If you’re heading to the top of the iconic mountain to look for a dead man and a 100-year-old camera, you want to start by running some tests on your device.
On the morning of July 10, 2018, a cook at K2 Base Camp in Pakistan was looking through his binoculars toward Broad Peak when he spotted something that looked like a body about 2,000 feet below the summit. The cook shared his discovery with Bartek Bargiel and his brother Andrzej, members of a Polish expedition hoping to make the first ski descent of K2, the world’s second-highest mountain. At first, the Poles thought they were looking at a corpse. But after more careful study they realized that it was a man in distress, clinging to the side of the mountain with an ice axe. There was no communication between the teams in the two separate base camps, so the Poles immediately dispatched one of their teammates, who took off running to the other camp, which was five miles down-glacier.
When he got there, he radioed back to the Bargiel brothers that the climber in trouble was the legendary British alpinist Rick Allen, who had set off on a solo attempt to push a new route up the mountain. His team hadn’t seen or heard from him for 36 hours. Bartek immediately thought of the recreational drone that he had brought with him to film his brother skiing down K2. It was a consumer-grade device called the Mavic Pro that weighed only one and a half pounds and fit in the palm of his hand. As far as he knew, no one had ever flown a small drone at this altitude, but he believed it was possible. If he could somehow reach Allen’s position, he might be able to see what was happening. And besides, it was a great way to test his theory before trying it on K2.
Several months earlier, Bartek had hacked the drone’s flight control software. Off the shelf, the Mavic Pro limits flight to only 1,640 feet above the launch point. This obviously wasn’t going to work for filming a skier descending from the summit of K2. Luckily, the manufacturer of the drone, a Chinese company called Dà-Jiāng Innovations, or DJI, had left a development debug code in one of its applications that had given Bartek a back door into the software.
Bartek quickly launched the drone, which sped over the glaciers toward Allen. When the drone was about three miles out, it suddenly stopped, reversed course, and began flying home. Bartek realized that while he had unlocked the drone’s height ceiling, the security controls on the battery were still in place and had directed the drone back home so that it wouldn’t run out of power midflight. While Allen languished at around 24,300 feet, nearly 8,000 feet higher than K2 base camp, Bartek plugged the drone into his computer and tried hacking into the battery’s security system.
His hack worked, and Bartek launched the drone again. Minutes later, he found Allen’s position on the steep slope and took a series of photos from about 100 feet away. The images showed Allen lying on his chest, hanging from his ice axe by both arms. A short distance below him, a gaping crevasse cut across the slope at the lip of a horrific 6,000-foot wall of sheer ice cliffs. Bartek recorded Allen’s location with the drone’s GPS and radioed the coordinates to Broad Peak Base Camp to guide the rescue team. It turned out that none of the climbers on Broad Peak had a working GPS device, so Bartek loaded a fresh battery into the drone and flew it back to K2.
He started flying it back and forth between Allen and the rescuers like a flying Saint Bernard with its cask of brandy. After several hours of tromping through the wind-sculpted snow, they found Allen at 7:30 pm and helped him descend to Camp 3 in the dark. It was later reported that Allen had fallen approximately 1,200 feet while descending alone from a summit attempt in a storm.
I happened upon the story of Rick Allen’s rescue at exactly the moment I was contemplating whether I could use drones on my own expedition to the roof of the world.
Climbing Mount Everest had never been a personal ambition of mine. I saw the world’s tallest mountain as a place overrun with inexperienced climbers who stacked the odds in their favor by outsourcing the most significant risks to the climbing sherpas, who carried the weight of everyone’s egos on their shoulders—and frequently paid with their lives. For me and many other climbers of my generation, the world’s highest mountain was not a worthy objective.
But that was before I found myself inexplicably drawn into an expedition that hoped to solve one of mountaineering’s greatest mysteries. It had been almost 100 years since George Mallory and Sandy Irvine were last seen at 28,200 feet on June 8, 1924, still “going strong” for the summit. Ever since, we have been left to wonder whether these two intrepid explorers might have stood on top that day, nearly three decades before the official first ascent by Edmund Hillary and Tenzing Norgay in May of 1953.
George Mallory’s body was discovered on Everest’s North Face in 1999, but his partner Sandy Irvine had never been found. Our plan was to search for his final resting spot and the pocket-sized Kodak camera that he is supposed to have carried. It was like looking for a needle in a frozen haystack. But if we could find the camera and the film was salvageable, it just might hold an image that would rewrite history.
I know it sounds crazy.
We weren’t the first to want to do this. A number of other teams had searched for that camera over the years. All had come up empty. But what if we came armed with a fleet of drones? We could cover more ground at 27,700 feet than any expedition before us … without ever having to leave camp. If we could pull it off, the device would give us the ability to quickly and safely search acres of terrain that would be nearly impossible (and possibly deadly) to cover on foot.
It was something that had never been done, mostly because until recently, the technology just wasn’t there. But the Bargiel brothers had shown that it was now at least theoretically possible. The more I thought about it, the more I became convinced that using drones to do the bulk of the searching wasn’t just a novel idea, it was essential to the success of the whole endeavor.
It wasn’t long after this revelation that I set about recruiting Renan Ozturk for the expedition. Renan was an old friend, a fellow member of The North Face Global Athlete Team. If you happened to see the film Meru, you’ll remember Renan as the guy who suffered a near fatal brain injury in a ski accident just months before he, Conrad Anker, and Jimmy Chin finally managed to snag the first ascent of the Shark’s Fin. Not only is Renan a genuinely humble badass and all around good dude, he’s also the best drone pilot I know. But I wasn’t sure he would bite. I knew that Renan wanted nothing to do with Everest. So I pitched the project as the “anti-Everest expedition.” I told him that we might not even try for the summit. It was all about solving a 100-year-old mystery, not conquering the mountain ourselves. As it turned out, I needn’t have downplayed our own prospects for climbing the mountain. Within days, Renan was in.
Renan wasn’t sure initially which drone would perform best on Mount Everest. The Mavic Pro 2, the latest iteration of the drone used by the Poles on K2, was an obvious choice. But Renan was more interested in a flying machine called the Inspire, which he described as the Mavic Pro’s big brother and one of the world’s most advanced drones. It weighs eight pounds and has carbon fiber arms that rise up like an eagle lifting its wings. It has a built-in heater that allows it to fly in extreme cold (something crucial for our expedition) and two batteries that give it a fly time of 27 minutes. Its top speed is 58 mph. The main advantage the Inspire has over the Mavic is that it comes with a more powerful camera that has better dynamic range, which means it can handle a variety of lighting conditions, including highly detailed exposure in low light. The only drawback was portability, and it was for this reason that Renan was unsure whether it would work for us on Everest.
“I just don’t know if we can even carry it up there,” he said, pointing to its case, which was four feet square and a foot and a half thick.
But even more troubling was the fact that our drone is lethally dangerous. Its long, stout blades, spinning at 8,000 revolutions per minute, are like mini Samurai swords. Since becoming a drone pilot, Renan has lost track of how many stitches he’s had from mishaps with drone propellers.
“A Mavic can fuck you up pretty good,” said Renan, “but an Inspire can kill you.”
Due to the lack of flat landing surfaces on Everest, Renan had decided someone would have to catch the drone in their hands each time it came in for a landing. “With the wind and other factors we’ll be dealing with … I’m honestly kind of terrified,” he said.Renan Ozturk and the author (left) launch a drone flight from the North Col, at 23,000 feet.Photograph: Thom Pollard
The entrance to National Technical Systems (NTS), on the outskirts of Anaheim, California, is marked by a sheet of four-by-eight plywood bearing a haphazard array of various metal signs.
DANGEROUS AREA TRESPASSERS WILL BE PROSECUTED
NOTICE ALL VEHICLES LEAVING THIS FACILITY ARE SUBJECT TO SEARCH
DANGER NEVER FIGHT EXPLOSIVES FIRES
Warning no photographic equipment is allowed on this facility. The taking of photographs by any means, long range, aerially etc., is prohibited due to national security
Duly advised, we proceeded up a gravel road in our SUV, which was chock-full of photographic equipment. We drove past barbed-wire pens filled with industrial detritus and pulled into a small lot outside a one-story metal office building. Our visit had been arranged by Renan’s wife, Taylor Rees, who is a filmmaker and climber who specializes in environmental storytelling. Rees’ friend, Christine Gebara, a Jet Propulsion Laboratories engineer who works on satellite and spacecraft systems, had pulled some strings to get us access to NTS. As the two women worked on getting us signed into the facility, Renan Ozturk started unloading the vehicle. Meanwhile, our tech guru and driver, Rudy Lehfeldt-Ehlinger, pulled a drone from one of the cases and began snapping on the blades.
We’d arranged to visit NTS to gauge the viability of flying our drone in extreme conditions like those we’d encounter on Mount Everest.
A few minutes later, a jovial walrus of a man with a massive potbelly, impressive jowls, and a thick gray mustache pulled up in a golf cart. He chuckled at our small mountain of Pelican cases. “Welcome to NTS,” he said. “I’m Randy.”
Randy Shaw, department manager and senior test lead, would be our handler for the day. As we introduced ourselves, a tractor trailer pulled into the parking lot.
“Ah, looks like we’re getting a missile today.”
“There’s a missile on that truck?” I said.
“Just a cruise missile.”
Shaw went on to explain that NTS was essentially the military version of Underwriters Laboratory, a global safety certification company. Scattered across the 160-acre campus of barren desert were various pressure chambers, centrifuges, drop towers, and shakers. Shaw pointed to a building large enough to park three Greyhound buses inside, and told us he could lower the temperature down to –200 degrees Fahrenheit. In another area, he boasted about shooting two-by-fours through a wall at 100 mph to simulate a hurricane, and of a device that could measure the shock wave that goes through the space shuttle when its rocket boosters ignite. It was clear that Shaw took delight in the incredible mayhem that he and his staff are capable of whipping up.
“Think of us as the redheaded stepchild of product testing,” he said.
While most of NTS’s clients are aerospace companies like Lockheed Martin, Northrop Grumman, JPL, and various NASA divisions, the facility can test nearly anything. It was once hired by a law firm to determine if tennis balls perform differently after being shipped in an unpressurized aircraft cargo hold. Shaw guessed that it was probably for some famous tennis player who lost a big match. (The tennis balls held up.) A Mexican potato chip manufacturer wanted to determine how many bags of chips it could cram into a box without having them pop open during shipment. In 2017, NTS tested the module in which a Kentucky Fried Chicken Zinger sandwich was launched into space. Some of the tests can take months of painstaking work to set up and last only tenths of a second. Others take years to complete.
As we spoke, a technician in a building about 300 feet away was preparing to test a 1,000-pound warhead on a mechanical shaker that would lift the weapon up and down, 100 times a minute, for days or weeks to see what might pop loose. All of this would take place in a concrete bunker buried 20 feet underground, just in case. That afternoon, another technician planned to test the newly arrived cruise missile on a drop tower. The missile would be strapped to a magnesium plate and then released from 40 feet in the air. Shaw’s guy would conduct the test, repeatedly, from inside a heavily armored bunker.
Shaw turned and looked at me poker-faced. “I haven’t had anyone in my department die in a few years, so I’m pretty happy about that.” After a few beats, he burst out laughing. “I’m sure you’re anxious to see the pressure chamber. I’ve put you guys in the corner of the facility to keep you as far away as possible from any explosives.” I caught his eye and waited for him to start guffawing again. But he climbed back onto the golf cart and, without another word, motioned for us to follow him.
The entire front of the chamber was a massive door with a circular glass opening in its center that looked like a porthole on a ship. The inside was polished steel, much of which looked as if it had been recently run over with a grinder. Two round steel plates the size of manhole covers were bolted to the outside of the right wall. Numbers were handwritten next to each bolt—torque settings, I’d later learn.
“What are these giant plates for?” I asked.
“No one here has any idea,” said Shaw. “We can only guess. This chamber is a Cold War baby. All we know is it was built for a big aerospace company.”
Shaw explained that within minutes he could depressurize the inside of this chamber to the equivalent atmospheric pressure of 85,000 feet above sea level and cool it down to –100 degrees Fahrenheit. The walls had to be solid steel a foot thick so the chamber wouldn’t implode. In other words, even though it was sitting on a concrete slab in Southern California, it could effectively mimic the conditions, at least in regards to temperature and atmospheric pressure, that the drone would face on Everest.
“By the way,” Shaw said, “I still don’t know what you guys are doing.”
“We want to fly a drone to the top of Mount Everest,” I told him.
“Really? Well, you’ve come to the right place.”
Shaw motioned for me to follow him around to the back of the chamber. Here, on a concrete slab, sat several pieces of heavy machinery. There was a boiler used to pump steam, a refrigeration unit, and two giant vacuum pumps connected to the back of the chamber with rusty four-inch pipes.
As the pumps sucked air out of the chamber, a numeric display recording barometric pressure began ticking downward. Renan and I peered through the porthole over Rudy’s shoulder as he worked the joystick on the controller like a teenager going for his high score on Grand Theft Auto. The drone, hovering about 18 inches above the floor of the chamber, veered wildly from side to side and snapped against its tethers like an angry junkyard dog. When the ticker hit 11.61 inHg—the equivalent of 24,000 feet above sea level—the drone went into a death wobble and flipped upside down. The propellers hit the metal floor and blew apart, spraying chunks of black plastic into the air like shrapnel. The Inspire 2 lay twitching on its back like a wounded animal.
“Shut down!” yelled Renan.
The test had taken only three or four minutes, but in that brief time Rudy had pushed the drone as hard as it would go. “As far as I could tell, it had plenty of thrust, which was the main thing I was worried about,” said Rudy.
“Why did it crash?” I asked.
“I’m not totally sure,” he replied.
The good news was that the drone had made it to 24,000 feet before it crashed. It was the highest Rudy and Renan had ever flown. The bad news was that the drone had only flown to 24,000 feet—4,000 feet below the height of the secret GPS coordinates where we hoped to find the long-lost remains of Sandy Irvine. And maybe, just maybe, an antique camera that could rewrite the history of the world’s tallest mountain.
All Rights Reserved for Mark Synnott