The topography of space communication is not as straight forward as people think.
Some highly profitable potential users of Starlink in Low Earth Orbit at 550km (LEO) concerned with the transmission of time-sensitive information, seem keen to engage in a developing race for speed maximisation and latency reduction. Starlink could likely achieve latency speeds of 43 milliseconds vs. long established 76 milliseconds transatlantic cable/fibre – a 77% increase, a massive gain for the time-sensitive industries that would use these networks and derive substantial financial benefit from such improved performance. Musk mentioned at a Satellite conference at the beginning of 2020 that they are even touting 20ms. One reason for this is because light travels 47% slower in glass on earth than in a space vacuum.
Space internet today is typically provided by High Throughput Satellites (HTS) using Ka- and Ku-bands usually in Geo- and Near- geostationary Orbit (GEO/NGEO) operated by companies like ViaSat, Telesat and Intelsat. Latency is the time delay over a communication link, and is primarily determined by the distance data must travel between a user and the server. LEO satellites are typically 35 times closer to Earth than traditional GEO Satellites. Satellite internet currently needs on average 638ms for a round trip from GEO.
The current fastest terrestrial connection on the planet is the privately owned Hibernian express cable at 58.95ms i.e. 39.4% slower than Starlink – between New York and London. The previous record was held by Atlantic Cable 1 (AC1) at 65 milliseconds, a $300m investment for no more than a 5 millisecond speed-up. One may imagine how valuable a 17 millisecond time-saving would be to those trading on the financial markets between London and New York. Such latency advantages increase rapidly with distance. For example a London/Singapore connection would generate a large latency advantage for time-sensitive industries.
Even though LEOs such as Starlink will likely struggle to compete against HTS GEOs on a cost per Gbps basis, industry experts believe the future of satcom will be a hybrid of GEO and non-GEO satellites with LEO satellites providing mobile backhaul, high speed consumer broadband and enterprise Very-Small Aperture (VSAT) connections.
Real Engineering has created a fascinating video that explains the impact such a project may have on internet speed as well as consequent financial and economic implications.
With the 12,000 Starlink satellites planned and FC approval in November 2018 for 7,518 satellites in addition to 4,425 satellites previously authorised, SpaceX launched the 1st batch of not fully functional Starlink satellites on May 24th 2019 to reach a 550km operational Low Earth Orbit (LEO) altitude – chosen chiefly to optimise speed and latency as against area of coverage.
The above performance selection preferences illustrate how Starlink can offer a formula that can attract time-sensitive customers. Elon Musk revealed that Starlink could eventually net perhaps $30 billion to $50 billion a year. The source of these revenues would, he said, be customers in the financial sector and others dealing in markets with rapidly changing prices. When speaking in these terms, Musk was in capital-raising mode rather than referring to the Starlink project’s original broad aim to connect to the internet, the 41% of the earth’s population still cut off from it.
The issue critical for the mass adoption of this type of technology is the production of radically cheaper ground station antennas. This was one reason why OneWeb failed. For their business model to be economically viable, One Web needed to reduce the average cost of ground stations to below $100. Starlink talks of $200 for a ground station, except that their current cost is around $1,500 and at scale could perhaps reach $1,000. To realise the world that Starlink or OneWeb seem to have in mind, would require installed bases of millions of ground stations costing billions of dollars. One estimate puts the cost of building out the Starlink network at $170 billion.
One UK Start-up that received £9 million of UK Space agency funding last summer, aims to achieve a $600 price point after the full development and testing of their product. This could happen in a few years time. Challenges of this order are generating scepticism about the immediate prospects for Low Earth Orbit (LOE) technologies. Some phased array technology costs between $30,000 and $250,000 – and that technology is not fully market ready with resonance issues at chip level - a problem that needs several years and >$10s of millions to resolve.
It is recognised that moving the US from a 4G to a 5G wireless network would cost $150 billion. This is set against global cumulative mobile operator capex spend between 2018-2020 of $479 billion. Roughly 80% of this is to be spent in less densely populated areas.
I have personal knowledge of the operation of a well-funded transformative satcoms company that does not need any ground stations and benefits from a first rate wholesale distribution model through existing MNOs (Mobile Network Operators) who understand the differing relevant regional regulatory intricacies. They are projected to generate several $billion revenue with only 10s of satellites. They need a very limited number of gateways to connect the satellites to the contributory MNO networks. They could generate upwards of $15 billion with a few hundred satellites. All at an infrastructure cost of maybe a couple of $ billion.
I would question the assumptions that the OneWebs and similar promoters seem to envisage as facilitators of the mass adoption of their technology. It seems from Musk’s comments at a Satellite conference at the beginning of 2020 that Starlink is not aiming for the mass adoption market because it will serve the "3 or 4 percent hardest-to-reach" customers, as well as those who "simply have no connectivity" right now. By appealing to those customers, Musk feels a "significant load" will be taken off from other Internet Service Providers – ISPs and by consequence the Mobile Network Operators (MNOs).
The other issue of course is physically launching and keeping 12,000 Starlink satellites in LEO (low earth orbit). LEO satellites have typically an operational life / stay up for only 5-7 years. This means that nearly ALL SpaceX rocket launch slots for the foreseeable future would be needed to meet the regulatory requirements to keep the license that Starlink has fought hard to secure.
This could cause a bottle neck for launching of other satellites which the likes of Skyrora and Rocket Labs have already recognised.
One needs also to bear in mind that regulatory requirements may set other challenges for those willing and able to continue participation as the technology develops. Perhaps there is some new technology in the offing that will provide answers to these important questions. As yet the mathematics seems to favour the sceptics.
Despite this, Mobile telecoms incumbents are coming under pressure to reduce capex spend while increasing revenue. Mobile tower operators are realising the limitations of terrestrial networks and fervently looking for new business models. The mobile telecoms players need to find new ways of reducing capex spend while increasing coverage. They need to be able to put a significant dent in to the cumulative capex spend of $479 billion while increasing revenues. Hence the MNOs are looking for new ways to monetise their existing subscriber base that roam in and out of terrestrial network coverage.
The Satcoms sector is facing an upcoming oversupply of satellite connectivity capacity and accelerating cannibalisation of TV broadcasting by internet streamers such as Netflix. These will pose fundamental challenges to the viability of current business models within the Satcoms sector. The satellite industry is building new satellites in GEO for High-Throughput bandwidth, and constellations in Medium Earth Orbit (MEO) and LEO have received $billions in funding to put up constellations of satellites that aim to provide broadband connectivity to high bandwidth users and specialised verticals, from enterprise to Government.
Large telcos are typically 50-100x orders of magnitude greater than satcoms players in terms of revenue. Over the last 5 years both satcoms & telcos P/E ratios have been hit hard. Innovative disrupters are poised to buck that trend and bridge the scale gap between satcoms and telcos.
I can think of a few who possibly could make it. OneWeb with only 74 of the planned 648 satellites in orbit is not one, unless it radically changes it’s positioning in the market. Investment by the UK government of a further £500m equity in to such enterprises is unlikely to provide an answer. This would add to $3 billion+ already invested in OneWeb.
Let us not forget Teledesic, a company founded in the 1990s to build a commercial broadband satellite constellation for Internet services in LEO, initially planning for 840 active satellites at 700km, then scaled back to 288 active satellites at 1,400km using Ka-band. It was an ambitious and ultimately failed project funded in part by Microsoft (the Musks’ and Bezos’ of the 1990s /2000), that ended up costing $9 billion. The commercial failure of the similar Iridium and Globalstar ventures (composed of 66 and 48 operational satellites respectively) and other systems, along with bankruptcy protection filings, played a primary role in halting the project. Teledesic officially suspended its satellite construction work on 1 October 2002.
Costly negative results can have severe deterrent effects on the development of new disruptor technologies at particular junctures.