by Tim Fernholz
Some questioned whether Musk, now running not just SpaceX, Tesla, and SolarCity but also two new start-ups focused on artificial intelligence and underground tunnels, was too distracted. Taking the opposite tack, others argued that he was driving his team too hard to do what could not be done. Stories of SpaceX employees burning out over time abound, but few regret their hard work or Musk’s intensity. “That’s how he thinks: ‘These guys are taking the easy way out; we need to take the hard way,’” Mueller, his propulsion engineer, said. “I’ve seen that hurt us before, I’ve seen that fail, but I’ve also seen where nobody thought it would work, [and] many times it was the right decision.”
Just as pressing as the operational and engineering questions were the concerns about money: Could SpaceX withstand another loss of $250 million or more—the missing revenues associated with six months of delayed launches after their last failure?
After the fire, SpaceX’s chief financial officer boasted that the company had $10 billion worth of future launches on its manifest, $1 billion in cash on hand, and no debt. Failures were part of the rocket business. While the loss of the SpaceX mission to the space station in 2015 punished the firm’s bottom line, it had little impact on its value. Fidelity, the enormous mutual fund company, had invested in SpaceX just six months before the accident. Its filings afterward showed that the rocket firm’s value had increased 15 percent in just eleven months, to almost $12 billion. Since 2012, Musk’s company had been in a special class—“unicorns,” venture-backed firms valued at $1 billion or more by their owners. Now the space start-up was considered twice as valuable as United Launch Alliance, which had rejected a takeover bid worth as much as $4 billion in 2015. That was less than Musk’s personal stake in his rocket firm, estimated to be worth more than $6 billion.
The most apparent reason for this disparity was that ULA was losing the competitive battle. Brett Tobey, the ULA vice president whose leaked 2016 remarks are a window into the Boeing-Lockheed venture’s thinking, summed up the launch market succinctly: “Along came Elon Musk and changed the game completely.” Describing ULA’s decision to bow out of bidding for a GPS satellite launch that year, Tobey said, “We saw it as a cost shootout between us and SpaceX, so now we’re going to have to figure out how to bid these things at a much lower cost. The government can’t just say, ‘You know, ULA’s got a great track record, they’ve done a hundred launches . . . [SpaceX’s] price points are coming down as low as $60 million. The best day you’ll see us bid is at $125 million or twice that number. Add in the capabilities cost, it eclipses $200 million.”
ULA wasn’t alone; Europe’s Arianespace faced questions about whether its next-generation vehicle was already obsolete. Orbital, which had replaced the engine in its Antares rocket and returned to flight, was having trouble marketing the rocket commercially, and would be purchased for nearly $8 billion by Northrop Grumman in 2017.
Yet SpaceX’s growing pool of backers were investing in more than just low-cost vehicles. They were investing in the transformative promise of reusability, which seemed tantalizingly close after several successful booster landings. Still, in 2016, rivals never hesitated to point out that the company had not actually flown one of its boosters for a second time, as Blue Origin had with its suborbital New Shepard. Many of SpaceX’s private investors were also counting on an entirely different line of business to justify the firm’s worth. In late 2014, Musk revealed that SpaceX didn’t just want to launch satellites that provided internet service. It wanted to build them, operate them—and profit from them. Naturally, Musk found a rival.
The catalyst for this internet satellite project was a gregarious entrepreneur named Greg Wyler who’d had a successful career as a telecom investor during the tech boom. In 2003, he had a chance meeting with an official in Rwanda’s government. Sensing opportunity, he started a new business that would build out fiber-optic internet infrastructure in the East African nation; it also became the major shareholder in the country’s largest telecom firm.
This was a feel-good project for Wyler and the Rwandans, who saw the internet as a way to establish a footing for their largely rural country in the twenty-first century. But the realities of installing expensive technology in a poor country quickly caught up with the scheme: providing internet access at a school without electricity is futile, and lowering the cost of internet access to less than $100 a month is useless if the average annual income is just a few hundred dollars. Progress slowed, and Wyler and his firm were scrutinized for promises unfulfilled. In 2006, Rwandan regulators alleged that Wyler’s company had tried to transfer its shares in the national telecom firm to a different company and fined them $400,000. Wyler stepped down as CEO in 2006, handing it over to new managers.
Wyler doesn’t comment on the specifics, except to say that Rwanda’s internet infrastructure was far stronger after he left. But he learned at least two lessons. One was that offering expanded internet to developing markets was a pitch that thrilled both potential investors and partners who were now ready to take risks on high-tech ventures abroad, especially those that would be welcomed by local authorities. The internet was not as controversial as resource-extraction concessions or low-wage manufacturing, if only because its power to shape political events was only slowly becoming clear. His second key lesson was that putting wires underground was costly and time-consuming. It was no way to bring the information age to sparsely populated areas or those, like Rwanda, that are far from the undersea cables that link the continents together. Why not just use satellites to do the job of linking the local network to the outside world?
In 2007, Wyler cofounded a company called O3b, which stood for the “other three billion”—a shorthand for the underserved chunk of the global population it targeted. It would take advantage of the fact that newer, more powerful, and longer-lasting satellites were more effective than their predecessors in providing internet links to telecom companies in countries from South Sudan to Madagascar. It found backers, including the European satellite giant SES, the sprawling global internet service provider Liberty Global, and Google. The search giant saw a chance to learn about driving internet access to remote areas and would later consider using the satellites to link up its Wi-Fi-broadcasting balloons. Wyler’s company endured a few years of tough fund-raising, borrowed $1.2 billion, and launched its first four satellites in 2014. It succeeded in attracting a diverse clientele, if not always one that aligned with its stated mission. Alongside Papua New Guinea and Pakistan, O3b found customers that one reporter cataloged through industry jokes: “the ‘other three billionaires’ for luxury yachts, the ‘other three barrels’ for offshore energy producers and the ‘other three battle groups’ for military customers.”
Whatever the source, investors saw that there was real demand for satellite internet access. SES exercised an option to buy a majority stake in O3b in 2016. By then, Wyler was already on his next step up the ladder, with new lessons learned and larger ambitions. O3b’s constellation would eventually grow to fourteen satellites in medium earth orbit, which limited its service area to a wide band around the equator. The benefit of flying that low was faster service, since the signals would need to travel tens of thousands of miles less than those from the geostationary satellites providing most terrestrial internet access. What if you doubled down on the concept to create a constellation of internet-broadcasting satellites in low earth orbit? To provide anywhere near constant coverage over a wide area would require hundreds, if not thousands, of satellites, swarming the globe so that at any given moment a handful would be within reach of a user on the ground. For context, there are just over fourteen hundred operational satellites in orbit right now, and the largest privately owned communications satellite constellation doesn’t boast more than a hundred.
This was not just any big idea; it was nearly the same big idea behind Teledesic and the other satellite constellations of the nineties, which had cost their investors billions of dollars and ended in bankruptcy. It had a
lso, indirectly, blown up the EELV program, which had counted on a burgeoning satellite business to keep its rocket makers in the black. “This is exactly the kind of pipe dream we have seen before,” satellite consultant Roger Rusch told the Wall Street Journal in 2014, predicting that the costs and delays of such plans would far outstrip early projections.
Wyler didn’t see it that way. Since the nineties, he told me, technical advances had cut huge chunks of risk out of a business plan that would still require billions of dollars to build and launch the satellites. He ticks them off: the miniaturization of chips and batteries, the improvements in solar panel technology, advances in satellite antennae, the lower cost of launches promised by companies like SpaceX, and the increased demand for and value of data transmission.
Wyler brought his idea to Google. The company had already shown an interest in satellite internet, and by 2013 it had $50 billion in cash on hand. If any firm had the hubris and the resources to launch its own satellite constellation, it would be the search kings of Mountain View, California. They hired Wyler to explore the possibilities of the scheme, and he worked with Google for months on a plan for a billion-dollar satellite scheme. He left in 2014, in an apparent dispute over the reach of the program and the software-driven company’s unwillingness to make the big up-front investments in advanced manufacturing technology needed to produce the satellites. Afterwards, Wyler made a beeline to another irrepressible salesman with a taste for big projects: Elon Musk.
The two men talked about how SpaceX, with its proven ability to innovate in space hardware manufacturing, could build its own satellite network to sell internet access around the world. Wyler had more than just his ideas to offer: through a company he created called WorldVu, he owned rights granted by the International Telecommunication Union to use a key chunk of radio spectrum. Part of the spectrum known as the Ku-band, it allowed super-high-frequency transmissions that enable the use of small antennae on the ground. If his company could put the frequency into operation by 2019, he would have exclusive rights to use it around the globe, pending permission from local telecom regulators. SpaceX could be just the partner needed to manufacture and launch satellites quickly and cheaply enough for the business case to close.
So the question that needs to be asked is: Why would SpaceX, in the middle of at least three major technology development projects—a crewed spacecraft, a reusable rocket, and a reusable heavy rocket—decide to invest in a whole new line of business? The answer, of course, is money. The earnings generated by satellite constellations shifting data around the planet dwarfed those of the rockets that launched them. The canny entrepreneur could see a path up the value chain, enabled by the savings of using his own rockets. In a sense, this would not be a departure; in the Dragon, SpaceX had already built an autonomous spacecraft that could survive in orbit and communicate with ground stations. It would simply need to make hundreds more that were simpler, smaller Dragons.
“Satellites constitute as much, or more, of the cost of space-based activity as the rockets do,” Musk said of the venture. “Very often, actually, the satellites are more expensive than the rocket. So in order for us to really revolutionize space, we have to address both satellites and rockets.”
But the partnership between Wyler and Musk was not to be; disagreements over how far to push the technology in the network drove their split. “Greg and I have a fundamental disagreement about the architecture,” Musk told a Bloomberg reporter at the time. “We want a satellite that is an order of magnitude more sophisticated than what Greg wants. I think there should be two competing systems.” They went their separate ways as competitors, each intending to make his own satellite vision a reality. Wyler formed a new company called OneWeb to develop a system for the WorldVu spectrum; in 2015, Musk opened a new office outside Seattle that would be dedicated to developing his satellite technology. The separation between Wyler and Musk was not without acrimony. Telecom insiders pointed to a filing registered at the International Telecommunication Union in June 2014 for Ku-band satellite spectrum; its particulars—including a four-thousand-satellite constellation—aligned with SpaceX’s plans.
“Part of the issue is, the original filings that Musk made were in late June last year, when he was still in discussion with Wyler about collaborating,” Tim Farrar, a satellite consultant who worked on Teledesic, told me.
Wyler’s team might have the upper hand: his rights to those frequencies give him an advantage for winning business and investment around the world. In the United States, an important market for any global telecommunications concern, telecom regulators will likely force competitors who can use the same frequencies to work out a plan to share them—but only once both competitors can demonstrate working systems. That means another race will take place, this time to get an operational constellation swarming in low earth orbit. In part to circumvent Wyler’s advantage in radio spectrum, SpaceX’s satellite team is working on linking its satellites together in an advanced laser communication network—another technical risk to introduce into an already complicated plan. A huge challenge for both endeavors will be creating the software to keep hundreds of satellites handing off signals between one another and with the ground.
Like everything SpaceX is doing, the satellite gambit has implications for reaching Mars in the decades ahead. If expanding into satellites made sense from a technical point of view, the possibility of turning the profits to fund the company’s larger ambitions in the solar system made it irresistible. “This is intended to generate a significant amount of revenue and help fund a city on Mars,” Musk said when he opened his “satellite office for satellites,” adding, with his usual dumb-it-down perceptiveness, “What’s needed to create a city on Mars? Well, one thing’s for sure: a lot of money.” Another thing that Musk will need is a space communications network that will allow his space vehicles to communicate with one another and with earth; this constellation could be the basis for that.
Musk and Wyler estimate that their systems will each cost more than $10 billion, an enormous sum to gamble on a high-tech venture, especially considering that if both succeed, they may very well split the market, to both companies’ detriment. The satellite community was already worried about the growing problem of space traffic management and the endless debris generated by decades of human activity in space. The US Air Force spends hundreds of millions of dollars to monitor orbital junk, sending a heads-up to satellite operators or the ISS whenever they forecast a collision. The kind of chain-reaction catastrophe depicted in the movie Gravity, where debris tears through the ISS, remains a real risk that space agencies around the world pay attention to. Experts say it is already apparent that new technologies to reutilize or remove space debris, and new practices to keep orbital lanes clear, will be needed. A surge in orbital satellites representing many times the current orbital population will stress an already stressed system of space traffic control.
Whatever the risks, both entrepreneurs’ track records allowed them to raise the funds for their projects. Investors were once again assigning fantastic values to technology ventures. In 2015, SpaceX won its largest single investment ever: $1 billion, in a round led by Google, which purchased 10 percent of the firm. While the money wasn’t earmarked for the satellite scheme, financial documents leaked to the Wall Street Journal at the time of the investment showed that SpaceX was expecting to generate $15 to $20 billion in revenue from the constellation by 2025. Those same documents showed how tight the company’s financial situation had been in the years before as its products began to come online; it would take years for SpaceX to recoup its investments in technology development.
Google’s decision to back Musk’s space dreams didn’t prevent Wyler from assembling his own cadre of industry backers. These included Qualcomm, which manufactures microchips used in satellites, and Airbus, the European aerospace champion, which wanted to develop mass-production techniques for satellites. (Like rockets, most satellites are assembled by hand in clean room
s, which is one reason for their expense.) Intelsat, another major operator in communications satellites, invested as well, perhaps hoping to ape SES’s successful incorporation of O3b into its collection of space assets. In 2016, OneWeb would win its own $1 billion investment from the Japanese conglomerate SoftBank, whose tycoon CEO, Masayoshi Son, was spearheading a $100 billion technology investment fund.
And, because every space venture needs a space billionaire, Richard Branson threw his hat in the ring, investing in OneWeb through Virgin Group and joining its board. The deal also entailed a contract to launch ten of the satellites on Virgin Orbit, a subsidiary of Virgin Galactic that would be spun out in 2017. Virgin Orbit intended to build the architecture to launch small satellites on rockets dropped at high altitude from a 747. A similar system, called Pegasus, had been operated by Orbital since 1990, but it had proven too expensive to win much business. The idea now was for Virgin to take this architecture and run it through the cost-cutting wringer SpaceX had used to disrupt ULA, in part with some of the same engineers who had helped SpaceX do that. Virgin Orbit’s rocket would aim to reinvigorate the small-satellite market that had attracted the Falcon 1, rather than face off against SpaceX or United Launch Alliance directly. They weren’t the only company aiming at this market: In 2016, Cantrell and Garvey founded their own firm, Vector Space Systems, to launch a new generation of privately funded small satellites, part of a resurgence in small rocket start-ups.
“I don’t think Elon can do a competing thing,” Branson said of the two internet satellite schemes. “Greg has the [spectrum] rights, and there isn’t space for another network—like there physically is not enough space. If Elon wants to get into this area, the logical thing for him would be to tie up with us.”
When I asked Wyler about such a linkup, the pragmatic entrepreneur offered a verbal shrug. “My crystal ball is broken. Nothing is off the table; the mission is to connect people.”