Emerging direct-to-smartphone capabilities are breathing new life into the mobile satellite services (MSS) industry.
For decades, Inmarsat and other satellite operators in this industry have been confined to providing services to specialized handsets with bulky antennas.
However, technological breakthroughs and work to standardize space and terrestrial communications promise to help break them free of niche products to serve mass-market smartphones and other devices.
It is still early days for the direct-to-device market. Services that have been launched so far and those on the near horizon are limited to low-bandwidth applications such as emergency messaging.
But the future shape of this market is still being determined as startups also seek to take it on with dedicated constellations.
For 44-year-old Inmarsat, direct-to-device is just one transformational opportunity to work through as the British company is in the middle of being sold to U.S.-based Viasat.
Both Viasat and Inmarsat operate communications satellites in geostationary orbit (GEO). Most of the players chasing the direct-to-device opportunity are in low Earth orbit (LEO), including Globalstar which has enabled an SOS feature on Apple’s latest iPhone in certain regions since November.
Satellites in LEO can provide lower latency connections than GEO because they are closer to Earth. As part of a multi-orbit constellation announced in 2021 called Orchestra, Inmarsat is considering satellites in LEO to bolster its global network as Starlink, OneWeb, and other LEO operators chase its connectivity markets.
Inmarsat also plans to add two payloads in highly elliptical orbit to the mix later this year for coverage over the globe’s northernmost latitudes.
Inmarsat chief technology officer Peter Hadinger, who joined the operator in 2011 and has spent more than four decades in the space industry, spoke to SpaceNews following last month’s launch of Inmarsat-6 F2 about plans to expand in the evolving MSS market. Inmarsat-6 F2, a twin of Inmarsat-6 F1 that launched in December 2021, promises improved services in L-band spectrum the operator uses for MSS services to handheld devices. The satellite also carries a Ka-band payload for providing high-speed broadband to larger terminals.
What do you make of all the activity going on in the direct-to-device market?
I’ve been in this industry for a very long time. I’ve seen the cycles, and I must say it’s been quite a long time since mobile satellite services were front and center of the business.
There was a flurry of mobile satellite systems toward the end of the 90s — Iridium, Globalstar, and others that came along and then disappeared — and then it got sort of set to this little kids’ table and disappeared at conferences altogether for a while.
Now it’s back, largely spurred by this direct-to-device capability.
There are a number of different pathways to direct-to-device. Some people are using mobile network operator spectrum and others are using satellite spectrum. Each one has its advantages and disadvantages.
For us, we’ve been doing direct-to-device for ages [with satellite L-band spectrum], although when people talk about direct-to-device they usually mean direct-to-cellphone. There, it’s really just a matter of whether the cellphone tunes the proper frequency band.
We have some partners like Bullitt and others who are building phones to speak in our L-band, and they’re using narrowband waveforms like NB-IoT to do that.
The big question then is how much bandwidth do you need and do you need to build a purpose-built fleet to do that? We’re not yet convinced that is the case, but we have been approached by everybody because we have spectrum.
Our challenge, and it’s a good problem to have, is our spectrum is largely being used. Unlike some who have spectrum that was pretty fallow. So it’s not just something where we can say ‘oh yeah, sure, let’s repurpose it.’
And you have to ask if the value of wideband spectrum is going to return the same as a lot of narrowband spectrum. I’ve often said that people will pay a lot more per bit for the first bits of connectivity than they will for the millionth bit of connectivity.
People talk about climbing mountains and doing FaceTime from the top of wherever, and the millions of bits it takes to do that, versus the 100 bits it might take to say ‘my car is broken down, I’m here, come get me.’ That takes fewer bits to do, but they’re probably more valuable bits.
We’ve spoken with everybody and we’re open to working with people who want to develop those kinds of things but, from an internal investment, I think we’re really focused on the narrowband market.
Iridium has partnered with Qualcomm to provide direct-to-device services. Have you been approached by chipset developers?
We can’t say who because of NDAs, but you can imagine the players that have spoken to us — everybody is kind of out there feeling the market out.
A number of people are starting things, but, in many ways, they’re taking careful baby steps to see if this is going to develop into something. It is a very expensive proposition to go to something that would be wideband.
There are some that have staked their companies on this and others that are just sort of dabbling around the edges. If you take a look, most systems are going to be starting off in the narrowband end of things because that’s where a fairly certain market is. The cost to deliver goes up radically as your bandwidth increases. It’s not clear if the revenue follows that same trend line.
The recently launched I-6 F1 and I-6 F2 satellites effectively double the amount of Inmarsat’s usable L-band capacity. Would that enable Inmarsat to deliver higher bandwidth direct-to-device services on top of narrowband?
Yes and no. Most of the increase in capability brought in by them is going to be improving the performance of our own wideband satcoms devices. These are not handheld phones. They’re shipboard terminals, airplane terminals, and so forth.
So it’s not really about having extra spectrum to do direct-to-device as much as it is improving services on these terminals. However, depending on the nature of the partner, they may wish to deploy something different. We’ve seen statements by Viasat that they’re interested in it. They may have ideas to bring to the table as well.
There’s also a debate over about whether these services are best delivered from GEO, or from LEO satellites like Inmarsat has been considering for Orchestra.
The GEOs are very efficient from an operational standpoint. You can put up a long-lived asset that’s up there for 15-20 years. You’re not having to replenish them continuously.
You can really target the capacity you have on a GEO to the places where it’s needed at a given point in time. Spacecraft are power limited, fundamentally, and you can take the power on a GEO spacecraft and have it in the east in the morning and the west in the evening, or where the hotspots are — it’s considerably different than a LEO where that power may be over water in the deep ocean for a considerable portion of its orbit, therefore unusable.
But LEOs also have advantages. We have done some experiments in LEO and we have filings in lots of different bands.
We have the luxury of being able to sit back and watch some of these markets mature. Inmarsat has had increased profits over the last two years, exceeding the industry in terms of revenue and EBITDA growth. It’s a healthy business so we’re in no rush to somehow change that radically.
We’re investing to essentially mature the different options that are out there. We said back when we introduced the Orchestra program that we’d make investments of $100 million through 2026 in LEO. That goes into maturing some of the new terminal technologies, some of the terrestrial technologies that we’re investing in, as well as the things we need to do to prepare for these options.
We don’t have a need for additional capacity and capabilities in the network until the end of the decade. We can afford to play a longer game and watch the industry consolidation, which we’re 100% convinced needs to happen in the current environment, and let that play out and then take our strengths in the market and put them where they’d best be used.
When do you think you’ll decide on a LEO strategy?
Inmarsat-7 will be deploying and coming into service in 2025, and we’ll be ready to shift investment and focus at that point for the capacity we may need by the end of the decade.
From a big investment question standpoint, that’s the timeline. Does it make sense in the interim to partner with somebody? Maybe. Would that partnership last a long time? Possibly.
Across the board people do recognize the value that Inmarsat brings to any relationship, just in terms of our understanding of mobility markets as the largest mobility player, and the market access, the relationships, the multi-spectrum bands — all the rest of it.
It’s a good time to be in this business as mobility is coming back into vogue. It’s nice to be able to talk about so much more these days. We’ve just been in the background, quietly doing our business for quite some time, and making good money out of it.
While Inmarsat has historically used L-band for voice and messaging, the spectrum is increasingly finding uses for connecting Internet of Things (IoT) devices — how has that market grown for your company?
We’re a big supplier to a lot of IoT markets today. Because we do that primarily indirectly, through partnerships and so forth, we don’t track and report the number of IoT devices that we have on the network, but it’s in the millions.
Viasat mainly sells directly to consumers. Could Inmarsat’s distribution strategy change if it is successfully sold to Viasat?
They’ve been able to go direct to consumers because it’s been a business based in the United States.
I think when they go internationally, they’ll find that partnerships are not only desirable but in some cases necessary because you need an in-country presence to have market access and so on.
Part of the advantage of [Viasat] buying Inmarsat is we’ve already got these relationships, partnerships, and market access.
One of the things Viasat bring, however, is technology. They are fundamentally a hardware vendor that has increasingly gotten into services with time. So they bring considerable technology to this that will complement our assets.
I still expect in the hardware sense that we’re going to work with the best of breed in the market. We have a number of long-standing hardware partners.
We’ve yet to have the opportunity to spend much time comparing notes. It’s a long-distance courtship in some sense, but I think each of us will have things to learn from the other as we get into the things we may or may not know about each other’s internal plans.
“It’s no longer GEO A versus GEO B. It’s GEOs versus the LEOs. For us, the consolidation that the market has needed for some time is sped up by this dynamic.”
We’ll get there. We’ve still got a few regulators to convince [to approve the deal] but I think our arguments have been pretty [compelling] when it comes to the profound changes that are happening in the market.
Especially the introduction of the LEOs. Having OneWeb, Starlink, etc., come in has really changed the nature of the market. It’s no longer GEO A versus GEO B. It’s GEOs versus the LEOs. For us, the consolidation that the market has needed for some time is sped up by this dynamic.
And these LEOs are coming with billions of dollars of investments.
The cost of developing and quite frankly sustaining a LEO network is massive.
Take Starlink’s 30,000 spacecraft, give them five-year lifetimes, and you’re launching 6,000 spacecraft a year just to keep it in business.
Even if you’re launching 60 of them on a Starship — you know, the larger, future style — that’s still 100 launches a year. Two launches a week of Starship. This is just a phenomenal amount of investment.
This is why it’s a good business for people with lots of money to spend, and they’re certainly taking some pretty big bets.
It may be the right thing for us to partner with somebody like that to do it, but they’re really addressing markets that are largely different from ours.
Our markets are not so big in mobility that you can afford to do that size of a constellation. You don’t need it from a capacity perspective.
The things we’ve put together conceptually are much more modest in size, but specifically designed to address mobility compared to fixed residential customers. And they are very different markets with different kinds of requirements. You have a high degree of expectation for certainty, as opposed to best efforts.
This article originally appeared in the March 2023 issue of SpaceNews magazine.
