May 13, 2014
Are robotics a disruptive military technology? Popular conception holds that to be obvious, but a team from RAND argues otherwise. In their new study “Armed and Dangerous? UAVs and US Security”, they observe that robotic aircraft cannot today defend themselves, but less permissive scenarios are now of greater interest in military planning. Thus, as Sandra Irwin writes at National Defense, the ‘shine is starting to wear off’ unmanned systems at the Pentagon. Moreover, RAND continues, robots today largely perform roles that manned systems can too, so we should be little troubled as potential enemies get their own. But even if this is all true today, in thinking about the future, I have concerns about dismissing concern. The question is not just whether unmanned systems can undertake manned missions, but whether they can do so with radically different cost-effectiveness.

As Zack Beauchamp reported from last Friday’s conference on the topic at the Center for Strategic and International Studies (CSIS), there may indeed be reason for worry. Michael Horowitz, an accomplished theorist at the University of Pennsylvania, argues that the United States is losing its edge: his article in Foreign Policy on the “looming robotics gap” was strong stuff. Yes, people can do the dull, dirty, and dangerous jobs if they must—but they need not, if robots can do them much less expensively. Radical reduction in cost lies at the heart of Clay Christiansen’s classic argument about what makes an innovation disruptive. Beyond that, as Mike Nemeth of Zyvex Technologies has put it, robotics take systems beyond the limits of physics and physiology, and in the process, make new missions possible. And as evidenced by the Stop Killer Robots conference getting underway today in Geneva, plenty of folks are freaked out. There’s little reason to ban things that you think will be unimportant.

So if robotics will prove disruptive, who will benefit? In our paper preceding tomorrow’s conference on disruptive technologies here at the Atlantic Council, Byron Callan and I identify five factors that make ideas into real military capabilities. Three of those—industrial readiness, systems integration, and organizational capacity—seem so far to favor the United States and its allies. Sam Brannen and others at the CSIS wrote in their recent report “Sustaining the US Lead in Unmanned Systems" of the "remarkable engineering, expert operator, and tactician community” built up over ten years of drone campaigns. All of this remains ready to build and experiment with whatever comes next.

Much of the hard institutional work of another factor—cultural receptivity—seems to have been done, with wide acceptance of robotics in many functions in the military services. But we should wonder whether that receptivity will hold when the next radically new application emerges. How, for example, will the Navy adapt to the possibility of unmanned ships and submarines? Rolls-Royce already thinks that crews on cargo ships are pointless, and is working to prove that. As I wrote with some colleagues for Proceedings in December 2011, putting to sea with "Robots in the Age of Pirates” could be very cost-effective. What’s less clear is whether the admirals will take a shine to ships without captains.

But counterinuitively, the biggest issue may be financial resources. The US spends much more than anyone on its military and their drones. Flying drones with global range seems out of reach for countries without expensive communications satellites. But will that matter? As Patrick Tucker writes at Defense One, there is some reason to think that most countries can have short-range armed drones within ten years. Frankly, any Russian hobbyist can do that today with a tablet computer, a quad-copter, and a submachine gun. The barriers to entry in basic robotics are already low, and may be getting lower. At the same time, the US government is restricting drone exports, and thus encouraging the development of industry overseas.

If robots can be taught to operate more autonomously—even against the demands from Geneva—the need for that over-the-horizon connectivity may abate. Robotics technology may then democratize, as other technologies are doing, possibly lessening incumbent advantages, and spreading some scary technologies around the world. On the other hand, if the software needed to build those systems remains on a Google-scale, advantage could remain with countries possessing similar scale in resources. Watching how the control technologies develop will be critical for forming strategies on how to invest in robotics, and how to invest in responses to their proliferation to possible adversaries.

James Hasik is a senior fellow at the Brent Scowcroft Center on International Security.