The collapse of OneWeb has sent shockwaves through satellite communications. Once a bright star in the sector with multi-billion dollar backing, the refusal of SoftBank to pour more money into the start-up led to OneWeb filing for Chapter 11 bankruptcy relief at the end of March. The timing of the announcement was particularly brutal, just hours before OneWeb launched more than 30 micro satellites into orbit from Kazakhstan, part of a planned constellation of 720 satellites which is now in serious doubt.

The demise of OneWeb has been attributed to both Covid-19 (coronavirus) and SoftBank tightening its investment belt. The Japanese giant has projected $16.7 billion in losses for 2019 for its Vision Fund, with OneWeb identified as one of its most notable losing investments not included in that fund. But industry sources suggest the satellite start-up has been struggling for years to adapt its business model as the consumer LEO proposition became increasingly unrealistic. And SoftBank has had enough. Covid-19 might have just been the accelerator of a foretold death.

The advantage of LEO satellites over their high-orbit cousins
Low earth orbit (LEO) satellites have been hailed as the next big thing in non-terrestrial communications, with the potential to boost data backhaul for operators, bring broadband connectivity to underserved regions, and further enable cloud and edge-based services. The key advantage of LEO satellites over legacy geostationary earth orbit (GEO) satellite services is the much lower altitude. This not only slashes the cost of launching and maintaining the constellation, but also improves downlink speeds and latency due to a much-reduced data round trip.

However, one of the key drawbacks of the LEO proposition is the expense on the ground. In order to maintain a strong and constant data link, LEO satellite antennas need to be steerable: simply put, the dish has to move to track the satellite. This means the antenna are considerably more complex and expensive than comparable terrestrial wireless receivers or the basic passive aerial on a GEO satellite phone handset. Clearly there is an option to simply throw more birds into the air to reduce the need for steering, but aside from the huge additional cost this raises other issues in terms of precisely planning the satellite trajectories and avoiding space debris in our already-crowded skies. So far, the focus has been on the space segment of the LEO constellation solutions, with most of the investment going onto the design, production and launch of the satellites, without much concentration on the end-user device segment (the stuff on the ground).

The cost of these antennas varies, but is likely be higher than the comparable cost of a fibre (FTTP) connection for more than 99 per cent of homes and small businesses. SpaceX reportedly said the cost of an individual satellite terminal was around $1,000 and that mass-market production could bring the cost down to $300. It’s likely to be many years before we’re even close to developing LEO antennas cheap enough to support a consumer proposition and, by that time, other solutions such as 3G and 4G are likely to have reached all but the most remote areas of the planet.

Despite the demise of OneWeb, the LEO proposition is still flying
This is not to say LEO is dead in the sky. Other operators such as Canadian communication satellite veterans Telesat have made a firm decision to target the wholesale and enterprise sector, even designing satellites with built-in flexibility via steerable antenna to target receivers with the highest data demand. There is a major market opportunity here in offering additional backhaul to existing operators as well as data-hungry industry verticals. And there will always be the sectors that need reliable connectivity in unserved regions such as aviation; maritime; oil and gas; and defence.

The renaissance of satellite has been driven by LEO constellations, allowing operators to throw hundreds of satellites into the sky at a lower cost. But whether satellite can ever offer a realistic consumer service proposition depends on cutting costs in three areas: networks, service and devices. The cost of putting satellites in the air isn’t likely to fall much (unless there’s a run on rocket fuel), while the amount the operators can charge the end-user is constrained by the usual RoI and competition from other technologies. So, until devices, in this case the customer premise equipment (CPE), come down in price, LEO is unlikely to be anything more than a very niche consumer proposition.

What’s next for OneWeb?
It doesn’t look like buyers are queuing up to acquire the company’s constellation assets, especially given the hardware has apparently been designed with a consumer rather than an enterprise focus. Each LEO satellite provider has designed their satellites with specific requirements which match the shared spectrum they’ve got a licence for and the planned business model. So the brutal truth is the dozens of satellites already launched look set to become more space junk, until their orbits decay sufficiently that they succumb to gravity and burn up on re-entry.

The asset OneWeb is flaunting as the most sellable one, its licensed spectrum, also has its limitations OneWeb’s spectrum is on the shared Ku-band, and cannot be repurposed easily. Not all OneWeb competitors have Ku-band spectrum. Secondly, being shared spectrum, any changes in the satellites to use the spectrum would require extensive testing to make sure there is no interference with the other constellations using the same band.

Having around 70 satellites in operation, OneWeb passed the first of the new ITU milestones approved in 2019. As per the new ITU requirements for non-geostationary constellations, spectrum licensees need to deploy 10 per cent of their constellation within two years of the licence being granted to secure priority status against interference from other constellation hopefuls. But whoever acquires the spectrum would have to fulfil the other ITU requirements: launch 50 per cent of satellites within five years and 100 per cent within seven. Otherwise the licence is revoked or limited to the satellites launched. An alternative would be a pre-emptive strategy from another LEO constellation operator to block other entrants. Considering how expensive, on top of the limitations imposed by ITU, this unlikely route would require deep pockets for anyone choosing it.

Will other players in the LEO constellation race be affected?
The demise of OneWeb could also cast shadows on the viability of the other three main LEO satellite constellation projects, namely SpaceX, Amazon’s Project Kuiper, and Telesat. However, there are some big differences between OneWeb and the other three. They all have recurring revenues from related or, in the case of Amazon, unrelated businesses. Of the three, Telesat’s main business is providing services to the communications and broadcasting market segments with GEO satellites, which positions it well with an existing customer base for its enterprise and backhaul transport plans for its LEO constellation. Meanwhile SpaceX and Amazon, being vertically integrated, are better positioned than OneWeb and have clear business models outlined. Amazon, with its vast resources, might be even willing to take a stab at an end user terminal, if anything as a reference design, as it has done in other consumer ventures. Ultimately the downfall of OneWeb clears a market which was potentially becoming overcrowded from the start.

Peter Boyland – lead analyst, Ecosystem Research; and Fernando Elizalde – senior analyst, GSMA Intelligence

The editorial views expressed in this article are solely those of the author and will not necessarily reflect the views of the GSMA, its Members or Associate Members.