We are at a crossroads, it would seem. Much has happened in the satellite space these past ten years or so, and all of it brings opportunity – to satellite network operators (SNOs), mobile network operators (MNOs), consumers, and, of course, those who rely upon connectivity in extreme and remote areas across the globe.
Using satellite connectivity for IoT applications has, traditionally, come with a hefty price tag. Inevitably, the costs to build, launch and maintain a satellite network has to be recouped. In 1995, launching a satellite would have cost $26,884/Kg.
That cost has come down more recently, largely because companies like SpaceX have made it possible to reuse launch components. The trend for launching Low Earth Orbit satellites – usually smaller, with less fuel needed to launch – has also played a large part. By 2020, the cost to launch a satellite network had reduced to $951/Kg, around one twenty eighth of that in 1995.
Today, we are in an era where SNOs have less outlay to recoup, given the lower capex, bringing satellite IoT connectivity costs down. This has made launching a satellite constellation an attractive proposition – but it still remains a costly pursuit. The NewSpace Index lists more than 70 companies set up to launch satellites to serve the IoT / M2M industry. Of these, more than half have been canceled before launching satellites or after prototypes were launched. Only three are complete: Iridium and Orbcomm, both of which are well-established ‘legacy’ players in the satellite industry, and Swarm, which has since been absorbed into Starlink, and is no longer being developed.
So, it’s important when we talk about the future of satellite IoT connectivity to bear in mind that very few SNOs aiming for the IoT / M2M market have been able to maintain the momentum needed to bring their vision to life.
For all the recent improvements, it is still expensive to build, launch and maintain a satellite network, and most companies need to be able to demonstrate that they will deliver a return on investment within as short a timeframe as possible.
Companies promising extremely low costs will in turn need to grow their subscriber base rapidly if they are to deliver shareholder value. This proved easy for Starlink as it disrupted the satellite broadband industry, but it will be more challenging for satellite IoT as the value per subscriber is so much lower than a conventional broadband model.
Furthermore, the growing number of new entrants and all of the accompanying hype may have left end users confused about what’s genuinely available. In the absence of certainty, decision-making may be postponed, or the trusted incumbent remains in situ. The rapid harvesting of new subscribers needed to make these new satellite networks commercially viable haven’t materialised - yet.
Proprietary vs. Standards-Based
To date, end users have been required to select and, to a large extent, stick with a single SNO. Services have required a proprietary modem to be attached to an IoT device, whether that be a drone, data buoy, tractor, or even a tracking collar for remote wildlife. The key issue here is that should you want to change suppliers, for whatever reason, that may be around price, speed, or better coverage, you'd have to change the modem, or indeed a whole host of them. By its very definition, Satellite IoT tends to be deployed in remote or even hazardous locations, adding a logistical ordeal to the already evident cost implication. The inherent 'stickiness' is attractive, which is likely why new entrants Swarm, Kinéis, SAS, Astrocast, and Myriota opted for a proprietary model.
The proprietary model is not in and of itself a barrier to lower costs. Although we're not seeing an abundance of new satellite networks emerging, more operators will enter the field, leading to greater competition and reduced costs.
The alternative is to launch new satellites that are compatible with unmodified terrestrial devices, now made possible by the 3GPP standard. This makes devices and satellites interoperable when they’re both designed and built. A single module will be able to carry multiple radio technologies – cellular, satellite, GPS – and critically, users should be able to switch between satellite network operators in the same way, conceptually, as an eSIM allows users to switch between cellular networks, which would potentially lead to greater confidence over airtime pricing.
Chiefly, it’s the newer satellite network operators, the ones who are in the process of establishing their networks, who have opted for this. But established player Iridium has indicated that they are in the process of enabling their constellation to support standards-based modules, although there is no word just yet on the timing of this.
There still exists challenges, however
Spectrum allocation is not an infinite source, and here's where we may see some 'trade-offs.' SNOs with licensed Mobile satellite services (MSS), the frequencies reserved for satellite services, require no further agreements with terrestrial networks (mobile operators) to send and receive data from a compatible device. Likewise, L-band, perhaps the best radio spectrum for satellite IoT applications, is also licensed and already allocated to established players, such as Iridium and Inmarsat / Viasat.
All this leaves new entrants with three options; they work with another radio spectrum, usually shared UHF / VHF; they work with terrestrial networks to repurpose cellular spectrum; or they work with the existing SNOs to rent their spectrum.
Shared UHF / VHF has congestion and capacity issues, and is only really viable where there isn’t a great deal of competition, and there’s no geographical hindrance, such as forests or rough terrain. Many new entrants take the second option, by working with MNOs, where a company can persuade the MNO to reallocate part of its spectrum for the SNO.
However, working with existing SNOs who already have licensed MSS spectrum and landing rights is the fastest way to achieving global service and a high-capability network. Given the huge IoT spectrum advantage they have, they’re going to be important in the success of future IoT networks, both proprietary and Standards based. There is an interesting dynamic here; if SNOs lease spectrum without requiring a relatively high fee, they run the risk of their new partners cannibalizing their own market share.
Satellite IoT opportunity beckons
The technology now exists that will enable the satellite IoT market to grow, with new entrants with a small number of satellites offering a new class of service: lower prices and lower capability. This is ideal for markets where receiving data once or twice a day would be a perfectly good outcome: agriculture and some environmental monitoring, for example.
Meanwhile the existing service class – higher prices, high capacity – continues to inch downwards in price, but is far away from parity with cellular. Everywhere that satellite IoT is utilised today is basically mission-critical: pipeline monitoring, tsunami, flood and wildfire alarms, asset tracking, weather reporting, remote security alerts - this list goes on, with a shared characteristic that the data is needed in real-time. And that urgency commands a premium.
The existing SNOs will have a huge influence on the rate at which these prices fall: either by adopting standards-based modules themselves, which should lead to a reduction in module pricing, or leasing their airtime at a competitive rate to start-ups. The tempting prospect of rate-switching between SNOs, however, is some way off, as the gulf in service proposition between the incumbent and new satellite networks is currently too great.
It benefits no one to see so many new entrants fail, so while we chase the dream of lower pricing brought about by rate-switching and standards-based modules, we need to bear in mind the financial realities of operating a satellite network. Prices for high-capacity service will come down – but in our view, not as far or as fast as some of the hype would have you believe.
Alastair Macleod is the CEO of remote connectivity provider Ground Control.
Related articles:
