Robot Submarines Go To War. Part 2: The Navy’s AUVs

Already, smart unmanned subs are set to replace dolphins as undersea mine sniffers. Next tech: mine detonation, remote sleuthing and robotic combat.
SPLIT STRIKE Deployed from a sub's hull, Manta could dispatch tiny mine-seeking AUVs or engage in more explosive combat. Illustration by John MacNeill

Sheathed in a chilling veil of
rain, under cover of darkness, a few Navy Seals descend from a ship into a small rubber boat. They motor to a nearby harbor, idle the engine, and gently lower three torpedo-shaped objects into the water. The mission? Locate that persistent nemesis of amphibious operations: undersea mines. But tonight, instead of the specially trained dolphins or human divers who would normally do this work, the Navy is relying on robots.

This scene has been enacted many times recently in San Diego, Seattle, Hawaii and the Mediterranean. It’s a test scenario, a dress rehearsal for the real thing. Naval experts won’t say whether their smart new unmanned subs, known as autonomous underwater vehicles (AUVs), will be sent into hostile harbors anytime soon — that’s classified — but they do express deep satisfaction with how the machines are performing.

The U.S. Navy has built its dominance on the conviction that bigger is better, but the nature of warfare is changing. In the post-USS Cole era it’s clear that the most ironclad expressions of might are vulnerable to a couple of guys in a rowboat. Early in the Persian Gulf War, the warships Tripoli and Princeton were blown open by mines — devices costing a few thousand dollars causing millions in damages. Unable to land, the Marines had to be airlifted in. Mines and other easily acquired technologies such as cruise and ballistic missiles mean that today any nation, no matter how technologically humble, can effectively bar U.S. forces from its coastline. In military parlance the problem is known as “broad-area denial.”

One way for the Navy to become more agile would be to employ unmanned submarines, which in the past decade have blossomed at university research labs and have been commercialized by small spin-off companies [see “Robots of the Deep Blue Yonder,” Feb.]. But though the Navy funded some early AUV development, like most large bureaucracies it was slow to incorporate novelties into its long-range planning. In April 2000 a Navy committee warned that robot-sub technology was available in the global marketplace and that if the agency didn’t wake up, it would fall behind. The committee’s report outlined four desirable “signature capabilities” for the AUV fleet of the future — reconnaissance, undersea search and survey, communications and navigation, and anti-submarine warfare — and the Navy has since stepped up its AUV research and development.

Building these devices won’t be easy. While a few AUVs — like those the Navy Seals used in their mine-hunting test — are already operational, their capabilities fall far short of what military planners ultimately envision. Better sensors, signal processing, energy sources and navigation methods will be necessary before AUVs can fulfill their potential role.

Operating unmanned submarines is far more challenging than flying unmanned aircraft like the Predator, which, controlled by a remote pilot, made successful sorties last year in Afghanistan. And the issue isn’t just autonomy. “Everything’s working against you underwater,” says Robert Wernli, a longtime AUV proponent at the Ocean Systems Division of the Space and Naval Warfare Systems Center (SPAWAR) in San Diego. In the ocean, visibility is much lower than in air. Currents pull vehicles off their plotted course. Saltwater corrodes them. Radio signals and GPS don’t penetrate deep water, which makes navigation and communication especially challenging.

But the military benefits AUVs promise almost certainly justify the effort. Nearly undetectable — they operate fully submerged and have low acoustic and magnetic signatures — they could be sent ahead to conduct surveillance or prepare for an invasion without tipping off enemy forces. They can be small enough to be launched from almost any ship, sub or aircraft — some are even light enough to be FedExed — and thus can conduct missions in water too shallow for conventional craft. They can be produced relatively inexpensively, so they wouldn’t need to be recovered in dangerous or inconvenient circumstances. They would act as “force multipliers,” taking care of programmable tasks and freeing up manned warships to take on more complex ones. And they could be sent on the riskiest missions, to help keep sailors and Marines out of harm’s way.

For now, the Navy is mostly developing AUVs as mine hunters. Long-term, though, they’re expected to do more than take the man out of the minefield. AUVs will map currents and the ocean bottom to help manned subs and ships navigate safely. They’ll be sent secretly to the coastlines of hostile countries to monitor enemy actions and report back via satellite. They will act as beacons, enabling other underwater devices to keep their bearings without surfacing for a GPS signal. And they will determine when an enemy sub is nearby, follow, and, if necessary, blow it up. “They increase the reach of Navy systems,” says Tom Swean, head of the Ocean Engineering and Marine Systems Program at the Office of Naval Research in Arlington, Virginia. “There may come a time, thanks to AUVs, when very few people are involved in violent action.”


Today’s Navy is a blue-water force — its strengths are in the depths, not the shallows — but it’s preparing for a brown-water fight. Seventy-four percent of Persian Gulf waters are shallower than 180 feet. And in what the Navy terms very shallow water — from about 40 feet to the 10-foot depths where waves begin to break — mines pose an especially acute threat. There, where ships and manned subs can’t venture, the Navy traditionally relies on dolphins to find and mark mines, and on human divers to set charges nearby, then swim away before they explode. The work is painstakingly slow and dangerous.

That’s why military planners would love to send in a team of AUVs instead. The AUVs in the Navy Seals’ tests are programmed before being dropped into the water. Each vehicle dedicates itself to a discrete portion of the harbor, covering it in a series of parallel runs — a tactic called “mowing the grass.” Scanning 150 feet in each direction with sonar, the robot subs note the location of all mine-like objects. To keep their bearings, they continually send signals to two transponders that the Seals have dropped in the water at predetermined locations. (The subs are programmed to know where the transponders are, so by assessing how long it takes for their signals to bounce back, they ascertain their own location.) After a few hours, when the robot subs have covered the entire harbor, they gather at an appointed place to be retrieved.

The AUV used in this operation is the Remus (Remote Environmental Monitoring Unit System), originally developed in 1994 at the Woods Hole Oceanographic Institute for non-military survey work. The institute’s commercial spin-off, Hydroid Inc., now produces special versions for military use.

The Remus is 63 inches long and 7.5 inches in diameter. It can stay out for up to 22 hours and specializes in depths of 10 to 40 feet, making it ideal for the shallows of the Gulf. The Navy teams several Remuses together in a system known as Sculpin. Each vehicle can be packed in a metal case that weighs just under 150 pounds, the FedEx maximum for rapid commercial delivery: Sculpin’s handlers travel lighter than most camera crews. For now the Sculpin system is designed only to detect mines; human divers would handle detonation.

The Battlespace Preparation AUV (BPAUV), being developed for the Navy by Bluefin Robotics in Cambridge, Massachusetts, operates in the next depth zone, between 40 and several hundred feet. Ten feet long, 21 inches in diameter and weighing about 500 pounds, the BPAUV is no minnow. But Scott Willcox, chief technology officer at Bluefin, says a smaller, more manageable version is in the works. Both the BPAUV and the Remus use side-scanning sonar devices to search for mines buried on or beneath the ocean floor. Transducers on the vehicles’ sides dispatch a thin beam of sound waves that extends out like a fan for 150 feet.

Acoustic sensors work best for the job because coastal waters are often turbid, according to Doug Blaha of Marine Sonic Technology in White Marsh, Virginia, which manufactures the Remus’s sidescan sonar. “If you walk into a smoke-filled room, you may not be able to see someone across the room,” Blaha says, “but you can certainly hear him.”

Shallow water isn’t just cloudy; it’s clogged with the detritus of humanity that sloshes against most of the world’s shorelines. “The big challenge is finding mines in these high-clutter areas, filled with human trash — cars, junk, old refrigerators,” says Rob Simmons, a program manager at the Navy’s Program Executive Office for Littoral and Mine Warfare in Washington, D.C. To help AUVs distinguish a mine from a hot water heater, the Navy is developing more powerful sonar that will provide higher-resolution acoustic images. Also under consideration for AUVs are magnetic gradiometers — sensors that use internal magnets to detect changes in local magnetic fields caused by metallic objects — and chemical sensors that would detect explosive materials leaching from poorly made mines.

Mines pose a hazard in deeper water as well. Typically this sort of mine floats near the water’s surface, attached to cables moored on the ocean bottom. Here, detection is currently done not by dolphins and divers but by ships or helicopters towing submerged sonar sleds. Neither system, though, is ideal: The ships put themselves at risk by navigating mine-infested waters, and the choppers, flying slow over the water, make an easy target for enemy gunners. The Navy hopes AUVs could one day do the work instead. The area to be covered is vast, so this sort of AUV will be required to stay out for significant amounts of time.

The 28-foot-long Seahorse, specially built for blue-water mine sweeping, will come online this year. It can traverse 300 nautical miles on batteries that will keep it going for up to 72 hours. While the Seahorse, which operates 100 to 1,000 feet below the surface, is ready now, the Navy is working to develop another blue-water AUV by 2004. The Long-Term Mine Reconnaissance System, currently under development at Boeing, consists of two 20-foot-long, 21-inch-diameter vehicles that will be launched and recovered through torpedo tubes. LMRS uses a forward-looking sonar — sending sound waves out like a movie projector — to spot floating mines, and to steer clear of them and other obstacles while it patrols. Each LMRS vehicle will go out for several days at a time, mapping 50 square miles a day as far as 100 nautical miles from the sub.

Because they are “value added” devices, increasing the Navy’s overall work force, AUVs will enable the military to be more proactive. In military parlance, AUVs would enhance “battlespace preparation.” They could be sent out to map the ocean floor in areas of potential conflict — identifying mines, dangerous currents and ideal landing routes ahead of time without alerting hostile countries that their shores are of interest. “We would hope to have the necessary mapping done before war breaks out,” says Martha Head, an oceanographer at the Naval Oceanographic Office at the Stennis Space Center in Mississippi. Today that prep work is done by large oceanographic ships, which makes it expensive, manpower-intensive, and far from stealthy. Transferring the job to small, unmanned, expendable vehicles that could get close to hostile shores without attracting notice would increase efficiency by at least an order of magnitude, according to Swean.


The purpose of mine seeking is to find access routes — both for ships to enter harbors and for special operations forces to land on beaches. In addition to detecting where mines are, it’s necessary to determine where they are not: Often the best strategy is to go around a threat rather than confront one. But when mines lie in strategic places, they must be removed, and usually the best way to do that is to blow them up.

Making an AUV that can spot objects that appear to be mines is one thing. That’s where the Navy is now. Making an AUV that can determine whether an object is a mine and not a million-dollar piece of oceanographic equipment, then destroy it if necessary, is much more complicated. That’s where the Navy ultimately wants to be.

One of the vehicles military planners believe might someday be able to destroy mines is the Surf Zone Crawler, an unusual AUV being developed at the Navy’s Coastal Systems Station in Panama City, Florida. These small, square, tank-like vehicles are less then a foot tall and have wide tracks that enable them to negotiate rocky ocean terrain. The crawlers aren’t as vulnerable to ocean currents as swimming AUVs, which makes them ideal for close-range identification of mines. But they’re a long way from being ready. “When it comes down to it, we put our robots in the water and sometimes we get stuck in the mud,” says Chuck Bernstein, a Coastal Systems Station engineer.

Another problem: AUVs that are used to blow up mines are likely to be destroyed in the blast. Wernli believes it makes sense to sacrifice one Remus or similar vehicle for every mine destroyed: “If I can take out a mine for $100,000, that’s worth it.” But until now the Navy has been accustomed to building complex manned craft it does not want to lose, so the idea of creating expendable vehicles represents something of a paradigm shift. And it becomes a problematic notion as AUVs get more sophisticated, sporting new kinds of sensors and communications systems. The more gear an AUV possesses, the more it costs, and the less appropriate it is for kamikaze missions. Simmons’ rule of thumb: “Don’t put it in the water if you can’t afford to lose it.”

One solution to the problem is a concept known as cooperative autonomy. In a mine-sweeping scenario, for example, high-end AUVs could release cheaper, more primitive machines that could install themselves next to the targeted mines and wait to be told to blow up. “You can still do the job, but each vehicle doesn’t have to be as smart,” says Bernstein.

Communications will be a key component of such cooperative arrangements. Willcox says Bluefin is outfitting the BPAUV to act as a subsurface server, capable of receiving new mission plans from a ship and relaying them to AUVs down below. The BPAUV will communicate with the submerged vehicles through short-range acoustics, but when it needs to transmit or receive mission-critical information, it will employ a new long-wavelength antenna that transmits radio waves just above the ocean surface. With this capability, the BPAUV could quietly talk to a ship up to 60 nautical miles away.

Another way to economize is to create modular AUVs that perform many functions, rather than a panoply of specialized craft. SPAWAR engineers are designing a mission-reconfigurable AUV. The goal is a versatile and far-ranging vehicle, launched and recovered from a sub’s torpedo tubes, that would be capable of anything from communications to reconnaissance work. If sub operators want to monitor radio signals near a hostile shoreline, “you equip it with an antenna to listen, collect intel, and then come back out to the safety of deep water to dump the information to those who can use it,” says SPAWAR engineer Barbara Fletcher. Change the payload and the vehicle could conduct mine neutralization or track enemy submarines. A preliminary version of this next-generation AUV is expected to join the fleet in 2009.


Autonomy is probably the most challenging aspect of the robot sub. AUVs will have to avoid potential threats, assess incoming information, and adapt mission priorities, all without direct human intervention. For all this to happen, sensors — sonar, magnetic and optical — will have to be improved so that AUVs can gather higher-resolution images at even greater distances. One technology that could get them there is synthetic aperture sonar, an underwater adaptation of a technique used in radar. For a given point on the ocean floor, a side-scanning sonar sends out and receives a single ping, but synthetic aperture sonars pelt the same spot over and over as the vehicle moves along. In effect, this widens the aperture of the sonar from 12 inches to 100 yards. Enson Chang of Dynamics Technology in Arlington, Virginia, says the new technique could resolve wallet-sized objects a kilometer away.

Another necessary area of improvement is in energy for propulsion. Currently, small AUVs like Remus can operate for up to 22 hours and larger ones like Seahorse for up to 72 hours, but military planners need vehicles that can stay out on their own for much longer. Devising better batteries for small AUVs and other power sources such as fuel cells for large ones is the key. To get an idea of how primitive current systems are: The Seahorse runs on 9,216 alkaline D cell batteries, the kind normally used in flashlights. For now that is the most efficient method, Head says, because D cells have higher energy density than currently available rechargeable batteries.

In an ideal world, Navy scientists would like to do much more. They want to build AUVs that have a salmon’s sensitivity to water chemistry, and polymer skins that propel them the way an octopus is moved along by its rippling hide. And they would like their AUV sonars to be more like the dolphin’s, whose broadband acoustic transmissions range from about 30 to 100,000 hertz. “A dolphin can detect a difference in an aluminum cylinder wall thickness of only 0.2 millimeters — that’s fantastic,” says Patrick Moore, who runs SPAWAR’s biosonar program.


Perhaps the most futuristic element in the Navy’s vision of wars to come — called Manta for its rounded delta shape — is taking form at the Naval Undersea Warfare Center Division at Newport, Rhode Island. Manta’s mother ship would be a new kind of submarine, its hull clad in “smart skin” like a huge metal shark, its weapons mounted externally. Several 100-foot-long Mantas would be streamlined into depressions in this new sub’s hull; while attached, they would be an integral part of the larger vessel. Once launched, however, they would be linked acoustically with the mother ship, extending her reach into canyons and shoals where a submarine can’t follow.

Standing well offshore, the mother sub would deploy her giant AUVs to run reconnaissance patrols and in-shore surveys, to act as defensive pickets protecting the main force, to raid enemy ship and shore facilities with a submarine’s full array of weaponry. But the Mantas would themselves be mother ships. Like stealthy Russian dolls, they would launch their own swarms of smaller AUVs and unmanned aerial drones. And, presumably, these smaller robots would be able to send out their own swarms of even smaller AUVs.

Toward this end, engineers at Newport have fashioned a one-third-scale Manta Test Vehicle. Shaped like a slightly overfed Concorde, the 8-ton vessel can make 10 knots, dive to 800 feet, and carry a variety of payloads. The craft began its at-sea trials in Narragansett Bay in 1999, and has since demonstrated such skills as deploying and communicating with smaller AUVs. Seen churning through the bay, the Manta Test Vehicle evokes images of the Loch Ness monster; but to undersea warriors, it looks more like the future. n

Carl Posey is a writer based in Alexandria, Virginia.