December 19, 2014

By standardizing too soon, the US Army risks missing the next new thing.


Citing budgetary restraints and interoperability challenges, the US Army last month announced plans to scale back its robotics program, and transition to a single type of ground robot sometime after 2020. Today's “mixed fleet of systems”, the Army says, has led to "high sustainment costs”; a single model could be operated more cheaply, in a common and efficient manner. But while this spartan standardization could fit the budget in the short term, it could unhelpfully constrain innovation in the long term.
It’s an ambitious program, to be sure. According to Defense News, the Army is looking to develop something it calls the
     Common Robotic System Individual (CRS-I). It’s intended to yield a new backpack-carried ground robot for surveillance missions, or bombs and hazardous materials, for soldiers on foot… The CRS-I, announced in an Army market survey in June, would weigh 20 pounds or less and allow a soldier to set it up in five minutes and operate it from up to 300 meters away. It would feature a joint plug-and-play architecture for sensors, claw arms and other peripherals, which allows the government to procure and service these separately.
It’s also remarkable that these plans were promulgated only days before the defense secretary announced his Defense Innovation Initiative, and the deputy secretary explained his technologically-intense Third Offset Strategy. The unpredictable nature of future threats will cause enough problems for these grand efforts. Commitment to any single concept limits the adaptability of the military in the face of evolving threats. Indeed, commitment to a single model of robot signals a singular strategy: the Army is betting that this One Robot, the CRS-I, will be adequate against all threats.
We recall that didn’t work well with Humvees in Iraq.
Indeed, it hasn’t worked well for a long time. The impulse to settle on an established pattern of manufacture in a technologically dynamic field is at least a century old. Katherine Epstein describes perhaps the first such conundrum in her new book Torpedo: Inventing the Military-Industrial Complex in the United States and Great Britain (Harvard University Press, 2014). As early as 1894, the Admiralty was receiving advice for “pattern unification,” in which only a single torpedo type would be bought for each size of tube. While that seems sensible and even essential today, settling on a single model too early could have left the Royal Navy with ordnance that was rapidly antiquating, just as the German High Seas Fleet was rapidly expanding.
Let us not forget the other side of the problem. For a classic science-fiction account getting more attention today, try Arthur C. Clark’s 1951 short story Superiority (neatly summarized here by our colleague August Cole). The great professor-general forgoes pedestrian technologies for endless research on the “irresistible weapon,” and in the process loses the war. Just recall all the effort poured into wunderwaffen in 1945 when the Germans couldn’t produce reliable equipment in meaningful quantities.
The tension at issue is found in what Charles Fine of MIT has called the Clockspeed of the technology. His early book on the subject was subtitled Winning Industry Control in the Age of Temporary Advantage (Perseus, 1998). Rapidly advancing fields just aren’t ripe for cost reduction through standardization. Modular architectures are alluring, as they provide operational flexibility, and can shift the focus of equipping from more expensive platforms to less expensive payloads. But if the architectures themselves aren’t settled, how do the managers know what to manage towards?
Technological clock-speeds do vary over time. Consider combat aircraft, for example. Today, the Joint Strike Fighter program is taking almost two decades to bring to fruition. In the 1930s, new aircraft concepts were appearing every year from designers’ hangar workshops. But with robotics, the pace of technological development hasn’t even begun to slow down. We fear that standardizing on a single model now would rather be like the RAF just choosing the best open-cockpit  biplane it could get in 1930, and cruising into war with that.
Worse, an entrenched single-model robotics program could discourage future procurement managers from revoking the Army’s commitment, even after an emergent threat laid bare any unexpected inadequacies. To compensate, the Army’s managers could very well attempt to load feature after feature on this CRS-I. But just as the doors of repeatedly unarmored Humvees would hardly swing open, that unified pattern could only be taken so far.
So, instead of investing in a single model, we recommend that the Army invest in a range of relatively inexpensive and readily disposable models. That’s the sort of acquisition strategy that will usher in a more dynamic and adaptable robotics program. By investing less in the standardization of technologies, the service could invest more in exploring the possibilities of new technologies, particularly at a time when the underlying technologies are advancing so rapidly.
Philip Thorell is an intern and James Hasik a senior fellow at the Brent Scowcroft Center on International Security.