At Everest Base Camp the silence breaks when a ladder falls out of the sky. On the other end of the flight is Milan Pandey, a drone pilot who has never worn crampons, sitting at a laptop and lowering an aluminium ladder onto the Khumbu Icefall by winch. A few kilometres above him, a team of Sherpas radios down a coordinate; minutes later the gear they need is at their feet. The scene, described by CNN this spring, is the part of the Everest drone story that the "8 minutes instead of 8 hours" headlines miss. This is not a stunt. It is the first real change to the deadliest chore on the mountain in seventy years.
The numbers are worth stating plainly, because they carry the whole story. Base Camp sits at 5,364 m, Camp 1 at 6,065 m, and the icefall between them is roughly 1.8 miles of shifting glacier. A Sherpa carrying a load crosses it in six to seven hours. A DJI FlyCart drone flies the same line in six to seven minutes. That is not an efficiency gain. It is the difference between spending hours inside the most lethal terrain on Everest and not being there at all.
The most dangerous commute in mountaineering
The Khumbu Icefall is a frozen river in slow collapse. Blocks of ice the size of houses — seracs — lean over the route and let go without warning, which is why crossings happen between roughly 3 and 5 a.m., when the cold holds the ice still. The Himalayan Database records 48 deaths there between 1953 and 2024. In 2014 a single serac killed 16 Sherpa guides in one morning. In 2023 an avalanche took three of Mingma Gyalje Sherpa's friends and fellow guides; their bodies were never recovered.
What makes the toll worse is repetition. To fix the route and stock the higher camps, Sherpas don't cross the icefall once — they cross it again and again. Mingma G describes teams going "up and down twenty times," first to find a line through the maze, then to ferry the ladders, ropes, tents, oxygen and food that a commercial expedition needs. Every one of those trips is a separate roll of the dice in a place that has killed dozens. That is the problem the drones were built to solve: not to make climbing faster, but to take a human out of the icefall as often as possible.
How a mapping startup ended up flying cargo on Everest
Airlift Technology didn't set out to reinvent Himalayan logistics. It is a small Nepali company — co-founders Milan Pandey, a computer engineer, and Raj Bikram Maharjan, an aeronautical engineer who built a DIY drone in Nepal over a decade ago and flew it in the aid effort after the 2015 earthquake. Their day job was 3D mapping.
The pivot came from a question. Bikram was mapping the Everest region for the Khumbu municipality when the local mayor asked how much weight his drones could actually lift. At the same time, Mingma G — still shaken by 2023 — had heard that drones were hauling gear on a mountain in China and wondered why not here. In April 2024, with two drones donated by DJI and the backing of the Sagarmatha Pollution Control Committee, Airlift started experimenting. It took them a month just to learn how the machines behaved in that air and that cold. Their first cleanup run brought about 500 kg of rubbish down from Camp 1 to Base Camp — and it took more than 40 flights, because they deliberately flew well under the drone's rated load to stay safe.
That caution is the whole engineering story in miniature.
Why the drones carry less than they can
The workhorse is the DJI FlyCart 30: a coaxial eight-rotor machine that weighs 42.5 kg empty, 65 kg with its two batteries, and is rated to lift up to 30 kg with a 95 kg maximum takeoff weight — at sea level. That last phrase is doing a lot of work. Lift depends on air, and near Camp 1 there is roughly half as much of it as at sea level. Thinner air means each rotor bites less, so the motors have to spend more just to hold station, and the payload you can safely add shrinks fast with altitude. A FlyCart 30 can reach 6,000 m with nothing slung underneath; ask it to carry a full load up there and the safe ceiling drops to a fraction of that. On Everest, the teams flew roughly 15 to 20 kg per trip, not 30.
Cold attacks from the other side. Airlift flew in temperatures down to −25 °C, where lithium batteries lose capacity, their internal resistance climbs and voltage sags under load — a pack that reads "full" can collapse the moment you pull current. So the batteries are pre-warmed in a heated incubator before every flight, and the drone launches into a window bounded by wind as much as by weather; one FlyCart was forced into an emergency landing when a gust hit 66 km/h. None of this is on the glossy spec sheet. It is the real envelope the pilots work inside.
DJI's newer FlyCart 100, launched in December 2025, is built for exactly these edges: 65 kg of payload over 12 km on two batteries (80 kg on one), a max takeoff weight of 170 kg, LiDAR and radar for terrain and obstacles, batteries that heat themselves, and a parachute that slows a failed drone's descent to 7 m/s. It is the difference between a proof of concept and a machine designed to do this for a living.
The season that made it stick
By 2025 the experiment had become logistics. Across the spring season Airlift moved about 2.5 tonnes of cargo with two drones and set a documented delivery record of 6,130 m. The load list is the part that convinced the mountaineering world: 444 kg of ropes and ladders for the Icefall Doctors who fix the route, 900 kg of equipment for one expedition operator, 150 oxygen cylinders for another. Rope-fixing teams saw their travel time cut by around 90 minutes a day. Airlift says the number of icefall rotations a support team has to make in a season fell from about twenty to five.
The other cargo goes downhill. Everest has a waste problem measured in tonnes — the Pollution Control Committee removed 83 tonnes of rubbish in 2025 alone, including nearly 32 tonnes of human waste — which is why the mountain gets called the world's highest garbage dump. In 2026 the bigger FlyCart 100 turned that into a routine, clearing more than 10,000 kg of waste in roughly eight-minute hops.
The economics are why it holds together. Each drone costs Airlift around $70,000 landed in Nepal, before the fuel, generators and camp costs of running it with no grid power. That still undercuts the alternative: helicopters can technically fly above Base Camp but are rarely allowed to for logistics, and where they are used, operators put drone transport at roughly a tenth of the cost.
Does this put Sherpas out of work?
It is a fair question, and the honest answer is: it depends which Sherpa. Some of the worry is real — low-altitude porters earn by the load, and a drone that carries the load doesn't pay them. But the people doing the most dangerous work tend to welcome it. Dawa Janzu Sherpa, 28, has been an icefall "frontman" for eight years and is the sole earner for his wife and two daughters. He calls the work risky and, honestly, more about the paycheck than the passion: "if there's a way to make it safer I welcome it." Drones, he says, have cut his time and risk roughly in half.
The likely shift is in the job, not the count of jobs. With drones carrying loads, Sherpas do less ferrying and more of the skilled support climbing that clients actually pay for — and they arrive at summit day less worn down by a season of hauling. Airlift is careful to say it isn't trying to replace porters; its priorities are emergency response, search and rescue, and garbage. Veterans of the mountain frame it the same way. Caroline Ogle of Adventure Consultants, five seasons at Base Camp, calls drones "part of the natural evolution" of a sport that already leans on weather forecasts and satellite phones. Lisa Thompson, a seven-summits climber, is blunter: "The mountain is still the mountain. The challenge is still real."
What Everest teaches a pilot at sea level
Strip away the altitude and two lessons on that mountain are the same ones that govern the drone in your bag. The first is physics: thin air and cold cost you lift and flight time, which is why DJI derates payload with altitude and heats its batteries — and why your own machine gives you shorter flights and softer margins on a cold winter day. The practical version for consumer batteries is the same discipline, scaled down.
The second is the rulebook. A 170 kg drone flying beyond visual line of sight above 6,000 m is nobody's idea of a casual flight, and EU law agrees: operations like these live in the Specific category, cleared only with a risk assessment and an authorisation from the aviation authority — not the Open-category rules a recreational pilot studies. Weight, distance from people and altitude decide your category, and the category decides what you may do — on Everest and over a field in Latvia alike. The mountain just shows the far end of that ladder, with the camera running.
The most dangerous errand on the highest mountain on Earth is no longer only a human's to run. That is worth more than the eight-minute headline — and it is a cleaner argument for what good drone work looks like than any spec sheet. If you want to learn to fly on the right side of the same rules, start here.
