China’s stealing an American ocean glider won’t stop the world from making a whole lot more.
In what Ankur Panda in the Diplomat termed an “exceptionally brazen and illegal move by Beijing,” the Chinese Navy this past week stole an American ocean glider. On Friday, Pentagon Press Secretary Peter Cook announced that the Defense Department had contacted the Chinese government to demand the glider back. After some nonsensical whining about American reconnaissance around its ships, the Chinese government agreed to return the little guy. After all, seizing another Navy’s boat on the high seas can be construed as an act of war. At the same time, seizing this one—indeed, almost any single one like it—provides no useful intelligence. While this particular stunt may not be soon repeated, incidents involving autonomous craft at sea may increase considerably, as robotic boats and subs proliferate quickly.
First, let’s recount the action, such as it was. The glider in question was most likely a Slocum, designed initially at the Woods Hole Oceanographic Institution, and now serially manufactured by Teledyne Webb Research. That class of submersible (one could almost call it an underwater drone—just don’t) is named after the famous Canadian-American explorer Joshua Slocum, the first person to sail on his own around the world. In November 1909 he disappeared from his 36-foot yawl the Spray. The ship was later recovered, but Joshua himself had never learned to swim.
Today, the US Navy has about one hundred and fifty Slocums, and quite a few analogous Wave Gliders from Liquid Robotics. The National Oceanographic and Atmospheric Administration (NOAA) has its own fleet. The largest type may be the Liberdade class (named for Josh Slocum’s earlier ship) from the Scripps Institution of Oceanography. Different designs function with a variety of propulsive means, including wave and, when surfaced, wind and solar. All move very slowly—at a knot or two—but what they really offer is endurance, with some missions lasting many months. All phone home their findings by satellite link to the Navy’s central control station in Bay Saint Louis, Mississippi.
Occasionally, they do go missing, often turning up in fishermen’s nets. For such contingencies, US Navy gliders carry a plaque bearing a dog-tag request: If found, please call Glider Operations Center, US number 228–688-5877, or e-mail firstname.lastname@example.org. That’s an indication of how routine the glider fleet’s activities have become. All these swimming robots stay busy on a daily basis collecting oceanographic data such as salinity, temperature, sound speed, and contaminant levels, all around the world. Some of that information, particularly when pulled from waters between the Philippines and Vietnam, is militarily useful.
Better, of course, that robots would erase or destroy whatever secrets they carry on board before they are captured. But in this case, as a former assistant secretary of the Navy put it to me, “don’t make a big deal out of” the loss of the glider and its information. Taking to Twitter, the president-elect wasn’t even sure that the Navy should want it back. There was little information to lose that the Chinese couldn’t collect themselves. And unfortunately for Chinese sensibilities, that research is entirely legitimate, particularly when it’s conducted 50 miles northwest of Subic Bay.
Expect more anguish, as there will be more ocean-going robots soon enough. As large a company as Boeing has noticed this market, agreeing earlier this month to buy Liquid Robotics. The two companies had been cooperating on payloads for some time. As CEO Gary Gysin stated at an Atlantic Council event last April, the smaller firm has generally felt that its “mission is to build a platform. Other people can put special-purpose defense sensors on it.” Now, Liquid Robotics will become a formal part of Boeing’s autonomous systems unit, but its headquarters and main facility will remain in Sunnyvale, California. Boeing may thus be planning to treat the little outfit as delicately as it has Insitu, the unmanned aircraft manufacturer it acquired in 2008. That loose relationship has permitted the smaller, more entrepreneurial outfit to flourish, but with the financial backing and institutional access of a huge military contractor.
Indeed, this otherwise-little deal is potentially a big deal, for Boeing already knows what it’s doing. Over the past few years, the company has been building progressively larger prototypes of unmanned submarines: the 18-foot Echo Ranger, the 32-foot Echo Seeker, and now the 51-foot Echo Voyager. Without a crew, but with a huge fuel fraction, this latest Boeing boat can stay at sea for up to six months, or travel 7,500 miles. That’s pretty impressive for anything without an atomic reactor. Just like a conventional submarine, it recharges its batteries with its diesel generator. It can also dive deeper than any manned military craft. Most importantly, the Echo Voyager does not require a nearby surface ship for launch-and-recovery. In theory, with the right sonar and software, it could trail Chinese surface ships and submarines for weeks.
And soon will theirs trail ours too. As a team at RAND recently wrote, Chinese efforts with autonomous systems have been intensive too. So what happens when both sides have legions of not just ocean gliders, but real robotic submarines? As Bryan Clark of the Center for Strategic and Budgetary Assessments (CSBA) wrote in January 2015, undersea military competition may have reached an inflection point, where autonomous systems play a hitherto unseen major role. In a few years, if a Chinese oceanographic “research” vessel tries to inspect an Echo Voyager the way it snatched that Slocum, the thing might bite back. What happens at that point is anyone’s guess.
James Hasik is a senior fellow at the Scowcroft Center for Strategy and Security. An earlier version of this essay described Liquid Robotics as feeling “an understandable allergy to government contracting.” After discussion with the company’s representative, and a review of CEO Gysin’s comments, that characterization may have been too strong.