Key Takeaway: Stratospheric drones, which developers see as capable of handling satellites’ telecom jobs, still require lighter batteries, more efficient solar cells and tougher structures.


The race is on to be the first to establish a long-term flying electronic platform, up to 70,000 feet above the earth’s surface, where it can generate enough solar power to keep flying for months or even years while connecting the entire globe to the internet and next-generation cell phone service. 

The effort is bringing social media, aerospace and large investors together — companies including Boeing, Airbus, Facebook, Google and SoftBank have been experimenting with the concept for years. 

They’re now tantalizingly close to being able to establish long-term “pseudo satellites” in the earth’s stratosphere. 

With a new breed of solar-powered drones delivering electronic connectivity to most of the earth’s land mass – even to the more than 80% that’s currently not reachable by cell towers – companies like Verizon and Facebook are looking to expand their reach and save costs.

The latest prototypes are aiming to establish a comfortable high-altitude setting for the new drones, where solar cells can provide a steady stream of energy to keep these pilotless vehicles aloft for months and years. Floating in the stratosphere, above the weather and commercial flights, winged electronic platforms could bring Internet service and 5G connectivity to the entire globe.

‘Eternal Flight’

The main obstacle these companies face now is making sure the drones stay aloft semi-permanently, providing continuous service to earthbound clients. 

“It’s the last great aviation milestone that exists — eternal flight,” says Virginia-based electrical engineer Justin Selfridge. 

Over the years, experimental models have nose-dived ingloriously back to earth, or run out of fuel after a short stay and returned to their bases.

What kind of flying machines are we talking about here? They come in a variety of shapes and sizes: a mono-wing, a modified fixed-wing, a sweeping 257-foot span powered by 10 propellers. They’re awkward-looking, with big, sprawling structures, unlike any other aircraft you’ve ever seen. 

When they take off, it’s not with a propulsive roar but often with just a gentle upward push into the air from their handlers. 

Increasing Competition

These new drones compete not only with each other, but with other aeronautical models like balloons and blimps. Commercial balloons, like World View’s Stratollite and Google’s Loon (funded by $125 million from Alphabet Inc.) have already established a market. “They now control nearly 80% of the market opportunity” for high-altitude platforms, says analyst Shivaprakash Muruganandham of research firm NSR.

“It’s the last great aviation milestone that exists — eternal flight.”

Drone-makers, though, are betting that high-altitude balloons’ Achilles heel is their lack of maneuverability and their vulnerability to turbulent weather; numerous  commercial balloons have already suffered some highly publicized crashes of their own. In blimp or airship platform development, only Thales Alenia Space, whose Stratobus could someday be a player in the field, has made a mark, says Muruganandham. “Severe competition,” he says, may limit its potential.

Among the high-flying drones that offer promise include the Hawk30 from HAPSMobile, an alliance between SoftBank and drone manufacturer AeroVironment. The craft officially launched for the first time in September, at the NASA Armstrong Flight Research Center (AFRC) in California, in a low-altitude test flight. Stratospheric tests will follow sometime before April, 2020, in Hawaii.

Boeing subsidiary Aurora Flight Sciences is developing the Odysseus. Its creators boast that Odysseus is “the world’s most capable, solar-powered autonomous aircraft” with “the biggest payload capacity available” in its field, though there’s no evidence yet that the craft, whose development has been largely cloaked in secrecy, has ever taken off.

Airbus’s Zephyr after six years of development has racked up some significant achievements. The development team has already claimed the record for longest flight by an autonomous aircraft, when Zephyr clocked almost 26 days (missing it by 3 minutes) of flight time in August 2018.

There are others, including a dark horse entry from China, Aviation Industry of China’s Qimingxing, which first launched in October 2018.

But the clear leader in the field is Zephyr, which, instead of the “eternal flight” model, now proposes a daisy chain system, with multiple drones lined up on a launching site, each ready to be rushed aloft as a predecessor fades and loses its place. Obviously, the closer each Zephyr drone comes to matching the record-breaking 26-day performance, the more successful Airbus will  be.

Zephyr is probably the closest to “operability,” says NSR’s Murganandham, whose firm expects a $2 billion “market opportunity in cumulative revenues” in the next decade for the new high altitude platforms. Airbus has already contracted with the UK’s Ministry of Defense for three Zephyr surveillance drones. 

Unfortunately, Zephyr’s endurance feat last year was followed this year by two devastating wipe-out crashes, destroying a pair of experimental models. The latest occurred in October when Zephyr encountered unexpected turbulence at its Australian testing ground, while the craft was trying to ascend to the stratosphere.

The incident highlighted one of the inherent weaknesses of the so-called “high-altitude pseudo satellites” (HAPS): structural fragility dictated by the need to keep aircraft weight as low as possible. It also cast doubts on the company’s rosy prognostications about Zephyr’s imminent prospects to establish worldwide electronic connectivity.

Obstacles to Growth

It’s not hard to figure out the appeal for the Big Tech companies of these long-lasting flying machines. Instead of investing billions in new earth-bound transmission towers and orbital satellites, the social media and cellular telephone companies get a cheap drone alternative, with some distinct benefits of their own.  Murganandham estimates the capital costs of a stratospheric drone at about $2.5 million compared to $25 million for an orbital satellite,in a back-of-the-envelope estimate.

The cost of delivering a satellite to orbit amounts to more than $10,000 per pound, according to Defense Department’s DARPA, with smaller satellites costing as much as $30,000 per pound. For a drone, there are no launching costs, little infrastructure, only the costs of taking off and landing like any other aircraft. 

Floating in the stratosphere, as much as a fiftieth of the distance to the earth’s surface as orbital satellites, they’re far better equipped to handle the floods of high-speed data that will be exchanged by the new communications technology. 

But planting a HAPS in the sky for long, workable stretch involves achieving a delicate balance of weight, energy and structural fortitude. 

The internal combustion engine, with its heavyweight components and its thirsty fuel demands, is worthless for this mission. The only way to do it, technologists like Selfridge have discovered, is with dozens of solar cells dimpling the wings and body of an ultra-lightweight aircraft, producing enough energy during the day to sustain it aloft through the night, until it can recharge under the sun’s powerful rays. 

The history of experimental high-altitude drones has been marked by modest successes and crushing failures, with high-profile companies like Facebook, Alphabet and others dropping out in frustration. 

High-altitude platforms are “still in the R&D phase,” contends John Robbins, associate professor of aeronautical science and coordinator of the unmanned aerial systems program at Embry-Riddle Aeronautical University. “We’re seeing just the tip of the spear for the kind of aircraft that will extend broad-band capabilities to much larger segments of the human population.” 

Selfridge, who is seeking funding for his own HAPS design while continuing research with an Air Force grant, says that flights like Zephyr’s 26-day flight have been carried out under optimal conditions, with maximum sunlight and a solar battery pushed to its charge-discharge limits (the number of times a battery can reliably recharge before failing). 

“What they don’t tell you is that where you’re flying makes a difference,” Selfridge says. “Think of the UK, where they only get about eight hours of daylight during the winter. And the sun doesn’t go that high — only about 30 or 40 degrees above the horizon. It’s a much more difficult problem to solve.” 

Selfridge proposes to build a big, gawky-looking drone using a helicopter design, the TURN (for Tethered Uni-Rotor Network), which, he says, solves the problem of too much weight devoted to the aircraft’s structure. The TURN is propelled by four light-weight propeller blades (the diameter of the entire craft is about 1,000 feet), each driven by a small, solar-powered engine at its tip.

“The rotors are so skinny they droop when at rest,” he says. “But once they’re spinning, you have physics helping you out. Centrifugal force provides a stiff lightweight platform.” 

“We’re seeing just the tip of the spear for the kind of aircraft that will extend broad-band capabilities to much larger segments of the human population.” 

Aside from the structural deficiencies, unmanned aircraft have still not solved the battery problem, with bulky lithium ion batteries limiting the payload capacity for some experimental HAPS. Some drone developers may be experimenting with a lithium-sulfur battery, which offers twice the storage capacity at the same weight, says Selfridge. Unfortunately, the experimental battery suffers from rapid degradation issues, though the race is on to solve that problem. Yale chemists are experimenting with various coatings to preserve lithium-sulfur cathodes, including a covering of super-strong, one-molecule-thick graphene. 

Similar experiments are underway at the University of Manchester in the UK to increase the efficiency of photo-voltaic cells, with a recent increase in the efficiency of silicon solar cells of 2%, researchers report. 

So when will there be drones tucked into the stratosphere, hibernating there long-term, like bears in their caves?

A spokesman for HAPSMobile says that the company expects the Hawk30 to be on the market sometime in 2023. Airbus plans a test next year of its multiple-drone approach, with four or five Zephyrs supplying service to an area. Other major drone manufacturers could not be reached.

But best estimates from technologists working in the field as to when stratospheric drones will be fully operative? Maybe four years.