Earth Observatory / NASA

For the last five years, a small Mars colony thrived in Hawaii, many miles away from civilization.

The Hawaii Space Exploration Analog and Simulation, or HI-SEAS, was carried out in a small white dome nestled along the slope of a massive volcano called Mauna Loa. The habitat usually housed six people at a time, for as long as a year. They prepared freeze-dried meals, took 30-second showers to conserve water, and wore space suits every time they left the dome. To replicate the communication gap between Earth and Mars, they waited 20 minutes for their emails to reach their family members, and another 20 to hear back. Sometimes, as they drifted off to sleep, with nothing but silence in their ears, they really believed they were on Mars.

In February of this year, something went wrong. The latest and sixth mission was just four days in when one of the crew members was carried out on a stretcher and taken to a hospital, an Atlantic investigation revealed in June. There had been a power outage in the habitat, and some troubleshooting ended with one of the residents sustaining an electric shock. The rest of the crew was evacuated, too. There was some discussion of returning—the injured person was treated and released in the same day—but another crew member felt the conditions weren’t safe enough and decided to withdraw. The Mars simulation couldn’t continue with a crew as small as three, and the entire program was put on hold.

But the habitat on Mauna Loa was not abandoned. While officials at the University of Hawaii and NASA investigated the incident, the wealthy Dutch entrepreneur who built the habitat was thinking about how the dome could be put to use.

Henk Rogers made his money designing computer games, but he is passionate about space exploration, and particularly the idea of constructing human settlements on other worlds. Life on Earth, just like his computers, needs a backup, he has said. It’s why he agreed to build the habitat, and why, when the latest Mars simulation came to an abrupt end, he saw an opportunity.

Under Rogers’s direction and funding, the HI-SEAS habitat will reopen this year—not as a Mars simulation, but a moon one.

“It’s my habitat, for chrissakes,” Rogers told me in a recent interview. “I don’t want to see it sitting there empty and do nothing.”

Rogers has long wanted to build another habitat on Mauna Loa specifically for moon simulations. For him, a moon colony is the next logical step in human space exploration, and a necessary milestone before a Mars mission.

“I describe it like this: You’ve just invented a canoe and you’re sitting on Maui and you’re looking at Lanai, which is right next door, and someone says, hey, let’s row to England,” he said. “I’m saying let’s row to Lanai first. Let’s learn how to live on the moon before we start trying to live on Mars.”

Over the summer, as NASA officials deliberated whether to maintain funding for the HI-SEAS program, Rogers and a small team got to work refurbishing the habitat. They installed new floors and furniture and upgraded the computer systems. They spruced up the interior design to make the habitat look more “space-y.” They replaced the aging space suits with sleek new versions.

The first test simulations begin next week, and will last about two days. The regular simulations will last several weeks. Rogers plans to solicit research proposals from scientists from around the world.

The makeover doesn’t mean that the Mars simulations are done for good. There will be a mission seven someday, said Kim Binsted, a professor at the University of Hawaii and the HI-SEAS principal investigator who, years ago, approached Rogers about building the habitat on Mauna Loa.

After the incident in February, NASA reviewed Binsted's grant. Human-research programs undergo similar reviews every year, but this one was especially "intense," she said. “They asked a bunch of questions.”

By mid-October, Binsted had obtained formal approval to continue her research and even received more funding. NASA did not assign blame or issue punitive measures. “Everything I’ve heard from them is that they are still confident in the work that we’re doing and eager to get our results,” she said.

But HI-SEAS had run out of time. A new simulation this year would take Binsted and her team beyond the deadline of the grant.

“At this point, because it’s taken so long to sort out the reviews and the funding, if we were to start recruiting now, we wouldn’t be able to even start a mission for another four months, and then there’d be the eight-month mission, and then there’d be the data analysis after that,” Binsted said. “Realistically, we wouldn’t have been able to get the results for roughly two years, and they wanted our results sooner than that.”

Instead, NASA asked Binsted and her team to spend the next year analyzing the data that has accumulated over the six missions, said Jenn Fogarty, the chief scientist at NASA’s Human Research Program, the office that provides financial grants to HI-SEAS.

After that, Binsted hopes to apply for another round of funding and organize another NASA-funded Mars simulation in 2020.

I asked Binsted whether she was disappointed that the simulations are over for now. “It’s actually a bit of a relief, to be honest,” Binsted said. Running back-to-back simulations was challenging and time-consuming. “Even under our original plan, we would have been on a really tight timeline to get NASA the answers they needed on their timelines.”

Plus, she’s got other projects on her plate. Binsted relocated from Hawaii to Washington, D.C. in September for a yearlong science and technology policy fellowship in Senator Sheldon Whitehouse’s office, a move that was in the works since before the incident. She plans to work on the HI-SEAS data remotely.

Members of the failed mission have also moved on. They’ve found new gigs in the United States or returned to their old positions. Michaela Musilova, an astrobiologist from Slovakia, just started a new job as the program manager of Rogers’s new moon simulations.

Musilova spent the summer in Hawaii, working at the Canada-France-Hawaii Telescope. She wondered whether she should return to Slovakia, where she is a professor and a chair of the Slovak Organization for Space Activities, a research organization. When she heard from Binsted in early September that there would not be another Mars mission, that move seemed likely. But then she caught up with Rogers, who had met with the crew before their mission in February. Rogers offered her the job.

Musilova made the bumpy drive up to Mauna Loa in September. The last time she had been at the habitat, she was collecting the belongings she had unpacked barely a week earlier. It was difficult to return, she said.

“It brought back memories of mission six, our time together there at the habitat, and our personal and collective plans for what was meant to be an eight-month-long mission,” Musilova said. “It was a sad realization that all those plans were not going to be put in place.”

In some ways, the moon simulation won’t be so different than a Mars one. Crew members will still need to be maintain the habitat and its various systems, like power, water, food, and the composting toilet. They will need to suit up before leaving the dome. And they will explore the rocky, rust-colored landscape of Mauna Loa, shaped by ancient lava.

The research focus will shift. The Mars simulations were meant to study how people would behave in long-duration, deep-space mission with minimal or no contact with Earth. Researchers instructed participants to wear devices to track their vitals, movements, and sleep; complete countless questionnaires about their own behavior and interactions with others; and journal several times a week about their feelings.

In the moon simulations, participants will be researchers instead of lab rats. They will conduct science experiments and test new technology aimed at surviving on the lunar surface, sometimes on behalf of researchers who provide the instruments but can’t reside on Mauna Loa. Like on the International Space Station, crews will rotate in and out as experiments hum along in the background.

“You could say that we’ve turned HI-SEAS into a high-end moon AirBnB,” Rogers said.

Rogers said the habitat is safe. He said the accident in mission six occurred because a crew member removed a safety panel from a circuit breaker, exposing wiring. ( Stojanovski, the participant who felt some conditions were unsafe, had said the circuit box was missing a protective panel, and officials have declined to describe the situation.)

But there will be some precautionary measures in place this time around. In the Mars simulations, when a cloudy or rainy day prevented the solar batteries that powered the habitat from charging, the crew had to suit up, go outside, and turn on a generator that runs on propane. Now, the switch from the solar batteries to the backup generator will either occur automatically or be initiated by mission control, which will run out of Rogers’s ranch on the Big Island.

“We’re trying to make sure that nothing could go wrong like it happened during my mission,” Musilova said.

Like their Mars counterparts, the moon participants will be left alone on the volcano during missions, but they will be far more in touch with the outside world. The dome will lose most of its disconnected coziness. The 20-minute communication delay will shrink to three seconds. Rogers has installed cameras in the common spaces of the habitat, and the team is currently debating whether to leave them on 24/7 or use them only in cases of emergency. Rogers is even thinking about live-streaming a mission to the public.

“We couldn’t do [that] during the NASA missions because having cameras around was considered an invasion of privacy, but we’re not running a honeymoon resort. We’re running a moon base,” he said.

Rogers’s plans come at a lively time for moon exploration. Under the Trump administration, the U.S. space program has put a strong focus on developing lunar technology, and Jim Bridenstine, the NASA administrator, has said the space agency will put robotic spacecraft on the moon as early as next year, and astronauts back on the moon within 10 years. ISRO, the Indian agency, plans to put a rover and lander on the lunar surface by next spring. Russia wants to land its first cosmonaut on the moon in 2030 and eventually establish a permanent outpost. The European Space Agency, a collection of 22 nations, announced in October that it will attempt to re-create the moon’s surface at a facility in Cologne, Germany, so that astronauts can train for future missions. Around the world, commercial companies around the world are proposing lunar missions at a pace the world hasn’t seen since the Apollo program.

“The quest for creating a permanent, sustainable settlement on the moon has gotten to be exciting,” Rogers said. “Everybody’s starting to talk about it now.”

Rogers plans to build a brand-new habitat specifically for moon simulations in Pohakuloa, another remote section of Mauna Loa, located less than 15 miles from the former Mars analog. A second facility would allow Binsted to return to the Mauna Loa habitat. And what if Binsted is ready to operate another Mars simulation, but Rogers is still carrying out the moon ones?

“That’s a matter of negotiation between us,” Binsted said. “The University of Hawaii has the permit to use the land. Henk owns the physical structure. So neither of us can get on without the other one.”

I asked Binsted whether it was strange, after all her years of work, to picture the Mauna Loa landscape as the moon, in part because the color of the rock, a lovely burnt sienna, is distinctly Mars.

“The thing that makes the HI-SEAS site good for [space simulations] is more about the isolation, the lack of visual stimulation, the lack of life outside the habitat,” she said. “Those are all things that the moon and Mars have in common.”

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