Swarm Technologies CEO Sara Spangelo was flying high in October 2017—piloting a small plane carrying John Krafcik, the CEO of Waymo, Google’s self-driving car company, and another entrepreneur on a pleasure flight. They passed over Tesla’s factory in Fremont, California. Spangelo joked later on Facebook that Krafcik was “checking out the competition.”
Like Tesla founder Elon Musk, Spangelo’s real interest is in space business, but unlike Musk, her company’s historic first in orbit threatens its very existence. In January 2018, Swarm launched the first satellites into space unauthorized by any government. People familiar with the business of launching satellites into space consider the situation odd, troubling and even dangerous: Access to space is supposed to be expensive, difficult and tightly guarded by nations under international treaty obligations.
How, exactly, did those tiny satellites go from Silicon Valley into a rocket at the Satish Dhawan Space Centre on the west coast of India, and then to orbit, without anyone asking for its paperwork?
No one at Swarm is talking, and some have hired lawyers. Whether a comedy of errors or a case of corporate line-crossing, the good news is that the experimental satellites were designed with benign intent.
What, some experts warn, if next time it isn’t?
Silicon Valley’s space fever
In 2013, Spangelo completed a University of Michigan Ph.D thesis on the potential for satellite networks to carry large amounts of data, then moved to Caltech’s Jet Propulsion Lab (JPL), where top space researchers design and execute space missions for NASA. Like many young space scientists, Spangelo next went north to Silicon Valley, taking a job at Google X, the company’s “moonshot factory” that prototypes advanced technology for potential businesses; Waymo was hatched there.
Spangelo worked on a scheme to use drones to deliver goods and on a team that chose which technologies to invest in. In 2017 she left to start Swarm Technologies with Benjamin Longmier, now the company’s CTO. Also a licensed pilot, Longmier is an assistant professor at University of Michigan’s aerospace department since 2015 and has long shuttled between Silicon Valley and academia. (Longmier and Spangelo did not respond to repeated requests for comment on this story.)
Through Swarm, the pair intended to deliver connectivity from space—”connecting the world with tiny satellites,” per Spangelo’s Linkedin Page.
A future flock of drones or a fleet of Waymo’s self-driving cars will need to be linked into a communication network to perform their work. For companies developing these and other “internet of things” applications, that could mean paying a fee to access existing wireless networks built by telecom firms like Verizon or AT&T. But others are betting it would be far better business for connected devices to have their own proprietary communications network in space, provided by companies like Swarm; established satellite companies like Inmarsat and Iridium already provide this service on a limited scale.
Venture capitalists have been throwing millions at new satellite companies that promise to perform more cheaply than existing providers. Swarm itself apparently had partnership interest from two Fortune 100 companies, won a $200,000 grant from the National Science Foundation, and inked paid research agreements with NASA and the US Air Force.
Size matters
“Cubesats,” satellites based on a form factor 10 cm3, have become a powerful vehicle for space technology in recent years, allowing companies to cheaply put powerful computers in space for earth observation and communication. Many start-ups build bigger satellites as they gain access to more capital, but Swarm’s plan was to go small. According to its Federal Communications Commission filings, it would use satellites one quarter the size of a regular cubesat:
Swarm requested an experimental license from the FCC in April 2017, less than a year ago. The first step for a US organization seeking to operate satellites in space is to gain permission from the FCC, which is charged with fulfilling US obligations to the Outer Space Treaty that new objects in orbit not collide with existing satellites or generate dangerous debris to threaten future spacecraft. With space debris capable of destroying billion dollar satellites or even costing the lives of astronauts on ISS, this is no small matter. The regulatory process proceeded along several months, as Swarm developed its technology.
Swarm hoped to launch in September 2017. They hired Spaceflight industries, a broker company that links satellite payloads with rockets, to handle the work of getting a spot on a rocket and “integrating” their satellite—essentially, installing it in the rocket so that it can be ejected at just the right moment in flight. Spaceflight, in turn, contracted with Antrix, the commercial arm of the Indian Space Research Organization (ISRO) to launch the satellites on their Polar Satellite Launch Vehicle, or PSLV. The PSLV had made a business in small satellites, launching more than 104 in a single flight in 2017.
A failed PSLV launch in August 2017 delayed the mission expected to carry Swarm into 2018. And a debate between the FCC and Swarm over the size of the satellites came to a head. The spacecraft, the FCC said, was “below the size threshold at which detection by the Space Surveillance Network (SSN) can be considered routine.”
In other words, the FCC feared that the tiny broadcasters would escape notice by routine radar surveillance, making them a potential danger in orbit. Swarm suggested adding special reflectors or using the satellites own GPS sensors, but in December 2017 the FCC gave the final word: You can’t launch these satellites.
A gap in the story
In January 2018, Swarm Technologies applied for a new license to launch satellites fitting the 10cm3 form-factor the FCC considered safe. Four days later, the PSLV rocket took off in India, launching the four smaller satellites the FCC said were too small into space.
What happened? The short answer is, we don’t know. Swarm, Spaceflight, and Antrix—and the government of India itself—all bear some responsibility, but attention has focused on Swarm itself. The company had hired a consulting engineer named Craig Scheffler to handle its FCC applications; he does the same work full-time for Planet, the leading small-satellite start-up. Scheffler wouldn’t speak to me for this story, but his attorney Jeff Carlisle told me he was simply a technical consultant and made clear to Spangelo that Swarm had been denied a license. He no longer works with Swarm.
Spaceflight’s job was bringing the satellite to Antrix. Last week, Spaceflight’s president Curt Blake told SpaceNews that “I always assumed that people wouldn’t launch something if they couldn’t”get a license, saying he operated under “a sort of a self-regulation function.” If so, that would be a remarkable admission.
“We work on [satellite licensing] and have to deal with three or four different parties, the launch operator, the launch provider, the satellite operator, the ground segment operator,” one satellite industry consultant told me of a typical launch. “At no point in time was there any way that you’d be able to sneak something like that on. Everything was quadruple checked by all parties. The people that do these things, they are extremely serious about paperwork and authorization, they do not take it lightly.”
Spaceflight refused to discuss its procedures for certifying that its payloads are compliant with regulations, but said it was reviewing and updating them. A person familiar with the controversial launch said that the company “was unaware the FCC raised any concerns with Swarm.” According to this person, under normal procedures Swarm’s spacecraft would have arrived in India about a month before launch, the same time the company received its rejection from the FCC, and would be plugged into the rocket several weeks ahead of lift-off. That raises the question of whether Swarm’s satellites were already being processed when their license was denied.
It’s not clear what reviews, if any, Antrix or ISRO performed on the launch payloads. (Antrix did not respond to multiple requests for comment.) They did, however, promote the satellites in their brochure for the day (pdf), noting that four SpaceBEE satellites, attributed only to “USA,” had flown on the mission, with a pixelated image of the spacecraft.
Who’s flying those satellites?
The informal community of amateur space trackers noticed first. Gunter Krebs, a German software engineer who has maintained a website tracking satellites for more than twenty years, tried to figure out who owned the SpaceBEES and recognized their depiction from regulatory filings. His e-mails to Swarm passed without reply, according to Jonathan McDowell, a researcher at the Havard-Smithsonian Center for Astrophysics, who also spotted the SpaceBEES and guessed they belonged to Swarm.
In February, the FCC granted a license to Swarm to fly their larger satellite design. But by March 8 the jig was up: The FCC revoked the license, telling Swarm they would assess its “apparent unauthorized launch and operation of four satellites, and related statements and representations, on its qualifications to be a Commission licensee.” Since then, the FCC has been mum about what’s going on, with a spokesperson only saying it is looking into the matter.
“I’ve never heard of that happening,” the satellite industry consultant says of an unauthorized launch. “If they are communicating with US based ground station on an unregistered satellite segment, they will get into a lot of trouble. The question is, without knowing the location of the Earth stations in the US or the frequencies being used, how do you prove it?”
The FCC knows, at least, the frequencies that the satellite was designed to broadcast on. The ground stations are registered at private homes in Silicon Valley owned by Spangelo and Longmier, which function as the main places of business for the still-tiny start-up, but they are small enough to be moved. Without literally intercepting transmissions, or witness testimony that they happened, the FCC can’t prove that the satellites were operated illegally, even if they were launched without permission.
Given how unusual the case is, some experts speculated the satellites might be authorized by some third country that had not announced itself through the slow channels of global space bureaucracy. Neither Swarm nor Antrix have said this. Indeed, Spangelo told the FCC that she would register her constellation with the global satellite regulator International Telecommunications Union in 2017, but an ITU spokesperson confirmed that Swarm Technologies did not file advanced notice for an experimental satellite constellation.
Emergency stop buzzer
As investors and companies jump into the space business, the pressure to be first in demonstrating technology is rising. “The industry is getting crowded and people are getting excited, because access is vastly reduced cost-wise,” the satellite consultant says. The SpaceBEEs in orbit are a sign of path-breaking technology. But we don’t know if they got to orbit because a Swarm felt it was better off asking for forgiveness than permission from regulators, or if a mistake was made in its communications with Antrix or Spaceflight.
This is a novel problem: Only a few years ago, a satellite small enough to slip past radar surveillance would not have been big enough to be useful. And the tools to detect them are coming from the private sector, too. LeoLabs, a start-up that wants to provide dedicated space debris monitoring in low earth orbit, says that it has no problem tracking the SpaceBEES. “They’re not too small to be seen,” Alan deClerk, an executive at LeoLabs, told Quartz. “It’s business as usual for us. We captured that and a lot of other things at the same time.”
Security experts have already been sounding the alarm about how easier access to space could have bad effects, too, warning that it could quickly create a space traffic management issue. This time around, the worst case scenario is an accidental collision, but future sneaky satellites could be operated by people with more pernicious aims. Concerns over lax supervision could mean trouble for India’s burgeoning commercial launch industry as well.
It remains to be seen what will happen to Swarm, which seems to have gained much of its initial start-up funding from the government. On March 15, Swarm was awarded an additional $740,000 by the National Science Foundation to continue its scientific experiments; the NSF has not responded to questions about what the FCC investigation means for the new grant.
When the FCC authorizes satellites, it asks for a contact for an “emergency stop buzzer”—someone who can shut down transmissions in the event of an unanticipated interference. For Swarm, that person is Alan Adamson, a former employee of BellSouth and ham radio enthusiast, who hosts a ground station at his home in a suburb of Atlanta, Georgia. Reached by phone on Friday, he said he couldn’t talk about what Swarm was up to.
“Direct all the questions to Sara,” he said.