November 01, 2014      

Technical chops Why did Andy leave the GooglePlex? Sometimes, rather than looking at the direction a ship is going, it?s often more advantageous to look back at its wake for insight on the reasons behind the ship?s voyage.

andy rubin

Such a look back may be in order when trying to understand the why behind Andy Rubin?s departure as head of Google?s robotics group. Business Insider?s Jillian D?Onfro offered up just such a look back today, one day after the GooglePlex released its startling announcement that the brains behind Google?s blitz-purchase of eight top-flight robotics companies last December and the leader of its newly-formed, top-secret robotics group has departed the company. What happened? Isn’t that what we all said at reading the news? Andy Rubin has more than adequately demonstrated his intelligence and knack for launching startups, just look at Android; and he obviously has an abiding love for robotics, having picked the target companies to acquire and more than likely a few ideas on what to do with them once acquired. The acquired companies and the ideas are still in place: Google didn’t KO the robotics group after Rubin left. Google is still passionate about robotics and its abilities to change the way we live. Robotics is still a Google ?moonshot?. See related: Inside Google?s Latest Series of Acquisitions Yum, Google Gobbles Up yet Another And Now for Something Completely Different: GoogleBots! Andy Rubin wasn?t run out of town. There are no tar and feathers scattered around Google?s campus. So, what?s up? Check out the wake D?Onfro, unlike the scads of other news writers who reported on the Rubin-Google split, didn?t hover over the departed; rather, she wisely looked back toward Google at Rubin?s replacement. It?s with James Kuffner, and his background in robotics compared with Rubin?s lack thereof, that an intimation arises as to why Rubin headed off into the world of incubating new technology. And the reason seems to take reasonable form. Kuffner is a roboticist, Rubin wasn?t. Maybe it was lack of technical chops in robotics that saw him leave in less than a year. This is how Kuffner described himself in an interview last December (2013) with MIT?s Technology Review:

?I?ve been working on humanoid robotics for 20 years and I was for seven years a professor at CMU [Carnegie Melon University] teaching robotics, and then five years ago moved to Google to work on the self-driving car, and I?ve always been interested in seeing forward-looking technology go from just an idea in a research lab to actually doing something practical and useful.?

So exactly what was he doing at CMU that could relate to the above and to Google?s plans in robotics? For his research interests at CMU, he writes:

?I am interested in developing algorithms and software for simulating and synthesizing motion for complex kinematic and dynamic systems. This research involves interdisciplinary work in robotics, computer graphics, and computational geometry.?

His interests in ?algorithms, computer graphics, and computational geometry? show why he fit in real well with Google?s driver-less car team. And this part fits more than well with Google robotics acquisitions: ?synthesizing motion for complex kinematic and dynamic systems,? which explains his presence at the DARPA Robotics Challenge where he was interviewed for Technology Review. But this, also in his CMU interests, really puts a stamp of where he?s always been headed and why the Google robot group is such a good fit: Humanoid robotics.

?For the past several years, I have been building general software components for autonomous humanoids based on planning, sensing, and control. This has concurrently involved researching techniques for automatically generating gross body motions for complex simulated models of human figures given high-level navigation or manipulation task commands, as well as generating motion trajectories for real humanoid robot hardware. Specifically, I have focused on path planning for obstacle avoidance, balance control, self-collision detection, footstep placement, and integrated sensor feedback systems.?

See his career direction toward humanoid robots in all of that? Andy Rubin, on the other hand, was not a roboticist. An engineer, yes; but not a roboticist, which is what Google?s robotics group needs. From Danger Inc. to Android, Rubin has always been an IT guy. Robotics, eventually, may have been too far out of his comfort zone.


After all, Google didn?t acquire just a bunch of machines, it acquired IP and human talent, all of which needs to be fashioned into a robot of sorts?preferably a human robot?because the acquired IP and human talent is all about humanoid robots. All of that acquired talent could also get very restless and irritable if Google deviates from the direction toward humanoid robots where their IP and careers are so embedded. Mutinies have been caused by less. Then too, it takes a strong personality to manage all that brilliant talent: someone strong, respected and capable of getting the job done. Talent management would have to be a must- have in any skillset when it comes to managing the founders and innovative geniuses behind these companies:

  • Holomni, a Mountain View, Calif. company that automates vehicle wheel motions.
  • Industrial Perception, a Palo Alto-headquartered business that uses 3-D vision to automate loading trucks and handling packages.
  • Redwood Robotics, a San Francisco-based company that focuses on creating robot arms for use in service industries. The startup was jointly ventured by Meka Robotics,
  • Meka Robotics, an MIT spin-off company that builds robot parts, like the Meka arm, that look ?friendly? to humans.
  • Schaft, a University of Tokyo spin-off that focuses on humanoid robots, Lead by a small team of Japanese roboticists, Schaft?s humanoid robot aims to fix one of the central problems facing most robots so far: They aren?t very strong. Core to that ambition is Schaft?s water-cooled, capacitor-powered motor system, which makes robots stronger while keeping them compact.
  • Autofuss, a San Francisco company that uses robots to create advertisements (like for Google?s Nexus 5).
  • Bot & Dolly, (sister company to AutoFuss) a company that creates robotic arms for film-making (most recently utilized in the movie Gravity), and
  • Boston Dynamics, a robotics design and engineering company best known for the development of BigDog, a quadruped robot designed for the U.S. military with funding from Defense Advanced Research Projects Agency (DARPA).

Kuffner would seem ideally suited to make something from this range of humanoid-oriented technologies and very talente
d human expertise. Will he need to have the management stones of a General Leslie Groves, famously contending with talented human expertise during the Manhattan Project? Probably. As a Kuffner buddy, robotics expert, Russ Tedrake related to the Wall Street Journal: “I know James well and honestly can?t imagine a better person for the job. He has the vision, the passion, and the technical chops to do great things with this opportunity.” For now, any future GooleBot seems to be in good hands. Good luck, Jim.