Nathan Cyprys/The Globe and Mail
Every spring and fall, Dean Wells has the job of telling Nunavut government employees the precise date and time they’ll be hurled back into the telecommunications dark ages. Internet connectivity for the territory’s government, and most of its economy, for that matter, is beamed via a single geostationary satellite locked in place 36,000 kilometres above the Earth, and over the course of several days twice a year, the angle of the sun overwhelms the satellite’s signal with thermal energy, shutting down communications for up to 12 minutes at a time. “It might not sound like much, but if you’re trying to make a phone call or send a file, or even if you’re standing in line at the grocery store to pay for your stuff, well, you just can’t,” says Wells, the government’s chief information officer.
It gets worse. While that satellite, Telstar 19 Vantage, launched by Ottawa-based Telesat in 2018, brought moderately faster internet speeds than an earlier one did, it suffers from high latency, or lag time, and its limited capacity means the government’s internet needs are six times greater than the satellite can provide. The solution, for this corner of the world’s ninth-largest economy, is internet rationing. High-priority files get transmitted during the day, but many other documents must be cached and sent overnight, when satellite usage is lighter. And never mind that getting a video call to work properly in Nunavut is a minor miracle. (It didn’t when I spoke with Wells in April.)
Yet, for a man who spends his life triaging fragile networks, Wells is bursting with excitement. A revolution is unfolding in the skies above him, with companies competing to launch tens of thousands of new satellites that will orbit far closer to the Earth’s surface and be capable of providing blazingly fast broadband internet to the most remote edges of the planet. “Nunavut,” says Wells, “will never be the same.”
Nunavut's chief information officer, Dean Wells
Universal high-speed connectivity is the promise at the core of the new multibillion-dollar space race that’s pitting two of the richest, most powerful rocket jockeys on the planet, Elon Musk (net worth: US$250 billion) and Jeff Bezos (net worth: US$148 billion), against each other.
Oh, and Dan Goldberg (net worth: considerably less).
Since 2006, Goldberg has been the CEO of Telesat, the world’s fourth-largest operator of fixed-position satellites. A few years ago, faced with the reality that its existing fleet of lumbering, high-orbit geostationary satellites weren’t up to the high-speed internet task, Goldberg announced plans to leap into the world of low-Earth-orbit (LEO) communications satellites, which whiz around the planet multiple times a day but at an altitude of 2,000 kilometres or less, allowing them to offer speedy and reliable internet on par with fibre. And so Lightspeed was born, a $6.5-billion constellation of 298 initial satellites aimed at serving enterprise customers like governments, telecoms, and companies in the marine and airline industries.
If ever there was a moment for the impossible dream of closing the digital divide that’s left one-third of the world’s population without online access—not to mention the vast swaths of rural Canada plagued by turtle-like download speeds—this is it. Amid the work-from-home revolution, the Internet of Things and the growing metaverse, demand for fast, remote internet that can’t feasibly be served by land-based fibre networks is exploding. By Telesat’s own estimate, the total addressable market for LEO satellites will hit US$430 billion in 2025, and the company believes it can capture 1% of that market. Even that sliver would imply revenue of US$4.3 billion for a company that posted sales of $760 million last year. “We’re spending a ton of money to do this, and it will be transformative for Canada, for the world and for the company as well,” says Goldberg, who notes the company has so far prepaid more than US$100 million to secure launches for its satellites.
The only problem is, Telesat hasn’t managed to build or launch a single commercial Lightspeed satellite. Its prime contractor for the project, France’s Thales Alenia Space, has been hit by the same supply chain delays for simple electronic components that have bedevilled car and home appliance buyers over the past year, and Telesat has had to push back the date for Lightspeed’s commercial rollout to at least 2026. The company was also expected to confirm during the release of its first-quarter results in early May (after this magazine went to print) that in the face of inflationary pressures it will “descope” the initial size of its constellation from 298 to 188. Nor had Telesat finalized financing for the ambitious project, even though the federal, Ontario and Quebec governments have committed $2.6 billion to the project. Since Telesat’s shares started to trade on Nasdaq and the TSX last November, their value has plunged nearly 70%.
Add it all up and Telesat is simultaneously faced with a galaxy of opportunities for its LEO satellites, and a world of hurt to get there. Nothing less than the future of the company is tethered to Goldberg getting the Lightspeed rollout right.
Like a cat with a bell on, you can hear Goldberg coming before you see him. Thud, rattle, drag. In March, while out for a run with a group of friends, Goldberg’s right foot came down hard on a patch of black ice as he turned his body. He heard the crack even before his brain comprehended what had happened to his shattered ankle, which broke in three places. Since then, he’s been on crutches (though he has since shed the plastic brace on his foot as he undergoes physiotherapy).
As if trying to do rocket science amid a pandemic wasn’t already hard enough.
Goldberg, a trim 57-year-old with an easy smile, isn’t actually a rocket scientist. He’s a Harvard-trained lawyer (one of his classmates was Barack Obama), but Goldberg has spent nearly his entire career around satellite companies in Europe and North America, first as in-house counsel at a company called PanAmSat and then at Netherlands-based NetSkies, where he rose to become chief operating officer. After that company was sold, he decided to move his family back across the pond. Goldberg was headhunted to join Telesat, a former Crown corporation then owned by Bell Canada, where he was tasked with leading the sale of the company. In 2007, Canada’s Public Sector Pension Investment Board and U.S.-based Loral Space & Communications teamed up to buy the company for $3.3 billion. (Telesat’s stock market listing last November followed its merger with Loral.)
It was during an airline flight in 2015, as he grappled with how to keep Telesat relevant in the high-speed digital age, that he pulled out a cocktail napkin and started jotting down ideas for what would eventually become Lightspeed. Earlier that year Musk had revealed his low-Earth-orbit plans for Starlink, a “giant global internet service provider” that would operate within his SpaceX rocket business. Telesat’s existing geostationary satellites, of which there are now 13, take about 800 milliseconds for signals to travel back and forth to Earth. With LEOs, that would be cut to well less than 50 milliseconds. But the new venture would be expensive to build and carry unforeseeable execution risks. In the end, the decision came down to what the company’s telecom, commercial and government customers were clear they wanted, says Goldberg: “A better, faster, more affordable, more resilient solution.” Only LEO satellites could offer that.
Today, the industry is centred on four main companies. In addition to Musk’s Starlink and Telesat, U.K.-based OneWeb has so far launched 422 satellites as part of its planned 648-satellite constellation, while Amazon’s Project Kuiper envisions an LEO constellation of more than 3,200 satellites.
Starlink, by far, is the most well-known of the bunch. That’s not a surprise when you have a showman like Musk as your founder. Starlink is also riding a wave of praise in the wake of Russia’s invasion of Ukraine after it worked with the U.S. government to send thousands of satellite internet kits to the war-torn country, allowing residents there to stay connected after Russia shut down other forms of communications.
But Starlink’s high profile also stems from the simple fact it’s already put so many satellites in space. At last count, Starlink has launched 2,400 LEO satellites out of a planned 40,000. So far this year, it’s successfully launched nine rockets filled with 448 satellites. In March, the company said it had signed up 250,000 customers in 25 countries.
Telesat’s planned satellite count may seem teeny by comparison to Starlink’s, but that’s by design. For one thing, the Lightspeed satellites will orbit at higher altitudes—most will circle the Earth at 1,000 kilometres, versus 550 for Starlink. That allows fewer satellites to cover a wider area on the ground, but not so high that it creates signal delays. Telesat’s satellites will also be interconnected with one another through optical laser links, creating a mesh capable of beaming capacity to areas where it’s needed most. A 2021 paper by researchers at the Massachusetts Institute of Technology analyzing the four main LEO contenders estimated Telesat’s satellite utilization rate will be 73.4% once the initial web of satellites is fully deployed, compared to 33% for Starlink.
Goldberg, who met with Musk and SpaceX president Gwynne Shotwell three years ago to discuss their respective LEO plans, has heard the Starlink comparisons many, many times before. While Starlink has recently begun to expand its offering into the enterprise sector Telesat plans to focus on, including deals to provide satellite internet to some airlines, Goldberg still sees plenty of room for both players. “We don’t delude ourselves. SpaceX has been a transformative company, but to be clear, they can be fully successful with their rocket and Starlink business, and Amazon can be successful with their Kuiper business, and we can be fully successful with Lightspeed because no one is going to dominate and own this entire market,” he says. “If we were saying we could get 50% of this market, people would be right to question it. But we feel pretty good about getting 1% or 2%.”
Telesat CEO Dan GoldbergJustin Tang/The Globe and Mail
It’s also not a stretch to say the popular view within Telesat is that Starlink’s approach of throwing tens of thousands of essentially disposable satellites into space (their shelf life is just five years, versus 10 for Lightspeed satellites) is madness. The company’s executives will never come out and say this, but it’s interwoven in how they talk about their own project. “We’ve built a constellation that’s optimized for enterprise users,” says Goldberg. “It’s not a best-efforts kind of broadband connectivity. It’s committed information rates. It’s providing service at a 99.99% availability standard. It’s what enterprise and government users demand.”
Dave Wendling, Telesat’s chief technical officer, puts it more succinctly: “Our spacecraft are not throwaways. They’re big, capable, leading-edge technology designed for graceful degradation. Because we can’t afford these things to be falling out of the sky.”
Telesat’s potential customers are awaiting its entrance into the market. “There’s an anticipating in the industry that Telesat has the technical potential to leapfrog some of the others, so we’re anxiously waiting for that,” says Rick Hodgkinson, owner and CEO of Galaxy Broadband Communications Inc., which provides satellite communications services to remote industrial customers like mines, oil rigs and large construction projects—many of them operating above the 57th parallel, which runs just north of Fort McMurray, Alta. At the moment, Hodgkinson is working with OneWeb to test its satellite internet system and has roughly a dozen terminals at various sites, including a construction site in Northern Ontario that had been relying on geostationary satellites for connectivity. He says he expects to roll out a full commercial launch this June.
But Hodgkinson says OneWeb doesn’t have enough capacity in its LEO constellation to meet the needs of his customers. What draws him to Telesat is its ability to beam capacity around its web of satellites to where it’s needed.
As for Starlink, Hodgkinson says its system design and focus on the consumer market doesn’t make it suitable for large commercial operations. “I think of it this way: If you’re a contractor building homes and you need a vehicle, you’re not going to buy a Toyota Prius,” he says. “It’s capable of going onto job sites, but it can’t haul anything there.”
That may be the case, but not every enterprise customer will be willing to wait for Telesat to finally enter the race. “I like the Telesat approach of making satellites that are more robust and capable, but the problem for them lagging behind is that if Starlink gets operational first, then Starlink can win enterprise customers purely by being the only one available,” says Dallas Kasaboski, a consultant with Northern Sky Research in Strasbourg, France. While Telesat can try to convert those customers over to Lightspeed once the constellation is up and running, that’s a slower and costlier prospect. “It’s a very difficult proposition to be the best but also be last.”
You don’t have to be in Nunavut to endure the grind of sluggish internet in Canada or to see the potential opportunity for satellite internet. You could, for instance, drive an hour and a half southwest of Toronto, the fourth-largest city in North America, to a country road in Norfolk County (my old stomping grounds). A little south of one of the hamlets there, at a spot where a log from the wagon-era corduroy road occasionally breaks through the pavement, is an invisible line where the latest rural broadband initiative is set to end. While the 200 residents in the hamlet are likely to see download speeds that meet the Canadian Radio-Television and Telecommunications Commission’s 50/10 target (50 Mbps download, 10 Mbps upload), the scattered farms to the south will be left with speeds in the range of one to three Mbps down.
According to the CRTC, only about 45% of rural households in Canada can obtain its target download speed. That figure falls to 35% for people living on First Nations reserves.
For rural development researchers like Helen Hambly at the University of Guelph, the endless promises about faster rural broadband make one thing clear: The status quo is failing. “At a time when you’ve got rising food prices and labour shortages on farms, you have this agriculture digital revolution happening without the underlying connectivity to support it,” she says, pointing to technologies like sensors and drones that help improve crop yields. “We’ve been working on these issues for, my goodness, 12 years now, and sometimes I have a hard time sustaining my energy to get people to buy into why we need more effort in this area.”
While Telesat will not be serving rural internet users directly, it does plan to work with telecoms to provide what’s known as backhaul connectivity to communities where fibre isn’t an affordable option. It turns out there are a lot of them.
At Telesat, it’s Manik Vinnakota’s job to map out the commercial opportunities for Lightspeed. Or in other words, to figure out how the massive project will pay for itself. That means compiling and analyzing millions of data points on things like commercial airline traffic routes, cruise and merchant ship traffic, urban-rural population divides and distances to local highways—an indicator of the ease with which an area can eventually be served by fibre optics on the ground.
When it came to Canada, Vinnakota used federal data that divides the entire Canadian landmass into a geospatial grid of 516,322 individual hexagons, with each 25-square-kilometre hexagon including sundry details about the population that lives there, what broadband speeds are available to them, and the distance to the closest land-based fibre optic cable. Their research identified roughly 1,200 rural communities where, over a 10-year time horizon, it would be more economical to connect them to high-speed through Lightspeed versus the cost of laying thousands of kilometres of new fibre optic cable. “That was a big light bulb that went off for us,” Vinnakota says. “It proved our initial numbers that showed we have a clear and viable business case, because the lowest-cost option should win.”
The question for Guelph’s Hambly then comes back to when Telesat will get its satellites to space, especially since the federal government committed nearly $600 million to the company’s Lightspeed project in 2020, and the launch date has now been pushed back. “I have no problem with this investment, but the folks who are not connected to high-speed internet needed this yesterday,” says Hambly.
In the meantime, rural customers are embracing Starlink, despite its high costs. (In March, Starlink’s price tag got even steeper due to what it called “excessive levels of inflation”—for new orders, its internet unit rose to $759 from $649, while monthly fees also rose from $129 to $140.) It’s a phenomenon Hambly is also tracking via customer reviews posted on the Reddit page for Starlink Canada. She recently launched a data analysis project to scrape comments from the 1,100-member community to understand to what extent the service is closing the high-speed gap.
It certainly worked for James Lindsay, a teacher and fire chief in Lynn Lake, a community of about 500 people in northern Manitoba. As soon as Starlink announced it would be available, Lindsay put his name on the waiting list, having endured 0.25 Mbps download speeds through Bell MTS’s “allegedly high-speed DSL,” he says.
On March 18, when Lindsay got his hands on his Dishy McFlatface—the nickname given by Starlink’s engineers to its internet terminal—he immediately set it up and tested it out. The results were staggering: 199 Mbps down, 13 Mbps up. At last count, at least 45 people in Lynn Lake and the nearby Marcel Colomb First Nation—the “raggedy edge” of Starlink’s current service area, as Lindsay puts it—have become customers. “The 21st century is finally going to arrive in Lynn Lake,” he says.
Using low-Earth satellites to provide broadband isn’t a new idea. In the 1990s, several companies—including Bill Gates–backed Teledesic—vowed to create “internet in the sky,” but all saw their plans go down in flames. In 2021, Musk acknowledged the industry’s poor track record in a tweet: “Every new satellite constellation in history has gone bankrupt. We hope to be the first that does not.”
A lot has changed since then. The technology behind satellites has improved dramatically, while the cost to build and launch them had been falling steadily, at least until the current wave of supply-chain disruptions and soaring energy costs began to bite. Most of all, the world’s appetite for high-speed broadband has only grown more voracious. In the early 1990s, global internet traffic amounted to 0.0001 exabytes a month, according to networking giant Cisco. (An exabyte is one billion gigabytes.) This year, IP traffic is forecast to hit 400 exabytes a month. Next stop: the zettabyte era.
That’s not to say the industry isn’t starting to look crowded, sometimes to bizarre degrees. This past October, Rwanda filed a request with the United Nations’ International Telecommunication Union (ITU) to put nearly 330,000 LEO satellites in space as part of that country’s quest to become the hub of Africa’s space industry. A month later, another Canadian startup, Toronto-based Kepler Communications, leapt into the fray with its own request to the ITU that would see it develop a megaconstellation of 114,852 satellites, though the company says it plans to launch just 200 itself to create an “internet in space.” The balance would come from other space-industry customers who would affix Kepler’s tech to their own satellites to tap into its space-based network. In March, Germany’s Rivada Space Networks proposed a 600-satellite constellation. Another Canadian company, Skytrac Systems of Kelowna, B.C., followed a month later with plans for a 700-satellite network of its own.
It’s not clear if any of those projects will receive regulatory approval and go ahead, but even if they don’t, space is filling up. Since 2011, the number of satellites has jumped to 8,600 from around 1,000. Telesat mission control, housed in a tinted-glass room in an office building overlooking Ottawa’s Elgin St., contains rows of computer terminals and monitors. One large screen in particular jumps out. It displays a feed of every satellite and object currently stationed in space or orbiting the Earth—a mass of tiny dots that combine to form a goopy mush around the planet.
That has scientists and astronomers alarmed. In a paper last month for the journal Nature Astronomy, a group of scientists warned that by 2030, the number of satellites in low-Earth orbit could blow past 100,000 and “significantly alter our whole perception of the night sky in the long term, appearing as ‘fake stars’ [and] rival the number of visible real stars.”
Even so, the LEO space race is showing no signs of slowing.
Space is a brutal, unforgiving place to do business. This past February, Starlink lost 40 satellites a day after launch when a geomagnetic storm caused them to fall from orbit and burn up. As Telesat notes in its prospectus about the dangers inherent in its industry, “these risks include in-orbit equipment failures, malfunctions and other kinds of anomalies [such as] circuit failures, transponder failures, solar array failures, telemetry subsystem failures, battery cell and other power system failures, satellite control system failures and propulsion system failures.” If that weren’t enough, the company continued, “magnetic, electrostatic or solar storms, space debris, satellite conjunctions [and] micrometeoroids” can take out a satellite at any moment.
But if all that sounds bad, it’s nothing compared to a gummed-up supply chain back on Earth.
In October, Goldberg was at his cottage recovering from COVID-19 when he got a call from the president of Thales Alenia’s space division, which has the contract to build the LEO constellation. A month earlier, Goldberg had been in Paris, his first trans-atlantic trip since the start of the pandemic, and had received hints suppliers were struggling to find parts. The phone call confirmed it: Telesat would have to push back Lightspeed’s launch date. “That was a kick in the teeth,” says Goldberg.
The delays don’t necessarily involve complex parts; instead, they’re the general building blocks of all electronic equipment: resistors, diodes and integrated circuits. Without them, the satellites can’t be completed.
In addition to announcing the delay of Lightspeed’s launch, Telesat is considering several steps to minimize further disruption, including switching some subcontractors, swapping components and ordering parts far earlier than it normally would.
Soaring inflation in the cost of parts and labour also presented Telesat with an uncomfortable decision—it would either have to raise more money or downsize its initial ambitions. As it stood in early May, Goldberg was still in talks with Export Development Canada and Bpifrance, the two export credit agencies the company is relying on to provide roughly 40% of the project’s $6.5 billion in capital funding. If either of those two agencies don’t come through, Telesat might have to find other ways to finance the project as the cost of borrowing gets much steeper. (Goldberg says he’s cautiously optimistic, but “if that doesn’t come to pass, we’ll cross that bridge when we get there.”)
That leaves the option of building a lighter Lightspeed. Wendling, Telesat’s chief technical officer, says even if the initial rollout has 188 satellites instead of 298, “the constellation will still provide many terabits per second of capacity, way more than is out there right now.” The long-term goal remains to expand Lightspeed to around 1,670 satellites, he says. It’ll just take longer to get there.
Investors are losing patience. While the recent rout has sunk many tech companies’ stock, Telesat’s has been in decline since November. As of early May, it traded for $15.50 on the TSX, down from $45 six months ago.
One problem, analysts say, is the company hasn’t done a great job developing relationships with equity investors. Its stock market listing didn’t come with the issuance of any new shares, so the Bay and Wall Street marketing machines that normally accompany an IPO have been absent.
At the same time, Goldberg has had to field repeated calls from Telesat bond investors whose debt is tied to declining revenue from the company’s legacy satellites. His exasperation came through on a recent conference call when one bond analyst kept describing spending on Lightspeed as a “drag.” When Telesat treasurer Michael Bolitho echoed the analyst’s wording, Goldberg jumped in. “Let’s not do this drag thing,” he said. “We are investing. We’ve spent our money pretty carefully. We don’t invest it, we hope, in a dumb way.” At another point in the call, Goldberg summed up his message to bond investors: “What we’re trying to say is, like, chill.”
The tensions reflect a half-century-old company that’s trying to transform itself in an industry now dominated by the world’s most disruptive personality: Elon Musk. “Elon is a very promotional person, and that’s part of the success of his companies, and that doesn’t work for everyone,” says Walter Piecyk, an analyst at New York–based LightShed Partners who covers Telesat. “But sometimes companies focus too much on making sure they have all their ducks lined up before they talk about what they’re doing, and in my opinion that’s the approach of a company that’s been talking to bond investors for the last decade.”
For starters, Piecyk hopes Telesat will spend more time at investor conferences and on the business-TV circuit. “Look, are you going to launch this thing or not? And if you are going to launch it,” he says, “you need to get out and talk about the target markets and what you’re going to do to address them.”
For his part, Goldberg says Telesat has signed up to do more investor conferences this year. But ultimately he ascribes to a “show, don’t tell” philosophy. “Actions speak louder than words, and I’m a big believer that folks need to see concrete milestones,” he says. “Investors need to see us close our financing, launch satellites, announce customer contracts, build a contractual backlog, and I’m very comfortable we can demonstrate that.”
In the meantime, in February it signed a deal with a contractor to build 30 landing stations worldwide that will link the Lightspeed constellation to terrestrial networks. It’s also letting potential customers see Lightspeed’s capability up close with LEO 1, a demonstration satellite launched in 2018. (A second demo craft is due to be launched later this year.)
Nunavut’s Dean Wells has already taken LEO 1 for a spin and was floored by the results. He believes the territory will ultimately sign contracts with several LEO satellite operators, with Telesat as one possibility. His team has already run fibre optic cables to government buildings in all 25 communities in the territory. Plans are also underway to install fibre linking Iqaluit to a destination farther south. Combine all that with a world that has embraced the remote delivery of vital services like health care and education, and the pieces are in place for life in Nunavut to be fundamentally transformed.
“You couldn’t have ordered it any better,” says Wells. “The stars are aligning.”
Your time is valuable. Have the Top Business Headlines newsletter conveniently delivered to your inbox in the morning or evening. Sign up today.