Today, Bug Labs interacts frequently with like-minded projects such as TuxPhone and companies such as Chumby and OpenMoko. But when Bug Labs began about two years ago, few of those efforts to develop open source hardware existed.
Instead, the inspiration for Bug Labs was personal, Semmelhack says. As he sees the situation, free software has increased productivity by lowering the barriers to getting involved. "You can do a lot today with very little code," he says. "You can put up Web sites for very little money, and you can put up ads on Web sites and start making money very quickly. If you are someone coming out of university today with a great idea for a new application, you can build it without a lot of investment."
By contrast, creating a new hardware product is far more costly and difficult. "If you want to do anything with hardware," Semmelhack says, "you have to spend a lot of money, because you have to buy materials. And what we've found is that you can't just go and order anything from anybody, because you have to order a minimum order in many instances. You have to buy 5,000; you can't just buy a couple. Or they won't sell to you because there's a waiting line, or you have to be on some approved list. In the world of bits, you just have to go to an FTP site, and you're done; in the world of atoms, you have supplies and inventories and investments that create a huge barrier for entry -- especially for a student who has an idea and just wants to go and build it. It isn't an engineering issue; it's an economics issue."
Manufacturing hardware also requires considerable expertise. "To come up with a pretty cool Web site, you only have to learn a language like Perl or PHP. If you wanted to build a gadget, you have to master arcane sorts of knowledge."
Semmelhack found inspiration from two sources: Memories of the Heathkits and Lego of his childhood, both of which lowered the barriers for entry in hardware and innovation, and his reading of Eric von Hippel's Democratizing Innovation, which argues that the largest source of product innovation is the needs of individual users.
"The mission for us," Semmelhack says, "is to create a platform that allows anyone to build any project they want by snapping it together like Lego. And to have the gadget be reliable, robust, and not too ugly or clunky-looking, and to make it easy to innovate in electronics in a way that hasn't been possible before." Software engineering expertise is still required to make use of Bug Labs' products -- specifically, expertise in GNU/Linux and Java -- but Semmelhack's plans are to remove as many of the hardware-specific barriers as possible.
The new Lego
The Bug Labs product line centers on the Bugbase, which consists of a CPU, RAM, a battery, and USB hubs and other connectors for adding modules. It runs a GNU/Linux distribution built around Openembedded. The largest modifications that the company has had to make to the Bugbase are device drivers, and improvements to make all of the modules hot-swappable.
The code for all these improvements, Semmelhack stresses, has been released. "We consider ourselves an active part of the Linux community -- and not just a taker, but a contributor. We chose Linux because everything we do is open source, software and hardware. Openness is very important for innovation."
Innovation starts with the device modules that you can attach to the Bugbase. Currently, the modules include a camera, LCD display, and a motion detector. Audio and wireless modem modules are due soon, as well as a "von Hippel module," which Semmelhack describes as "a module for building other modules." He adds that the module is named for von Hippel because it was inspired by von Hippel's comment that "you can't have a system that's open if you don't give people the ability to build their own stuff" -- and, he adds jokingly, because "von Hippel is a great mad scientist kind of name."
Bug Labs' plan is to add "20 or 30" more modules by the end of next year, so that users can build any sort of device that they can imagine. "Our goal is to come up with as many of these modules as we can," Semmelhack says. "One of the things we like to say is that we would like to have our unfair share of developers' imaginations."
With these building blocks, the Bug Labs community is starting to release its own devices. So far, they include an Asterisk server and a radar detector that locates police speed traps and posts them on an online map. "A lot more applications are coming out of the community than we would have thought of for ourselves, which is exactly what we want," Semmelhack says.
The plans for these community-built devices are being posted on the company site. So far, Semmelhack says, everyone is using the GNU General Public License or another free license, although he suspects that might change as other companies become interested in Bug Labs.
Making money from open source
Asked why a company should buy Bug Labs products, Semmelhack replies, "We have a very low transaction cost model, meaning you can use all of our intellectual property for nothing. You can download and start working with it for nothing, so the barrier for entry is very, very low. I think we appeal for the same reason as Linux. There's no cost to experiment, and no cost to start playing around."
And what is Bug Labs' business model? "Our biggest market going forward is selling hardware, obviously," Semmelhack says. "But we'll also be selling services for hardware around that stack." Possible revenue sources include custom programming and devices, brokering deals for companies that need specific expertise from the community, and a fee for listing commercial devices on the Bug Labs site. Hardware maintenance and software update services are also possibilities.
Semmelhack acknowledges that nothing in Bug Labs' business plan prevents another company from using its intellectual property and building cheaper modules. However, by the time that happens, he intends for Bug Labs to have a large enough community that it remains at the center of things.
To that end, Bug Labs gives considerable attention to meeting the community at demos and Maker Faires. "Because this is a device, a physical object, people like to hold it and play with it in the proximity of others," Semmelhack says. "So, for us, the more physical we can get in terms of building the community, the better."
Such attitudes may prove as difficult for hardware manufacturers to accept as free software has been for software vendors. However, Semmelhack suggests that such models are the wave of the future. "In the same way that the recording industry has had to be dragged into the future, proprietary models for protecting innovation are going to go away. And I think that, in the future, big companies are going to be built, not on patented, proprietary models, but on open areas of innovation -- areas where communities have been harnessed to create whole new ways of doing things.
"What we're trying to do is identify a market. We're trying to bring something to the world that hasn't really existed before, and that's choice. In the past 20 years, control has been moving from the hands of manufacturers to those of the end users. When I was growing up, you watched TV according to what NBC, ABC, and CBS had to say about it. But today, you can watch it any time you want, even without commercials, and that has completely upended the broadcasting community. I think the same thing is happening everywhere with control of all the products in our lives. Ten years from now, we're going to look back at this time and ask how we endured letting big companies dictate to us what we could buy. It's going to seem bizarre. And we're going to make that change happen in electronics."
Whoever gets the job will be working with Obama administration's new cabinet-level position: 
Last week we reported that 
