Robots not smart enough yet for Navy deep-ocean missions.
Since a U.S. Navy remotely operated vehicle successfully recovered a nuclear bomb in 2,800 feet of ocean water nearly four decades ago, sensors and other technologies have advanced significantly--even though the basic techniques remain essentially the same.
The big breakthrough that the robotics industry has been seeking for the past 20 years--autonomous vehicles that can think independently--is finally here, but the technology is not yet mature or affordable enough for most users of deep-ocean salvage equipment.
Among the most cumbersome features of current underwater vehicles is the umbilical cable, which slows down the search process, particularly when the area being probed is 10,000 or 20,000 feet deep, says Tom Salmon. He is in charge of salvage operations and ocean engineering under the Navy's supervisor of salvage and diving, known as SUPSALV.
The agency has been involved in high-profile salvage operations, such as the recovery of the Japanese trawler Ehime Maru--sunk in the Pacific last year after a collision with the submarine USS Greenville--and the lifting last month of the 160-ton gun turret of the Civil War vessel USS Monitor.
The current generation of tethered sonar search systems, says Salmon, might be replaced with autonomous underwater vehicles within a decade, if the technology matures.
"We've probably built our last deep-ocean search system" that operates with an umbilical, he says. "Sometime in the next 10 years, I would predict that we'll be going untethered, at least for search vehicles."
There is more than one definition for vehicle autonomy, Salmon explains. Under one concept, the vehicle dives and conducts a preprogrammed search. It then comes back to the surface, so operators aboard the ship can download the data and images of what the submersible saw.
Under a more advanced form of autonomy, the vehicle would be linked acoustically, through the water, to the ship. If the submersible found a potential target, the human operator could instruct it to hover over a spot, take pictures and transmit them. "That's the true autonomous system, Salmon notes. That technology is not ready for prime time. Nevertheless, he says, "I think we are getting a lot closer than I ever thought we would during my career."
In the short term, Salmon says, SUPSALV officials would like to see increased use of untethered search systems, such as side-scan sonar or TV cameras. Current tethered devices commonly are referred to as "tow fish."
The advantages of not having a tow cable are numerous, says Salmon. The tether makes the operation highly dependent on the weather and the sea conditions, he explains. During a typical salvage mission, "We are trying to figure out where that tow fish is behind us. If we are towing in deep water, we may have 20,000 or 30,000 feet of cable strung behind the ship."
Knowing the exact location of the tow fish is important, Salmon notes, in order to plot an estimated position of the target.
The existing remotely-piloted hydraulic-operated vehicles used in salvage operations are relatively old, but quite dependable, he adds. Some recent upgrades primarily involve the sensors, which the Navy has incorporated over time, as the industry developed more sophisticated TV cameras and sonars.
Salmon expects that the next step in the modernization of SUPSALV equipment is the introduction of untethered sensor vehicles that can function autonomously. The batteries available today make it possible to send a search vehicle to swim around, probe the area and come back. "It does not take a tremendous amount of power to do that," he says. "You can get enough hours out of the battery to make it worthwhile to do that."
But that is as far as robotic and propulsion technologies can go today. The heavy-duty ROVs--the actual work vehicles that dive deep, identify targets and figure out a way to grab them and raise them to the surface--cannot operate autonomously, given the current state of technology. "That requires some muscle, some heavy work," says Salmon.
The Navy's work-class ROV used to pick up aircraft debris operates in 20,000 feet of water. These are known as Cable-controlled Underwater Recovery Vehicles. The Navy's first CURV dove 2,800 feet to recover an atomic bomb that fell from an Air Force B-52 bomber in the mid 1960s, off the coast of Spain. The Navy subsequently upgraded the CUR Vs, to support torpedo testing in depths of up to 20,000 feet in West Coast ranges. In later years, the Navy lost at least two CURVs in mishaps. The current model, the CURV III, was built 15 years ago. SUPSALV also owns a so-called Advanced Unmanned Search System, which became a test bed in the 1980s for further research on autonomous underwater operations.
CURV-type vehicles will not be autonomous, for the foreseeable future, says Salmon. "The battery technology isn't there to really support a work-class remote vehicle s ystem," he adds. "I'm nor sure when we are going to get around using the umbilical for those guys."
However, the Navy's Orion Search System--a side-scanning sonar tow-fish--eventually will be untethered, says Salmon.
Small recovery ROVs, such as the Navy's Deep Drone, which operates in 7,200 feet of water, conceivably could be converted to battery powered mode, since their missions tend to last only a couple of hours. Currently, the Deep Drone is electrically powered by a diesel generator or the power system of the host vessel.
Small ROVs get widespread use throughout the Navy. They are the workhorses for mine-hunting and ordnance-disposal units. Scientists use them to map the ocean. Most ROVs used by the commercial off-shore oil, gas and civil engineering industries are much smaller than the Navy's CURVs, and operate in less than 1,000 feet of water. For SUPSALV operations, routine depths range from 4,000 to 5,000 feet.
As long as umbilicals remain standard equipment in ROVs, the Navy could benefit greatly from power-conducting technologies, says Salmon. "We need the R&D [research and development] guys to come up with a more efficient way to transmit power down the cable."
Water currents also present problems during salvage operations, especially near the Gulf Stream, where many aircraft-recovery missions take place. During a recent helicopter recovery off the coast of Norfolk, Va., says Salmon, "We would be working one day with minimal amount of current, and the next day we would have 5 knots of current. ... We couldn't launch the [CURV] vehicle."
Aircraft recovery demands both scientific and creative skills sometimes, he says. "F-14s don't look the same after they crash. ... These airplanes go into thousands and thousands of pieces. We have to look, inspect and then figure out how we are going to salvage it."
No two operations are the same. "I can't tell you today how we would salvage a plane that crashes tomorrow, even though we done hundreds of them," says Salmon.
The difficulties created by the water current could be alleviated somewhat if the umbilical cable were smaller in diameter, he says.
Most Navy ROV cables are about an inch and a half or more in diameter. "If we are working in 10,000 feet of water, you got a cable that is 10,000 feet long and 1.5 to 2 inches wide, that's a lot of funnel area for the current to impact," Salmon says. "There is a lot of drag on that cable. If we can reduce the size of the cable, then we can operate much more efficiently." Modern commercial ROVs, by comparison, tend to have smaller cables, with diameters of 1.25 inches or less.
Sensor technologies, meanwhile, are improving, he adds. But shortfalls remain. TV systems and acoustic imaging capabilities are "getting better," says Salmon. TV cameras, however, can only see as far as the lights will project in water, about 25-40 feet. Beyond that, sonar is needed. "As we dose in on it, we try to get to the spot with our TV cameras," Salmon explains. "Right now, that takes a lot of time."
Salvage crews, he says, would like to have more advanced capabilities in acoustic imaging and sonar sensors that work from standoff ranges.
With acoustic sensors, images are created by the return from the burst of energy generated by the acoustic equipment. The process is comparable to underwater radar. With advanced computer graphics and imagery, the sonar data can produce three-dimensional derailed pictures.
"If we can do a better job of imaging the target acoustically and identify it with a sonar, then get a really good acoustic image from 200-300 feet away, you could differentiate, for example, a piece of pipe from a torpedo or a lost bomb, or a 55-gallon drum," says Salmon. Those 55-gallon drums are a dime a dozen in the ocean. "We spend a lot of time identifying 55-gallon drums, when we are really looking for a torpedo."
Autonomous underwater vehicles are available in the commercial industry, but they are largely experimental systems, says Kenneth Cast, vice president of Deep Ocean Engineering, Inc., in San Leandro, Calif.
The biggest technical challenge in this arena is inertial navigation, which helps vehicles figure out how to avoid obstacles. Most of the time, "they get stuck," says Cast. "They run out of ideas of what they are going to do to get out." Autonomous vehicles, additionally, are "extremely expensive, compared to an ROV."
It is up to the U.S. government to decide whether SUPSALV crews will be called upon to salvage an aircraft or ship. At times, the Navy chooses to fund an underwater salvage operation so it can give divers an opportunity to train.
The deeper the recovery, the higher the price tag, says Salmon. "We don't do many operations that are deeper than 10,000 feet." When greater depths are involved, the costs can be astronomical. "The biggest cost factors are mobilization and demobilization," he says. That means moving the equipment from a warehouse in the Washington, D.C., area, by air or by truck to a port, then on to a Navy or commercial ship.
Cast of Salvage
The cost of recovering an airplane ranges from $300,000 to $1.5 million, depending on the transportation arrangements. A crew of nine to 12 typically is needed, without counting the ship operators.
A key consideration in aircraft recovery is whether the airplane had a pinger, a device that produces pinging noises, so it's easier to detect. "If we detect the pinger, then we very quickly shift to sonar mode and verify that we have a good debris field that looks like the crashed airplane," Salmon explains. If the pinger can't be found--because it burned in the crash or has been buried by debris--the recovery takes much longer. In such cases, the airplane can be located using Global Positioning System satellite navigation.
Commercial airlines as a rule do not seek the Navy's help to lift aircraft wreckage from the ocean. SUPSALV only gets involved when "it's in the best interest of the U.S. government or if there is no commercial outfit available to do the operation," Salmon says. Otherwise, "we are not allowed to do this. We'd be competing with industry. We'd be taking government-owned equipment and we would put the commercial guys out of business. That is not our intent."
The raising of the USS Monitor off the coast of Virginia was funded by the Navy to create a training opportunity for its divers, who rarely get a chance to conduct live operations in deep waters, Salmon notes.
Earlier this year, SUPSALV recovered an F-14 from 10,000 feet of water, after a March 2 training accident in the Mediterranean, which killed the pilot. The recovery of that airplane cost $1 million, but it helped the Navy figure out that there was corrosion on the outer nose landing-gear cylinder, which caused the mishap. After the investigation, the Navy grounded the entire Tomcat fleet for several weeks.
This summer, SUPSALV was scheduled to salvage the wreckage of another F-14 off shallow waters in the Atlantic. It appears that the aircraft contained classified material, which the Navy wanted to retrieve as soon as possible. Navy officials also wanted to prevent the airplane from becoming a tourist attraction for the local divers.