The Lyceum: Space Economy Weekly — Mar 20, 2026

The most concrete human spaceflight mission in a generation is sitting on its pad, four astronauts are in quarantine, and everything from Artemis III onward is still being redesigned in real time. Meanwhile, the week's quieter moves tell a sharper story: Canada canceled its lunar rover to fund surveillance satellites, Europe locked in a €248 million Arctic weather constellation, and the FCC is becoming the arena where the entire direct-to-device market gets litigated into shape. The theme isn't launches — it's triage. Governments and agencies are making hard, visible choices about what they can actually afford to build, and those choices are revealing where the real money flows.

• Starlink Group 17-24 (SpaceX): 25 Starlink satellites deployed from Vandenberg SLC-4E on March 17, pushing the active constellation past 10,000 satellites for the first time.
• Starlink Group 17-30 (SpaceX): 29 more Starlink satellites from Cape Canaveral SLC-40 on March 18 — SpaceX's 30th Starlink mission of 2026.
• Synspective Strix-8 SAR satellite (Rocket Lab / Synspective): Eighth operational synthetic aperture radar satellite deployed to LEO from Māhia, New Zealand on March 20.
• Yuxing-3 06 (Hunan University of Science and Technology): Technology demonstration smallsat launched to sun-synchronous orbit on March 16.
• Progress MS-33 (Roscosmos): Three-ton ISS resupply cargo launched from repaired Baikonur Site 31 — first use of that pad since November damage.

Four astronauts are in quarantine. The SLS rocket began rolling to Pad 39B at 12:20 a.m. EDT on March 20 — the second time this stack has made the four-mile crawl from the Vehicle Assembly Building, and the one that's supposed to stick. NASA returned the rocket in February after discovering a helium flow problem on the upper stage during a fueling test. Engineers refreshed flight termination batteries, replaced batteries on the upper stage, core stage, and solid rocket boosters, and swapped a seal on the core stage liquid oxygen feed line while the vehicle was inside. The launch window opens no earlier than April 1.

Here's the context that makes this more than a rollout story. NASA officials recently announced that the first Moon landing has been postponed from Artemis III to Artemis IV, and that Artemis III will instead become an Earth-orbit demonstration of Orion docking with either SpaceX's Starship lunar lander or Blue Origin's Blue Moon Mk.2. NASA said the public would get greater clarity about Artemis III specifics within 60 to 90 days — meaning international partners like ESA are about to hear the new architecture formally explained for the first time.

What changes if this succeeds: Artemis II flying cleanly in April validates the SLS-Orion stack for crewed deep-space flight and gives NASA political cover for the restructured program. What failure looks like: A wet dress rehearsal anomaly or launch scrub that pushes past April creates scheduling pressure on pad availability at Kennedy and hands critics more ammunition against SLS's cost-per-flight. Watch the rehearsal data in the next ten days — that's the real gate.

Ten thousand active satellites isn't an arbitrary milestone. It's the point where the network stops being a broadband constellation and starts being communications infrastructure that governments, militaries, and carriers cannot easily route around.

SpaceX crossed the threshold following a March 16 launch from Vandenberg, with another 29 satellites added from Cape Canaveral two days later. Independent trackers confirm roughly 54 satellites across two missions this week alone.

But the real story is what comes next. Starship V3 is expected to lift 100 metric tons to orbit, unlocking a V3 Starlink satellite that industry analysts describe as carrying a terabit of capacity — "just far and away more than any other low Earth orbit satellite out there." Quilty Space forecasts as many as eight Starship launches with Starlink satellites during 2026. Meanwhile, regulatory pressure is rising: the FCC's latest Starlink authorization includes explicit conditions on post-mission disposal, and the American Astronomical Society has filed a formal petition arguing that per-satellite casualty risk standards don't scale to tens of thousands of reentries.

The current 10,000-satellite network is the last Starlink generation competitors have any realistic chance of matching. V3 Starlink on V3 Starship is a different class of problem entirely. If the FCC starts attaching constellation-level casualty risk caps, that changes satellite design and insurance economics for everyone — not just SpaceX.

One week after Canada announced $337 million for a sovereign launch site and small launch companies, its space agency canceled the only rover it had been building for the Moon. The contrast appears deliberate.

The Canadian Space Agency's 2026–27 Departmental Plan officially terminated the Lunar Rover Mission — a blow to developer Canadensys Aerospace, which had the mission approved in 2022 and manifested to launch with Firefly in 2029. Canadensys had assembled a large team of subcontractors and built a Canada-wide science team. Budget cuts begin at $6.66 million in 2026–27, scaling to $14.36 million by 2028–29.

What Canada is keeping tells the real story. The CSA is advancing a new satellite for the RADARSAT Constellation Mission — SAR infrastructure handling roughly 300,000 imagery requests annually for the government and military — while pushing forward on WildFireSat, targeting a 2029 launch with an estimated economic ROI of up to $5 billion over five years. The agency is also launching QEYSSat, a quantum key distribution satellite, in 2026–27.

Canada is choosing surveillance, wildfire intelligence, and quantum-secure comms over exploration prestige. If the pattern holds, expect other mid-tier space agencies to make similar triage calls as budgets tighten. The signal to watch: whether Canadensys's institutional capacity finds a new home or simply dissipates — that team doesn't reassemble easily.

Northern shipping routes and offshore energy operators just got a weather lifeline with a contract number attached to it.

EUMETSAT signed a €248 million contract with OHB Sweden to build EPS-Sterna — a constellation of 20 microsatellites carrying microwave instruments designed to fill critical observation gaps over the Arctic. The polar region is notoriously underserved by existing weather satellites, most of which are optimized for mid-latitude coverage. As Arctic sea ice retreats and commercial shipping through the Northern Sea Route increases, the gap between what operators need and what forecasters can deliver has become an operational risk, not just a scientific curiosity. Deployment is targeted for 2029.

What changes if this works: insurance and route-planning firms get storm and sea-ice forecasts accurate enough to reshape commercial shipping decisions in real time. Arctic offshore operations — oil, gas, wind — get the same forecast quality that mid-latitude operations take for granted. What failure looks like: microsatellite constellations have struggled with calibration and cross-platform consistency before; if EPS-Sterna's microwave instruments can't match the precision of larger heritage instruments, the data may not meet operational thresholds. Watch whether EUMETSAT publishes validation benchmarks against existing polar assets within the first year of operations.

The fight over who gets to beam signals directly to your phone from space just escalated from regulatory maneuvering to open combat.

Ligado Networks — the company that spent a decade fighting the GPS lobby over its L-band spectrum rights — filed to deploy a payload called SkyTerra Next on AST SpaceMobile's LEO constellation. In plain terms: Ligado wants to use AST's satellites to exercise its spectrum license, bypassing the need to build its own birds. The pushback was immediate. Iridium filed a petition asking the FCC to deny the request, accusing Ligado of spectrum hoarding and arguing the arrangement effectively transfers license control to AST SpaceMobile. The GPS Innovation Alliance piled on, arguing the proposal raises substantial questions about harmful interference to GPS services.

Zoom out and the picture gets bigger. The direct-to-device (D2D) market — satellite signals reaching unmodified phones without a cell tower — now has AST SpaceMobile, SpaceX (via its T-Mobile partnership), Ligado, and Globalstar (backed by Apple with a planned 54-satellite constellation) all contesting the same regulatory space simultaneously. Under current FCC leadership, the commission has moved toward a streamlined "assembly line" licensing approach for mega-constellations.

The FCC is effectively becoming the arena where D2D market structure gets decided. If Ligado wins, AST gets L-band capacity in the U.S. without filing for it directly. If it loses, Ligado's spectrum position — valued at billions — may be stranded. The observable signal: watch whether the FCC treats this as a routine modification or triggers a full interference review. That procedural choice will tell you which way the wind is blowing for every D2D player.

A rocket that's been to the pad twice and back sitting under Florida rain while four astronauts eat quarantine meals; Canada writing a $337 million check for a spaceport with one hand while canceling its only Moon rover with the other; and four companies simultaneously suing each other at the FCC over who gets to beam L-band signals to your phone from space.

The direct-to-device market is about to learn what the airline industry learned decades ago: the real competition isn't in the sky — it's in the docket.

Clear, satisfying skies —

If someone you know tracks orbits, spectrum fights, or budget triage for a living, forward this their way.

The FTC made AI hiring bias a federal enforcement target this week — relevant for any team using automated screening in recruiting. Read → The FTC Just Made AI Hiring Bias a Federal Enforcement Target