Where are the great robot companies?

Robotics shows are fun. There’s genuinely interesting stuff happening at lots of booths, like robots picking things up, or showing how they defuse bombs, or telling pretty good stories. All of these were happening when I walked around the show floor at RoboBusiness in Boston last Friday. I stopped and took videos of a few things, not something I would do at most trade shows. Robots are visceral in a way most other technologies are not.

Yet underneath the fun and fascination is the reality of the robotics business: There isn’t much of one. Colin Angle, the CEO of iRobot, summed it up well in his keynote speech. He thought back to the first RoboBusiness event 10 years ago and said “if you’d asked us, ’10 years from today what is this conference going to be like?’ I think we would all have imagined bigger.”

iRobot is probably the most successful robotics company (depending on how you define robot), yet it has mustered a market cap of only a billion dollars. Younger companies like Amazon and Facebook are worth hundreds of billions of dollars. HubSpot, a modestly-sized marketing software company that just went public, is worth about the same as iRobot. People expected more from robot companies.

Angle himself said as much in his keynote. He co-founded iRobot in 1990. “You could kick yourself because you missed very important industries being birthed in the 1990s that you were expert in (like mobile). But so be it. We’re building something even bigger that I believe will take longer.”

I am not saying Angle and his company have been failures. They just aren’t a huge financial success. They have not delivered the massive return on investment that a software or Internet company can create.

Not yet, anyway. Angle thinks the biggest returns lie ahead, as he told me when I spent some time with him after his talk. He told me that as an industry, robots are “still in the bootstrapping phase. Enough exists that a VC could feel like he could hit a single or a double with a robot company.” But home runs will take years, he thinks. They will require new technology, like 3d printing, to mature, and accelerate.

“Moore’s Law doesn’t work with gears,” Angle said. 3d printing will make it simpler, faster and cheaper to produce gears and other robot parts. Then we will see inexpensive robots in manufacturing, in communications, in oil exploration, in defense, in mining, and perhaps especially in health care, where robots of various sorts might help people live on their own longer. Those are high value, big return markets for robots.

Angle’s words were on my mind Friday morning when I watched Cynthia Breazeal, a visionary MIT researcher whose lab has done fascinating work on making robots seem empathetic and emotional. She thinks these sorts of robots will make a dent in the swollen amounts American spend to manage chronic diseases (her stat: 75 cents of every healthcare dollar goes to manage a chronic disease). Her data say robots are better than human coaches for physical fitness and weight. I haven’t checked to see if anyone’s been able to replicate her results, though I’m willing to concede that a robot could be programmed to be less annoying than people perhaps even more supportive.

And yet, what gets people excited about Breazeal’s work is JIBO, a family robot that, judging from the trailer below, can manage our lives and almost read our minds. I thought Jibo was fascinating; I’m impressed that 4,800 people have pre-ordered one at $500 to $600 each — Breazeal said on stage that one-third are developers, whose apps will be essential for bringing JIBO to life. But I also found myself thinking back to General Magic, a captivating early personal digital assistant that failed. I don’t wish a similar fate on JIBO. But, it isn’t set to launch until 2016. That’s both a long time from now and a much shorter time than Angle’s 10-year delay.