Read our COVID-19 research and news.

The Swiss team tests a laser mapping submersible in Lake Annecy in France.


Just 18% of the ocean floor has been mapped. XPRIZE drones could change that

Next week, a small yellow and white–striped boat will slip out of port in Kalamata, Greece, and motor away from shore. The vessel won’t carry a captain or crew, just an array of electronics that will tell it where to go, and when to drop the torpedo-shaped pod lodged in its stern. Once released, the sonar-equipped vehicle will descend several kilometers into the frigid abyss of the Hellenic Trench, the deepest part of the Mediterranean Sea, and map the sea floor with pinging pulses of sound. The team behind the effort is the first of eight competing over the next few months in the finals of the $7 million Shell Ocean Discovery XPRIZE. “I’m not sure if we are crazy or not, but we decided to go first,” says Rochelle Wigley, a marine geologist at the University of New Hampshire in Durham, who leads the XPRIZE team of the Japanese Nippon Foundation and the General Bathymetric Chart of the Oceans (GEBCO), an international organization.

XPRIZE, a nonprofit based in Culver City, California, runs competitions to spur innovation, and in 2015, it turned to the problem of mapping the ocean floor, says contest director Jyotika Virmani. The catalyst was the disappearance of Malaysia Airlines Flight 370 somewhere over the Indian Ocean, and the stark realization that recovery teams knew little about what lay below the surface of the search area. “Instead of the airplane, unfortunately, they did find two new volcanoes, one of which is bigger than Mount Vesuvius,” she says.

Sharper pictures of the ocean floor could help companies look for resources such as oil. (The energy company Shell is the prize’s sponsor.) But researchers want a clearer view, too. For example, Dave Clague, a geologist at the Monterey Bay Aquarium Research Institute in Moss Landing, California, studies volcanic activity along midocean ridges—submarine mountain chains that generate new ocean crust—by identifying lava flows. But scientists have fine-scale maps for only a tiny fraction of the 65,000-kilometer-long system, limiting their understanding of how new crust forms and what happens to it as it moves away from the ridge.

Biologists also need better maps, to manage fisheries and identify deep-sea habitats. They have already discovered new colonies of cold-water corals just by looking for structures rising from the sea floor, says Craig Brown, a mapping expert at Nova Scotia Community College in Halifax, Canada. “They usually have quite dramatic topography,” he says.

So far, just 9% of the seafloor has been mapped in detail with modern sonar technology, Wigley says, and only 18% of the world’s ocean bottom has been surveyed at all, often at resolution so coarse that jumbo jets—and volcanoes—would have no trouble hiding. The rest—four-fifths of the two-thirds of the planet covered by water—is virtually unknown. As usual, the limitations are money and time. The research vessels that do high-resolution mapping cost up to $100,000 a day to operate. And they move so slowly that it would take centuries for them to chart the world’s oceans, Virmani says.

Race to the bottom

The eight teams competing for the ocean mapping XPRIZE use a mix of uncrewed surface vehicles and autonomous underwater vehicles (AUVs). 

Team NameCountrySurface OpsNumber of AUVs
ArggonautsGermanyFive shipsFive
Blue Devil Ocean EngineeringUnited StatesTwo aerial dronesTwo
GEBCO-Nippon Foundation alumniInternationalOne shipOne
KuroshioJapanOne shipTwo
PISCESPortugalOne ship, two acoustic beaconsOne
Team TaoUnited KingdomOne shipFive
Texas A&MUnited StatesOne shipOne

Satellites can also map the sea floor, by measuring slight variations in the ocean surface caused by the gravitational pull of massive seafloor features. But the resolution is crude. In recent years, researchers have turned to autonomous underwater vehicles (AUVs). They follow preprogrammed paths using inertial navigation systems that precisely track their speed and direction, and carry miniature multibeam sonars. By cruising close to the ocean bottom, they can detect contours in the seabed smaller than a meter—a vast improvement over the 50-meter resolution of a typical ship-based system working in the deep ocean, says Clague, who is not involved in the XPRIZE contest. But the AUVs are still slow. Efforts to add batteries and extend diving time only bulk up the AUV, requiring bigger ships to launch them, “which kind of defeats the purpose,” Clague says.

XPRIZE hopes its competition will spark faster, cheaper autonomous systems. Starting from shore, the eight finalists must map between 250 and 500 square kilometers in 24 hours, at depths down to 4000 meters and resolutions of 5 meters or better. They must also carry instruments to collect images of 10 interesting features and find a trophy stashed on the sea floor. The technical challenges include building instruments to withstand enormous pressure, balancing battery life against speed, and making the robots smart enough to carry out the whole operation without human guidance. “Everything is hard,” says Martin Brooke, an engineer at Duke University in Durham, North Carolina, and leader of its XPRIZE team.

Brooke’s group—mostly engineering students—will try to gain time by using heavy-lift aerial drones to carry buoys that will lower tethered mapping pods into the ocean. Most teams use an autonomous surface vessel to save their AUV’s precious power and to serve as a communication hub. The Swiss CFIS team, led by Toby Jackson, a financial trader–turned–inventor, will send 20 lightweight, 3D-printed AUVs directly from shore. Instead of sonar, they will use lasers, which can bounce light off the sea floor because they are at such close range.

Team Tao will also use a swarm approach, launching five custom-built AUVs from an autonomous catamaran it calls the “vending machine.” Eventually, the system will carry two dozen subsea drones, says team leader Hua Khee Chan, an engineer at Newcastle University in the United Kingdom, allowing half to work while the others charge. Each AUV will follow a simple vertical path, enabling it to sample the temperature and salinity of the water column as it descends. Chan says it’s “extra data that we get for free while it’s traveling.” Both Chan and Jackson say they aim to produce their AUVs for less than $25,000 a pop—a bargain compared with the sophisticated models used today, which can cost $1 million or more.

Cheaper, more flexible systems could help researchers rapidly fill the gaps in seafloor maps—and enable repeat surveys to monitor changes over time. Clague would like to measure how much lava is produced during a single eruption on a midocean ridge, which gives clues about magma generation in the mantle. Repeat mapping could also track movement along offshore faults that generate earthquakes, and in seafloor sediments after major weather events.

As XPRIZE’s sponsor, Shell reserves first rights to negotiate with each team for use of its technology, which it could use for oil and gas exploration or to monitor production wells and pipelines. Companies hoping to mine the seafloor for minerals are also eager to get a better look. But Wigley says mapping could also aid in marine protection. “If we understand the seafloor better, we can manage where it’s happening better and understand the impacts better.”

For now, that’s a long way off, and most teams are just scrambling to prepare for the competition in Greece. A Portuguese team still hasn’t tested its acoustic positioning system, which relies on a constellation of floating beacons, in deep water. “From the math, it should work,” says team leader Nuno Cruz, an engineer at the University of Porto in Portugal. “But you go into the ocean and things are not like math.” Some teams already know they won’t win, but they are fine with that. Most entered for the challenge, not the purse, and XPRIZE is pleased with the progress they’ve made, Virmani says. “We’ve already shifted the field.”

*Correction, 31 October, 4 p.m.: An earlier version of the story misstated the resolution and coverage rate of typical AUVs.