Matthew Travis / Reuters

On a balmy night in late October 2014, Rachel Lindbergh and dozens of others stood on the grass at the end of Arbuckle Neck Road in Virginia, staring across the bay. Their eyes were trained on a spot on Wallops Island less than two miles away, where a 14-story-tall Antares rocket stood ready to blast off into space, loaded with food, supplies, and science experiments, including one that Lindbergh had been working on for two years.  

The group ticked off the seconds together as the countdown from mission control came over the portable speakers. The engines ignited, shooting thick curls of smoke from the launchpad, and the Antares began its ascent, bound for the International Space Station. For a few seconds the rocket shone like a yellow jewel against the dark sky, and then it was gone, consumed in a ball of fire. The shock wave that followed the explosion knocked some of the spectators on Arbuckle Neck Road to the ground.

“I didn’t really believe what was happening,” Lindbergh said.

Now in her second year at the University of Chicago, Lindbergh is a member of a an exclusive group no one actually wants to be in: people who have seen their work destroyed in a failed rocket launch. Resupply missions to the International Space Station, like the one carrying Lindbergh’s experiment, are routine at this point. Cargo is launched every couple months, usually from Russia or the United States. The commercial spaceflight companies NASA uses to carry out these missions are, for the most part, good at it. But everyone knows that something could still go wrong, and sometimes it does. Orbital ATK, the company that owns the Antares rockets and Cygnus spacecrafts, eventually concluded that something—they couldn’t be sure exactly what—caused the main engine system to explode, and engineers were forced to hit the self-destruct button before the rocket fell to the ground.

Lindbergh and her fellow student researchers left the water’s edge that day, got some ice cream, and got to work recreating their experiment, a study of “tin whiskers,” potentially dangerous hair-like filaments that can develop on soldering spots on metal circuitry, under spaceflight conditions. The original experiment had taken them about two years to build. This time, they hustled, made some extra tweaks, and got it on a Dragon capsule bound for the ISS atop a Falcon 9 rocket, developed by SpaceX.

Lindbergh drove from her hometown of Charleston, South Carolina, to Cape Canaveral, Florida to watch the launch in June 2015, less than a year after the original attempt. The rocket took off—and then blew up two minutes later.

No one was injured in either explosion, nor in a third incident in September 2016, when a Falcon 9 rocket exploded on the launchpad two days before its scheduled resupply mission. That’s key here, that there’s no loss of life in these kinds of flights. In the grand scheme of things, in the record of space exploration, which has claimed lives, the mourning period for cargo is quite short. Failures may cause delays in research, but experiments can often be rebuilt and supplies replenished.

Still, there’s a certain anguish in witnessing your precious work or equipment vanish in an instant. “I don’t think my brain could process it right away,” said Mike Safyan, the director of launch and regulatory affairs at the San-Francisco based company Planet, which lost 26 imaging satellites in the 2014 explosion and eight more in the 2015 incident.

Putting the experience into words proves difficult. Lindbergh considered her tin-whiskers experiment to be her “baby,” and “it’s definitely an incredibly, unbelievably traumatic experience to see your baby destroyed twice in a row.”

Stacy Hamel, the flight operations manager at the Student Spaceflight Experiments Program, which lets students like Lindbergh send experiments to the ISS, said it’s feels like “someone who writes a novel and loses their entire work because of a computer failure.” Trevor Hammond, a spokesperson for Planet, likened the experience to “getting a phone call on Sunday that all your servers that your company [uses] are on fire.” Chris Mason, a geneticist at Weill Cornell Medicine in New York City who lost some equipment for a NASA study on astronauts in the 2015 explosion, said it feels like being punched in the gut. In fewer, and perhaps the most relatable, words: It sucks.

Again, they all knew the worst could happen, and that awareness helped to cushion some of the pain. At Planet, leadership comes out before every launch to remind employees to brace for both success and failure. After all, private spaceflight companies like SpaceX and Orbital ATK have been in the game for fewer years than more established players, making their launches risky. Space is hard, the old saying goes, and the members of this unlucky group looked to this adage for some comfort. A few days after he lost about $5,000 worth of equipment in the 2015 incident, Mason received this letter in the mail from NASA:

(Courtesy of Chris Mason)

Some critics say “space is hard” is a cliche, an excuse NASA and news organizations rolls out after a launch failure to deflect attention from the pace of development in commercial crew efforts. If routine resupply missions are hard, they argue, how can humans make it to Mars? There exists a strange dichotomy when it comes to the ease or difficulty of spaceflight, one that’s sometimes perpetuated by a single entity. Just a few days ago, SpaceX, less than six months out from its last explosion, announced it’s working with NASA scientists to identify potential landing sites on Mars.

But Mason, along with the others I spoke to, believes the platitude is accurate, and not an excuse. “It gives people a sense of the necessity of failure,” he said. “That things have to explode to be corrected.”

None of them blames the spaceflight companies, either. It may be because the losses, while disappointing, can be restored, at least in most cases. Mason’s equipment reached the ISS on another flight, and he’s currently poring over the data he got back to study how spaceflight might affect astronauts’ DNA. Planet Labs, which can produce 20 of its Dove satellites per week, recently deployed 88 satellites into orbit atop a rocket developed by India’s space agency. Lindbergh and her team rebuilt their experiment—again—and put it on another rocket last April, a Falcon 9 launching from Cape Canaveral. She watched as the engines ignited, the smoke billowed out, and the rocket rose.

This time, it didn’t blow up.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.