An artist's conceptualization of DARPA's ACTUV sub hunting robo-ship in action. (credit: DARPA)

One of the biggest problems the US Navy faces today is the threat of quiet, short-range diesel-electric submarines. When running on batteries alone, these subs are extremely difficult for the Navy's ships, subs, and patrol craft to detect with passive sonar. In war games with US allies—and in recent propaganda-generating "drills" by Iran's navy—US ships have consistently ended up in the periscope crosshairs of diesel submarines that have gone undetected or that the Navy has simply lost track of. The best way to keep tabs on diesel submarines is to literally stay on top of them, tracking them with subs or ships from the moment they set out to sea until they return home.

This is the sort of thing the US Navy used to do with Soviet submarines operating off the coast of the US. The problem is that doing it for diesel submarines in distant parts of the world would tie up ships and sailors, pulling them away from other missions—and the US Navy doesn't have the same sort of resources for antisubmarine warfare that it had during the Cold War to bring to bear on the diesel-electric subs of potential adversaries. This is especially true for those operating in waters far from the US. That's a problem that a Defense Advanced Research Projects Agency project now nearing fruition seeks to solve with drone ships.

During a roundtable with media last week, Darpa Deputy Director Steve Walker announced that the agency and its contractor Leidos would launch the first full prototype of an autonomous ship designed to hunt submarines and trail them for weeks at a time. Eventually, autonomous vessels could be deployed to track the latest generation of quiet diesel-electric submarines over distances of thousands of miles, providing targeting information to US Navy submarines, ships, and patrol planes—or simply harassing the subs with relentless active tracking to deter them from carrying out their mission.

Read 4 remaining paragraphs | Comments