Hey, Silicon Valley: Wake Up and Smell the Robots
Investors and entrepreneurs in the Bay Area pride themselves on being the first to identify and exploit new technologies with huge commercial potential. And they’ve earned the right to be a little cocky. Since 1960, Silicon Valley companies have been the pacesetters in four consecutive infotech revolutions (semiconductors, personal computers, the Internet, and mobile).
It’s strange, then, that the Bay Area innovation community seems content to let the next big technology revolution happen elsewhere.
I’m talking about robotics, the subject of Xconomy’s next big public forum on May 3. After decades stuck behind bars—or rather, behind safety cages on assembly-line floors—robots are finally getting out on parole, and are starting to mix with humans in a range of real-world situations. Just look at the iRobot vacuum cleaners battling our dust bunnies, the self-driving cars Google is sending up and down Highway 101, the InTouch health medical telepresence bots rounding in hospitals, or the prototype rescue and disaster-response robots that inspired DARPA’s latest Grand Challenge competition. It is clear that the developed world is reaching some kind of threshold, and that before this decade is out, many of us will be living, working, and driving alongside robots.
But for the most part, this is a market that’s being opened up by companies outside Silicon Valley. Boston boasts a major cluster of robotics companies (iRobot, Kiva Systems, Bluefin Robotics, Boston Dynamics, Foster-Miller, etc.) and so does Pittsburgh (Carnegie Robotics, Cardiorobotics, RE2, Robomatter, etc.). Even Southern California has its share of bot makers (InTouch Health, Evolution Robotics). But while Silicon Valley has seen one or two big success stories in robotics—notably industrial robot maker Adept Technology and Intuitive Surgical, maker of the Da Vinci microsurgery robot used today in more than half of all prostatectomies in the U.S.—the local robotics industry is still mostly a story waiting to happen.
Part of the blame can be placed on Sand Hill Road investors, who’ve been too busy chasing startups in areas like games, social networking, enterprise software, and e-commerce optimization to pay much attention to actual hardware lately. It’s a significant blind spot, because these are the technologies that will determine how we work, how we get around, and how other things gets made. “It’s pretty clear that Boston has a savvy group of venture investors who know how to look at robotics,” says Rich Mahoney, director of the robotics program at SRI International in Menlo Park, CA. “But the Silicon Valley venture community doesn’t understand the field that well.”
Investors haven’t been the only missing ingredient. Robotics in the Bay Area has also lagged due to talent shortages, a lack of collaboration between companies, a dearth of options for engineers looking to change companies, and other systemic limitations that aren’t seen in the local software and Internet sectors. For the most part, Bay Area roboticists have been laboring in lonely silence, in garages, university labs, or industrial parks scattered from Berkeley to San Francisco to Palo Alto to Pleasanton.
But all that may be starting to change. Mahoney at SRI is one of the organizers of Silicon Valley Robotics, an informal industry association whose first major meeting in January was attended by more than 100 people representing 30 to 40 organizations. This month, to help celebrate National Robotics Week, the group threw a robot block party at Stanford University’s Volkswagen Automotive Innovation Lab that attracted 1,500 visitors. “There was a solid community there,” says Mahoney. “The group’s role in the valley is getting established, at least at a modest level. And there are more companies that have emerged.”
I’ve had robots on the brain lately because I’ve been preparing for The Future of Robotics in Silicon Valley and Beyond, the event we’ll stage this Thursday on the SRI campus. One of our goals for the event—which we designed with Mahoney’s input—is to assemble ideas from inside and outside the Bay Area about what robotics researchers and entrepreneurs here can do to maintain and increase the early momentum that Mahoney is talking about.
To the extent that this momentum is rooted in old-fashioned technological progress, it may gather on its own. Like all areas touched by computing, robotics benefits from advances in semiconductor manufacturing. “At Precise Automation in San Jose, they now sell vision systems for less than $3,000 that I used to sell at Adept Technology for $25,000; that is just Moore’s Law, applied,” says Charlie Duncheon, a veteran of the local robotics scene who now heads Grabit, a maker of novel gripping tools for industrial and materials-handling robots. On top of that, concepts from the open source movement, which gave rise to so many of the other key components of the Internet explosion, are finally taking hold in robotics. A few weeks ago I profiled Willow Garage in Menlo Park, where engineers are pouring much of their effort into a common platform for robotics innovation called the Robot Operating System (ROS). Willow Garage’s Brian Gerkey (who, like Mahoney and Duncheon, will be speaking at our event) calls ROS a “LAMP stack for robotics,” referring to Linux, Apache, MySQL, and PHP/Perl/Python, the basic underpinnings for most Web services today.
Certainly, there’s growing demand for better, faster, cheaper robots. Foxconn, the Taiwan-based electronics manufacturer now famous (or infamous) as one of Apple’s biggest manufacturing contractors, says it wants to install as many as 1 million robots on its assembly lines over the next three years, partly because human labor is getting too expensive. That’s the kind of scale that spurs innovators to discover new efficiencies. “I certainly see the economies of scale improving in robotics,” Duncheon says. “Not to the extent that they have on the Internet, but the economics are going in the right direction.”
Nirvana, for the robotics industry, would be reaching a plateau of standardization and interchangability like the one the PC business reached in the early 1990s. Mahoney thinks we’re getting there. “In the PC industry you didn’t just have a computer, you had the operating system and software that ran on it, and you had the hard drive and the motherboard and the peripherals,” he says. “You are starting to see the same thing with robotics. You may have a robot arm, but you can’t do anything with it unless you have a perception system and the right motors and the right software to operate it. All of these pieces are starting to fall into place, and what I’m hoping is that a development community with emerge, so that people who aren’t core roboticists can go out and buy a robot for $5,000 and build their own applications on top of that.”
Hardware? Check. Software? Check. People? Both Stanford and UC Berkeley boast outstanding robotics laboratories that are churning out new PhDs and engineers at a rapid clip. Added to that, Bay Area companies have never had much trouble stealing technical talent from other regions.
Which brings us back to the first ingredient: investment.
Charlie Duncheon says he started showing his Grabit presentation to Silicon Valley VCs since last summer. “It was pretty clear to me that I was not going to get Sand Hill Road interested,” he says. For Valley VCs, “it just appears that there’s a declining interest in hardware, manufacturing, and industrial technology,” Duncheon says. To win backing, Duncheon says he expects he’ll turn to strategic investors: big makers of manufacturing equipment whose products might be enhanced by Grabit’s novel electroadhesion technology.
None of which was a surprise. Duncheon says that even before starting his fundraising quest he consulted with Kiva Systems founder and CEO Mick Mountz, who made a similar trek to Sand Hill Road back in Kiva’s early days and came up dry. Mountz did ultimately win venture capital for his fleets of warehouse robots, but it came from two firms back East: Bain Capital Ventures in Boston and Meakem Becker Venture Capital outside Pittsburgh.
The two firms put about $18 million into Kiva all told. Now, of course, they’re whistling all the way to the bank: Amazon bought Kiva Systems last month for $775 million. Mahoney thinks that this sizable exit may finally begin to wake Silicon Valley up.
“I know just from the number of people who are reaching out to us and the reaction of the VC community that there is a buzz around robotics now,” Mahoney says. “The biggest thing is Kiva Systems [getting acquired]. At the same time, though, iRobot has lost around 30 percent of their market cap lately, so VCs are still going to be skeptical. But they will pay attention. They are seeing guys like Jeff Bezos at Amazon and Scott Hassan from Google [the founder and main backer of Willow Garage]—people that are first movers and have the cash to do interesting things—and are starting to feel like, ‘I better get involved here or I am going to be left behind.’”
Fear of missing out: for a VC, it may be one of the strongest motivations of all. So when it comes to the robot races, maybe there’s hope for Silicon Valley yet. To find out for sure, you’ll have to join us at SRI next week.