In vast swaths of the United States and the world, there are millions of people who don’t have reliable internet access. These unconnected people aren’t just in far-flung places like rural America or New Zealand or sub-Saharan Africa, either. There are plenty of people living in dense city centers who struggle to access affordable broadband. The pandemic has brought new urgency to the problem, and while companies like Google and Facebook have floated far-out ideas for solving this problem, the internet technology that’s most promising is also the one that’s already proven: satellite broadband.

In early March, just days before cities across the United States shut down due to the Covid-19 pandemic, Elon Musk shared the latest details about his plan to build a satellite broadband service called Starlink. Speaking to an audience at a satellite conference in Washington, DC, Musk described how a constellation of Starlink satellites will “blink” when they enter low-Earth orbit. As described, they almost sound like streaks of glitter in the night sky, or magic bands of flying gadgets that can beam internet down to anyone on the planet.

Combined with improvements to existing technology like DSL, cable, and fiber — not to mention 4G and 5G cellular networks — futuristic satellite broadband stands to bridge the digital divide in the US and elsewhere. And because the pandemic has prompted explosive demand for better, more widely available internet connectivity, fast progress seems more inevitable than ever.

Musk’s new satellites went online in early September, giving beta testers download speeds that rival those of terrestrial broadband. SpaceX has now put 700 Starlink satellites into orbit in the past 16 months and has plans to deliver as many as 30,000 more in the next few years. More satellites mean more bandwidth and faster speeds, and eventually, SpaceX says its low-Earth orbit satellite constellations could deliver high-speed internet to the entire US. Amazon, Facebook, and several startups have made similar promises in recent years.

The concept of satellite-based internet service is actually decades old. However, the innovative low-Earth orbit satellite technology being developed by SpaceX and others could be essential, if not transformative, for everything from telemedicine to remote learning in places that aren’t already connected. Satellite broadband could also be very profitable for whichever company figures it out first. One could imagine Amazon using satellite broadband to boost its Amazon Web Services (AWS) business, or Facebook using it to ensure that more people get online and look at Facebook. And if Musk gets his way, his Starlink constellations will generate billions of dollars in profits to fund his mission to colonize Mars.

This all sounds futuristic, but satellite broadband is already a very real thing. In fact, if you’ve ever connected to the wifi on a plane or cruise ship, you’ve probably used it. The basic idea is that ground stations connected to the internet, known as gateways, can send data up to a satellite which then relays that data to antennas somewhere else on the ground — or on a ship or an airplane.

The problem with this technological feat is that it’s all very expensive. It can cost hundreds of millions of dollars to launch satellites into space, and that’s not even taking into account what it takes to get over regulatory hurdles. Plenty of companies have tried and failed to crack the business model in the past 20 years, but rather suddenly, the space internet game has changed.

“The Covid-19 crisis has significantly accelerated attention to and investment in satellite technology,” Babak Beheshti, dean of the College of Engineering and Computing Sciences at the New York Institute of Technology, told Recode. Beheshti added the number of launches had gone up tenfold from last year to this year. “Why? Because schools, local governments, and others suddenly needed to have broadband internet access in areas where there was really no infrastructure in place.”

This might sound like proof that satellite broadband is finally on its way to solving the digital divide, but the situation remains incredibly tenuous. As SpaceX started firing up its Starlink satellites, Amazon in July received approval from the Federal Communications Commission to launch 3,236 low-Earth orbit satellites for a constellation of its own called Project Kuiper. Meanwhile, longtime satellite broadband industry leaders like Viasat can’t seem to get new satellites into the sky fast enough to keep up with demand. And along the way, the federal government is pledging billions of dollars in subsidies to companies that bring broadband to rural America.

In some ways, the dream of connecting everyone on Earth has never been closer. In other ways, it’s hard to tell if the latest innovative ideas will suffer the same pitfalls as those of years past.

Satellite broadband, briefly explained

Satellite broadband is exactly what it sounds like: broadband internet access delivered via satellite. The basic idea hasn’t changed much since the heyday of satellite TV in the late ’90s, when companies would beam internet connectivity to the same dish that received your HBO signal at speeds that were faster than dial-up but still slower than today’s broadband.

In 2020, there are two main ways companies deliver satellite broadband. The key difference between them is how high the satellites orbit. Geosynchronous satellites, which orbit about 22,000 miles above a fixed place on Earth’s surface, is an older technology that companies like Viasat use for broadband connections. You’ve probably used this tech for airplane wifi. Then there are low-Earth orbit constellations, which are made up of hundreds, if not thousands, of smaller satellites that orbit between 300 and 1,200 miles above the earth. This is the approach that’s getting all of the buzz lately, and the one that SpaceX and Amazon are taking.

Geosynchronous satellites are the more mature, more proven technology. Viasat and a company called Hughes, which is the former parent company of DirecTV, have been around for decades. (DirecTV actually used its dishes and infrastructure to offer a satellite internet service called DirecPC back in the late ’90s.) Viasat and Hughes are also the two companies that most likely offer satellite broadband in remote parts of the US right now. If you’re someone who lives in the New Hampshire wilderness, where there are no terrestrial broadband options, you can get a version of DSL, which operates on existing copper telephone lines, that’s essentially as sluggish as dial-up. Or, you can sign up for geosynchronous satellite broadband through Viasat or Hughes and get speeds comparable to basic broadband: about 25 megabits-per-second. Plans start at $40 to $50 a month and get more expensive if you want more bandwidth.

Though they are dependable, these geosynchronous satellite systems have some issues. The main one is latency. The satellites are thousands of miles above Earth’s surface, so it takes time for data to travel — and that might mean a slight delay between sending and receiving. This isn’t a problem if you’re just browsing the web. It’s a significant problem if you’re trying to stream video games or do video calls, something we’re all doing more than ever before. Just think about remote TV news correspondents who have to wait half a beat between when the anchor in the studio asks the question and when they hear it in their earpiece, as the signal travels up to a communications satellite and then back down to the surface.

Low-Earth orbit constellations, like the ones SpaceX and Amazon are building, promise to solve the latency problem. Because the satellites are closer to the ground, the data doesn’t have to travel as far. Musk says this means SpaceX’s Starlink satellites, which will orbit at around 340 miles above the surface, will offer high latency, thus reducing the risk of lag. The latency question is a big deal to the FCC and its decision to hand out billions of dollars in subsidies, by the way. The agency says it will prioritize networks that offer low latency when giving out funding.

Still, there are other unanswered questions about just how fast and dependable newly designed low-Earth orbit constellations will be. Unlike geosynchronous satellites, which are fixed above one spot, low-Earth orbit satellites circle the planet every 90 to 120 minutes. They’re designed to stay connected to the ground station and to the end user by staying connected to each other, but if this chain gets broken, it would disrupt the connection. These constellations are also made up of thousands of relatively small satellites — Starlink satellites weigh less than 600 pounds — which means they require multiple launches, which are expensive.

“As more satellites go up, they optimize the network architecture,” explained Manny Shar, head of analytics at Bryce Space and Technology. “In the next couple of years, we should see decent improvements in rural areas where there’s really limited capability, and there’s limited competition to improve that. So at the very least, there will be an alternative option that those rural users can take advantage of.”

Shar’s point about limited competition is an important one. Many parts of the United States, for instance, have access to slower DSL connections thanks to telephone lines, but because upgrading that infrastructure is so expensive, the telecom companies that serve those areas often have little incentive to do so. That leaves residents depending on a mix of poor wired connections and often spotty cellular networks.

New technology like 5G could ostensibly bring faster cellular speeds to remote areas, but again, building that infrastructure takes time and money. Satellite broadband, meanwhile, can beam fast, reliable, and potentially affordable internet access down to nearly anywhere on Earth. This also requires time and money, but what we’re seeing in 2020 is that the pandemic is attracting all kinds of investment in the technology, which means more satellites are launching.

Both geosynchronous and low-Earth orbit satellite broadband systems have pros and cons. The former is already viable, albeit not perfect. The latter holds promise, albeit unfulfilled. Inevitably, though, to get to that goal of connecting more people, it will all come down to money.

The slow march of progress

The future of satellite-based broadband largely depends on who can get the most bandwidth into space for the least amount of money. Each individual satellite, by design, can offer a limited amount of bandwidth, so companies are either making lots of satellites to launch at once — this is what SpaceX is doing — or they’re investing technological improvements and launching new satellites every few years. This is Viasat’s strategy, and the company plans to launch a new satellite called Viasat 3 next year that’s expected to vastly improve its network. This satellite and others like it weigh tens of thousands of pounds, so these launches are expensive.

So one could see the appeal of launching lots of smaller satellites over time, especially if you’re a company like SpaceX and own your own rockets. Amazon and its Project Kuiper, similarly, have the benefit of being owned by Jeff Bezos, who also owns the rocket ship maker Blue Origin. It’s so far unclear how Blue Origin might factor into Project Kuiper, however. In fact, Amazon has revealed very little about the project other than it plans to offer affordable high speed, low-latency internet service through low-Earth orbit satellites.

“There are still too many places where broadband access is unreliable or where it doesn’t exist at all,” Amazon senior vice president Dave Limp said in a statement following the FCC’s approval of the first Project Kuiper launch. “Our $10 billion investment will create jobs and infrastructure around the United States that will help us close this gap.”

It’s worth pointing out a difficult truth here. Selling affordable satellite broadband to individual customers in rural areas will not generate enough revenue to send the needed satellites to space. Again, each launch costs hundreds of millions of dollars, and selling service for $40 a month to individual households can’t cover the startup costs. And even then, not everyone who needs internet access can afford that. This economic challenge is part of the reason why the dream of offering satellite-based internet service to anybody on Earth — or any other kind of reliable, high-speed internet service — has been so elusive.

This is also why companies that have been successful at building satellite broadband networks have approached the challenge from different angles. Viasat, for instance, spent years building out an enterprise business, selling bandwidth to the military and governments, not to mention helping you get wifi on airplanes. Now, the company says demand from the consumer market has been on the rise and has simply exploded since the pandemic hit. And that demand isn’t necessarily coming from the most remote areas.

“It turns out that a lot of the demand tends to be around the major metro areas,” said Viasat CEO Mark Dankberg. “In the highest demand markets — in the Midwest, in the Southeast — we’ve been out of bandwidth for two years. So we can’t have that many more customers until we get our next satellite.” Dankberg added Viasat is developing technology that would involve connecting its existing geosynchronous satellites with its own low-Earth orbit satellites, as well as cellular networks, for faster, lower latency connections.

As Recode’s Emily Stewart recently explained, broadband access isn’t just a problem in rural Montana. Even in city centers and suburbs, the infrastructure to offer high-speed internet access either doesn’t exist or is too expensive for many people to afford. This means that new options, including space internet, could stand to connect millions of Americans more quickly than it would take to expand existing terrestrial infrastructure.

That doesn’t make providing access to those in far-flung regions any less of a priority, and government subsidy programs are helping to make this happen, albeit slowly. Coincidentally, just as the pandemic pushed the country into lockdown, the FCC launched its Rural Digital Opportunity Fund, which will provide up to $16 billion to telecom companies that expand internet access in rural areas. SpaceX has applied for funding, although it must prove its service offers the low latency and high speeds required by the agency to get the money. Viasat received $87.1 million in funding from a similar FCC program last year.

Again, in the absence of government funding, companies like SpaceX and Amazon are in a unique position to take the lead in the satellite broadband industry because building such an infrastructure will come in handy for other reasons. SpaceX is in a unique position to deliver its satellites into low-Earth orbit. The benefit of Amazon owning its own satellite broadband network also seems apparent. When it goes online, Project Kuiper could be an immediate boon to the company’s AWS business.

“Amazon is essentially, effectively going to be its own biggest customer to really prime the pump for the revenue stream,” said Beheshti, who is also a senior member of the Institute of Electrical and Electronics Engineers. “And then, obviously, the additional revenue streams would come from the residential individual consumers.”

The benefits of satellite-based internet services have been obvious for years. However, for years, companies have struggled to make those ambitions meet reality. It’s not for lack of trying — and trying creative approaches, too. Alphabet continues to pursue a project called Loon, which started out as a Google experiment about 10 years ago. Loon involves using high-altitude balloons that beam internet access down to rural areas. After being deployed in Puerto Rico after Hurricane Maria, a fleet of Loon balloons started delivering service to millions of people in Kenya in July, marking the first commercial application of the technology.

Meanwhile, Facebook has had its own far-fetched plans. Its initiative called Internet.org that aims to connect the entire planet suffered a big setback in 2016, when a SpaceX rocket carrying a satellite designed to deliver internet access to sub-Saharan Africa exploded on the launch pad. There was also Project Aquila, which involved sending solar-powered drones 60,000 feet into the atmosphere to connect rural areas. The company abandoned the project in 2018.

Big internet companies like Facebook and Google have also faced backlash for their lofty connectivity projects. While projects like Loon and Internet.org are billed as charitable initiatives to serve the public good, critics say they stand to violate the principles of net neutrality and serve the companies’ best interest, rather than the public’s. After all, a free or low-cost internet service from Facebook or Google could simply steer billions of people to Facebook and Google’s products and services, Balkanizing the internet as we know it.

With all of these efforts, there are bound to be more failures, and possibly more backlash, in the future. Elon Musk’s goal of offering high-speed broadband to everyone on Earth is a lofty one. We do know that such a thing is technically possible. It’s expensive, and plenty of smart people are figuring out how to pay for it, while other promising tech, like 5G, continues to roll out. But if anything would motivate such a tremendous disruption in the internet service business, the pandemic should do it. Never before have we depended so much on connectivity. We might just have to leave planet Earth to get it.

Help keep Vox free for all

Millions turn to Vox each month to understand what’s happening in the news, from the coronavirus crisis to a racial reckoning to what is, quite possibly, the most consequential presidential election of our lifetimes. Our mission has never been more vital than it is in this moment: to empower you through understanding. But our distinctive brand of explanatory journalism takes resources. Even when the economy and the news advertising market recovers, your support will be a critical part of sustaining our resource-intensive work. If you have already contributed, thank you. If you haven’t, please consider helping everyone make sense of an increasingly chaotic world: Contribute today from as little as $3.