EP74 — Unicorn Files— OPEN COSMOS L’Europa può costruire una rete spaziale sovrana senza copiare Starlink?

Open Cosmos bets Europe's sovereignty on a shared satellite network

Open Cosmos, founded in 2015 by Rafael Jordà Siquier, started out building satellites to order and is now betting it can become Europe's sovereign space data network. The wager is that governments will pay into a shared constellation one satellite at a time, and that sovereignty means specialised capacity and control of the data path rather than Starlink's scale. The constraint that decides the outcome is regulatory, since the frequencies the network needs can be lost over delays the company does not control. Its first network satellites flew on Rocket Lab, so the independence on offer is not yet complete.

Takeaways

• Open Orbit made satellite missions turnkey. Coordinating a mission was infernal work, involving solar-panel suppliers, sensor makers, thermal testing and launch negotiations. Open Orbit sells the whole chain from initial engineering to in-orbit operations, with flight heritage from QB50 in 2017 and PhiSat-2 for the European Space Agency in 2024.
• Onboard AI turns satellites into edge computers. PhiSat-2 runs AI on the satellite itself. It inspects each image in orbit, notices when it contains only clouds and deletes it, so only useful data is downlinked instead of clogging the ground segment.
• One satellite buys the whole network. Governments pay for a single satellite and gain access to every other member's data, catalogued through Data Cosmos. Security is handled by tasking priority, since the payer keeps absolute priority over its own areas of interest and only excess capacity is shared.
• Sovereignty is not the same as scale. Against OneWeb's 600-plus satellites, the EU's IRIS at almost 350 and Starlink's 10,000, Open Cosmos argues institutions are buying resilience and cryptographic certainty that data never transits third-party servers.
• Radio spectrum, not orbit, is the bottleneck. On 5 November 2026 the ITU, acting on filings made through Liechtenstein, extended filing 3-1 to 2027 but rejected 3ECOM-3, apparently after a failed launch. Under use-it-or-lose-it, losing frequencies kills the roadmap and the ability to repay the €300m.
• The €300m round hides its structure. The September 2026 raise and a reported valuation near $2bn rest on the network being built. The split between equity and debt is not transparent, and lenders financing the capex want certain cash flows rather than press coverage.
• Europe's launcher dependence undercuts the pitch. The company promising European sovereignty launched its first network satellites on Rocket Lab. Ottavi argues total strategic independence remains half an illusion until Europe's launcher crisis is resolved, because Europe depends on others for material access to space.