20,000 Robots Under the Sea
Jules Jaffe of the Scripps Institution of Oceanography is developing an army of underwater explorers that researchers hope will produce a fine-grained, real-time map of the movements of the sea.
If you’re a scientist who wants to study animals in their natural habitats, the process is simple enough: get a pair of binoculars, find a shady spot to sit, and watch the critters.
But what if your quarry lives deep in the ocean — and is so tiny it’s barely visible?
Jules Jaffe, a research oceanographer at the University of California, San Diego’s Scripps Institution of Oceanography, thinks he’s got a solution. With the help of a few million dollars in National Science Foundation funding, Jaffe is developing an army of small, networked, underwater robots that will drift passively along with the ocean’s currents and unobtrusively keep tabs on other things doing likewise, from algae to fish larvae to globules of spilled oil. The ’bots will relay data on what’s happening around them to human beings on the surface, providing unprecedented insight into how tiny organisms and objects travel in the complex welter of sub-surface ocean currents.
“In the ocean, there are these very fragile interactions, and when we bring stuff up to the surface, we disturb those interactions. So what’s really happening down there could be very different from what we think,” says Jaffe, a merry 60-year-old with a thatch of thick silver hair that looks like it was carefully combed a day or two ago. We’re having a lunch of salmon and spinach at a restaurant near his lab on UC San Diego’s coastal campus. “In the ocean, there’s no shortage of things to learn. It’s very complicated, and the diversity of animal life is poorly understood,” Jaffe continues. “But we can now build instruments to actually go down there, look at things in situ, and really figure things out.”
The Jan-Feb 2012
This article appears in our Jan-Feb 2012 issue under the title "20,000 Robots Under the Sea." To see a schedule of when more articles from this issue will appear on Miller-McCune.com, please visit the
Jan-Feb 2012 magazine page.
Developing the robots — officially dubbed “autonomous underwater explorers” — poses formidable technical challenges. But Jaffe is used to that; he’s been devising new technologies to gather images from underwater for nearly two decades.
He stumbled into this rarified niche almost accidentally. The son of Eastern European immigrants, neither of whom made it past high school, Jaffe grew up in Queens, New York, taking full advantage of the pleasures available to a teenager in the 1960s. “There was plenty of sex, drugs, and rock ’n’ roll in high school,” he says. He may be the only Hendrix fan in the world, though, to have walked out of Woodstock. “A friend and I snuck in,” he says. “Everyone was covered with mud, and there wasn’t enough food or bathrooms. I thought it was going to be a disaster. So we left without seeing anyone play,” he says. “I’m really sorry about that.”
As a graduate student in biophysics at the University of California, Berkeley, Jaffe started working extensively with computers. He was hired by a biomedical company that processed digital images of angiograms. But Jaffe found he didn’t like having a boss. He did like sailing, though—a hobby he’d picked up in Berkeley. So when he saw an ad with a picture of a sailboat in an engineering journal calling for someone to do image processing at the Woods Hole Oceanographic Institution in Massachusetts, he thought it would be a great way to get back to academia and have fun.
At Woods Hole, Jaffe worked on a range of systems to gather pictures of life beneath the waves, from adapting then-new digital cameras to work underwater to developing a three-dimensional sonar system he dubbed “Fish TV.” Over the years, Jaffe has applied his techniques to track zooplankton in the Red Sea, protect manatees off the coast of Florida, and to help find the sunken wreck of the Titanic. “There’s so much cool stuff out there, and people just don’t think about putting it in the ocean,” says Jaffe. “And the oceanographers typically aren’t trained in technological innovation.”
Jaffe has been at Scripps since 1988 and has been working on the underwater explorers for the last 10 years. Most of that time was spent just figuring out how to build the things. Now, he and his team are finally ready to put some prototypes into the ocean.
After lunch, Jaffe gives me a whirlwind tour of his lab, a cluster of machining shops, research cubicles, and offices populated with scruffy young engineers, in a low modernist building overlooking the Pacific. Then, wondering aloud where his notes went, he buzzes off to teach a class, leaving his senior engineer, Robert Glatts, to explain the machines to me.
“Whatever projects he comes up with, I have to develop them,” sighs Glatts, a leathery Harry Dean Stanton-esque man in jeans and sneakers. “My first response is usually, ‘That can’t be done.’ But he’s very persistent. Eventually, we reach some kind of compromise.”
There are other free-floating, data-gathering machines being used by oceanographers, foremost among them the Argo project, an international collaboration that has scattered more than 3,000 floats throughout the world’s seas. Like them, Jaffe’s explorers will track the movement of currents in three-dimensional space plus time and gather other data like temperature. But the explorers will have two advantages: the ability to be tracked while they are underwater, and a price low enough that thousands of them can be deployed in a single area.
“It’s the most fun but the most difficult project I’ve ever worked on,” says Glatts. “There’s nothing you can buy off the shelf for it. We’ve had to custom-make everything.”
The latest model explorer is about the size of a small flowerpot. Atop its bright yellow epoxy-foam body is a miniature skyline of protruding sensors — gizmos to record pressure, temperature, acceleration, and sound. A tiny piston can bring water in or out to change the explorer’s buoyancy. Inside, the device is densely crammed with batteries, a fingernail-sized flash memory card to store the data, and a GPS signaler so that the explorer can be found when it surfaces at a mission’s end.
Since satellite signals can’t reach underwater, the idea is that swarms of these submerged robots will communicate with each other via acoustic transmissions, then relay their information to buoys floating on the surface, which will then pass along the data to human controllers on shore or ship. The result, hopefully, will be a fine-grained, real-time map of how things are moving underwater.
That would be a huge boon to ocean researchers and could have all kinds of practical benefits. The explorers could help track algae blooms and oil spills, and determine the best sites to cordon off as marine-protected areas. “You could sprinkle these things on the site of a plane crash, and they’d drift with the currents along with the debris so you could figure out where the black box is,” Jaffe says.
“It’s a high-risk, high-return type project,” says Emmanuel Boss, an aquatic physicist at the University of Maine. “The ocean is a very corrosive, harsh environment. There’s all kinds of reasons why these things might not work as expected. But if Jaffe can make them work at the price he says he can, it could revolutionize oceanography. We’ll be able to observe the ocean in ways no one ever has before.”