BugLabs and open-source hardware innovation

Lunch talks at the Berkman Center usually promise a challenging room, filled with smart people asking tough questions. But it’s rare that speakers have as tough an act to follow as Peter Semmelhack of Bug Labs whose lunch talk fell directly on the tail of President Obama’s inaugural address, watched on lossy streaming video by a room full of Berkfolk.

Semmelhack is the Founder and CEO of Bug Labs, a group focused on “bottom-up, community based innovation” in the hardware market. His talk, “The Long Tail of Gadgets”, focuses on how open source hardware is enabling this bottom-up innovation. Semmelhack’s vision for hardware is that we’ll move away from a small number of companies building gadgets with markets of millions, to millions of innovators creating devices for the few.

He offers one of these niche devices. As a New Yorker, in the immediate aftermath of 9/11, he found himself wanting to know the physical location of his wife and child. He wanted to build a GPS tracker which sent a signal every five minutes allowing him to track his family on a website. He’d hoped to find a device that served the job. When he didn’t, he looked for a Heathkit-like solution that would allow him to hack a device together. He wasn’t able to find a platform he could build on.

It’s possible that no one created the device because it was only interesting to a small niche of people. Or because, as Semmelhack explains, consumer electronics is “a crap business”. Due to falling profit margins, the tendency for competitors to knock off your product, it’s not sufficient to have just one hit product – you need a supply chain of hits. Apple’s done this, as has Nokia, but few other companies are capable of doing this.

To change the model so that any business can build devices and so customers can innovate, we need lower barriers to entry. Hardware design needs to work a bit more like software. Innovators in software can use a wealth of open-source software to build their new products using the LAMP (Linux, Apache, MySQL, Perl/Python/PHP) stack of tools to roll out web services. As a result, innovators in the space are able to start new companies with very little upfront investment, which is challenging existing models of venture capital.

Innovating with hardware is sigificantly harder. You can’t simply start building a product. Instead, you need to design schematics, put together a bill of materials, design Gerber files to print circuit boards… Beyond that, to successfully design a product, you need to master outsourced supply chains and manage relationships with vendors in many different countries. But there are ways to make this easier – Lego Mindstorms is an amazing systems to build certain types of tools. They’ve open sourced the code and the hardware, and now offer a service called Lego Factory which allows fans to build novel models and pieces and have them produced by Lego’s real-world factory.

The ambitions of BugLabs are similar – they’re building a platform that offers pieces and parts, a mix of software and hardware, designed to be a platform for building new tools. The major opportunity Semmelhack sees is around the wireless/GSM space. He points out that the spread of PCs and phones has allowed wireless networks to grow. But now these markets are pretty thoroughly saturated. Network operators want new devices that use their networks. It’s possible that these things will be built by small compaies and consumers, rather than by large manufacturers.

BugLabs is in the right place, Semmelhack believes, because we’re seeing the rise of “extreme personalization”. You can already design your Mini Cooper online or your Nike shoes. With open source apparel manufacturers and open source shoe manufacturers, it’s possible that new, disruptive innovation will come from the bottom up. BugLabs makes money supplying this bottom-up innovation, selling kits to individuals and research groups at companies. The design decision, in some cases, is between soldering something yourself and buying devices from BugLabs, using their technology to scale from an idea to a production level. In this sense, the company is deeply old-school, as they make their money selling atoms to people.

I wondered whether LAMP was the right analogy for a hardware company. There’s another form of software development – game development – which has enormous startup costs, a tendency towards large teams, and budgets that can exceed that of Hollywood films. LAMP-based products tend to succeed based on network affects. Consumer electronics, on the other hand, tend to be highly personal and based around elegant interface… and interface is what open source tends to do worst. Isn’t consumer electronic design more like game design than LAMP, and inherently hard to scale?

Semmelhack makes it clear that his model isn’t to replace the consumer electronics world, but to add choice to a world that has very little right now. He notes that consumer devices focus on the increase of choice. It’s hard for young consumers to imagine TV that you can’t pause, once they’ve experienced Tivo. Perhaps this choice will extend into the world of consumer devices and people will want the choice between elegant, closed systems and bottom-up innovation.

Some of the most interesting examples of what people are building with BugLabs are coming out of the developing world. His group is working with Grameen Foundation in Uganda, and has been talking with open source hackers in Kenya. He tells us about a group in Kenya that’s been building a collective pothole mapping project, using an accelerometer to measure the incidence of a pothole and GPS to map the locations on a collective map. It’s a very cool idea, and one that mainstream electronics companies are unlikely ever to build. If BugLabs can make it easier to build devices like this, it’s probably a deeply good thing, in my opinion.