In the Hunt for Ocean Treasures, Robots Lead the Way

EarthzineAutonomous Underwater Vehicles 2017, In-Depth, Themed Articles, Writing Fellowship 2017

Shell Ocean Discovery XPRIZE teams are building autonomous robots to map the seafloor and hunt for unknown signals in the hope of winning $7 million.

Some of the Aggie Ocean Discovery student team, pictured in front of Marlin. Image Credit: Aggie Ocean Discovery Team

In the frigid depths of the ocean this winter, swarms of sleek, free-swimming robots will be hunting the deep sea in packs. Without any human guidance, these metal bloodhounds will be searching the water for a signal that will lead them to a $1 million jackpot. One of these robots, Marlin, is a plucky, lemon-yellow, torpedo-shaped bot. He will travel the competition space, map the seabed and scan the water for chemical or biological signals. These signals are clues to the whereabouts of this jackpot, a hidden treasure on the seafloor.

Marlin is an Autonomous Underwater Vehicle (AUV). He belongs to the Texas A & M University’s (TAMU) Aggie Ocean Discovery Team, comprised entirely of students. They’re one of the teams vying for an extra $1 million bonus prize, to locate an unknown target, within the $7 million Shell Ocean Discovery XPRIZE. This global competition, among more than 20 teams, involves using advanced underwater robotics to push the boundaries of ocean exploration.

The teams competing in the XPRIZE are tasked with: delivering a high-resolution (5 meters per pixel) bathymetric map; providing images from the seafloor of a specific target; and, identifying archaeological, biological or geological features at depths greater than 2,000 meters. The bonus prize, sponsored by the U.S. National Oceanic and Atmospheric Administration (NOAA), ups the ante, as the team’s robots will have to sniff-out a pre-determined chemical signal and locate its source on the seafloor. All of this must be done without any human intervention. The robots have to “think” for themselves.

The AUV Marlin is on the hunt for a hidden signal in the ocean. Image Credit: Aggie Ocean Discovery Team

XPRIZE hopes the competition will serve as a catalyst for increased ocean exploration and discovery, by incentivizing the development of low-cost, scalable AUVs, equipped with the most advanced artificial intelligence. By making them cheaper, smarter, and faster, the goal of a high-resolution seafloor map might finally be realized. And if Marlin finds the extra jackpot, this clever robot and his so-called “smart sniffer” will radically transform our understanding of the ocean.

The ocean, the Earth’s biggest ecosystem, plays a fundamental role in regulating climate, supplying oxygen, absorbing carbon dioxide, and contributing to global economies. All these factors are linked to biodiversity, the variety and abundance of species and populations in the ocean and a strong indicator of ocean health.

Still, the ocean is under-protected. The United Nations recently laid out a goal of protecting at least 10 percent of the world’s oceans by 2020. The General Bathymetric Chart of the Oceans (GEBCO) in 2016 also laid out a goal of mapping the entire sea floor by 2030. While satellite-derived maps of the Earth exist, the resolution doesn’t allow scientists and policy experts to fully identify areas for study, conservation, and management. Obtaining maps of a fine enough scale is, at this stage, uneconomical and the primary driver for the Shell Ocean Discovery XPRIZE, organizers say.

Biodiversity in the ocean occurs in hotspots, spread out over the vast ocean floor and through all layers of the water column, hidden by an inky thick blanket of water. These hotspots include deep-sea coral communities, hydrothermal vents, and cold methane seeps, all teeming with life. In particular, hydrothermal vents support extensive communities of extremophiles.

Hydrothermal vents spew high-temperature metal-rich fluid into the cold deep ocean, forming chimneys. This hydrothermal vent, surveyed during NOAA’s 2016 Deepwater Exploration of the Marianas, is 30 meters high. Video credit: NOAA Office of Ocean Exploration

Extremophiles are capable of surviving in one of Earth’s most hostile environments – a place of extreme temperatures and pressures, and devoid of light. These qualities make extremophiles interesting not only for conservation and biomedical researchers, but also for investigating how life first began on our planet billions of years ago and, possibly, where we might find life on other planets. By building robots which can intelligently track biological, chemical or physical signals in the ocean to their source, these unique communities can be found, and then protected, without having to wait for a treasure map.

Venus flytrap anemone perched on a bamboo coral at Kingman Cone. Image Credit: NOAA Office of Ocean Exploration and Research

While recent advances in machine learning will make it possible to program the competition robots to hunt down the mystery jackpot without a map, it won’t be easy. Twelve months ago, Dylan Blakeslee had $2,000 from an unconvinced university department and a cool idea. Today, he heads the Aggie Ocean Discovery team behind Marlin – 32 multidisciplinary students at the forefront of innovation.

“It’s a lot of hard work. 80 hours a week, all day, every day, working from 4 a.m. to 9 p.m. most days. It’s really a passion for all of us,” says Blakeslee, a senior engineering student and Marlin’s project manager. They’re capitalizing on this passion to develop Marlin, an autonomous vehicle depth rated to 6,000 meters and equipped with artificial intelligence to allow it to map its own path underwater, rather than taking instructions from a human operator. They’re also developing a topside vehicle, Wahoo, a dynamically-positioned 25-foot catamaran, which will stay put in the ocean, launch and recover, and constantly monitor and communicate with Marlin to accurately determine his location.

Marlin will communicate with Wahoo, a 25-foot catamaran, for positioning. Wahoo is also able to launch and recover Marlin, replacing the need for a ship-based deployment. Image Credit: Aggie Ocean Discovery Team

Their focus however is not only on the competition but how this technology will survive and thrive beyond the testing zone. Alongside industry partners – including Deep Down Inc. and Oceaneering – Marlin’s creators want to develop “technology to benefit humanity,” while inspiring people to think more about the ocean.

“Our long-term goal is to have an open, app-enabled platform so that any student, anywhere in the world, can drop a pin and ask Marlin to research part of the ocean,” said Blakeslee. “Initially this will provide a virtual reality world they can interact with, while generating a bunch of data. We then want to take that data and make it open to all universities around the world.”

The XPRIZE competition can influence the AUV field and student lives and future careers. To Blakeslee, working with industry teaches so much more than can be learned in a classroom or lab alone. “We have an incoming freshman, and they’ve got more knowledge about AUVs and sub-sea vehicles than most graduate students,” he said. As the need for ocean exploration increases in the future, these skills will become vital.

“I think this is by far the most important thing any of us has ever done.” – Dylan Blakeslee, Project Manager of TAMU Aggie Ocean Discovery Team

Long-term focus is key to all XPRIZE competitions, as the excitement doesn’t end with the announcement of the victors. Once the check is handed over, XPRIZE partners with the winners to apply the technology globally, by creating user-friendly components for a wide range of end-users.

Chris Sabine, NOAA/Pacific Marine Environmental Laboratory director and Shell Ocean Discovery XPRIZE scientific adviser, says, “A lot of what XPRIZE does is adding value. There’s a lot of follow up on the prizes, to develop the market that goes with that competition or that new technology.”

The technology behind the “smart-sniffer” is not new, however. As Chris Scholin, president and CEO of Monterey Bay Aquarium Research Institute (MBARI), explains, “adaptive sampling is a highly-evolved field. These machines monitor what’s happening outside and then when some condition is met it triggers a sample acquisition event.”

Scholin developed MBARI’s Environmental Sample Processor (ESP), the first in situ DNA sampler. A self-contained “lab-in-a-can,” the ESP is capable of sequencing biology in the ocean on the fly, and has been used extensively to study harmful algal blooms (HABs), toxic colonies of algae that grow unimpeded, causing health problems in humans and widespread ecosystem damage.

MBARI researchers Christina Preston and Scott Jensen assembling a prototype of the third generation Environmental Sample Processor (ESP) inside the hull of one of MBARI’s long-range autonomous underwater vehicles (AUVs). Image Credit: Jim Birch

In 2015, the ESP aboard MBARIs Long Range AUV was used to monitor a large HAB in the Pacific Ocean, and many are now permanently moored to the ocean floor, providing near-instantaneous data to scientists back on shore. This technology once required collecting samples individually and then analyzing them in a lab. The ESP completely changed this approach.  

As adaptive sampling, which uses DNA to guide an AUVs trajectory, is someway off, Scholin expects competitors will use a standard suite of scientific sensors, which measure chlorophyll, backscatter, temperature, or pH, among others, and the real challenge lies in effectively programming these robots to think for themselves.

This sentiment is echoed by Peter Gurguis, Bonus Prize lead of OCEANZUS, another XPRIZE competitor. “Studying such an immense habitat requires, at times, incredibly high-tech solutions. Then again, it sometimes requires one to deploy the simplest, and often lowest-tech, solution in a manner that best serves our scientific needs,” he said. “Our efforts are a mix of those two: use technology to its greatest advantage and avoid it when you don’t need it.”

Programming how the AUVs will respond to all inputs and perform autonomous decision-making will be the biggest hurdle the teams will face. If successful, the NOAA Bonus Prize may precipitate a step change in ocean exploration, as the future of ocean science and our understanding of the deep-sea will be closely linked with artificial intelligence (AI).

Artificial general intelligence, the AI required for these robots to truly think for themselves, is many years away, so intelligent human operators will still largely be a part of the picture. Woods Hole Oceanographic Institute’s (WHOI) AUV Sentry used adaptive sampling while assessing the Deepwater Horizon spill, though humans still played the major role in decision-making.

This “co-exploration” is something WHOI is focusing on, said Carl Kaiser, Operations Manager at WHOI’s Deep Submergence Laboratory. “We’re improving how humans will interact with these vehicles over the next five years, with more sophisticated levels of human and robot interaction. This will bridge the gap between what we have now and full autonomy, which is quite some distance in the future.”

NOAA’s Alan Leonardi, director of the Office of Ocean Exploration and Research and XPRIZE scientific adviser, agrees, Advances in autonomous technologies and artificial intelligence will likely never fully replace humans in undersea data collection, but they are changing the nature of our involvement and expanding the pace and scope of our efforts both at sea and ashore.”

This increased pace and scope, while a boon for conservation, has some suggesting that our expansion into the ocean will lead to greater exploitation of marine resources. The ocean hides many secrets, one of which is deep-sea minerals, which are attractive targets for mining companies. Sabine however views things differently: “There will always be a bad side to science and exploration – but we can’t protect what we can’t see.”

By learning about and understanding seafloor communities, conservation of these precious environments is more likely, Sabine said. Increased data and understanding of the ocean is the only way these deep ocean communities will be protected.

“Our focus is on collecting the highest quality data and information to support ocean science,” said Leonardi. “The data itself is neither intended to encourage nor discourage exploitation of ocean resources but instead support effective decision-making in resource management.”

Erica Spain is a Ph.D. candidate at the Institute for Marine and Antarctic Studies in Hobart, Australia. She’s using AUVs to explore extreme environments in the Southern Ocean and Antarctic. Follow her on Twitter @xSmerica.

Editor’s Note: This article was updated on Dec. 20, 2017, to correct Sabine’s title.