At the bottom of the Mariana Trench, at a place called the Challenger Deep near Guam, 36,000 feet beneath the surface of the ocean, the pressures from the water above reach a crushing eight tons per square inch—about a thousand times the standard atmospheric pressure at sea level. Some comparisons ask us to picture 100 adult elephants standing on your head, which would no doubt be painful if you even survived long enough while exposed to that kind of pressure to feel anything at all.
But that has not prevented humans from venturing to this inhospitable place. Explorers have done so only a handful of times, with both crewed and uncrewed systems, in specially-designed craft that can handle the pressure. However, they’ve never done so in a long-term, systematic way—but that is about to change.
At the Woods Hole Oceanographic Institution (WHOI) in Massachusetts, a small team of scientists and engineers are developing a new class of autonomous robotic systems called Orpheus, named for a figure in Greek mythology who ventured through the depths of Hades. They will soon be able to reach any part of the deepest, darkest reaches of the sea.
The idea is to develop a small fleet of autonomous craft that can stay down for hours, perhaps days, collecting vast amounts of data and samples that will help researchers better understand everything from climate change to how life can survive at such extremes. It’s no wonder that when scientists contemplate the next frontier where such technology might be used, they look to space.
Exploring oceans here—and on other worlds
The planet’s deep ocean is the least explored and understood place on Earth. In the so-called hadal zone, from 20,000 to 36,000 feet beneath the surface, the pressures are immense, the darkness absolute, and the mysteries of life are bountiful.
It is believed that if life exists elsewhere in the solar system, it will be found on one of the ice-caked moons of Jupiter (Europa) or Saturn (Enceladus), both of which also have oceans beneath their ice. If there is life, scientists say, most likely it would take the form of microbes like the ones found at great depths in our own oceans. Or maybe there would be something else entirely. A new form of life. We simply have no idea.
NASA is currently contemplating missions to find life on other moons by landing craft and exploring beneath the ice, but they likely remain decades out. In the meantime, as we develop the technology that will be required for a craft to land and submerge in an icy foreign ocean, we are looking to the great depths of our own seas as the closest analog. That work is being done now.
Diving in with Orpheus
Two Orpheus vehicles have already been built and are being tested at various deep ocean spots around the United States, including a vast, underwater canyon along the New England continental shelf, and off the coast of Florida at a place called the Blake Plateau. Each robot is about the size of a Harley Davidson motorcycle, is shaped like a hoagie sandwich, and costs less than $200,000 to build—far cheaper than other underwater robotic systems with similar capability. They can operate autonomously, which they’ll do at the hadal depths, but can also work while connected to a tether.
The Orpheus project is part of a larger program to explore the deep ocean at WHOI called HADEX, short for Hadal Exploration. The project promises to open up a new frontier in deep ocean science and to help marine biologists better understand how organisms can survive the deep ocean’s crushing pressures. The conditions are so different down there, the adaptations required to survive must also be very different, scientists believe. The exploration of the deepest part of the ocean may lead to the discovery of entirely new life forms, possibly new Kingdoms of life.
“Once you get past 6,500 meters, everything seems to change,” says Timothy Shank, a deep-sea biologist at the Woods Hole Oceanographic Institution, and head of the Orpheus project. “The microbial system seems to change. The animal life that’s in the water column changes. The things on the seafloor change.”
“I mean, they’re reminiscent sometimes of what we see elsewhere,” he continues. “There are worms, there are shrimp. There are things like that, but they have different adaptations. You have different bio-molecules inside their bodies. They do different things in order to live there. And so there’s a whole host of questions about how they live there with their metabolism, their physiology, and what we can learn from that.”
This will not be Shank’s first time exploring the deepest part of the ocean. A previous robot called Nereus, designed within the Deep Sea Submergence Lab at WHOI, ventured down to the Mariana Trench back in 2009 and was expected to be a workhorse for deep-sea/hadal exploration. However, on a dive at the Kermadec Trench near New Zealand in 2014, the Nereus vehicle lost contact with its mothership and disappeared. It is thought that the staggering ocean pressure likely caused the robot to implode.
“It was a $14 million project,” says Shank. “We’d had over a dozen publications of novel findings just based on four dives that we had. Tremendous. New species, all kinds of stuff discovered.”
Since the loss of Nereus, the deepest part of the ocean has largely been closed to long-term exploration by researchers. Until now.
“The idea now is to build an autonomous underwater vehicle [that’s] lightweight, [and] cheap to make. We can have a fleet of them and throw them out, and they would go and survey the seafloor in the deepest parts of our ocean,” says Shank. “Let them traverse vast distances and bring that information back to us.”
It’s not easy creating vehicles that can navigate under immense pressures, in total darkness, in undersea canyons that can sometimes be just a few hundred yards wide. To do so with autonomous robots that act on their own is even harder. That’s why Shank turned to the folks who have vast experience building autonomous robots that can handle extremely harsh conditions: NASA’s Jet Propulsion Laboratory (JPL).
Sensing the landscape and seascape
For decades, scientists at JPL have been developing systems that can operate autonomously on other planetary bodies. They’ve sent them to Mars and distant asteroids, but most of those systems have been relatively simple by the standards of what they hope to eventually accomplish. None will be as challenging and complicated as those we want to send beneath the ocean of another planet’s moon. The challenges are so enormous that only recently have researchers had both the computing power and the expertise to write algorithms that may be able to deal with all the contingencies and complexities of exploring in such extreme environments.
“Autonomy is a major multiplier for what we’ll be able to do with exploration,” says Andrew Klesh, the lead Orpheus systems engineer at JPL. “It will allow us to be more audacious in our scientific questions and queries and our goals of returning data back.”
Part of that effort will depend on a sensory capability not often employed by autonomous vehicles in the deep. Most underwater systems depend on sonar for navigation, measuring the reflection of sound waves hitting an object to determine location, size, and shape. But the Orpheus system uses visual capability, employing small cameras that can map and record the local terrain.
“It’s essentially a set of eyes that can look across the ocean floor for features we can recognize and then use these features to determine how to move forward,” says Klesh. “And not only moving forward, but how our attitude and orientation changes. We do this all the time on Mars.” But another incredibly useful feature of the system is that it is built to remember where it’s been.
On Mars and at the bottom of the ocean, navigation systems like the Global Positioning System (GPS) are obviously unavailable, so spacecraft exploring other bodies in space or at ocean depths need to use different techniques to determine where they are. By memorizing various features in an environment and accessing them while in motion, these new vehicles can understand where they are, and provide that information to future expeditions.
As Orpheus traverses the bottom, it will take high-resolution, overlapping images that create a three-dimensional picture of the seafloor. It will also carry a vast array of sensors that will allow it to detect the chemical signatures of hydrothermal vents or low-temperature seeps.
“So then, as we talk about doing revisits to areas, we’re identifying interesting sites for scientists to send samplers down and then get the most interesting things back,” says Klesh.
Right now, the state-of-the-art set of algorithms for autonomous vehicles is called Terrain Relative Navigation (TRN), and it is currently working on the surface of Mars. The Mars 2020 Perseverance Rover, which landed on Feb. 18, 2021, at the Jezero Crater on the Red Planet, was programmed to allow NASA to land the rover safely by quickly and autonomously sensing its location relative to the Martian surface and modifying its trajectory as needed during descent. It also allows the vehicle to navigate hazardous terrain on the surface by building a 3D visual map of the surroundings.
By using a similar system in Orpheus vehicles, which can perform many missions and have that data available for analysis, the engineers are able to further train the system for the future, a future that they hope will include missions to space, including a potential mission to Europa.
“The environments are very similar between Europa and Earth,” says Russell Smith, the lead JPL software engineer on Orpheus, referring to the pressures in the two oceans. “After getting through the ice, the exploration is going to be very similar to what we are doing now with Orpheus.”
Plunging down to enormous depths
So far, the Orpheus platform is still in its testing phase, but Shank says that a dive at one of the planet’s deep trenches could take place as early as 2022. The tests that have been performed so far offer hope that the vehicle will perform well at great depths. The results of the most recent test, launched from the National Oceanic and Atmospheric Administration’s (NOAA) Okeanos Explorer in May, were encouraging, says Shank.
“We had eight dives with a [Orpheus] vehicle. It performed fantastically,” he says. “We were looking at altitude-hold, turns, just the ability of this thing to hold heading. And it was extremely successful. We went over 24 kilometers [15 miles] on eight dives. We imaged the seafloor, making 3D mosaics of it. We had a chemical sensor on board that senses things in situ, and was sensing differences in the oxygen content of the water as we were driving.”
As the Orpheus system continues to make strides, the scientists at JPL are beginning to look skyward, but making the leap from the deep ocean to deep space will not happen overnight.
“We are several decades out from being able to send vehicles to explore the oceans of Enceladus and Europa,” says Klesh. Several projects are underway, such as the Europa Clipper program, which will conduct a detailed survey of Europa to determine whether it harbors the conditions suitable for life. But, says Klesh, “There’s a lot of work yet to come.”
According to Shank, there is so much to learn, and now is the time to do it with a system that is cheaper and more flexible than those of the past.
“There are so many questions—it’s mind-boggling,” says Shank. “The plan is to usher in a new era, and it’s not just an era of discovery of science, but also of technology.”