Deep daytime darkness, lack of oxygen, near or total weightlessness, the possibility of hostile beings, vast mystery—explorers of deep space and deep seas share some similar experiences. Both realms intrigue not only researchers, but all those compelled to stare at the ocean or sky, wondering about their origins and phenomena. Increasingly, these two worlds are beginning to yield some of their secrets through automated explorers. Into space, we send unmanned gadgets like the orbiting Hubble Space Telescope. And under the waves, oceanographers are increasingly finding uses for autonomous underwater vehicles (AUVs), highly sophisticated instruments capable of defining previously uncharted ocean regions.

AUVs offer capabilities not possible with more traditional oceanographic tools. One mainstay for years has been drifters, which track and analyze currents. With special radio and sonic transmission instrumentation, drifters can relay real-time information about salinity, depth, temperature, and other parameters from remote locations.

Other widely used systems are remotely operated vehicles (ROVs), such as Jason, which researchers from the Woods Hole Oceanographic Institution used to probe the wreck of the Titanic. Deployed via a cable extended from a ship, such unmanned, tethered vehicles can descend to great depths, record sound and video, take temperature readings, and more.

However, these technologies have their drawbacks. Drifters can be swept away by strong currents. An ROV requires a costly attendant research vessel and can be cumbersome with its tether. An AUV, on the other hand, is a relatively inexpensive vehicle, has no tether, needs no research vessel to attend it, and can be programmed to fit the unique needs of varied missions. As a result, these highly versatile platforms are becoming an increasingly important part of the oceanographic research arsenal, says Jim Bellingham, principal research engineer with MIT Sea Grant’s AUV Laboratory.

Bellingham began his work with the AUV lab in 1989, when deep-diving AUVs were still clunky, million-dollar contraptions. But at the time, Chrys Chryssostomidis, MIT Sea Grant director, had a different vision, one of sleek, economical vehicles plunging beneath the sea. And through a 10-year odyssey of developing vehicles called, well, Odyssey, Bellingham and colleagues have helped give shape to that vision.

With funding from MIT Sea Grant and the U.S. Office of Naval Research (ONR), the AUV lab has been at the forefront in the revolution of AUV technology. Bellingham and his team have developed AUVs that can take on the most daunting ocean conditions and still bring back accurate information. Measuring seven feet long and weighing about 400 pounds, the latest bright yellow iteration of the vehicle, Odyssey IIB, can travel 60 kilometers on one charge of its silver-zinc batteries and dive to 6000 meters, or more than three miles beneath the water’s surface. At the same time, it can be equipped with sensors that send back information about temperature, salinity, water depth, pressure, current speed, and other measures, all in real time. Best of all, the combined components only cost about $50,000. And though it’s hard to put a price on all the hours of labor that have gone into development, it’s easy to see that, once perfected, Odyssey could put the tools of coastal resource management into the hands of countries with limited budgets for environmental efforts.

Making Decisions

What’s special about these AUVs is a sophisticated system of "layered" computer software that gives them an extraordinary decision-making ability. "For humans it’s logical thinking, but a machine has to be told what do to," says Robert Grieve, AUV lab operations manager. "Layers are a way of designing priorities, of predicting what will be most important and the wisest course of action in a given situation, and that is not an easy task in computer programming."

But it’s essential when you have AUVs operating in the often hostile environment of the open ocean. For instance, a vehicle that is programmed to follow a straight course has to be able to make a decision to turn if suddenly confronted with an underwater mountain.

But MIT Sea Grant’s AUVs are being developed to do a lot more than dodge cliffs. As the vehicles evolve, Bellingham hopes to see them expand their functions to follow regions defined by temperature, known as thermoclines. Odyssey II might also sniff out a specific chemical, follow it, and map its presence. The AUVs are also expected to be able to discern when they have wandered out of an area of interest and to search until they rediscover the desired sampling feature. This adaptive sampling capability makes these vehicles extraordinarily useful to all kinds of oceanographers.

"These devices are able to map out large three-dimensional areas of the ocean," Bellingham says. "We don’t understand much about the ocean at all, and we need to get a better handle on what generates the ocean’s internal weather, what causes plankton to bloom under certain conditions and at certain times of the year, and why fish move from one habitat to another. A lot of our understanding comes from just the kind of data that AUVs can provide."

The lab is getting more and more calls from people who want to use its AUVs. In fact, the group has to turn down missions, Bellingham reports. "We’ve done 20 cruises at this point, and increasingly they are focused less on development of our technology and more on actually acquiring useful data. It’s a rapidly changing field, and AUVs are becoming more and more accepted by the scientific community."

Scientists are particularly interested in using AUVs to map the dynamic characteristics of the ocean’s major fisheries. Fishing in the North Atlantic has undergone dramatic shifts in the past few decades, with stocks of groundfish such as flounder, cod, and haddock now at all-time lows. Coastal managers and researchers agree that closer monitoring will be needed to maintain stocks at current levels or to attempt to increase them.

Odyssey II with the Woods Hole research vessel Oceanus in Massachusetts Bay.  Photo courtesy of MIT Sea Grant.

Helping Fisheries

Kenneth Sherman, director of the National Marine Fisheries Service Laboratory in Narragansett, RI, is optimistic about the future role of AUVs in that monitoring. "Although we now use research vessels to tow our systems, we look to taking advantage of ships of opportunity, such as container vessels, in the future," Sherman says. "Autonomous vehicles would allow us to be even more independent in our sampling and data collection. I would say it’s an exciting prospect to consider small flotillas of AUVs five to 10 years from now being utilized in coastal monitoring projects, and it’s not beyond the realm of possibility."

Today, scientists are already using AUVs to collect data in settings where other systems are too expensive or insufficiently maneuverable. Working with Bellingham and his team, Henrik Schmidt, MIT ocean engineering professor, began using Odyssey II in 1996 to study Haro Strait tidal fronts, off the coast of British Columbia. In that region, extremely vigorous mixing occurs on a daily basis. Tides flow through the strait at speeds of up to four knots, often forming whirlpools large enough to swallow a small boat.

While inhospitable to small craft, the region has historically been welcoming to marine life because of its rich food sources. Waters of different salinities and temperatures mix as if in a giant Cuisinart, causing microscopic plankton to bloom. The plankton feed many kinds of sea creatures, and either directly or indirectly impact the entire food chain. For instance, the area’s high concentration of salmon nourishes one of the densest populations of killer whales in the world. But that’s hardly the end of the buffet line. Haro Strait is also home to harbor porpoises, harbor seals and Dall’s porpoises.

Unfortunately for these long-time residents, Haro Strait and neighboring waters have also become increasingly popular with tourists, fishermen, and large vessels, such as oil tankers.

"There are about 30 tankers going through the strait each day," Schmidt says. "If one runs aground and the oil spills and gets into the water column, it will quickly mix into the water, and there will be nothing we can do about it. It would be similar to the Exxon Valdez disaster, if not worse. It would quickly become a food-chain issue."

In order to find out more about tidal fronts in Haro Strait, Schmidt wanted to send Odyssey II directly across the front as the freshwater runoff from the mainland collided with the saltwater ocean tides. That’s not unlike asking a dinghy to stay on course a few paces behind Moses and the Israelites as the two halves of the parted Red Sea rejoin each other. No float could have performed the task; even the most sophisticated ones would have been flushed out of the strait by the tides within minutes. Likewise, a towed sensor would have been too difficult to steer, since changing course with a long cable trailing behind a ship can require miles of turning radius.

"With an AUV you eliminate the hassles of the tether," says Grieve. "Just like a boat or a submarine, the AUV turns and follows the next track line so you can ‘mow the lawn’ in an area of interest. It also eliminates the surface effects of the tether, so the sensors can be stable within centimeters."

When researchers begin envisioning the potential of AUVs for monitoring fisheries in Haro Strait and other regions, the possibilities grow quickly. Counting fish stocks is currently done with sonar, which can only give a rough estimate of the amount of fish in an area and cannot accurately distinguish among species. AUVs could follow fish tagged with acoustic beacons and estimate their numbers.

AUVs might also help out in monitoring shellfish. Currently, wildlife personnel monitor the level of dangerous bacteria in shellfish manually. This method is time-consuming and does not allow continuous readings across wide areas. With a resident AUV, or better yet, a roving band of AUVs, researchers could receive data continuously from large shellfish beds. Such information might even help them predict when bacterial levels are about to become toxic. In addition, scientists on shore could program AUVs to follow and monitor endangered or rare aquatic species and then relay those movements back to shore.

"You can even install video cameras and thoroughly document the sea life in an area," says Grieve. "Even better, you can deploy the vehicle from a land station and download all the data without going out and recovering it. We can sit here in the office and collect data from the Labrador Sea."

And that, in fact, is what the MIT Sea Grant team did as part of its Autonomous Ocean Sampling Networks (AOSN) program, which has used AUVs to observe and track the temperature effects of massive freshwater runoff into the Labrador Sea. The long-term goal of the AOSN program is to create a network of AUVs that can be deployed throughout the world to track and model trends in ocean temperature and weather. Work in this vein will continue in the Atlantic Layer Tracking Experiment (ALTEX), for which an extra-large, four-meter–long AUV is now being designed. This device will dive beneath the layer of ice in the Arctic Ocean to relay information about temperature and salinity. The method of information recovery is somewhat ingenious: The AUV will release floats that literally burn their way up through the ice until they break through to the air above, where they begin broadcasting data back to the MIT laboratory.

Understanding Global Climate Change

In addition to widening the envelope of AUV potential, researchers hope the ALTEX project will provide important new information about the poorly understood phenomenon of global climate change. Setting sail off eastern Greenland, the ALTEX AUV will head north toward the Arctic Ocean and then follow the continental shelf off the Bering Sea. As it does, it will map plumes of warmer, saltier Atlantic water as they pervade the colder Arctic ocean.

"The real question is the heat flux question," Bellingham says. "The Arctic Basin climate is changing, and the question is, is this a product of a global warming trend, or part of natu-ral background variation? To answer these questions, we have to understand the processes that cause change in the ocean. Currently, our understanding of heat flux through interchanges in the Arctic Basin is poor."

The use of AUVs throughout the world could contribute to an understanding of ocean and weather dynamics leading to improved climate predictions. Allan Robinson, Gordon McKay professor of geophysical fluid dynamics at Harvard University’s division of engineering and applied sciences and department of earth and planetary sciences, has received funding from ONR to use Odyssey to refine his system of predicting dynamic changes in the ocean. The Harvard Ocean Prediction System (HOPS) is a sophisticated model of flow, temperature, plankton growth, acoustical propagation and other parameters that allows Robinson to predict the movements of temperature and pressure fronts in the ocean just as a television weatherman might predict snow, rain, or hail. Robinson has been collaborating with the AUV lab for over three years, and has used data from the Haro Strait mission to do forecasting in that region. He has also used AUVs to collect data and forecast "water weather" in Massachusetts Bay.

"If we look at just one feature at one place and time, it’s hard to interpret because it’s just one small piece of the action," says Robinson. "If AUVs see mixtures of high and low temperatures and chlorophyll, that means that plankton is blooming, probably because some new nutrients have arrived."

Nutrients from sewage outfall pipes in Massachusetts Bay cause such blooms to occur unpredictably, and the frequent use of the waterway for shipping and boating causes additional unpredictable currents. Robinson programmed the probe to divert from its prescribed mission and explore these intense plankton blooms while research ships in the area collected coordinated sampling of related phenomena.

"This was an experiment at the cutting edge of research," Robinson says. "It was multidisciplinary and multi-scale and, in my opinion, the impact of the data that AUVs collect is enhanced by placing it in the context of supporting data from other scales."

Predicting weather beneath the ocean’s surface will allow oceanographers to gain more insights into management of fisheries, endangered species, and other natural resources. In this way, AUVs are opening a window to a new level of understanding about life and the environment under the sea. But the applications of AUVs may spread even wider than that. Bellingham envisions AUVs that will provide site surveys for offshore oil fields or accomplish route surveys for telecommunications cables. MIT Sea Grant AUVs have already been loaned to a pilot project at a NATO marine laboratory in La Spezia, Italy, to detect underwater mines. The experiment was so successful that it led to a five-year program to develop this capability further.

"Oil, telecommunications, and mine detection are all areas where I think we’ll see AUVs working in the future," Bellingham says. "But I think what I can predict is that AUVs will outstrip our imaginations. You may see AUVs that will be able to do complex manipulations, perhaps even run factories on the sea floor. It’s hard to guess, but one thing is for sure: The ocean is a huge part of our planet, and there are a lot of resources locked away down there. AUVs can help us get in the door."