Beneath the ice-covered surface of Jupiter's giant moon Europa, a vast ocean circles the globe. It is one of the few places in our solar system where, scientists believe, life may have taken hold. Various ideas have been proposed about how life could survive in that frigid environment.

One suggestion is that hydrothermal vents, like those that dot the floor of Earth's oceans, are also present in the Europan deep. NASA hopes one day to send a spacecraft to the moon's icy waters to learn their secrets.

But it won't be easy. Not only will engineers need to figure out how to burrow through kilometers of ice to get to the underlying ocean, they'll also have to teach an underwater robot how to explore a world about which neither the robot, nor the scientists who plan its mission, will have any advance knowledge.

No mission so complex has ever before been attempted. Which is why NASA wants to get in a little practice here on Earth first.

The goal of NASA's ASTEP program is to support the development of technology for exploring other worlds, and to test that technology on Earth under conditions that, to the extent possible, are similar.

One such project is Sentry, an autonomous underwater vehicle (AUV) built by the Woods Hole Oceanographic Institution (WHOI) to discover and explore deep-sea hydrothermal vents. Sentry is the successor to WHOI's Autonomous Benthic Explorer (ABE), an older AUV. ABE, too, has assisted in the discovery of previously unknown hydrothermal vents. But ABE is not able to make discoveries entirely on its own.

ABE "functions autonomously once we launch it," says Dana Yoerger, the PI of the Sentry ASTEP project, "but there's a lot of pre-programming that goes on in terms of selecting the exact survey technique, and site." Plans for Sentry call for the vehicle to discover and explore vents entirely on its own.

Without Sentry, "Right now it takes us at least three dives to locate and survey a vent site. We'd like to actually launch the vehicle and have it find the vent site, take close-up pictures and then finally get samples from the vent." All without human intervention.

To that end, Sentry will incorporate a number of improvements over ABE. Already, Sentry "can go faster, it can go farther" without needing a battery recharge and "it's easier to operate," Yoerger says.

"It's got better performance in almost every aspect." Future enhancements will include a more-precise chemical sensor, a more-intelligent imaging system, sampling tools and advanced software that will enable it to make decisions on its own.

Sentry's chemical sensor, a mass spectrometer, will detect specific chemical compounds, such as methane, dissolved in ocean water. Hydrothermal vents spew out plumes of superheated water rich in chemical compounds that are found only in trace amounts elsewhere in the ocean.

By first trolling an area for signs of unusual chemistry, and then honing in on the spot where the highest concentrations occur, Sentry will be able to locate previously undiscovered vents.

The vehicle's optical imaging system will be able not only to take pictures, but also to identify organisms that inhabit vent communities. "It would for example be able to recognize a field of mussels or a bacterial mat, or a clump of tube worms, and separate that out from geological background," Yoerger says.

Although ABE also has an optical imaging system, it is used solely for conducting photographic surveys of vents. Scientists must then interpret these images and use another vehicle to collect samples. Sentry will use its pattern-recognition system to decide on its own where to sample. And it will be able to collect those samples itself.

Yoerger points out, however, that Sentry won't "sample the vent the way, say, the human-occupied vehicle Alvin or the ROV [remote-operated vehicle] Jason would sample." Those vehicles, which are also part of WHOI's fleet of deep-sea explorers, have robotic arms with pincer-like tools on the end for collecting large vent animals, such as tubeworms, crabs and mussels.

Sentry won't try "to actually grab a tube worm, or vacuum up some of a bacterial mat." It will collect tiny stuff. Its sampling system will consist of a pump and a set of filters, which will trap microbes and the larvae of larger animals that live near the vents rather than the animals themselves. That material then will be brought to the surface and its DNA analyzed "to understand the spectrum of animals that live there," Yoerger says.

These innovations are what will enable Sentry to perform its tasks, from locating a previously undiscovered vent to sampling the life forms that inhabit it, without any human intervention whatsoever. The level of autonomy Yoerger hopes Sentry will achieve represents the baseline functionality that will be required of an underwater vehicle sent to search for life in Europa's ocean.

"If you want to go to Europa, and you want to find active hydrothermal sites and then you want to home in and sample the biology, this is an analog for that," Yoerger says.