Google finally proves it won’t pursue military contacts, pulls leading robot from DARPA competition

Share This article

When it comes to fully functional humanoid robots with versatile real-world dexterity, there are basically only two big games in town: ATLAS, a stompy kill-bot created by Boston Dynamics, and SCHAFT, a more abstract but well-balanced man-bot created by a Japanese company of the same name. If you pay much attention to industrial robotics news (and why wouldn’t you), you’ll know that both of these companies were recently purchased by Google, which has made a rather abrupt overall turn toward robotics research. Though the search giant had nothing to do with actually designing either platform, Schaft’s decisive first place and ATLAS’ strong second-place finish in the first leg of DARPA’s Grand Challenge are now basically Google’s by corporate marriage. That’s a problem for DARPA, since Google has vowed not to pursue any military contracts in the future — and this week, that policy finally came into effect. Google has withdrawn the Schaft robot from DARPA’s competition, robbing the challenge of its strongest competitor.
DARPA and Google have two competing narratives as to why this happened. The story favored by DARPA spokespeople is that Google simply wants to focus on the technology’s consumer applications first and foremost. This does make some sense, since Google definitely got into the robot business to corner the home robot market early, and Schaft’s mastery of domestic challenges like stairs and ladders makes it well-suited to that purpose. DARPA’s statement almost seems to imply that in this case, corporate greed is overriding the fine and noble intentions of the people at DARPA, who of course only want to create life-saving technology in line with their humanitarian mission statement.
On the other hand, Google has been very open about its feelings on military funding: it doesn’t want any. Keeping the Google team working to win the DARPA challenge would have meant receiving indirect funding from the Pentagon and (hopefully) prize money, too. DARPA, of course, hates this analysis, since its whole strategy with the Grand Challenge has been to claim that totally non-violent civilian applications are the one and only aim of this competition. This is supposed to be about building robots that can burst into burning American homes to save American lives; any resemblance these actions may bear to the sorts of military maneuvers that are DARPA’s actual raison d’être are purely coincidental.
This almost insultingly obvious falsehood has been so thoroughly swallowed and regurgitated by the mainstream media that a slew of new international teams are about to be announced as new competitors. Think about that: some of these teams will almost certainly be receiving subsidies or tax-breaks from their own governments to do research that will help American dominance in military future-tech. Whatever you think of their goals or motivations, you have to sit back and admire the sheer audacity it takes to try — let alone pull off — something like that.

ATLAS

Probably the most oft-repeated and annoying mistake made with respect to these robots is that they could be used to help clean up places like Fukushima Daiichi — even though a lack of manual dexterity is in no way what has barred robots from delving deeper into that reactor. Sure, something like Schaft will probably be used to clean up dangerous spots like that in the future (once we have radiation-proof robotics) but the robot in question will come from companies like Toshiba or Google itself — not on loan from the US military.

Toshiba’s stab at a robot that can brave the heart of Fukushima Daiichi.

Bear in mind that existing contracts mean that the ATLAS platform, created by the now Google-owned Boston Dynamics, will still be used by four outside teams in the Challenge finals. Funding freed up by Schaft’s withdrawal should allow several other, smaller teams to come in, but this is still a major blow to the veneer of inoffensiveness that DARPA has created around this project. Schaft was awarded $1 million in extra funding because it won the Challenge’s first round, and now that winner will be absent from the finals.