Satellite IoT is coming of ages with a number of competing projects. Recent announcements triggered this opinion paper about technology and goto market for the various IoT initiatives.
As a simple description, the main hurdles to overcome for a satellite IoT project (or a telecom project) are the following:
- Technology: just make it work, low power and easy to use being key drivers.
- Frequencies: without spectrum, no business, simple as that.
- Cash: cash is king, and a lot of cash is required to build a system but also to operate it and (often overlooked) develop the business especially when the market needs to be created from scratch.
- Market: IoT in general has been over-hyped in the past (remember Sigfox?) and market uptake has been slower than promised. Satellite IoT is only a fraction of terrestrial IoT and market uptake might take some time as it will probably remain a niche market. In any case, satellite IoT ventures often focus mostly on technique rather than on market which is an issue as history has shown (see below)…
Where it all started…
As a matter of fact, the concept of short messaging by satellite is not new and dates back to the 90s. This is when the first constellation bubble started in the USA with several projects: Iridium pushed by Motorola for global voice communications, Orbcomm an Orbital Sciences project to provide 2-way satellite messagging and Globalstar, similar to Iridium but less ambitious.
What do they have in common? They all failed, filed for chapter 11 but are still up and running and doing rather well now.
All projects got their frequencies all right, only Orbcomm landed poor VHF spectrum. All projects raised enough money to design, build and launch the first version of their system.
All projects, although not perfect, were great technological achievements.
And all projects had (almost) no customers when they started operating. The problem here is that when operational, the countdown starts as space infrastructure has a limited life time. And any lost or delayed customer is an additional nail in the coffin.
Simply put, with too much CAPEX, an infrastructure that degrades immediately after launch and slower than planned market ramp-up, these systems were doomed.
So nothing happened for 20 years, the time to forget…
And here comes the NewSpace …
And with NewSpace comes the second wave of IoT constellations. The promise of NewSpace is simple: reduce the cost to space by a factor 10 or even 100 and make things faster. This makes it possible to build satellite constellations quicker with less CAPEX. As IoT is less complex than voice or broadband, this paved the way for new projects pushed by start-ups.
Several companies decides to take the risk to build a system and go to market: Kineis in France, Hiber in the Netherlands, Astrocast in Switzerland, Swarm in the USA, Myriota and Fleet Space in Australia…
These projects all successfully demonstrated their technological implementation, they managed to get some authorizations and frequency bands and raised some cash for their projects. To my knowledge, only Kineis is fully funded, Hiber completely pivoted towards the downstream market et Swam was bought by Starlink.
The gotomarket still remains the major issue for these project – nothing new here. However, another issue popped up, on the technology side now.
… but they forgot about technology…
These new IoT projects forgot that technology is evolving and the big push came from the telecom market (which is 100 times bigger than the commercial satellites market). To make things short, this is called 5G NTN (Non Terrestrial Network) or the capability to connect 5G devices directly to satellites. The 3GPP that standardizes the cellular world made provisions in the 5G implementation for satellite networks.
Several new start-ups decided to follow that route: Lynk, AST Space in the USA, and OQ and Sateliot in Europe. All plan to provide NB-IoT connectivity from their satellites with standard IoT devices: this makes it possible to have a 5G IoT using a common terrestrial network, say Orange or Vodafone, and switch to satellite whenerever these networks are no longer available. The benefit is to have cheaper devices and the possibillity to easily switch network, for example in case of failure. This greatly secures the investment of customers: after all, why choosing a proprietary network when standards are available, that reduce the risks.
In addition, another standard is being used, Lora with companies such as Lacuna Space that has already several satellites in LEO orbit. While LoRa was invented by Cycleo, a French company purchased by Semtech, it is becoming an IoT standard as well. LoRa benefits from a large ecosystem of device makers, service providers and network operators.
And now we are back to basics…
However, recent developments may demonstrate that even these new IoT projects based on standards can have a hard time as there may be potential game changers in this still moving satellite IoT space. And the surprise may come from established companies…
Echostar Mobile, an established satellite operator is offering a new service on its European GEO satellite. Echostar has chosen to use the LoRa standard for its new offering. Using a small module, it is possible to seamlessly use terrestrial LoRa or the Echostar satellite.
Similarly, Inmarsat just announced a partnership with Mediatek to provide 5G IoT connectivity via its fleet of GEO satellites. Mediatek, a major Taiwanese chip maker, will provide a specific chip that can communicate directly with the Inmarsat satellites. This chip can be used in devices, smartphones or vehicles and leverages the new 5G NTN standards..
The game changer here is that Echostar and Inmarsat not only implement standards but also reduce the CAPEX to a minimum as they do not need any new infrastructure. They use existing satellites, reprogrammed to accomodate the standard protocols. In addition, they leverage the frequencies (L and S band) that they already use for their other services. This reduces significantly the risk of launching this new business and will make them significantly more competitive than other operators.
One unknown is if Starlink and Kuiper will enter the IoT market. OneWEB will probably be limited by its technology, but Starlink V2 and Kuiper may choose also to go the 5G NTN path and could offer IoT as a side service to their general broadband offering.
So what’s next?
As always, future is difficult to foresee. However, history has shown that standards are key for large and fast market update and that the lowest CAPEX is key to success (remember Geostar vs Qualcomm Omnitracs?).
Non standard operators still can make a business, but they probably will have to find their own niche operations or try to go vertical to increase their margins.