I’ve been logging the ones I find on my blog, zero-waste.blogspot.com) although they do tend to be lo-tech for the very reason of ease of replication. But there are some with bigger ambitions. I guess that things are sure to become more hi-tech as the ability to synthesise products directly from computer becomes more ubiquitous. In the mean time, all that need be done is to share the existing source code: engineering drawings and schematics, sketches, anything that defines the creation of an object and you’ve got an open source project. There are plenty of people who can interpret them, and I don’t think it need necessarily be any harder than it is for software programmers.
Collaboration and the tools needed for it is indeed an area which needs to be developed, but I think instructables and the like are already close. Patents, as you say are an existing standard for communicating concept but a nightmare to navigate. The media needed is already there whether it be engineering drawings, CAD files, photographs, but it is a meaningful framework in which to exchange them that is needed. A kind of sourceForge equivalent perhaps? (he said, not knowing sourceforge that well at all)
May 3rd, 2006 at 7:19 pm
What are the tools needed to assist in hardware development for non-professionals? I for one would love to see good options for a logic analyzer(something like the digiview) or oscope using the computer as a front end, or an AVR studio interoperable ICE clone, I dont see why these couldn’t be realized with a huge reduction in cost over proprietary or even some kit options(especially with the dropping cost/rising performance of DSPs). What else would people out there find useful?
A lot of these costs are considered expected cost of entrance to work on these types of projects, you’ll have a much larger base of people able to work on and understand future projects if you can create an affordable standardized toolset.
Leave a Reply
squid-labs.com is proudly powered by WordPress
Entries (RSS) and Comments (RSS).