These days, whenever a group of roboticists gets together to talk shop, the subject almost inevitably turns to Google and its secretive robotics division.What are those guys up to?
The curiosity is understandable. It’s been nearly three years since Google made its huge move into robotics by acquiring an impressive and diverse group of companies, including Meka and Redwood Robotics, Industrial Perception, Bot & Dolly, Holomni, Autofuss, Schaft, Reflexxes, and, most notably, Boston Dynamics. Google’s robotics division, which has some of the world’s brightest robotics engineers and some of the most advanced robotics hardware ever built, has been working quietly at various secluded locations in California, Massachusetts, and Tokyo, and details about their plans have been scarce. Earlier this year, following Google’s reorganization as Alphabet, the robotics unit became part of X, Alphabet’s experimental technology lab, or as the company calls it, its “moonshot factory.”
Now the robotics community’s curiosity has reached a new height after news broke that Alphabet is reportedly selling Boston Dynamics, according to sources who spoke to Bloomberg. The revelation surprised many observers, but what they say was most confounding were reports suggesting thatAlphabet wants to get out of robotics altogether. Is the company done with robots? Is the robotics group being disbanded?
That’s not what is happening, the company says. “X has always been the home of long-term projects that involve both hardware and software, and we have a number of projects related to robotics already (e.g. Project Wing and Makani), as well as the lab facilities that are helpful for this kind of work,” Courtney Hohne, an X spokesperson, said in a statement provided to IEEE Spectrum. “We’re now looking at the great technology work the teams have done so far, defining some specific real-world problems in which robotics could help, and trying to frame moonshots to address them. We want to fall in love with a problem, not with a specific technology.”
The search for moonshots will probably continue to be a matter of intense debate internally. For those outside the company, it’s a captivating issue, and everyone seems to have a different opinion on what Alphabet could or should do with its robots. In fact, some roboticists encouraged us to gather those opinions in a public place to stimulate the debate. We liked the idea, so we contacted nearly 50 robotics people with a variety of different backgrounds and asked them the following question:
If you were in charge of Google’s robotics division and you had all those robotics companies at your disposal, what would you do? What kinds of robots would you build and for what markets?
Many declined to comment, citing ties to Alphabet. Others said they didn’t have a good answer (as one Japanese robotics executive put it, “I know exactly what I want to do with my robot business. Sorry, but I have no idea about Google.”) And despite our prodding for respondents to dream up wild ideas, no one suggested anything too crazy (Meka-BigDog hybrid centaur robot to deliver mail, anyone?).
But we did receive over a dozen thoughtful comments, which we’re including below. Alphabet and its robotics group certainly face huge challenges as they decide what to do next. We hope the ideas here will help advance the discussion about how to take robots out of labs into the real world. Commercial success in robotics, after all, is important not only for Alphabet, but for the future of our entire industry.
After you read through these comments, we’d love it if you could let us know what you think, too: contribute to the discussion by sharing your opinions or ideas in the comment section.