Photographing an alien planet is incredibly difficult. Now try doing it from a massive balloon high in Earth’s atmosphere.

That’s the task Christopher Mendillo, an astronomer at the University of Massachusetts Lowell, and his colleagues have tackled with a project called PICTURE-C, which made its first test flight in September 2019. That flight was never going to capture an image of an exoplanet — but it could pave the way for a future mission to do just that. Mendillo recounted the successes and challenges of the test flight earlier this month at the 235th meeting of the American Astronomical Society in Honolulu.

“We were able to demonstrate everything we wanted to demonstrate,” Mendillo later told Space.com. “We met all of our success criteria, we had a great flight, everything worked.”

Scientists have never seen most of the 4,000 exoplanets discovered to date: these worlds are just passing shadows or invisible dance partners of their stars. Directly imaging a planet is difficult because such objects are so much fainter than but so close to their stars: Astronomers have to block out the starlight without losing sight of the planet.

That’s the promise of direct imaging, and the sort of work that Mendillo hopes a system like PICTURE-C could do, perhaps about five years down the line if all goes well. That future instrument would need to be able to latch very precisely onto a specific star, blot it out, revisit it after studying other targets and reveal a planet, all through some of Earth’s blurring atmosphere.

From, you’ll recall, a massive balloon. The direct imaging that has been done to date has used massive telescopes on the ground. And many dreamed-of future space telescopes will follow the same approach, but not for at least a decade. Suborbital, balloon-lofted instruments might bridge that gap, Mendillo and his colleagues say.

“It’s sort of unanswered question,” he said of the potential for such research. “We know we can do it from space, we know what we can do on the ground. No one’s ever done a balloon before. We think that we can do really compelling science from a balloon, but we’re not sure.”

And so they’re finding out. Mendillo first worked on two similar projects that launched on sounding rockets that don’t reach orbit. But such a flight can only ever give an instrument a few minutes to work, which isn’t long enough for direct imaging.

So the team turned to a balloon unlike any you’ve ever seen. It inflates to 400 feet (122 meters) across and takes 3 hours to climb to an altitude of about 127,000 feet, or 24 miles (39 kilometers). And there, it hovers.

Well, it’s more complicated than that. The launch site, which is located in New Mexico, only operates in the late summer and early autumn. The weather has to cooperate with winds that won’t snatch the balloon away. The balloon must set off around dawn, but the instrument can’t study the stars until night falls, making the flight a 28-hour workday for the team.

Retrieving the data requires that the PICTURE-C apparatus jettison the massive balloon and parachute toward land, ideally someplace soft. “We can watch it falling, which is really fun,” Mendillo said. “It’s nosediving and swinging all around and doing all this violent, very fast swings that you never thought your telescope would do as it skydives.”

But PICTURE-C’s first flight went smoothly, Mendillo said, and overall the team is pleased with how the telescope performed. (An onboard camera even spotted a meteor as a bonus.) For next flight, they’ll turn on a system they flew and tested but did not connect, which will block additional starlight and improve the telescope’s images.

The scientists also discovered one key issue they’ll need to fix, an interfering vibration that prevented the rig from gathering more than 6 minutes of data at a time. With that addressed, the telescope should be able to stare at any given star for a few hours at a time.

That should be long enough for PICTURE-C to spot features like asteroid belts, dust rings and perhaps particularly bright planets. The team has a list of a half dozen nearby, bright stars that seem to sport such accessories that the telescope may be able to image during its next flight. “No one’s actually taken a picture of these dust rings or planets before,” Mendillo said. “That’s the goal.”

PICTURE-C and its successors could not just fill the time before space-based direct imaging is a reality, Mendillo said, but could also prepare engineers for those missions. NASA’s Wide Field Infrared Survey Telescope (WFIRST) will carry a technology demonstration mechanism to block starlight, and two of the four missions NASA is considering to launch in the 2030s would rely on such technology to meet their science goals.

“Actually being able to build something that’s autonomous,” Mendillo said, “that can fly on its own, that can survive the environment and have all the systems operating together in a place where you can’t go in and turn a screwdriver if you need to — all the lessons you learn developing a system like that are applicable to building a space mission.”