ONE MORNING IN JUNE, Daniel Roa and Max McLaughlin pull over near the dusty intersection of County Road 80C and Deadman Road in northern Colorado. Their white pickup truck, which their employer got on Craigslist, carries a few tanks of helium, a black trunk packed full of instruments, and a yellow barrel stuffed with a deflated balloon.
They flip up a mast set in the vehicle’s bed, hook one side of the balloon to it, and clip the other to a carabiner near the cab. They stick a hose attached to a helium tank under the balloon’s skirt, and the gas inflates the material till it looks like a kid in a ghost costume, if that kid were the size of a Volkswagen Bus.
While Urban Sky is not the only company using balloons to observe the Earth, its product’s combination of agility, reusability, relatively low price, and onboard processing is unique. Urban Sky’s floaters are also nearing readiness: They can go where they’re supposed to when they’re supposed to, get hauled back to headquarters, and return to the field again. But the outfit is still working out occasional kinks with the sensors and needs to build a larger fleet so it can say yes when would-be customers call.
Out at the launch point, near a spot called Jimmy Creek, Roa and McLaughlin are preparing to try out Urban Sky’s “Hotspot” system, which takes infrared data of Earth’s surface and processes it, precisely mapping unusually warm areas for people on the ground. Nothing’s on fire today, but with this test flight, Urban Sky wants to see if it can get the camera to distinguish between the temperatures of, say, a sunbaked parking lot and a shady pasture.
Once the balloon is fully inflated, the men grip the bottom and remove it from the truck.
At 8:11 a.m., they simply let go.
URBAN SKY BALLOONS spun out of the co-founders’ previous jobs. Jared Leidich and Andrew Antonio, chief technology officer and CEO, respectively, met while working on a project called StratEx. The venture sent Google executive Alan Eustace via a large balloon some 136,000 feet into the upper stratosphere, from which height he would parachute back to Earth in 2014, breaking the world record for the feat. Antonio, and later Leidich, also worked at a company called World View, founded by the same people leading StratEx and collaborating on the effort. It makes stratospheric systems—the stadium-size kind—for both human flight and research. “These balloons are so big, we could launch them from one place on planet Earth,” says Leidich, an aerospace engineer who led the design of Eustace’s flight suit. “That was Roswell International Airport.” The behemoths they worked on stretched more than 400 feet high, and the New Mexico facility had the necessary infrastructure, while being far enough from other infrastructure, as well as large population centers, to make sure nothing landed on a building or a human.
At the same time World View was launching mega-flyers, StratEx was also sending up smaller weather balloons with instruments and cameras for testing. “The footage from those flights was incredible,” says Leidich. “You could see half of the state all in the same frame.” Once, when it rained, the flooded oil wells across the state appeared like hundreds of strange glints on the ground.
“It gave us an epiphany,” says Leidich. “This information is useful. Someone probably wants to know this.” In 2019, he and Antonio both left World View to start Urban Sky. Eustace was their first investor.
The team quickly converged on collecting insights about wildfires, specifically their boundaries and outgrowths. It was a smart choice. Steve Brumby, cofounder of Impact Observatory, which uses geospatial data to monitor environmental risks for customers in government, private industry, and NGOs, says fire managers could use these kinds of observations to make detailed maps showing “where the fire really is, fast enough and with high enough resolution that you can use it to plan where to put fire crews.”
It’s information that can be challenging to get with conventional methods. The US government has satellites, including those with instruments called MODIS and VIIRS, that spot sparks from space. Their data, which comes from NASA and the National Oceanic and Atmospheric Administration, is free and public, and sometimes leads to fire identification before anyone calls 911. But it isn’t always detailed, or frequent: The satellites see flareups only once they’re big, or don’t pass over a given spot often, so they’re not good at precise real-time monitoring.
Planes and helicopters can do the job, but the flights are expensive. And sometimes, if multiple sets of flames are tearing through a region, there simply aren’t enough winged resources. Even if firefighting agencies are willing to spend money, “they don’t have the actual capacity to go do it,” says Antonio, whose background in sales and marketing makes him Urban Sky’s more public-facing presence.
Additionally, as you might imagine, piloting through a disaster area is dangerous. In June and July of 2022, for instance, there were three fatal crashes of helicopters in firefighting operations. “It’s hard to see what’s going on,” says Leidich. “And there’s a lot of other aircraft in the sky.” Drones fly at the same altitudes as planes in these zones, increasing the risk of collision—with each other and with the other aircraft.
Urban Sky’s microballoons, on the other hand, can take off and land outside restricted areas like airports. The Hotspot technology, whose development was funded in 2020 by an Air Force grant of $50,000, also has a couple of one-ups on satellites and other craft: Its camera has a resolution of around 16 feet—meaning it can see a fire that’s about the size of a canoe, many times smaller than what the federal orbiters typically pick up; and, thanks to its lofty vantage, it can do so continuously, as a plane would, but for a sweeping area normally spied by satellites.
WHEN MCLAUGHLIN and Roa release the balloon this June morning, it travels up, loose and billow-flapping like a jellyfish. It soars above the scrubby plants and green pine trees at Jimmy Creek, then ascends over the still-snow-sodden mountaintops.
Soon it is gone, rising to nearly 11 miles above sea level. From there, the Hotspot device ideally will show the Air Force—by pinpointing places that are warmer than others—that Urban Sky will be able to tell where a fire is and isn’t. Roa and McLaughlin start driving eastward, to where Urban Sky software predicts the balloon will land.
Meanwhile, data pings back to headquarters, an old warehouse in an unhip area next to Denver’s hip River North neighborhood. Unlike the former industrial sites that have transmuted into co-working spaces, breweries, coffee shops, and long-lined brunch places along alleyways graced by street art, the building where Urban Sky has set up shop remains a warehouse. The air outside smells like the Purina pet food produced in a nearby plant. Still, the rent is low.
Antonio and Leidich sit in the kibble-scented air, watching the position of the top of the payload, the battery life, system and gas temperatures, and the data on terrestrial areas radiating heat.
The Hotspot system stares at Earth, soaring over a national forest, then the edge of the Front Range metropolitan area. On board, a credit-card-size computer takes in the vast information and immediately distills it—the capability that interests the Air Force. The results shoot up to an Iridium communications satellite, which bounces them back down and into the computers at HQ.
What shows up is a constellation of purple dots overlaid on a regular map, indicating where the temperature crosses a certain threshold. “Here’s a hot thing, here’s a hot thing,” says Leidich, pointing to the dots. While there’s no active fire, the system can measure the temperatures of, say, a cool green field and a hot tin roof of some rancher’s outbuilding, determine that the grass remains below the alert level and the roof rises above, then set a purple point corresponding to the latter spot.
When the first purple dot pops up on the screen, it sends happy chills through the co-founders. But halfway through the flight, they realize the attitude sensor—which tells the Hotspot where on Earth it’s pointed—isn’t working. The first clue is that a blip of heat shows up in the middle of a field, something that wouldn’t make sense unless there were a fire. They reset the sensor; they reset the camera. Nothing fixes the problem.
TODAY’S FLIGHT is Urban Sky’s inaugural test of the full system, which the team hopes to perfect soon so it can get more paying clients. Theoretically, there’s plenty of good work out there: By August, nearly 600 active blazes will have burned across the US, scorching around 5.3 million acres.
Actively burning spots aren’t the only notable parts of the inferno equation. The before and after are also telling. For example, taking the lay of flame-prone land can reveal potential fuel. And once a blaze dies out, it’s worth watching how the environment bounces back or changes. “As you would expect, the middle bit—the crisis response—gets almost all the attention and almost all the funding,” says Impact Observatory’s Brumby.
Urban Sky is currently researching the “after” picture for the United States Geological Survey (USGS) through the agency’s National Uncrewed Systems Office. That office, headed up by Lance Brady, uses more than 200 no-pilot aircraft to gather data about Earth and its hazards.
When he heard about Urban Sky, he was intrigued—though not by Hotspot, yet. The microballoons can also carry stratosphere-capable cameras that can focus through the thin air up there and take natural-color images—just regular pictures—with pixels the size of an index card. The shots cover a lot of ground at a relatively sharp resolution and for a relatively low cost. It was, says Brady, “very, very, very competitive.”
So he contracted the team to fly over a burn scar near Rocky Mountain National Park in the fall of 2021—194,000 acres left charred by the East Troublesome Fire in 2020. USGS was gathering data on how troubled the land remained: whether and how vegetation had grown back, how the ground was faring against erosion, what habitats might have recovered and become available to animals.
On the October day of the first test, the balloon had gone up where Urban Sky had predicted, at the correct time, followed the projected path, and not landed on anyone’s house. “We were very happy with that,” says Brady. One problem: The camera’s on switch never activated. The team plans to swoop over the space again. “Hopefully before the snow flies,” says Brady, and definitely sometime before December 2022.
The Hotspot technology’s camera has a resolution of around 16 feet—meaning it can see a fire that’s about the size of a canoe, many times smaller than what the federal satellites typically pick up.
Urban Sky is also successfully snapping some “after” pictures for Denver Water, the city’s hydration utility. When fire tears through a landscape, it takes out trees and plants that keep sediment—dirt, rock—in check. Freed up, that sediment washes down now-bare slopes into reservoirs and streams, dirtying the water supply and clogging flow. “We have to adjust our treatment operations to fix those issues,” says John Nolte, the utility’s geographic information systems manager.
Denver Water typically does an annual survey of the area to understand what it needs to fix where. After 2021’s Platte River Fire, it first used a drone for part of that work. But because Urban Sky’s microballoons were more affordable, Denver Water could manage more launches to catch changes over a shorter time and determine how well mitigations like erosion barriers were working. The flyers also offered more imagery from a single launch. According to the utility, the drone consultant had nearly lost its craft to the wind and was fine with the switch.
So Denver Water commissioned pilot flights on May 13, 2022, and four days later. Both had only minor glitches. To see how well Urban Sky did, the utility compared the microballoon images over the burn scar to other aerial images of the same space. “It stacked up pretty well,” Nolte says. The utility hopes to continue the collaboration, monitoring the scars with improving technology as time goes on. “They’re still perfecting a lot of things.”
ON THE DAY of the June Hotspot test, though, the proverbial winds are not blowing in Urban Sky’s favor. The attitude sensor never sets itself straight. About two hours into the flight, still without the data the team desires, Antonio and Leidich trigger a cap at the top of the bubble to open, letting out the gas and sending the balloon falling toward Earth.
Over the next half hour, the device drops down, down, down, finally landing 75 miles east of where it started, right where mission control had told McLaughlin and Roa to wait, in open grassland—big-blue-sky country owned by the Bureau of Land Management.
At least the balloon’s path was stable and predictable, and the data processing worked, even if the output itself let the team down a bit. Nothing else to be done, Roa and McLaughlin load the pile of rumpled material onto the truck and begin the journey back to Denver.
In a year or so, the building that houses Urban Sky is slated to be torn down. The company hopes that by then it will have raised its profile enough to afford a new space. If organizations like Denver Water and the USGS want to take advantage of little high flyers, that might come to pass. But Urban Sky won’t be alone in trying for customers. Its small systems float with those of other companies, like Near Space Labs and Raven Aerostar. Urban Sky’s advantage is that its reusable system is further along than other efforts and allows it to keep costs low and charge customers less than a company that needs to build a balloon for every mission.
Last summer, Aerostar equipped one of its more “tactical” (i.e., smaller) balloons—which hover and remain aloft for days—with a fire-sensing infrared camera that sends back up-to-the-moment data. On its own dime, it sent the floater to several Western blazes to test it out. “That caught the attention of a few people that were on-site,” says Mark Ketcham, an aerospace engineer at Aerostar. Now the company is having “conversations” with interested managers from the National Interagency Fire Center, which coordinates responses from different groups and regions. Aerostar also has a balloon significantly bigger than Urban Sky’s model hovering over the ongoing Moose Fire in Idaho for its own research and development. Its design and air-ducting system mean it can stay in one place, which Urban Sky’s can’t—but Aerostar’s can’t deliver data or images quite as cheaply or launch as location agnostically as the startup.
Beyond making sure its Hotspot system works reliably, the Urban Sky team—like any potential competitors—faces another hurdle before it can become a fire manager’s go-to: It has to scale up and be able to deploy balloons to a wide area, on little notice. “Not only is it like, ‘It needs to be imaged this night and not tomorrow night and not the night before,’ but you don’t have any lead time,” says Leidich. “It’s a fire.” That requires more balloons, and perhaps cross-continental on-call crews located near nodes stocked with balloons.
It might be easier to teach firefighters and USGS personnel to launch the balloons themselves—something both Urban Sky and the government agency are considering. Then they could just keep the devices on hand, a possibility that interests USGS’s Brady. Regardless, Urban Sky is going to need more than its current 10 gasbags. And more than a warehouse destined for destruction. But for now, this is where it toils.
After a flight like today’s, that toil involves a sort of medical checkup. At this point the team calls the balloon a carcass, and it requires resurrection. “You could land on a cactus or a thorn bush or something,” says Antonio. To discover breaches, team members inflate the sphere between two trusses and pump smoke inside. Lights off, they shine flashlights on the sides to look for little lesions. Antonio says this carcass has a few small ones, which the team marks with a Sharpie.
The next step is to lay the balloon on a long table and patch it with the original material, as you would mend a tent or an inner tube. The material—proprietary—is shiny and iridescent, like a deli slice of prom dress. It’s a composite, layered substance about six times thinner than the painter’s plastic you’d typically get at Home Depot. Although you can poke a hole in it, you, or a branch, can’t tear it. The team has tried.
Once repaired, the balloon can go out on another mission—perhaps looking into how fires, once extinguished, keep affecting the landscape. Or perhaps just taking a snapshot of the world as it is right now.
A picture of this very spot hangs above the work area, taken by a balloon over Denver that captured a third of the metro area in one swoop. “It doesn’t look like much,” says Antonio, “but to us that sums up basically three years.” He’s high, obviously, on that success.
This story originally appeared in the High Issue of Popular Science. Read more PopSci+ stories.
