Space Economy Weekly — Mar 09, 2026

NASA broke up with Boeing's moon rocket upper stage and handed the job to hardware that's already flying — the clearest signal yet that "flight-proven" is now procurement doctrine, not just a SpaceX talking point. Meanwhile, Artemis 2 rolled off the pad again, Spain's PLD Space closed a round big enough to make European launch a real conversation, and the FAA's legacy launch licensing framework transitioned today, meaning operators who have not completed the formal Part 450 transition are not authorized to conduct launches under the old rules as of March 9, 2026.

If you've been wondering how serious NASA's new leadership is about cutting legacy cost and schedule risk out of Artemis, this week answered that question.

On March 6, NASA published a sole-source procurement filing formally determining that ULA's Centaur 5 upper stage — the one already flying on Vulcan — is the only viable replacement for Boeing's Exploration Upper Stage (EUS), the chronically late and over-budget piece that was supposed to power SLS from Artemis 4 onward. The filing covers Centaur stages for Artemis 4 and 5, plus a flight spare. The estimated cost was redacted.

NASA leadership had already announced in late February that EUS work would stop entirely. Centaur 5 has flown on multiple Vulcan missions since 2024, using RL-10 engines and liquid hydrogen/oxygen — proven hardware doing a proven job. Under current law, NASA is technically still required to develop the Block 1B SLS with the EUS. But a NASA authorization bill approved by the Senate Committee on Commerce, Science, and Transportation (full committee) on March 4 would let the agency replace the EUS if it is "unlikely to achieve mission goals" — legislative cover coincident with a procurement decision.

The business implication is blunt: Boeing loses a major production contract, ULA gains a guaranteed second customer for hardware it was already building, and suppliers down the lunar logistics chain may see this as a signal that tying into flight-proven architectures could be the fastest path to a contract. That's a doctrinal shift, not just a vendor swap.

The most talked-about rocket in the world can't stay on its pad.

Engineers hit problems repressurizing the SLS upper stage's helium tanks overnight, forcing yet another rollback to the Vehicle Assembly Building. The mission — humanity's first crewed flight beyond low Earth orbit since Apollo 17 — slipped from a March window to at least early April on the schedule. NASA leadership publicly ruled out a March window. This was the fourth delay (after February 5, February 8, and March 6), each for different reasons.

The commercial read isn't the delay itself — it's the pattern. Every rollback generates fresh scrutiny of SLS launch cadence assumptions, which directly affects the business case for lunar logistics companies like Intuitive Machines and Astrobotic that are sizing operations around an Artemis flight tempo that keeps slipping. Two rollouts and three pad days to not launch a rocket — that range utilization pressure matters for the entire Kennedy Space Center manifest.

European small launch has been a "promising but not yet" story for years. This week's news from Valencia suggests the "yet" may be arriving.

PLD Space raised what press reports variously describe as $209 million or €180 million in a Series C led by Mitsubishi Electric with significant Spanish government backing. The capital is for scaling production of Miura 5, an orbital small-lift vehicle targeting 300–500 kg payloads to LEO — the same bracket as Rocket Lab's Electron. PLD flew the suborbital Miura 1 in 2023, making it the first private Spanish rocket to reach space.

For context, Rocket Lab raised roughly $140M in its 2020 SPAC round. A $200M+ pre-operational raise signals that institutional investors — including a major Japanese industrial partner — are genuinely betting on European sovereign launch capacity. Europe's political appetite for not depending solely on SpaceX is a real demand signal, and PLD is positioning as the mid-market answer.

Adding urgency: Ariane 6's April manifest slipped slightly this week (early March 2026), creating a temporary bottleneck for European EO and government payloads — exactly the window PLD hopes to fill. The question is execution: Miura 5 still needs development completion, vehicle certification, and a customer manifest. That's 18–36 months of risk. Watch for a launch site announcement at El Arenosillo in southern Spain, which would be the next concrete milestone.

While most LEO broadband coverage focuses on Starlink vs. Kuiper, a quieter but strategically significant move happened in Europe.

Pan-European provider Open Cosmos announced ConnectedCosmos, a LEO constellation designed as a sovereign connectivity backbone for European governments and companies. The system will use optical inter-satellite links — laser crosslinks that let satellites pass data without routing through ground stations — to create what the company calls a "gatewayless sovereign space mesh."

The sovereignty framing is the story. Post-Ukraine, European governments became acutely aware that connectivity infrastructure routed through non-European ground stations is a national security liability. ConnectedCosmos isn't positioning as a cheaper Starlink — it's a system where European governments retain end-to-end data control.

Open Cosmos also secured a €60M contract to build Greece's first Earth observation constellation: seven satellites with multispectral and hyperspectral cameras for marine pollution monitoring and precision agriculture. This is early-stage in terms of flight hardware, but the contract pipeline is real. Watch for ESA co-funding announcements, which would significantly de-risk the build schedule.

The headline is the satellite count. The real story is what the booster reuse numbers reveal about operational tempo.

The Starlink 10-40 mission added 29 more broadband satellites to LEO, marking the 600th Starlink satellite launched in 2026 — nearly 10 per day across 62 calendar days. Booster B1080 flew its 25th mission. The constellation now exceeds 9,900 spacecraft, with the 10,000-satellite threshold likely within weeks.

That number matters beyond bragging rights. At 10,000+ active satellites, Starlink alone represents a dominant fraction of all tracked objects in certain orbital shells, which amplifies both the debris risk debate and spectrum coordination pressure on every other non-geostationary constellation operator. Meanwhile, civil-society groups filed supplemental materials with the FCC opposing SpaceX's proposed "orbital data center" mega-constellation — a sign that organized, technically sophisticated opposition to large filings is maturing fast.

Separately, SpaceX launched a ~500-satellite batch on March 7 targeting a Southeast Asian regional shell, and a Vandenberg flight highlighted direct-to-cell V2 Mini hardware moving closer to field trials. The production machine isn't just big — it's layering new services atop the same reuse fleet.