Battle of the (constellation) titans
With the upcoming launches of Starlink and OneWeb, we are getting closes to the big showdown of many LEO communication constellations, that might reshape the world wide reach of the Internet itself. Not much will change for the majority of the world population. But those currently off-grid or near the edge of the grid will see a major change. Together with solar and batteries adoption, universal cheap and low-latency Internet access could decrease or even reverse huge urbanization shift of human population world-wide.
Note that in this blog I rely on combination of publicly available information and speculation from me personally.
The race is on.The contenders are OneWeb, SES (with mPower), SpaceX (with Starlink), Amazon, Telesat, Leosat and Viasat. There are also variety of smaller GEO players with their HTS offerings, but they do not really have the publicly expressed ambition to lead the revolution, and are trying to play it safe with combination of HTS and FSS services.
This is a winner-takes-all global race, with huge rewards. But definitely many will be burned. Much like mobile communications revolution of the last decade, where top-ten manufacturers from 2000 exited the market except the market leader (Nokia) which reaped most of the profits, and then smartphone revolution where (again) top-ten manufacturers from 2010 were completely replaced with except the market leaders (Apple and Samsung) which reaped all the profits.
Lets start the analysis with smaller players of the race. Viasat is in many aspects betting on traditional evolution of GEO HTS. It has existing customer base and proven technology, which reduces the risk significantly. Although high cost per each Viasat-3 satellite, its 2-billion dollar investment and its experience in building gateways and POP devices makes it really serious contender. Its achilles heal is the communication latency - its position in GEO brings huge latency that can be solved by scaling the same type of satellites to MEO. This move would require gateway upgrades and POP antennas replacements, but the same issue is common to all other players in the system.
Leosat posted a white-flag, as a first smaller player that was burned.
Telesat, as primary Canadian company is more focused on polar regions. it has existing GEO customers, but its constellation is much smaller than OneWeb. Which, by the way, plans to address polar customers from the start. So they are not in good position and will probably exit the LEO constellation market. The absence of news regarding the constellation seems to confirm that.
Amazon is an interesting prospect. As major Internet company, and the largest cloud provider, they definitely have an interest in securing both dark-fiber/high speed connectivity, super low latency and address the last mile problem for its subscribers/customers. Plus they could secure launches for Blue Origin/New Glenn, Jeff Besos pet company. But in reality, Blue Origin is years behind SpaceX regarding launch capability. New Glenn will economically outperform Falcon Heavy by a small margin, in the five year time frame. But by that time, Falcon will have hundreds of launches, with high rate of reusability. And next generation Starship will already be flying. So Kuiper project will be hindered and late for the party.
SES mPower, like Viasat, is low-risk approach to the constellation. It is moderate investment, providing gradual migration path for existing customers from GEO to MEO markets. SES is well positioned to survive LEO constellation bankruptcy wave 2.0. It solves main drawback of GEO latency, eases antenna pointing accuracy and much slower satellite tracking compared to LEO satellite tracking. It uses similar high frequency reuse approach like Viasat. Per bit costs are similar or better than OneWeb. With existing and growing large customer base, it is well positioned to survive in one form or the other.
OneWeb is the latest casualty in the race, using COVID-19 outbreak as an excuse for fundraising failure. But the truth is, their costs were just too expensive. Just on the launch cost basis, they were a billion dollar more expensive than Starlink. While Starlink focuses initially on single market (USA) and high margin, direct to consumer service, OneWeb had very complex web of investors, creditors and technology partners which made it very hard to cut costs. Simply put, Starlink moved much faster and provided much better integration between launch vehicle capacity (mass and volume). Due to its software-first, Silicon Valley attitude and experience in building mass produced products coming from Elon Musk/Tesla, I estimate that cost per bit of the space segment was order of magnitude more expensive for OneWeb than for Starlink. So even if OneWeb managed to launch the complete constellation, it would still have problems competing with Starlink. And due to the timing and looming expiration of OneWeb spectrum rights, it is likely that the company will be shutdown.
So that leaves Starlink. Its success is not guarantied. The good thing is that they essentially dont have any real competitors at least until New Glenn becomes operational. But ground segment costs, especially frequency reuse and phased array antennae might cause additional funding crunch. If this constellation becomes profitable, competition will come fast, either from new LEO constellations or from accelerated buildup of terrestrial networks. In some way, Starlink was designed to give Starship initial market, so it recoup (some) development costs.
Note that in this blog I rely on combination of publicly available information and speculation from me personally.
The race is on.The contenders are OneWeb, SES (with mPower), SpaceX (with Starlink), Amazon, Telesat, Leosat and Viasat. There are also variety of smaller GEO players with their HTS offerings, but they do not really have the publicly expressed ambition to lead the revolution, and are trying to play it safe with combination of HTS and FSS services.
This is a winner-takes-all global race, with huge rewards. But definitely many will be burned. Much like mobile communications revolution of the last decade, where top-ten manufacturers from 2000 exited the market except the market leader (Nokia) which reaped most of the profits, and then smartphone revolution where (again) top-ten manufacturers from 2010 were completely replaced with except the market leaders (Apple and Samsung) which reaped all the profits.
Lets start the analysis with smaller players of the race. Viasat is in many aspects betting on traditional evolution of GEO HTS. It has existing customer base and proven technology, which reduces the risk significantly. Although high cost per each Viasat-3 satellite, its 2-billion dollar investment and its experience in building gateways and POP devices makes it really serious contender. Its achilles heal is the communication latency - its position in GEO brings huge latency that can be solved by scaling the same type of satellites to MEO. This move would require gateway upgrades and POP antennas replacements, but the same issue is common to all other players in the system.
Leosat posted a white-flag, as a first smaller player that was burned.
Telesat, as primary Canadian company is more focused on polar regions. it has existing GEO customers, but its constellation is much smaller than OneWeb. Which, by the way, plans to address polar customers from the start. So they are not in good position and will probably exit the LEO constellation market. The absence of news regarding the constellation seems to confirm that.
Amazon is an interesting prospect. As major Internet company, and the largest cloud provider, they definitely have an interest in securing both dark-fiber/high speed connectivity, super low latency and address the last mile problem for its subscribers/customers. Plus they could secure launches for Blue Origin/New Glenn, Jeff Besos pet company. But in reality, Blue Origin is years behind SpaceX regarding launch capability. New Glenn will economically outperform Falcon Heavy by a small margin, in the five year time frame. But by that time, Falcon will have hundreds of launches, with high rate of reusability. And next generation Starship will already be flying. So Kuiper project will be hindered and late for the party.
SES mPower, like Viasat, is low-risk approach to the constellation. It is moderate investment, providing gradual migration path for existing customers from GEO to MEO markets. SES is well positioned to survive LEO constellation bankruptcy wave 2.0. It solves main drawback of GEO latency, eases antenna pointing accuracy and much slower satellite tracking compared to LEO satellite tracking. It uses similar high frequency reuse approach like Viasat. Per bit costs are similar or better than OneWeb. With existing and growing large customer base, it is well positioned to survive in one form or the other.
OneWeb is the latest casualty in the race, using COVID-19 outbreak as an excuse for fundraising failure. But the truth is, their costs were just too expensive. Just on the launch cost basis, they were a billion dollar more expensive than Starlink. While Starlink focuses initially on single market (USA) and high margin, direct to consumer service, OneWeb had very complex web of investors, creditors and technology partners which made it very hard to cut costs. Simply put, Starlink moved much faster and provided much better integration between launch vehicle capacity (mass and volume). Due to its software-first, Silicon Valley attitude and experience in building mass produced products coming from Elon Musk/Tesla, I estimate that cost per bit of the space segment was order of magnitude more expensive for OneWeb than for Starlink. So even if OneWeb managed to launch the complete constellation, it would still have problems competing with Starlink. And due to the timing and looming expiration of OneWeb spectrum rights, it is likely that the company will be shutdown.
So that leaves Starlink. Its success is not guarantied. The good thing is that they essentially dont have any real competitors at least until New Glenn becomes operational. But ground segment costs, especially frequency reuse and phased array antennae might cause additional funding crunch. If this constellation becomes profitable, competition will come fast, either from new LEO constellations or from accelerated buildup of terrestrial networks. In some way, Starlink was designed to give Starship initial market, so it recoup (some) development costs.
Comments
Post a Comment