Sea Power in the Robotic Age


The Military of the Future

BRUCE BERKOWITZ

Sea Power in the Robotic Age

The military has demonstrated the feasibility of airborne drones; the next development is likely to be a variety of remotely controlled marine vessels.

Robotic weapons are revolutionizing warfare. Anyone following the media knows about so-called aerial “drones,” and soldiers and Marines are using tracked and wheeled robots in ground combat to probe enemy positions, disable bombs, and carry out other risky activities. But R what about the three-fifths of the world covered by water?

Sea-based robots—unmanned maritime systems, or “UMSs” in current Pentagon parlance—receive less attention, even though they have been around a long time. They will likely have as great a role in shaping how the United States fights future wars. The challenge is to resolve the technical, operational, and policy questions UMSs raise: specifically,

• What technologies make UMSs more capable than before, and how can the United States keep its edge?

• What military requirements can UMSs address, and how do these requirements fit into U.S. strategy?

• How can we deliver UMSs quickly and integrate them into U.S. naval forces and planning?

• And, finally, what are the legal, policy, and diplomatic issues?

The first hurdle is that there is no universally accepted definition of UMS, but most would agree a “robot” is any machine that carries out complex tasks via remote control, or using an internal computer that integrates inputs according to pre-loaded instructions. Sea-based robots can be described according to four dimensions:

Autonomy. UMSs can be either free-swimming or tethered to a surface vessel, a submarine, or a larger robot. Tethers simplify providing power, control, and data transmission, but limit maneuverability and range. Recently developers have built highly autonomous systems that can navigate, maneuver, and carry out surprisingly complex tasks.

Operating mode. UMSs can operate on the ocean’s surface, at or just below the surface, or entirely underwater. Operating above or near the surface simplifies the power and control, but compromises stealth. The U.S. Navy has devoted particular attention to unmanned underwater vehicles (UUSs) during the past 10-15 years. Its unmanned surface vehicles (USVs) are much less far along; the Navy has put a higher priority on using automation to reduce crew size in U.S. warships.

Size. UMSs range from small or “man-portable” (25-50 pounds); to midsized (500-3,000 pounds, often designed to operate from existing submarine weapon launch tubes and airlocks); to large (projected in the future to be up to 40 feet long and 10 tons displacement), designed to operate from a pier, off a surface vessel, or from a large weapon tube.

Function. UMSs can be equipped with a variety of manipulator arms, capture devices, sensors, communications packages, and warheads.

The lineage of sea-based robots dates to the nineteenth century. Indeed, the very first autonomous vehicle—sea-based or otherwise—was the naval torpedo. Robert Whitehead, a British engineer working for Austria in the 1860s, invented a gyro-stabilized torpedo propelled by a compressed air motor, making it the first UUS. During World War II the U.S. Navy developed the Mk 24 self-propelled mine, a sonar-equipped torpedo that rested on the sea bottom until it detected an enemy submarine. The sonar activated the weapon and guided it to its target, making it the first intelligent UUS.

Tethered vehicles trace their origins to the early 1920s, when Howard Hayes, a Navy research physicist, invented the towed sonar array, a string of hydrophones played out from a ship’s stern. The array was long enough to clear the ship’s propeller noise, allowing it to detect submarines and measure sea depths more precisely. Such “tails” soon appeared on both surface ships and submarines. At about the same time, navies also developed “sleds” towed by ships (and, later, aircraft) to locate and identify explosive mines. Tails and sleds gradually became more sophisticated with additional sensors and controls allowing them to maneuver as they were towed.

The first autonomous unmanned surface vehicle appears to have been Coast Battleship No. 4, a target ship the Navy converted from the USS Iowa in 1919 by adding a radio-controlled helm. The Navy converted several other ships in this way during the 1930s and, after World War II, built unmanned “drone boats” to photograph nuclear weapon tests at close range. Since then it has experimented with small, unmanned surface minesweepers, though its surface drone efforts never matched those for underwater vehicles.

These systems evolved steadily through the mid-1900s, and technologies for towed and autonomous vehicles began to merge and overlap. For example, the Navy’s Mk37 torpedo, introduced in 1960, combined autonomous gyro stabilization with a wire control system; the wires transmitted guidance commands and spooled out as the torpedo closed on its target. Similarly, the Cable-Controlled Underwater Recovery Vehicle (CURV), which the Naval Ordinance Test Station designed in the early 1960s to retrieve torpedoes, was tethered, but also had thruster motors to maneuver and featured cameras, lights, and mechanisms for attaching lines. (CURV gained fame when it recovered a nuclear weapon lost off the coast of Spain in 1966.) The Royal Navy had used a similar system, CUTLET, a few years earlier, and before that underwater photographer and technologist Dimitri Rebikoff developed one for archeological research in 1953.

The big technology leap occurred in the 1980s, when small, rugged electronics made possible sophisticated, versatile tethered maritime robots, often called “remote operated vehicles,” or ROVs. The Navy funded some of the early development using ROVs for a variety of missions, including surveying submarines lost at sea. Scientists also used ROVs for oceanography and marine archeology. But the biggest push came from the deep pockets of industry, which used ROVs to service and repair offshore oilrigs and sub -marine cables.

The first practical autonomous UMSs appeared in the late 1990s for oceanographic surveys and clearing mines. Several technologies have made it possible today to build even more sophisticated UMSs. Some of these include:

• Improved navigation systems using sonar for higher-fidelity sea bottom mapping and detection of obstacles;

• Miniaturized, solid-state inertial navigation units that allow vehicles to track their position with greater precision via dead reckoning;

• Fiber optic communications, both to link ROVs to their parent ship, and small ROVs to larger host vehicles;

• Global Positioning System (GPS), providing location data via either direct satellite links (in the case of USSs, and UUSs that periodically surface), or through a fiber optic link to a parent ship or buoy (in the case of ROVs);

• Acoustic modems that can transmit digital data through water, albeit at a lower rate and shorter ranges than their electronic counterparts, and acoustic positioning systems;

• Side-scanning and synthetic aperture sonar, electro-optical cameras, magnetometers, and other sensors; and

• Power innovations such as lithium batteries and fuels cells.

As important as these individual technologies are, system integration is even more critical. The biggest challenges for underwater vehicles remain communications and power, along with ocean pressures that stress battery housings and other structures, and deep sea cold and darkness that make any kind of work challenging. It is the combination of sensors, communications, data processing, and mechanical design that produces an effective package.

The United States leads in basic technology and certain categories of UMSs, but the industry is global. Texas-based Oceaneering is the largest ROV manufacturer ($2.8 billion in total revenue, 10,000 employees, but this includes operations as well as manufacturing). A company spokesman told a reporter during the Deepwater Horizon cleanup in 2010 that Oceaneering built about half of about 500 industrial ROVs then in service worldwide. Other players include FMC Technologies and Forum Energy Technologies (both also Texas-based); C&C Technologies (Louisiana); Global Marine Systems and SMD (Britain); ECA Robotics (France); Ageotec (Italy); ISE (Canada); Kongsberg Maritime and Subsea 7 (Norway); and SAAB Seaeye (Sweden).

Kongsberg Maritime is also the current leader in autonomous UUSs. Kongsberg started its UUS program in 1990 and bought its main competitor, Massachusetts-based Hydroid, Inc., in 2008. (Hydroid commercialized technology developed at Woods Hole Oceanographic Institute, much of which had been funded by the U.S. Navy.) Another U.S. company, Bluefin Robotics, a Battelle subsidiary, is also in Massachusetts and historically tied to MIT. Others include ISE, which builds autonomous UUSs in addition to its ROVs; and Teledyne Gavia, which Teledyne Technologies created after buying Hafmynd, an Icelandic company, in 2010. China’s 863 Program, associated with microelectronics and space flight, has experimented with autonomous UUSs since at least the 1990s. Many other labs and companies around the world build small vehicles for marine research.

So while the United States has a strong hand to play, there will be lots of competition as navies explore UMSs. Governments fund research and development, but most UMS manufacturing expertise is in the private sector. Most key components are commercially available. Navies can buy the capability they seek.

Potential missions

The U.S. Navy’s recent interest in UMSs is partly tied to the “strategic pivot.” Officials believe that U.S. military forces will more likely be used in the Pacific and Indian Oceans, rather than in Central Asia, the Middle East, and Europe. This presents the Navy with new challenges.

One is distance. Operating in Asia takes personnel further from home, stretches logistics, demands greater operating range from ships, and so on. Also, potential hot spots in Asia are highly dispersed. Looking north to south, they include the resource-rich Arctic (Russia is making new territorial claims); Korea (Pyongyang, along with its traditional invasion, submarine, and artillery threat to South Korea, has added a nuclear missile threat to nations throughout the region); the perennial Taiwan Strait flashpoint; ocean claims by China against Japan, the Philippines, and Vietnam; military and pirate threats to Indian Ocean sea lanes; and so on. The result is greater demand for a U.S. presence over a much larger, more distant area.

As the saying goes, even the best warship cannot be in two places at the same time. But under current plans the Navy will shrink, not grow. The five-year budget the Navy proposed to Congress in 2013 cut $58 billion from its FY 2012-13 proposal. Current plans will leave the Navy’s budget essentially flat at $152-156 billion annually through FY 2018-19. Few experts believe the Navy will top 300 surface ships during the next decade; some think it may have fewer than 250. By comparison, at the Cold War’s end it had almost 560 ships.

Sea-based robots cannot fill the gap completely, but they can fill important missions and add new capabilities to confound potential adversaries. Some scenarios discussed include:

Counter-mine warfare. Mines are a particular concern for the Strait of Hormuz and Strait of Malacca. As noted, mine clearing has been a primary mission of UMSs. Robotic systems can often clear a larger area faster and usually more safely than older methods. Countermine UMSs could be kept on station in such hotspots and also to serve as a deterrent.

Anti-submarine warfare. Countries such as Iran and China are deploying new submarines that many Western analysts believe could threaten maritime trade or prevent the United States from delivering supplies and reinforcements to its regional allies in wartime. UMSs could assume part of the mission of detecting, tracking, and following them, partly offsetting the Navy’s shrinking submarine fleet.

Information operations. Because UUSs are usually quieter than submarines, it is straightforward to equip them to mimic the sounds of their manned counterparts. Again, this capability could help as the U.S. submarine force is stretched thin. Such “spoofing” complicates targeting for an adversary in wartime, and its ability to develop reliable signatures and databases of U.S. forces before a conflict begins.

Strike warfare and area denial. It would be possible to adapt short- and medium-range precision-guided missiles to UMSs, just as missiles were added to unmanned aerial vehicles (UAVs) such as Predator and Reaper. In one scenario, the Navy could position armed UUSs close to a potential hotspot such as North Korea or the Taiwan Strait. During a crisis some might be surfaced as a “show of force,” and then the weapons would be concealed to serve as a deterrent. UUSs could also attack enemy ships in port or en route. They could provide greater endurance than manned vessels for remaining on station, and greater ability than traditional torpedoes or mines to adapt to changing conditions.

Infiltration and payload emplacement. Prior to amphibious operations in World War II, Navy frogman surveyed approaches, emplaced demolitions to destroy enemy defenses, and planted navigation beacons to guide forces to their targets. SEALs perform similar activities today. UMSs could allow them to work from a greater distance.

Infrastructure inspection and servicing. The Navy maintains an extensive underwater infrastructure for command, control, and communications. UMSs could be used to maintain this infrastructure, much as the oil exploration industry uses them today to service offshore rigs.

Harbor policing. The Navy has performed this function in operations overseas and as part of its force protection responsibilities at home. They could use UMSs to detect intruders and inspect vessels, as could the Coast Guard, Customs and Border Protection, and local police.

Ad-hoc sensor and communication networks. Today when U.S. forces go to war, one of their first tasks is to establish “tactical intranets” to move information among units. UMSs offer a means to establish such networks quickly in a combat zone offshore or at sea, using a combination of radio frequency, acoustic, and pre-laid fiber optic links.

Environmental monitoring and oceanography. UMSs can be used to determine conditions in littorals just before a military operation and collect information on ocean conditions for long-term planning and research.

Some missions proposed for sea-based robots do seem, let us say, ambitious. The recent “Hydra” solicitation from the Defense Advanced Research Agency (DARPA) requested proposals for UUSs that could host both smaller undersea and aerial vehicles. The hybrid submarine/aircraft carrier idea has been around since the 1930s, when the French put a seaplane hanger on the 3,000 ton Surcouf, and the technology might be feasible. But a more likely solution is to integrate UMSs with UAVs operated from land and surface ships. Indeed, a major challenge for making UMSs useful weapons is integrating them into the force structure. Building less complex robotic systems and integrating them incrementally reduces risk, since the failure of one program or technology would not set back the entire UMS concept.

Legal and policy concerns

New military technologies often lead to questions about rules of engagement. Submarines, for example, raised issues about whether international law allowed no-warning attacks on merchant vessels. More recently, precision-guided munitions able to destroy specific buildings, vehicles, or persons have put lawyers into the military targeting process.

Sea-based robots will raise similar issues, and several new ones. They seem to fall into three categories: How much autonomy can be built into UMSs? How to deal with the inherently clandestine nature of UMSs? And when can countries preemptively defend against UMSs?

Autonomy. As noted, some modern UMSs can operate with little or no direct human control. Treaties defining rules of war are ambiguous about when a human operator needs to be in the loop. The closest thing to a restriction on sea-based robots seems to be the 1907 Hague Convention’s ban on “automatic contact mines” laid “with the sole object of interrupting shipping.” Various bans on indiscriminate targeting of civilian populations may also apply.

The main constraints on autonomy for Navy UMSs are policies that the Defense Department has adopted for itself. Department of Defense Directive 300.09 (November 21, 2012), “Autonomy in Weapon Systems” applies to all U.S. weapon systems, maritime or otherwise. It defines two classes of robotic weapons:

Semi-autonomous weapons, or those that are launched by an operator and then are guided to a specific target. Examples are “fire and forget” anti-tank missiles or torpedoes that lock onto a targeted ship or submarine.

Autonomous weapons are those that can select targets without additional action by a human operator once they are activated. Phalanx, the Navy’s close-in missile defense system, is an example. Anti-ship missiles approach faster than humans can react, so Phalanx combines a Gatling gun mated to a radar aiming system. Activate Phalanx, and when a missile approaches, Phalanx automatically shoots at it.

The directive gives wider latitude to semi-autonomous systems because they have an operator in the decision loop to engage a target. It puts tighter limits on autonomous weapon systems, which can engage targets on their own. For example, it allows semi-autonomous weapons but not autonomous weapon designed to destroy manned targets. The directive also assigns responsibility for making sure the design, testing, and evaluation of robotic systems meet these requirements.

Thus, the Navy cannot deploy UMSs that automatically identify and destroy vessels it presumes to be hostile. But it allows UMSs that autonomously clear mines from a harbor or a sea passage. Autonomy will be controversial, if only because it is hard to predict future technologies and how U.S. forces will need to use them. Officials should have analysts think about these issues now, and include outsiders since these controversies will inevitably become public.

Clandestine military operations. UMSs, and especially UUSs, are inherently stealthy, which means that their operations are inherently clandestine.1 This is nothing new; submarine operations are often clandestine. But such operations may become more common as UUSs become more widely available and more capable. U.S. officials should anticipate more oversight requirements.

We should also anticipate other countries—including adversaries— will carry out similar operations. With so much of the technology being developed in the private sector, and much of it overseas, other countries will likely develop their own UMSs. Keep in mind that U.S. operations will create precedents other nations might cite. Again, this is why U.S. officials should promote discussion of policies on the use of UMSs now, before the issues get heated.

Rights of passage. National security wags joke that a ne’er-do-well might seize an autonomous sea-based robot as an abandoned vessel. It’s a assessment, even if salvage law is clear; vessels novel threat assessment, even if salvage law is clear; vessels are not abandoned as long as their owners claim them.

But UMSs do raise issues about how the United States will respond when others interfere with their operations. Defense Department policy on autonomous systems would prohibit a self-defense capability. Also, sea law often favors the defender. Intercepting a UMS from one’s territorial waters— especially if it is armed—fits the definition of “removing a hazard to navigation,” though with a certain amount of irony.

The legalities of intercepting an autonomous UMS operating clandestinely on the high seas are more ambiguous, and it would be hard for the Navy to prove the loss resulted from an attack and not an accident. It is likely that, at some point in the future, an adversary will hold a press conference to display an American UMS that it has hauled from its territorial waters. It is also likely that some American UMSs will be “lost at sea,” without a trace. Because no crew would be lost, officials will likely treat the capture or loss of a single UMS with much less alarm than that of a submarine. But if it turns out later that foul play was at work, things could get complicated. The capture of a U.S. Navy UMS also raises the possibility that any sensitive technologies it employs might be compromised, undermining the U.S. edge in UMS design.

How to proceed

The Navy has published several UMS master plans over the past decade. Unfortunately, these plans are only concepts. They propose how UMSs might be developed and used, but they are not connected to dollars.

This is a problem. UMS development is roughly where unmanned aerial vehicles were in the 1980s—full of promise, but not a practical weapon, not yet integrated into U.S. military planning, and lacking a sponsor. Other countries— Israel in particular—had begun to use UAVs and DARPA sponsored efforts that led to an Army-led program, Aquila, in 1979. The system became too complex, partly because it was required to fill too many requirements. As Aquila ran into technical troubles, costs rose, Congress expressed doubts, and it was cancelled in 1983.

Then the Navy began to buy much simpler, Israeli-built Pioneer UAVs for directing naval gunfire against land targets. These were successful, and eventually other services also bought off-the-shelf UAVs. Building on this experience, the Defense Department developed more sophisticated systems such as the Predator, which it used for tactical surveillance in the Balkans. Ten years later—accelerated by the Afghanistan and Iraq wars—UAVs were fully assimilated into the U.S. force structure. Today the Air Force trains more pilots for UAVs than for manned fighters and bombers. Under current plans UAV numbers will grow, while manned fighter and bombers will decline.

If we adopt some of the approaches used with UAVs—and avoid the wrong turns—it might be possible to deploy militarily useful UMSs in 5-10 years and integrate them into naval operations the way UAVs are integrated into air operations within, say, 10 to 15 years. A strategy to achieve this would:

Focus on the practical. Long-term plans depending on yet-to-be-proved technology are high risk. The Navy and industry have already demonstrated basic functions for existing UMSs. This provides a foundation for more ambitious programs. Getting systems out quickly will give crews real-world experience, build a cadre of UMS specialists, and provide feedback for developers. Military and commercial users should also share this experience, as each could benefit the other.

Avoid requirement creep. Systems that try to do everything for every commander are also high-risk; this was one lesson the UAV community learned from Aquila. Focusing first on systems with limited functions will provide technology that can be applied to more ambitious systems, and integrated with other systems. An open architecture approach will allow developers to add new technology when the risks and costs are acceptable.

Identify specific technology gaps that developers can address. Another reason to field working UMSs fast is that the early systems make clear what technologies are needed to provide the next increment of capability, reliability, or operational flexibility.

Build constituencies.When histories are written, most acquisition success stories focus on the developers. But the users are often even more important. Most successful systems have had users who said they needed the new capability. This makes officials (and legislators) more apt to commit funding. Fielding an “80% solution” UMS fast will give users a taste of the possible and win their support.

Fund analysis to develop operating doctrine, policy, and legal options. The UMS community has several trade organizations, but these deal mainly with technology and market assessments. The government should encourage companies, research institutions, and civilian government agencies such as NASA and the Department of Energy to work together and think through the broader issues UMSs will present and ways to deal with them.

To succeed the UMS community must find alternatives to the usual defense acquisition process, which is too complex and too slow. UAV proponents were creative in using options in DOD regulations such as “Other Transaction Agreements” (OTAs) and technology consortiums that allow for developing new systems,.

An OTA allows a company and a government agency to create what is in effect a joint venture. The company bears part of the cost and acquires commercial rights to technologies developed under the agreement. OTAs may be especially suited for UMSs, which already have wide commercial use and industry investment. The Office of Naval Research suggested an OTA as an option in a recent solicitation to develop non-nuclear power sources for UUSs.

There are additional arrows in the contracting quiver. Predator was developed in part under the Advanced Concepts Technology Development program. ACTD projects are exempted from much of the usual DOD acquisition process and provide early capabilities from prototypes. The Reaper UAV used another tactic: Industrial Research & Development (IRAD) funds. Defense contractors are allowed a surcharge on their existing contracts to accumulate funds that they can use to develop technologies for future DOD projects. General Atomics, which developed Predator, believed the Defense Department would need a larger, longer-ranged UAV, and used its IRAD funds for early development.

Another option is to carry out some UMS development as Special Access Programs (SAP). These programs, which are more tightly held than other classified programs, have often been faster and often more efficient because security requirements demand that they use smaller staffs. Smaller staffs also usually generate less paperwork. The F-117A stealth fighter is one example.

There are downsides to these approaches. Systems developed outside the usual Defense Department acquisition process often are a poor fit with DOD logistics; this can hurt their readiness. Systems developed under tightly compartmentalized SAPs can become “orphans” that are not updated regularly and thus get retired earlier than their mainstream counterparts. Also, the companies that developed the early UAVs resemble the companies developing UMSs today. They are smaller than the typical defense prime contractor, and more responsive. But small companies may lack the capacity of the big, traditional defense contractors. This became apparent when General Atomics struggled to keep up as demand for Predators grew.

Even so, these challenges can be managed. The technology for delivering useful UMSs to the Navy is available. Focusing on delivering UMSs quickly will drive development, and practical experience will resolve the operational and policy issues.


Bruce Berkowitz is an independent national security and intelligence analyst and the author of The New Face of War: How War Will Be Fought in the 21st Century.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>