In the Böddensell forest 167km west of Berlin, on clear, starry nights - when the position of the Earth to the sun is just right - you can look up and see the future. White dots pass over the treetops in slow motion, like a string of pearls being towed across the sky. At a speed of 8km per second, the celestial bodies circle the earth every 108 minutes - at an altitude of 550 kilometres.
What could be mistaken for UFOs are actually Starlink satellites. Elon Musk has already launched about 1,000 of them into orbit with his SpaceX rockets. All-in-all 12,000 are planned, with the aim of providing internet access everywhere on Earth. Musk hopes to surround the globe with a dense network of satellites by 2027 and is promising bandwidths of up to 150 Mbit/s, turbo-internet compared to speeds in most German villages. Starlink says it will be offering Germany internet access from space nationwide from mid-2021.
The 235 villagers of Böddensell have been waiting for fast internet for what seems like forever. A few days a week, Telekom technicians can be seen working in the forest. Centimetre by centimetre, they dig their way through the sand along the old flagstone path. Rumour has it an optical fibre connection is coming soon.
Across Germany, around 450,000 households are in the same boat: they have to get by with 10 Mbit/s or slower. Even in Friedrichshain the bandwidth often falls short of enabling a stable Zoom call while your flatmate watches Netflix, even on major streets. In the capital of the world's fourth largest economy, simultaneous streaming and conferencing is impossible for many households. For decades, we've been told that fibre optics are the future, fibre-to-the-home is important! But the transport and digital infrastructure ministry's Broadband Atlas shows only 14 per cent of households in Friedrichshain-Kreuzberg are hooked up to the fibre network despite having a higher population density than Hong Kong.
For 40 years, broadband rollout in Germany has been a story of procrastination. The journalist Sascha Lobo once described it as the "BER of the internet". It's a tragedy in which former chancellor Helmut Kohl and Deutsche Telekom played leading roles as obstructionists. Back in 1981, the SPD-FDP coalition under chancellor Helmut Schmitt developed plans for a nationwide fibre-optic network. A year later, the new chancellor, Helmut Kohl, put those plans on ice and promoted cable TV instead.
In the 1990s, Kohl promised the people in the east of the country “flourishing landscapes”. Soon after, workers tore up the fields and the forest track in Böddensell for the first time. Instead of laying fibre-optic cable, copper cable and cable TV were installed as a supposed infrastructure upgrades. So much for e-government. Finally, Ossis could watch Tutti Frutti, an erotic game show!
What Deutsche Post began at that time was later completed by the privatised Deutsche Telekom. In 1996, the so-called T-Share went public. The new Telekom was a hybrid entity, neither a fully state-owned enterprise nor a proper private corporation. Even today, with more than 30 per cent, the German state is Telekom's largest shareholder - directly and indirectly through the state-owned Kreditanstalt für Wiederaufbau (KfW). Insiders say in retrospect the Telekom board chairmen distributed profits generated by the copper cable infrastructure for too long - under pressure from their majority shareholder. Hence, for years, copper was a cash cow for the German government and the Telekom executives Ron Sommer, Kai-Uwe Ricke and René Obermann who collaborated with it.
This pleased generations of German chancellors, their finance ministers and their state secretaries. Helmut Kohl and Theo Waigel. Gerhard Schröder and Hans Eichel. Angela Merkel and Wolfgang Schäuble, later Olaf Scholz. And for most of these years: Werner Gatzer, a top player in the finance ministry.
The four horsemen of the apocalypse
Instead of investing in optical fibre early enough, Telekom - bound by the interests of its shareholders - kept upgrading its outdated infrastructure with a technology known as "vectoring". Compared to expensive fibre-optic cables, this was a "cost-saving" method that could achieve higher bandwidths with minimal investment, but it was also a short-sighted one. In the long run, only fibre-optic networks can handle the amount of data streams that will soon be required. For two years, the German Broadband Communications Association (BREKO) has been pointing out the need for gigabit networks with speeds of 1,000 Mbit/s - which is expected to be the average bandwidth required by businesses and households by 2025.
Anyone who wants to lay a fibre-optic cable in Germany needs a lot of staying power. German public procurement law is complicated: if the cable is to be laid under a road, it needs the approval of the highway planning authority. If you want to lay a cable through the forest or fields of Böddensell you need permission to use forest and farm roads. The district administrator, mayor, building authorities, road authorities, county and municipality - the number of parties involved in a supposedly simple undertaking seems endless. The new Telecommunications Act (TKG) is supposed to somewhat simplify the processs. In Börde county - where Böddensell is situated - broadband management officer Holger Haupt in Department IV of the Broadband Unit still has his work cut out for him, despite the updated TKG. (www.giganetz-boerde.de - 503 Service Unavailable, editor's note).
It gets really strenuous when broadband expansion is publicly subsidised. Then, alongside the federal government, the EU comes into play. From the preliminary funding decision to the approval procedure and actual installation can take six or seven years.
Many network operators therefore complain, at least internally, about too many subsidies, because in practice subsidies are also granted where they're not needed, resulting in windfall profits. Companies that invest without public subsidies lose out economically. This leads to a delay in broadband expansion even in places where expansion could be economically attractive for network operators.
A special monopolies commission that watches over the federal government's holdings and competition issues has said this creates false incentives. While isolated farms are being connected to fibre-optic networks, broadband expansion in larger municipalities is stagnating. The subsidies are not intended for areas with high demand. The result is exorbitant spending on connecting rural homes. While connecting a household to the fibre-optic network in a city costs a few hundred euros, in a rural area it can cost up to €20,000 per household, complain voices familiar with the matter - and up to 60 per cent of the costs are borne by the taxpayer.
Telekom, as a network operator, is the primary beneficiary of this arrangement. Last year, the group reported profits of €4.1b on a turnover of €101b. At the same time, Telekom has debts worth €130b.
For current Telekom CEO Timotheus Höttges, his CFO Christian Illek and Chief Innovation Officer Claudia Nemat, the system of technical dithering has paid off. And the core business in Germany subsidises acquisitions in the United States.
In fact, for Höttges, the US is probably more attractrive as a growth market than the core business in Germany. After tough negotiations, Telekom bought its competitor Sprint there for a sum in the double-digit billions. "The merger in the United States is a historic milestone for the group," Höttges said after the takeover in 2020. In the next few years, Telekom plans to invest $40b stateside. Fear of the future at home finances expansion on the other side of the Atlantic.
Starlink didn't reinvent satellite internet
Elon Musk, on the other hand, has little to do with telcos and the complicated conditions on the ground. He has long since turned his attention to space. For years, Space X has been launching cargo into orbit for NASA. Today, the company is the world's leader in private satellite launches, with a market share of 65 per cent. SpaceX offers commercial rocket launches for as little as $60 million. A European Ariane launch can cost three times as much.
Even though it hasn't launched officially in Germany, Germans can already sign up for Starlink internet. The Starlink Kit including the antenna costs a one-time fee of €499, plus €59 shipping. The monthly fee is €99, making Starlink significantly more expensive than a regular internet connection.
The idea isn't new. European industry leader Eutelsat already offers customers satellite internet with a bandwidth of 100 Mbit/s - and charges only €69.99 per month.
However, bandwidth isn't the only factor when it comes to the speed of an internet connection. Also important is latency - the time it takes for a data packet to travel from the sender to a server on the internet and back. Because geostationary satellites orbit at an altitude of 36,000km, satellite internet users can experience delays despite high bandwidth since the signal has to travel from earth to the satellite and back again. Latency times of 600 milliseconds are not uncommon, unfavourable conditions for interactive applications like online gaming. But it's also not reliable for factory robots or aircraft.
The far lower orbit of the Starlink satellites (550km) makes for significantly lower latencies. A data packet sent from Earth to a Starlink satellite and back is supposed to take no more than 25 to 35 milliseconds. As soon as a satellite moves out of the user's reception range, another satellite takes over. Thus, Starlink combines the advantages of low orbit with the reliability of geostationary satellites.
In future, the satellites will even be able to communicate with each other. Satellite internet should be available in regions that currently have no open internet access for political or economic reasons. The internet from space could result in a whole new level of globalisation. Neither the jurisdiction of nation states nor technical or geographical limitations could hinder access.
Musk: we're not competing with cable internet
Elon Musk emphasises that Starlink is not competing with classic network operators in urban areas. Nonetheless, satellite internet isn't just some niche product. Musk expects Starlink to have an annual turnover of $30b as early as 2025. Not bad for an initial investment of $10b.
Musk need not fear competition - at least not in the short term. The British company OneWeb is the only competitor that has launched satellites into space on the same scale as Starlink . However, after the stock market crash of 2020, the company was on the verge of insolvency and had to be rescued by the British government. The plan to provide the world with internet from space remains highly speculative, even for Starlink: "Every new satellite constellation in history has gone bankrupt. We hope to be the first not to," Musk commented on Twitter in February.
The space junk problem
The greatest danger for Starlink, however, lurks in its own success. If it actually succeeds in putting 12,000 satellites into orbit by 2027, that will be more than ever before. According to the European Space Agency (ESA), over 9,000 satellites have enterd Earth orbit in the last 60 years, of which about 2,000 are still active. Another 3,000 are drifting aimlessly. The littering of space is a serious problem: the smallest pieces of debris pose a great danger to satellites - they can pierce their outer skin like bullets and, in the worst case, cause a cascade of further collisions. Many satellites carry extra fuel just to be able to dodge dangerous situations.
One particularly gloomy vision of the future that astronomers are concerned about is the Kessler syndrome, first proposed by astrophysicist Donald Kessler. Kessler hypothesised that space pollution could reach a tipping point: A single collision could produce so much debris that it would trigger a chain reaction of collisions. In this scenario, hundreds of satellites would be destroyed for good, creating a ring of debris that would make launching new satellites impossible. For this reason alone, the OECD estimates that the cost of satellite launches will go up by 10 per cent in the future.
Even if it succeeds in having its own satellites burn up in an orderly manner at the end of their lifespan in orbit, Starlink is dependent on its competitors catching up. International standards governing commercial use, especially of low orbit, are crucial to Starlink's success.
Internet from space sounds like the future, but it's still a marginal phenomenon. Beyond scenarios such as the "Kessler syndrome", perhaps the real question is whether it's wise to leave an issue as critical as internet infrastructure to Elon Musk instead of developing one's own ideas. Critical infrastructure - power grids, water supply, and now internet - have always been core tasks of the state.
If we want to understand why the rollout is progressing so slowly in Germany and what might have to happen to ultimately succeed in establishing network infrastructures that are independent of Musk's satellites, it's worth thinking about the German fibre-optic dilemma beyond the culpability of individual market players . How we get our internet is ultimately a question of digital sovereignty and shouldn't fail because of false incentives from a former state-owned company or the German administrative jungle.
The satellite alternative, according to the government and private sector companies alike, is a viable option for providing access to rural areas, but it cannot replace broadband expansion on the ground.
Recently, the companies involved in broadband expansion have been showing more support for new technologies and a desire to significantly accelerate fibre-optic rollout. The will is there. And so is the willingness to invest. But too many technical norms and too much bureaucracy is still hampering the process.
In the meantime, users are sharing their experiences with Starlink internet on Reddit. Internet access is still often intermittent and even slower than advertised, some users say. Other customers are satisfied. Some German users, however, are plagued by more earthly problems: Their antennas are said to be stuck in the DHL warehouse in Leipzig.
Disclaimer: None of the writers involved holds shares in any companies named in the article.