Max Bhatti and the four other engineers at Basalt Space worked 22 hours a day in March to assemble the startup’s first satellite so it would be finished in time for a launch deadline. “It makes 996 look like a vacation,” says Bhatti, the CEO. To keep electronics free of contamination, the team operated in a well-ventilated tent that Bhatti boasts is more dust-free than a hospital. It sits in one of three adjacent apartments the company leases in San Francisco’s Lower Nob Hill neighborhood.

The apartments have been home and office for the Basalt team for the past two years, replete with all the staples of a hacker house, including a laundry machine, an outdoor gym, and stacks of ramen. Employees, who are all in their 20s, feel a sense of urgency as the third and largest-yet wave of satellite development unfolds across the US.

Basalt is part of a generation of startups aiming to broaden reliable and secure access to satellite imaging, navigation, and communication services. As they envision it, more of the world will be continually photographed, more items will be tracked, and customers won’t have to fear gatekeepers like Starlink cutting off their transmissions.

From the first satellite launch in 1957 until the past couple of decades, governments and defense contractors largely controlled access to data from space. Alternatives followed, including Globalstar, Planet Labs, and Skybox Imaging, which launched a few low-cost satellites and conveyed specific data to paying customers. But Basalt wants to go further, providing any client with their own set of five to 15 satellites in a similar fashion to how cloud computing firms give companies access to data centers full of sophisticated servers. Faster satellite data could help farmers stop pests and diseases before they spread widely. Fewer restrictions and increased reliability could enable news organizations and investors to better understand migration and trade.

“The question that I posed to myself when I started the company was like, ‘What's the most fundamental thing that we could change about the aerospace industry?’” Bhatti says. “And I think it's the idea that the end user should just be able to directly task a constellation, not even just one satellite.”

Operating the satellites using AI in place of people is an essential but unproven part of Basalt’s business plans. But the startup already has been aided by the rapid decrease in costs of manufacturing and launching satellites over the past five years. The Trump administration’s recent decision to relax some regulatory hurdles has also helped, according to Bhatti. “A lot of the hoops that you would jump through are gone, and that's welcomed by everyone in the industry,” he says, declining to get into specifics.

The war in Iran has also provided a golden opportunity to pitch the technology. Planet Labs and other satellite imagery providers recently restricted access to feeds from the Middle East, citing concerns about misuse amid the conflict. For the thousands of dollars a month customers pay to a satellite imaging provider, they could lease or own their own constellation through Basalt, Bhatti believes. “No one can cut you in line. No one can turn off the data,” he says, though it’s unclear if the latter promise will hold up to future regulatory pressure. “Especially in times like this, it's kind of obvious how important it is to actually see what's going on in the ground. What is really the truth?”

But Basalt and other new companies face questions over whether customers will materialize at the volume they expect. Satellite startups that a few years ago bet on eagerness for climate change data turned out to be wrong, with their services now largely serving militaries. Backlash against space junk, light pollution, and other environmental impacts of satellites could grow as the industry booms. The increased opportunity for intrusive surveillance from space also could draw civil liberties activists into a fight.

Roughly 15,000 satellites are orbiting Earth today. Various forecasts estimate more than four or five times that number by 2030 as both militaries and companies accelerate their space endeavors.

Reaching Liftoff

Attempts to get small satellites into the air surged from 2009 to 2016. But roughly half of them failed, according to data compiled by NASA. Failures with the rockets carrying them up accounted for some of the issues. But most of the problems came from satellites performing inadequately, the NASA analysis found. In the years since, SpaceX has developed a dependable and affordable launch vehicle. The billions of dollars it is seeking to raise through a public stock market debut this year are poised to further advance its capabilities.

Satellite makers also became more skilled at carefully selecting cheaper components that could withstand the unique elements of space, including high radiation. They found that parts used in servers and other common gear on Earth could be sufficiently hardened with shields, chemical sprays, and redundant systems. These components are also smaller than traditional gear, allowing for lighter satellites, which in turn, means companies can launch more of them for the same price.

These factors prompted Bhatti to drop out of UCLA in 2023 and pursue Basalt. The following year, he moved to San Francisco to be part of the Y Combinator startup accelerator program. Spending on satellite manufacturing and data services reached $130 billion globally that year, according to the industry’s main trade body.

Now, Basalt is among eight companies WIRED identified making satellites in the San Francisco Bay Area, one of the country’s top centers for the space industry. Last month, Xona Space Systems unveiled a 25,000 square foot production floor at its bayside office. CEO Brian Manning says Xona wants to eventually replace GPS for some customers. While the legacy technology is free and run by the US government, it is also vulnerable to spoofing and jamming attacks. Xona’s Pulsar system is meant to fix all of that—for a fee. “All these things need to know where they are and we can give them the information of where they are with better accuracy and in more places than GPS can today,” Manning says.

It’s just a vision for now. Xona has one satellite in orbit, and the first satellite completed at its new facility is headed up soon. Over a dozen companies that make some of the 2 billion new GPS receivers put into the world each year are testing with Xona, along with US government agencies, according to Manning. The goal is for the company’s Pulsar system to work with existing, everyday devices. Xona has announced deals already with governments and companies supporting farms and power plants with precision tracking and timing.

A nearby competitor, Earthtraq, plans to use radio signals to provide satellite-based positioning, navigation, and timing. Customers would affix a tracker the size of a small sticker to items, says Isobel Porteous, the startup’s head of business development. The compact chip will make it possible to follow items and animals that couldn’t be tracked before because of the cost, complexity, or weight of GPS receivers.

Other startups are focused on data collection and communications. Like Basalt, Muon Space is building and running constellations of remote sensing satellites for specific customer projects, such as spotting GPS interference and wildfires. A satellite with the necessary infrared sensor and high-speed internet for monitoring fires would have cost $1 billion not too long ago, says Muon CEO Jonny Dyer. Muon developed its initial version for millions of dollars, with Starlink eventually handling the internet. The company has completed 15 satellites. More are being manufactured at a new 130,000-square-foot facility in San Jose.

Meanwhile, Astranis has announced more than $1 billion in customer commitments as it works on satellites to provide the Philippines, Oman, Taiwan, and the US with their own satellite internet networks as an alternative to Starlink and undersea cables. “Countries need private, secure networks to know their data is secure and in the right hands,” Astranis CEO John Gedmark recently wrote on LinkedIn. The startup operates a 150,000 square foot San Francisco factory.

Each of these companies still has much to prove. Basalt is just beginning to solicit potential government and commercial customers. The company also has to show that its bet on using AI to automate flying satellites works. At other companies, humans handle the sometimes around-the-clock task of ensuring satellites aren’t inundated with instructions or data, knocked off course, or overheating. If all works out, Bhatti expects to move to another aerospace hub such as Los Angeles and produce 100 to 200 satellites annually there by 2029.

For now, Basalt’s San Francisco apartments are factory and mission control. Bhatti says the Federal Communications Commission allowed Basalt to put up a satellite ground station on the building’s roof to communicate with its prototype. Basalt managed to finish that first satellite and deliver it to SpaceX on April 1—Bhatti’s 23rd birthday. It launched on a Falcon 9 rocket from Vandenberg Space Force Base in Southern California on Saturday.