Estimates put global internet penetration at just above 60% of the world’s population. That means between 3 billion and 3.5 billion people are still without any internet access at all. The number without access to high-speed or broadband service is even higher. If we agree that bringing people online is generally a pretty good thing, how do we connect broader segments of the population given that so many are outside of highly developed or densely populated areas?
The traditional option is to build the ground infrastructure to wire millions of miles of high-speed cable.
Another is that we could launch hundreds of rockets to place thousands of satellites into orbit, such that every spot on Earth has direct line of site to an internet providing spacecraft. In recent years, theres been a massive investment in this space-based approach.
This post will seek to discuss the following topics around how that’s happening and what that means:
How Does Satellite Internet Work?
Existing Satellite Broadband Options
New LEO Satellite Internet Constellations
The Companies Competing In This Market
How Big Is This Market?
First though, a quick note: “broadband” is classified as an internet connection with minimum download speeds of 25Mbps and upload speeds of 3Mbps. I had to look that up and in case anyone else didn’t know, thought I’d save you a Google as well.
How Does Satellite Internet Work?
The above diagram provides an illustration of how users can get online using satellites. A ground station connected to the internet beams information up at a satellite orbiting high above the Earth. That satellite then has a direct line of site to a user terminal (the dish antenna on the house) and the satellite transmits the information coming from the ground station back down to the user terminal. The user terminal is then connected to a modem and router which provide the information from the user terminal to your internet enabled devices. You can then send information back to the internet from your home devices by transmitting right back to the ground station through the satellite.
Existing Satellite Broadband Options
Broadband satellite internet actually already exists. The most popular provider right now is HughesNet. The HughesNet satellites are placed in Geostationary Orbit, which means they are far enough away that they orbit Earth once every 24 hours. This conveniently allows them to stay over the same spot on the Earth’s surface so that ground user antennas never have to change directions to keep pointing at the same HughesNet satellite, similar to your satellite television dishes.
In the next sections we’ll talk about why companies are developing internet satellites in low Earth orbit, but before we do it’s worth noting that internet satellites in geostationary orbit aren’t necessarily going away anytime soon. There are even new entrants into that market. For example Astranis is a venture capital backed organization that has raised over $350M to provide more affordable and much higher bandwidth internet with geostationary satellites.
The Development of Low Earth Orbit Satellite Internet Constellations
A number of organizations are looking to improve the quality of satellite internet by bringing the satellites much closer to Earth, from geostationary orbit to low Earth orbit. The primary disadvantage to having internet satellites relatively far from Earth in geostationary orbits (about 35,000 km away) is that it takes .25 seconds to communicate with a HughesNet satellite and for the satellite to then connect back to servers down on Earth. That means that for your computer to get a response from the servers, it takes another .25 seconds, or .5 seconds total. For some applications like email or Netflix streaming that’s not a huge deal, but for others like playing video games where the user is rapidly responding to feedback from the internet, that delay is an issue.
For this reason and a few others, some organizations are working to provide that broadband internet service using satellites in low Earth orbit that are 1/35th the distance to geostationary orbit. Satellites in low Earth orbit travel around the Earth once every 90 minutes and since they are so much closer to Earth, each satellite only sees a small portion of the Earth’s surface at a time. As a result, you need a lot more satellites in low Earth orbit to be able to have global coverage, often hundreds or thousands of them.
In addition, since each individual satellite is passing by overhead so quickly, the ground antennas actually need to continually be changing directions to follow the satellites as they zip across the sky. This requires a new type of antenna called a phased array antenna which directs its signal electronically rather than having to constantly be changing the direction the antenna is pointing.
The Companies Competing in this Market
Starlink
SpaceX’s Starlink is currently the most deployed of the low Earth orbit satellite constellations. SpaceX began launching its first prototype Starlink satellites back in February 2018. At the time of this writing, SpaceX has placed about 1,800 Starlink satellites into orbit and has expressed interest with the FCC for potentially up to 42,000 (although it is unclear if SpaceX really plans to use that many). Starlink initiated a paid beta program in November 2020 and SpaceX recently told the FCC that beta speeds are “exceeding” 100Mbps. The costs are $499 for the physical ground antenna that connects the user to the satellites and then a $99 per month subscription for the service. Furthermore, SpaceX has announced partnerships with Microsoft Azure and Google Cloud as the cloud computing partners for the Starlink network.
One of Starlink’s biggest advantages is that SpaceX is the only player in this market that is both building the satellite constellation and the rockets to launch them. By not having to pay an outside organization for a ride to orbit, SpaceX’s costs are likely anywhere from $20M to $50M lower than its competitors for each launch (depending on which competitor you compare it to and what SpaceX’s margins are per launch). It probably takes at least 50 launches to deploy a constellation and potentially up to a few hundred. This translates to billions of dollars in cost savings for SpaceX. Furthermore, managing the launches in house means that SpaceX can get on a ride much more easily and quickly than others. In 2021 so far, SpaceX has launched a Starlink mission about once every other week, a blistering pace. All while still launching satellites for its commercial and government customers.
Project Kuiper
In the Spring of 2019, Amazon announced their low Earth orbit broadband satellite constellation, Project Kuiper. The full constellation is planned to consist of 3,236 satellites. While Project Kuiper might not have the internal launch infrastructure to leverage, like SpaceX does with Starlink, there are a number of benefits that come from being part of the Amazon ecosystem. For one, Amazon has expressed plans to invest $10B into Project Kuiper, suggesting that the project has no uncertainties about funding. This is in contrast to competitors that have had to go through the effort of raising external capital to fund their constellations.
Furthermore, Project Kuiper is being developed within Amazon Web Services, meaning that Amazon plans to combine its category-leading cloud computing services with the Project Kuiper internet service to provide a vertically integrated user experience. Furthermore, Amazon may be leveraging its experience in consumer electronics to support the design and development of its user terminals. Amazon has announced specifications of those terminals to be under 1 foot in diameter, an impressively small form factor.
It’s worth noting that while there is sometimes assumed to be a relationship between Amazon and Jeff Bezos’s rocket company Blue Origin, the two have no formal partnership. The Project Kuiper team has been clear that they will launch their satellites with whatever launch company helps them best meet their goals. In fact, Kuiper has secured nine future launches on Atlas V rockets built by launch provider United Launch Alliance.
OneWeb
OneWeb was founded in 2012, originally under the name WorldVu Satellites. The company plans to launch a constellation of 648 satellites to provide global broadband internet coverage. OneWeb’s first major fundraise took place in 2015 when it raised $500M, mostly from major aerospace entities including Arianespace and Virgin. In 2016, the company raised an additional $1B from SoftBank. In 2019, OneWeb formed a joint venture with Airbus Space and Defense named OneWeb Satellites which would manufacture the satellites and was intended to leverage Airbus’s satellite development experience.
OneWeb filed for bankruptcy in March 2020 after being unable to raise further capital to fund the project. In November 2020 it was announced that the company was purchased at auction by a consortium led by the Government of the United Kingdom and Bharti Global, an Indian conglomerate with a large telecommunications business. Since coming out of bankruptcy, OneWeb has re-continued launching its satellites into space and is again pursuing its plans for a full internet-providing constellation
OneWeb began launching its first satellites in 2019 and has so far launched 254 satellites aboard eight separate Soyuz rocket launches. Current launches are deploying 36 satellites on each rocket. OneWeb plans to begin offering internet service in the UK before the end of 2021 and global service in 2022.
Telesat Lightspeed
Telesat is a satellite communications company based in Canada that has existed under varying ownership since 1969. The company currently provides satellite television and geostationary broadband satellite services to users in Canada and the US. In 2016, the company announced Telesat Lightspeed, a constellation of low Earth orbit satellites to provide low latency broadband internet. What was first revealed as a constellation of 120 satellites has since grown to plans of over 1600. Telesat launched the first prototype satellite in 2018 and is planning to start launching the full constellation in 2023. Telesat has also stated that they plan to begin sending up their satellites using Blue Origin’s upcoming New Glenn rocket.
Telesat Lightspeed’s service will initially prioritize customers in Canada but the constellation will grow to global availability. Telesat has stated that they seek to target enterprise business customers whereas they expect other providers like Starlink and Project Kuiper to focus more towards consumer customers.
Telesat has also announced plans to be listed on the NASDAQ exchange sometime in 2021. The full Lightspeed constellation is projected to cost $5B for which the company will need to raise capital to develop.
How Big Is This Market?
As stated earlier, nearly 40% of the global population is still without access to the internet. To say though that this entire over 3 billion population is going to pay for a service that might require up to $500 in upfront user terminal costs and then potentially another $99 per month (this is SpaceX’s pricing, the other providers haven’t announced pricing yet) is obviously not a reality. However, while not everyone is going to be able to afford satellite internet immediately, there are likely many rural communities in nearly every nation that can afford it and would jump at the opportunity to pay for a premium broadband service.
In the US, according to the FCC’s 2020 Broadband Deployment Report about 30% of the rural populations or around 18 million Americans did not have access to broadband internet in 2018. With an average US household size of 2.54, that corresponds to about 7 million US households. If LEO internet providers can serve 10% of those households at $99/month, that comes to about $850M in yearly revenue. However, at these early stages it’s difficult to determine if anything like that 10% adoption rate is a reasonable assumption. There is also the likelihood that the satellite internet service providers will capture some segment of the approximately 126 million US households that already have access to broadband internet but are unsatisfied with their current service.
Beyond that, the US only accounts for about 5% of the global population. In fact the markets that seems to excite many of these companies are developing ones. For example one particular area of interest is in bringing online broader populations of the continent of Africa. Just as much of that continent skipped past landline technologies and went directly to cellular communications, maybe it is possible for much of Africa to skip laying millions of miles of high speed cable and instead go straight to satellite internet. A key caveat to that will be if these services can be made affordable. The ability of these companies to serve nations without existing high-quality low-latency broadband options is going to depend significantly on how much the providers can bring down prices as the services scale.
It’s important to also note that there are significant international regulatory challenges to overcome as well. These companies will need to negotiate ground station infrastructure in every nation they hope to serve and some governments might have concerns regarding taking consumer internet traffic outside of their country. Additionally, some nations might also be protective of their own domestic projects as well. China (which accounts for about 18% of the world’s population) recently announced plans for its own 13,000 satellite constellation and its hard to imagine the Chinese government not wanting internet traffic to travel through that constellation rather than through one developed by a private western company.
Moreover, the traditional ground-based telecommunications providers are not sitting idly by while the LEO satellite constellations make their way to market. Internet global penetration seems to be growing at about 3% every year, so traditional internet providers are looking to sweep up that unserved market as well.
The LEO internet companies themselves claim the total market to be about $30-$40 billion and if those numbers are true, the upside is massive. SpaceX has recently stated that it has around 90,000 user terminals in their beta program in less than a year of operation. That would correspond to about $106M yearly recurring revenue and rising quickly.
Overall, high speed internet is only becoming a more critical part of the way we live our professional and personal lives. The benefits of bringing underserved populations online will likely be enormously positive in raising productivity and standards of living around the world. If low Earth orbit satellite providers can create a high-quality and affordable service, then it seems theres a tremendous opportunity for these companies to thrive.
Additional Reading
Liked this read? Follow me on twitter for more frequent updates on the space world.
Want to read more about satellites? Check my previous post about investment opportunities from the rise of satellite constellations.
I wonder how hard it will be to scale up the whole constellation to a higher speed. Say 10 years from now, the constellation is set, but the average internet speed has gone up considerably. Will these companies need to launch newer satellites, or is there any bottleneck (e.g. ground infrastructure) that they can keep upgrading without new launch?
Very informative article, Ian. Question: What level of government support/cooperation is required for all of this to happen? For example, are additional frequency allocations from the FCC required? If so, will it be an auction process and could this materially alter the economic model? Are there other ways in which in which the government or other powerful third-parties might alter how all of this rolls out?