French communications satellite network provider Eutelsat has finally announced its ELO constellation, which will use a low earth orbit (LEO) to provide Sigfox connectivity to devices on the ground. However, at only 25 planned nanosatellites, this is a small constellation, and one that is going to face availability pressure from the other nanosatellite upstarts.
There is also the question of motivation for Eutelsat, a company with annual revenues of €1.32bn and a net income of €357mn. Why does it want to expand into the LPWAN market if it is already making bank in conventional satellite communication? Well, the answer seems to be that this is a very low-cost project, from its perspective, with very high potential upside.
We first learned of the Eutelsat-Sigfox project back in February, when speaking to Sigfox CEO Ludovic Le Moan. For Sigfox, as with all other IoT startups, one of the biggest hurdles to signing enterprise-scale customers is convincing them of your longevity and credibility. With Eutelsat, Sigfox gets the backing of a recognized operator with a good track record, and potentially a new parent, should the project be successful.
Of course, terrestrial infrastructure will provide better indoor and underground penetration than satellite-based Sigfox, but in terms of covering outdoor space, satellites have extremely compelling capex advantages. To this end, it is notable that filling in the gaps in satellite coverage with terrestrial base stations would have been the better way to go about this, rather than using satellite to fill in the terrestrial gaps, but we imagine Eutelsat would not have signed up if it were starting completely from scratch.
Rodolphe Belmer, CEO of Eutelsat said “we are delighted to be completing the first steps of our ELO constellation project, aimed at positioning Eutelsat in the nascent IoT segment. Our partnership with Sigfox enables worldwide coverage through the combination of satellite and terrestrial IoT. This relatively modest investment at Group level, which is fully scalable, enables Eutelsat to access an additional potential growth lever in the context of its Connectivity strategy.”
Eutelsat bought its first satellite from Tyvak International last year, and will be launching this one in early 2020. This will be followed by four satellites supplied by Loft Orbital and Clyde Space, with each vendor providing two. This constellation will provide commercial services upon launch, and, Eutelsat says that if the service proves successful, it will deploy the full fleet of 25 nanosatellites by 2022 – at a capex cost of no more than €1mn per satellite.
The Tyvak nanosatellite will be used to test how Sigfox signals perform when sent to and from orbit. Due to there being far less interference when traveling upwards, the Sigfox ecosystem is pretty confident that regular Sigfox radios will comfortably be able to reach the constellation, with some believing they could reach the much larger satellites in higher orbits too.
However, the big difference between the nanosatellites and the larger communications satellites is that the big boys are almost always fixed in a geostationary orbit, meaning they maintain the same relative position above the earth at all times – moving in sync with the earth’s orbit, at an altitude of around 36,000 kilometers.
The nanosatellites almost always move relative to the earth, roving around their orbit to cover more of the earth’s surface at an altitude of around 600 kilometers. This means that one satellite could cover much more surface area than one in a geostationary orbit, but the trade-off is that it will not always be available. In geostationary deployments, the end-device can always send messages upwards, while in the low-volume LEO constellations, there will be periods of time when there is no coverage.
That problem is solved by simply adding more satellites to the constellation, so that when one nanosatellite is straying too far from an end-device, another satellite is already following behind its orbit, ready to take over. Alphabet’s Project Loon has figured out how to do this using weather balloons at a lower altitude, but the nanosatellite operators are doing much the same – working out to always keep a nanosatellite above a particular area, to ensure coverage. NASA has a pretty good explanation of the different types of orbits here.
But this is where the number of satellites comes into play. Eutelsat’s ELO seems quite low, on just 25 units. Sure, initial launches are smaller in number, with the nanosatellite operator launching a few devices and being able to promise one message per day, as the limited number of nanosats travel across the earth’s surface. In time, adding more satellites means that the availability reaches 100% of time, but in their conversations, the nanosat crowd seem to expect to need many more satellites than Eutelsat is planning.
Fleet Space Technologies is planning to launch 10 more satellites by 2020, after successfully launching four last year, and plans to launch 100. Astrocast is planning to launch 80 by 2023 or 2024, upping the number from the 64 initially planned to meet equatorial coverage. Audacy has talked of 2,000 units, supported by larger MEO satellites, Hiber has plans for 48 nanosats, and Sky and Space Global (SAS) says it will launch 200 units.
All these startups are aiming substantially higher than Eutelsat, in terms of numbers of satellites. Old rival Iridium has also just finished its Next constellation of 75 LEO satellites, but these larger units have been rather costly – part of a $3bn operations upgrade. It is worth noting that both SpaceX and OneWeb have plans to launch thousands of LEO nanosatellites, but these are more focused on conventional broadband delivery, rather than IoT connections. Currently, SpaceX has launched 60 of its planned 12,000, and OneWeb has asked the FCC to authorize 1,200 extra satellites, on top of its original plan for 900.
However, the difference in constellation size can be explained by the different sort of use-case that Sigfox has been chasing. Its ‘0G’ branding positions itself way down the table of applications, to things that only need intermittent messaging, rather than the sort of IoT application that needs 100% availability and massive bandwidth.
These nanosatellite launch costs are also getting to the point where a large enterprise could justify owning their own satellite constellation. Once you’ve started down the conversation of owning network infrastructure, there’s not all that much difference between WiFi, private LTE, or indeed a handful of nanosats.