FEATURE
TOOLS OF THE (EXTREME) TRADE
Engineers and researchers in search of answers anywhere from the darkest ocean depths to the very edge of the atmosphere require tools that must operate in inhospitable conditions.
Written by Leslie Nemo
Since 1988, the ROV Ventana has been making dives off the California coast to help expand humanity’s oceanic understanding. Photo: Marike Pinsonneault/MBARI
IT’S EASY FOR SCIENTISTS TO LEARN where they would like to do research. The hard part is getting there. Far-away observations of natural phenomena make closer examinations tempting. But when the destinations themselves are too hot, too far away, too risky, or too expensive for researchers themselves to access, scientists can design instruments that reach difficult destinations on their behalf.
If researchers anticipate conditions correctly, their equipment should withstand the kinds of destructive forces the gear will encounter on its mission—or at least, last long enough to take some measurements.
Three research teams are designing the gear they send into extreme conditions: A drill meant to carve through Antarctic ice with unprecedented speed and depth, a submersible with the ability to sneak up on deep sea creatures, and a device that sends thousands of lightweight sensors drifting through severe storms. With the information each piece of engineering retrieves, researchers assemble more accurate ideas about the environments themselves and the way they shape life elsewhere on Earth.
Deliberate choices about materials and power go into each device. But since each piece of equipment gets sent into places where humans have spent little to no time themselves, experiments and creative solutions become part of the designs, too.
“I’m an engineer purist. I love to design stuff from the ground up and analyze it all ahead of time,” said Terry Hock, a researcher developing meteorological equipment that gets launched from planes. “But there’s the practical side, and that’s not how we did it whatsoever.”
The ROV Ventana deploys from the R/V Rachel Carson, the Monterey Bay Aquarium Research Institute’s (MBARI) coastal vessel, to begin a dive. Video: MBARI
Driller’s station, with augers being raised by drill mast. Custom augers are used to make an open borehole in the porous firn, which must be cased to prevent fluid leakage. Photo: John Goodge, University of Minnesota Duluth
A custom ‘swivel,’ which connects pipe of the drill string (held by the orange drill chuck) by hoses to the fluid recirculation system. The metal parts have frost buildup due to the cold temperatures. Photo: John Goodge, University of Minnesota Duluth
DRILLING DEEP
Trapped in the ice of Antarctica are tiny air bubbles holding pockets of the Earth’s atmosphere as old as the frozen water itself. The deeper the ice, the older those miniscule samples, with some dating back hundreds of thousands to millions of years. Scientists look to pull up vertical cylinders, or cores, from as far down as possible in the ice to assemble a timeline of how the global atmosphere has shifted over millennia—as well as how climate change affects the ice today.
But the glassy Antarctic sheets are thicker in some places than others, so where should researchers drill to retrieve the oldest material they can? The Rapid Access Ice Drill (RAID) was built to answer that question.
This reconnaissance tool cuts up to 3,000-meter-deep holes into ice within a few days. By quickly digging deeper than any other drill that American research teams have in Antarctica, RAID can scope out the thickest ice and flag sites for follow-up teams to investigate, explained John Goodge, the lead principal investigator on the grant-funded project. While the equipment ran successful tests near the National Science Foundation’s (NSF) McMurdo Station near the South Pole, it will be several years before the technology becomes fully operational.
To stay speedy, RAID isn’t precious about the ice it cuts through. The design, which draws inspiration from mineral exploration drills, chips the ice as it cuts with steel or tungsten-carbide cutters. (The time-consuming process of extracting a series of one- to two-meter-long research cores is left to the drills that follow in its wake.) RAID can go beyond ice, too. When a wire system runs through the middle of the drill pipe, it can exchange the ice blades for diamond bits that carve bedrock.
This downhole circulation figure shows the main elements and fluid pathways of the RAID system, from the surface to the cutting face (not to scale), explained John Goodge. The FRS is a module on the surface that removes ice cuttings from the drilling fluid so that it can be recirculated in the drill hole. The diverter, casing, and packer together make a sealed unit through which drilling fluid can flow without loss to the firn. This three-part system keeps fluid contained. While cutting an ice borehole, fluid moves down the annulus, through openings in the drill bit, and then up the inside of the hollow drill pipe, carrying ice cuttings with it. By moving fluid up the drill pipe, cuttings can be rapidly flushed out of the hole—this is the key to fast drilling in ice (and why RAID can cut through ice at 3 meters per minute). Credit: John Goodge, University of Minnesota Duluth
While researchers have retrieved chunks of the continent beneath Antarctic ice before, those come from thinner, more peripheral parts of the ice. With RAID, researchers hope to see if the ice is melting or sliding over the center of its foundation. As a geologist and former professor at the University of Minnesota-Duluth, Goodge is particularly excited about these possible samples. “There’s really no limit to how deep we can go,” he said.
Plumbing extreme depths requires planning for the harsh conditions that could derail RAID. Where the equipment will operate, temperatures hover around -40 °C and the wind can blow in the tens of miles per hour. To get the deepest cores, equipment has to go to the highest ice elevations, bringing crews more than 10,000 feet above sea level. The environment makes material brittle and people coping with altitude sickness or cold might be clumsier, so Goodge and his colleagues prioritized off-the-shelf components that can be more easily replaced if broken.
Like other drills, the RAID drill bit is wider than the pipe that comes after it. A gap will form between the chilly walls and the metal components, space where the ice could start to close in. And when it’s time to cut rock, the blades produce enough heat to melt ice, which could refreeze and trap the equipment. RAID addresses both problems the way other ice drills do: by repurposing a cleaner as a drill fluid. Estisol 140, made by the Danish company Esti Chem, was formulated as a booster to paint and adhesive removers. It also has the same density as ice and a freezing point below the conditions in Antarctic bore holes.
As a drill fluid, the clear liquid surges down the outside of the equipment—creating enough pressure to keep the cylindrical walls in place—before carrying ice chippings up the center of RAID. Rock operations reverse the flow, with the drill fluid moving from the center of the drill out toward the walls. The fluid brings rock dust and chips with it, all while keeping RAID drilling heat from melting nearby ice.
Though other ice research drills turn to Estisol for smooth operations, RAID is otherwise fairly one-of-a-kind. Other rock-cutting drills available to U.S. research teams in Antarctica or Greenland only handle much shallower ice, said Mary Albert, the executive director of the NSF’s Ice Drilling Program, which provides the equipment for different scientific crews. As a result, the components are comparatively lightweight and much easier to get on site. If those drills can be likened to a shovel or a snowblower, “RAID is like if you’re removing snow from a highway,” Albert said.
RAID weighs over 100,000 pounds and needs a military cargo plane to reach Antarctica. To get around the ice, it needs a fleet of six tractors—equipment the U.S. keeps on site but that recently fell into disrepair. It will be several years at least until funding for the heavy-duty hauling gear is restored and RAID can be deployed. “Whether it will happen in my active time or not, I don't know,” Goodge said. “At this point, I'm cautiously optimistic.”
Inserting an inflatable packer unit, which will seal the auger borehole at the bottom. Black rubber membrane is inflated with N2 gas, and balloons against the borehole walls. Photo: John Goodge, University of Minnesota Duluth
This drill string figure shows the different types of materials and borehole processes through the working depth range of the RAID drill. Near the surface of the ice sheet (white at the top) is glacial firn, which is the compacted but porous snow undergoing progressive metamorphism into dense, non-porous ice (blue), explained John Goodge. Since the firn is porous, the researchers prevent drilling fluid from leaking into the firn with a metal casing (shown in green). The brown section in ice is the open borehole, and the orange section is the ice and rock coring part of the system. The orange coring system is inserted into the brown drill string and mechanically latched at the base of the drill string. An open borehole is about 3.5 inches in diameter through most of the ice, and then coring tools are inserted in order to recover sample cores. The Burj Kalifa is shown for depth scale comparison. Credit: John Goodge, University of Minnesota Duluth
Nick Underwood, programs engineer at NOAA, prepares to release a dropsonde during a NASA-NOAA joint mission in 2022. Photo: NASA
SKIES ABOVE
To better understand hurricanes and other extreme weather, it helps to see exactly what’s going on inside. That’s why researchers have been shoving tubes full of meteorology equipment called dropsondes out of planes and into storms since the 1960s.
If launched correctly, dropsondes fall between 10 and 15 meters a second, sending pressure, temperature, wind speed and humidity readings to the plane until they hit the ocean. “We really slice the atmosphere,” said Terry Hock, the sensing facility manager at the NSF’s National Center for Atmospheric Research (NCAR), which originally developed dropsondes. About 3,000 of the instruments get launched around the world every year. The center still owns the patents and develops new generations of dropsondes.
Today’s dropsonde casing is phenolic tubing, essentially cardboard with resin running through it. Hock and his team landed on the design after seeing what materials would cave under a 100-pound weight in the lab. Dropsondes are shot out of aircraft as high as 45,000 feet in the air and in some cases, with 75 to 80 g-force, so the casing has to stay lightweight without collapsing on ejection, Hock said. Cutouts at the bottom of the tube provide protection for the temperature and humidity sensors sticking out the end while also encouraging airflow around the probe so it picks up accurate readings. Other data, like wind speed, comes in thanks to a GPS tracker communicating where the dropsonde gets blown across the storm as it falls.
An overview of the various technologies that NOAA uses to monitor severe weather events. Chart: NOAA
The data is useful for programs like the Atmospheric River Reconnaissance project at the University of California, San Diego and Scripps Institution of Oceanography. The partnership is trying to improve atmospheric river forecasts and estimates of how much rain each one might dump on the Western U.S. Since the bands come with heavy clouds, satellites have a hard time gauging how much water is inside, said Anna Wilson, assistant director with the project. When the Reconnaissance program finds an atmospheric river that hasn’t yet made landfall, pilots can take a launch crew to release about 30 dropsondes across the width of the clouds per flight. “The atmospheric river itself has a lot of moisture and wind—knowing the exact amounts are important for the program,” Wilson said.
NCAR never gets the dropsondes back after launch. If equipment fails—if it falls too fast or stops transmitting data—fixing the problem often requires experimentation. The electronics on board occasionally cut out while sailing through electric fields in storm clouds, for example. Rounding any metal edges, like the corners of the rectangular telemetry antenna sending information back to computers on the plane, helps reduce the odds that electrical fields would damage the equipment.
The latest dropsonde addressed another common failure: Bungled parachute release. Previous generations relied on a ribbon to unwrap during descent, pulling open a parachute sewn with four, squared-off sides that keep the dropsonde relatively stable. If the unraveling didn’t go as planned, the parachute wouldn’t open and the dropsonde would fall too fast to collect data. The NRD-41 dropsonde, however, takes a different approach. The tubes carry an electrical resistor wrapped in string, a duo paired with a small sensor that detects magnetic fields. When the launch tube is lined with magnets, the sensor picks up on their presence as it passes through, kicking off a timer. When the countdown ends, the dropsonde microprocessor allows the resistor to overheat and melt the adhesive, freeing the string to pull open the parachute cap.
The NRD-41 is also the lightest dropsonde yet at just under six ounces, a fraction of the first generation’s 4.5 pounds. As new designs and modifications progress, Hock and his colleagues will keep following their unusual take on development. “I’m an electrical engineer, so I’ve done a lot of analysis and system design before building anything,” he said. “Dropsondes are almost the opposite.”
Skyfora airborne expendable bathythermographs, minisonde, and dropsondes on a table before a NOAA Lockheed WP-3D Orion at the NOAA Aircraft Operations Center, Lakeland, Fla., on June 27, 2023. Photo: NOAA
A view of the vessel while in the process of submerging. Photo: Marike Pinsonneault/MBARI
Sometimes a little suction is needed to draw in creatures of the deep. Photo: MBARI
DEEPEST DEPTHS
A machine bigger than most couches gets lowered into the ocean off the coast of central California about once every other day. It’s called the Remotely Operated Vehicle (ROV) Ventana, and its job is to swim down just shy of 6,500 feet and learn more about life in water too deep for sunlight to reach.
The Ventana has been making these dives since 1988, when it was built for the Monterey Bay Aquarium Research Institute (MBARI). Recently, it’s helped collect fish, crustaceans, and more delicate creatures that live between 800 and 3,600 feet underwater for the Monterey Bay Aquarium’s Into the Deep exhibit. “Many of those animals spend their entire life without touching anything,” said DJ Osborne, the chief pilot of the Ventana. Reaching these flighty organisms and scooping them up requires a suite of equipment that ranges from the highly-specialized to the extremely simple.
Osborne and his crew operate the Ventana from aboard the research vessel the R/V Rachel Carson. A power and fiber optics cable runs from the boat down to the craft, which navigates with thrusters. However deep an ROV might go, its connection to the ocean surface must run just as long, meaning an extensive cable can add significant weight to the boat, said Jason Williams, an engineering senior manager at the Schmidt Ocean Institute, an ocean research nonprofit with its own ROV called SuBastian. ROVs also have to drag power cords through the ocean as they maneuver around, so cable weight and capacity changes how much horsepower gets dedicated to the task. “When I first designed the ROV SuBastian, that [cord] was literally one of the first thing to finalize on,” Williams said.
An underwater view of the ROV Ventana on the move. Video: MBARI
At around 6,500 feet, an underwater vehicle is under roughly 200 times more pressure than it is at sea level. Housing built to withstand these conditions can get thick, heavy and expensive, Osborne said. An easier—and popular—solution is an oil-filled, pressure-compensating system. Rubber bladders holding oil can encase equipment so that as the ROV descends, the pressure balances in the liquid, keeping the gear from collapsing. “It’s one of the most important things to have on an ROV or any kind of submersible vehicle,” Williams said. “That oil compensation circuit is how we make it happen.”
Depending on where the Ventana is going, Osborne and his team attach different equipment. If the Ventana is pursuing midwater research—depths starting at about 660 feet—it carries containers that are aboard all kinds of deep-sea collection projects. One, the detritus sampler, is a canister where the top and bottom slide open. Once snuck under a target, hydraulic pistons shut the ends. For hardier species that can handle more pressure, the midwater equipment also includes a suction sampler. Osborne and his team will navigate the Ventana so that a nozzle sucks up a fish and deposits it in one of a dozen sample containers on a rotating rack.
For benthic, or seafloor, collections, the Ventana carries an off-the-shelf hydraulic robotic arm. It doesn’t have force feedback telling Osborne and his co-pilots how hard it’s squeezing, so instead of picking up items directly, the arm picks up tools to interact with its surroundings, like a shovel or a chisel to take off a piece of coral.
Capturing deep sea creatures requires the use of several different attachments on the Ventana, ranging from the highly-specialized to the extremely simple. Photo: MBARI
The equipment in the grip of an ROV robotic arm can be as basic as a mesh bucket or even a $20 apple picker, like the one Williams bought for a researcher interested in collecting rock scrapings. For components submerged repeatedly, material considerations get more advanced. The near-freezing temperatures of the Twilight Zone hold more dissolved oxygen than warmer waters, which can make extreme depths more corrosive. “We’ve had stainless steel just eat away from corrosion,” Osborne said.
Constant exposure can lead the Ventana team to choose titanium—which withstands the deep ocean conditions well—for gear like the benthic collection arm. The metal is also challenging to weld and drill into, so the complicated frames of ROVs are more likely to be aluminum, Willliams said.
Once it has an organism in its grip, the Ventana arm can drop the collection in a drawer that extends several feet out to where the arm is working. Not everything wants to stay—anemones will want to float away, while gorgonians—a soft coral—will flop out. But here, too, the fix can be very simple. To keep crabs from scampering away, an upside-down toilet bowl lid did the trick. “They couldn't climb past the inverted lid,” Osborne added.
Leslie Nemo is an independent writer in Brooklyn, N.Y.
Crews prepare the ROV Ventana for its next mission on the deck of the R/V Rachel Carson. Photo: Marike Pinsonneault/MBARI
A basket attachment scoops up a starfish on the ocean floor. Photo: MBARI
© 2025 The American Society of Mechanical Engineers. All rights reserved.