“We weren’t going to announce we had a lunar data center until after we’d already landed on the Moon.” That’s what Lonestar CEO Christopher Stott said in an interview with SpaceQ this week.
Lonestar just emerged from “stealth mode” announcing that they would be building and supporting data centers on the Moon. Lonestar will be joining two Intuitive Machines NASA sponsored Commercial Lunar Payload Services (CLPS) missions to the Moon: the IM-1 mission set to land near Oceanus Procellarum on the dark side of the Moon later this year, and the IM-2 mission landing on the lunar pole at some point in 2023.
“Stealth” or no, Lonestar and its missions aren’t vapor. They’re very real. The IM-1 mission will be “a series of advanced service tests.” The IM-2 mission will feature a proof-of-concept data services payload build, which Skycorp has already been contracted to build on Lonestar’s behalf. Lonestar has also already completed data tests on the ISS in conjunction with Redwire and with Canonical, developers of Ubuntu Linux.
In fact, Stott and his co-founders have been quietly building the company for years now. They’re leveraging extensive satellite industry experience: Stott founded ManSky over 20 years ago, Mark Matossian was the CEO of ICEYE, Carol Goldstein was with ABN AMRO Bank, and Del Smith was the Senior Space Business Counsel at Dentons. It’s no surprise that they’d be aware of the growing discussion regarding space-based data centers; Seraphim Capital CEO Mark Boggett talked about it in his recent interview with SpaceQ as one reason for their investment in Bamboo and other “space infrastructure” companies.
Yet Stott said that it was the customers, however, that started the ball rolling for him and his team. Stott explained how the idea took shape at the 2018 TED Conference in Vancouver. He said that he was approached by prominent conference attendees that were becoming increasingly concerned about the environmental impact of data centers. The data centers were consuming an ever-increasing amount of terrestrial electricity and producing an ever-increasing amount of greenhouse gasses, especially with the rise of power-hungry blockchain tech. They were looking for ways to “find a place to safely save our data, our mission-critical data, outside of the biosphere.”
Stott and Lonestar were confronted by the same observation driving Boggett and others to invest in orbital data center development: that robust satellite communications can ensure that a data center doesn’t need to be on Earth to service Earth. In Stott’s words, data is “fragile…but also agile”; it can be easily damaged, but is just as easily placed in multiple places and moved around just as easily. Why not put it in orbit on a bunch of satellites?
So it makes sense for Stott and Lonestar to take data centers into orbit. But why go all the way to the Moon?
“We’re all satellite communications people…and the Moon is Earth’s biggest satellite,” Stott explained. He said that they had looked at low Earth Orbit (LEO )and Geostationary orbit (GEO) options, and realized that the Moon would be a good choice thanks to NASA’s CLPS program, and due to several physical factors. Stott said the Moon is “weather-free, outside the biosphere, 400k away…and yet still only five seconds away from Return to Operations (RTO).” It isn’t as crowded as LEO and doesn’t face the deorbiting issues, and is actually cheaper than GEO, as it doesn’t require the same level of radiation-hardening and necessary redundancy of large, expensive GEO satellites.
It’s also a way of resolving issues with data sovereignty laws and other national-level issues. Customers facing laws and policies against storing their data on servers located in other countries can securely store their data 400 kilometers away from any country, with no concerns about either terrestrial conflict or the growing problem of anti-satellite weaponry.
While five seconds is somewhat high latency for, say, up-to-the-second financial transaction data, Stott emphasized that a big part of their service is “disaster recovery as a service” (DRaaS). Short communication delays aren’t an issue for long-term lunar disaster recovery backups. Data could sit securely on the Moon for years, even decades, to be put to use when needed if and when a disaster strikes.
Both customers and investors have expressed interest: several high-profile customers are set to announce their interest in the coming months, and Lonestar is “finishing out” (in Stott’s words) a successful $5 million USD seed round led by Scout Ventures that also includes Seldor Capital and 2 Future Holding.
What surprised Stott and his team was the demand for lunar edge computing. Stott said that “we started getting phone calls from people asking ‘can we do some machine learning with you? Can we process some edge data with you?'” Stott couldn’t go into details on any of these potential customers, but said that “now we’re technically almost full.” (Stott told SpaceNews earlier that the IM-2 payload will have 16 terabytes of capacity.)
It was those same customers that pushed Lonestar out of “stealth mode.” Stott was adamant that they were not looking for the limelight, and certainly not for publicity. They’re a B2B enterprise that is busily reaching out to enterprise customers, he said, and Lonestar’s whole focus is on fulfilling their needs. They really were planning to stay “Stealth” until they hit the lunar surface, since it wasn’t affecting their ability to attract funding and potential customers. Since their customers wanted them to go public, however, they made their announcement, and Stott does grant that the announcement does “help us to engage with the larger market.”
Before they can start serving customers, though, they need to actually get those computers working on the Moon. And while the IM missions will likely be fine, truly long-term data storage will need to contend with lunar dust, the threat of meteorite strikes, potential radiation damage, and the issue of temperature regulation—always a priority with terrestrial data centers.
Stott was confident that they’d overcome those challenges. He believes that radiators will handle data center cooling as easily on the Moon as they do in orbit, but that the intense cold of the lunar surface means that the warming effect of the data centers’ components will be necessary and useful.
Radiation is a concern, but one they’re able to manage from their experience with GEO satellites, and there’s less radiation on the Moon than there is in GEO. Regolith is also a concern, but their current payloads are “self contained and not open to vacuum”, and they’re working with architectural firm BIG on the lunar data center’s external design. They also plan to avoid meteorite damage through a combination of on-site redundancy and (eventually) by moving their servers down into the lava tubes beneath the lunar surface.