Throughline Desk · May 2, 2026

The Lyceum: Space Economy Weekly — May 02, 2026

Week of May 2, 2026

The Big Picture

The gap between ambition and operational reality is widening for everyone except SpaceX. Falcon Heavy returned after 18 months to finish Viasat's decade-long global broadband bet; Amazon Leo crossed 300 satellites this week and still needs roughly 1,300 more in 90 days; Blue Origin sits grounded after putting the BlueBird 7 mobile broadband satellite into the wrong orbit. The quieter story underneath: the FCC quietly killed a 1990s power cap that has been throttling LEO broadband for years — and industry observers say the bottleneck on national security space is shifting from rockets to ground software and payload processing.

What Just Shipped

ViaSat-3 F3 (Viasat / Boeing): 6-ton GEO broadband satellite launched April 29 on Falcon Heavy, completing Viasat's three-satellite global constellation with Asia-Pacific coverage adding more than 1 Tbps to the network.
Amazon Leo batch (Ariane 64) (Amazon): 32 broadband satellites deployed April 30 from French Guiana on the second-ever Ariane 64 flight — Amazon Leo's first heavy-lift mission with Arianespace.
GPS III SV-10 (Lockheed Martin / U.S. Space Force): The tenth and final GPS III satellite, launched April 21 on a Falcon 9 from Cape Canaveral after a mission reassignment from Vulcan to protect schedule.
H3 F6 launch window opened (JAXA): Reserved June 10–30 launch window announced April 24 for the H3-30 test configuration carrying VEP-5 and six microsatellites — Japan's return-to-flight after December's second-stage anomaly.
Amazon Leo enterprise beta (Amazon): Constellation reached 300+ active satellites following the dual-continent launch week, formally the third-largest constellation in orbit.

This Week's Stories

Falcon Heavy Returns — and Viasat's Decade-Long Global Bet Is Finally Complete

A triple-core Falcon Heavy lifted off from Kennedy Space Center on April 29, sending the 6-ton ViaSat-3 F3 toward geostationary transfer orbit. It was the rocket's first flight in 18 months — the last was NASA's Europa Clipper in October 2024 — and it completes the third and final satellite in Viasat's globe-spanning Ka-band broadband network.

The program's history is a cautionary tale in GEO risk. The first ViaSat-3, launched in May 2023, suffered a partial antenna deployment failure and reportedly delivered roughly 10% of its design capacity on that mission, per CBS. The second satellite launched successfully last November. Now the third is healthy and headed to 155.58° East to anchor Asia-Pacific coverage. Viasat's VP of Satellite Systems told Spaceflight Now that airline customers in the region are "really anxious to get this capacity online."

What changes if it works: Viasat finally has the global, low-latency-for-GEO footprint it has been selling to airlines for a decade — more than 1 Tbps of additional capacity targeting in-flight connectivity over the Pacific, exactly where Starlink Aviation has been racking up airline wins. What failure looks like: roughly two months of electric orbit-raising still lie ahead before F3 reaches its operational slot, and any anomaly during that climb — see F1's antenna — would gut Viasat's APAC monetization story. The signal to watch: airline contract announcements over the next two quarters. If they don't come, the LEO-vs-GEO broadband question is functionally settled over the Pacific.

Amazon Leo Hits 300 Satellites — and Needs 1,300 More in 90 Days

The most consequential deadline in commercial broadband isn't a product launch — it's a regulatory one, and Amazon is running out of runway.

Amazon Leo (formerly Project Kuiper) crossed 300 active satellites this week after two launches on two continents in seven days: an Atlas V from Cape Canaveral, then 32 satellites on the second-ever Ariane 64 from French Guiana on April 30. Amazon is now operating the third-largest constellation in orbit — and is still roughly 1,300 satellites short of the FCC's July 30, 2026 milestone of 1,618 operational satellites. The company has filed for an extension, per coverage in Gagadget.

What changes if the FCC grants a clean extension: Amazon gets to keep scaling on its own timeline while the enterprise beta — launched April 8, targeting Verizon, Vodafone, and maritime/aviation customers rather than households — matures into revenue. What failure looks like: an extension with conditions (deployment milestones, spectrum givebacks, coverage mandates) that hands Starlink another 12–18 months of clear runway in the enterprise tier. The signal to watch: not just whether the FCC grants the extension, but the conditions attached. The terms become the template for every other constellation operator with milestone obligations — and there are many.

The FCC Kills the 1990s Speed Limit on LEO Broadband

On April 30, the FCC voted to replace its Equivalent Power Flux Density rules — late-1990s caps on how much power LEO satellites can beam toward Earth, originally written to protect GEO operators from interference back when LEO broadband was theoretical.

In practice, EPFD has been a throttle on every modern LEO constellation. Removing it lets satellites transmit at higher power, which translates directly to faster speeds and better service in marginal-geometry environments — exactly the maritime and aviation cases where Starlink, Amazon Leo, and OneWeb compete hardest.

What changes: Starlink's existing 10,000-plus satellites get an immediate performance bump without new hardware; Amazon Leo's still-deploying satellites can perform closer to design spec from day one. What failure looks like: the GEO incumbents — who lobbied hard to preserve the old caps — push the fight to the ITU, where international coordination filings could re-impose constraints by another route. The signal to watch is the next round of ITU coordination disputes. If GEO operators start filing aggressive interference complaints, the regulatory war just moved venues.

SpaceX Sweeps Another Round of Space Force Launch Orders — and the Concentration Risk Is Now Structural

The Space Force awarded SpaceX nine NSSL Phase 3 Lane 1 task orders worth $739 million combined, covering missions for the Space Development Agency and the National Reconnaissance Office. Per GovCon Wire, the SDA-2 task order alone covers two launches of 18 Tranche 2 Tracking Layer satellites built by L3Harris and one launch of eight Fire-control On-Orbit Support to the Warfighter (F2) vehicles built by Boeing's Millennium Space Systems. First launches begin in Q4 of FY26.

What changes: SpaceX captures the bulk of FY26 national security launch revenue while ULA's Vulcan ramps and Blue Origin's New Glenn remains uncertified for these missions, per SpaceNews. What failure looks like — and this is the part the Pentagon is openly worried about — is exactly the present situation: a one-provider-deep manifest for the country's most sensitive payloads. The signal to watch is twofold: New Glenn's grounding investigation timeline, and the GAO-flagged payload processing capacity at the Cape. Per Congress.gov's defense primer, Space Force officials describe payload processing as "the greatest challenge facing DOD's space launch efforts." The rocket isn't the bottleneck anymore. The integration facility is.

New Glenn Is Grounded — and Its Customer's Satellite Is Being Deorbited

Blue Origin's third New Glenn flight on April 20 looked, from the outside, like a milestone: first reuse of a recovered booster, clean liftoff. Then the upper stage underperformed. per Ars Technica, the BlueBird 7 mobile broadband satellite was deployed into the wrong orbit and will be deorbited rather than salvaged. New Glenn is grounded pending investigation.

What changes if Blue Origin closes out the investigation quickly: the company stays on its certification path for NSSL Phase 3 Lane 2 — two flights down, two to go, per Air & Space Forces Magazine — and the Space Force gets the third heavy-lift provider it has been engineering toward since 2020. What failure looks like is an investigation that drags into Q3, which would push New Glenn's certification window past the FY27 task-order cycle and lock in the SpaceX-ULA duopoly for another year. The signal to watch: the public root-cause readout. If Blue Origin can't name the failure mode by July, certification timelines slip with it.

⚡ What Most People Missed

The SDA's Tranche 1 satellites are in orbit but troops still can't use them: Per Payload Space, six months after launch, SDA Acting Director GP Sandhoo said warfighters still aren't getting the better comms because of "small issues" that forced a "strategic pause." For an agency whose entire pitch was speed-to-capability, the gap between hardware in orbit and capability in the field is the real story — and almost no one is covering it.
A second Starlink "internal energetic source" anomaly in five weeks: Aviation Week reports another Starlink contact loss, with LeoLabs noting the cause was "remarkably similar" to the December 17 Starlink 35956 event. Per Scientific American, Jonathan McDowell warned that if the fragmentation arose from a design flaw, "the risks go up, a lot." SpaceX says software mitigations are deploying — translation: it may not be hardware. Independent root-cause confirmation has not appeared.
Feds move to start charging launch companies for federal range use: Per Gizmodo, the FAA and DOT are moving toward a fee structure that would end decades of implicit federal subsidy for commercial range operations at facilities including Cape Canaveral. High-cadence operators amortize the cost; small launchers feel it disproportionately. Watch for a Federal Register notice to confirm this is rulemaking, not a trial balloon.
NASA quietly raised the CLPS contract ceiling from $2.6B to $4.2B: A Johnson Space Center procurement notice reported by GovCon Wire on April 29 lifts the Commercial Lunar Payload Services ceiling by $1.6 billion. That's not new task orders — but it's NASA telling its contracting shop to expect to buy substantially more lunar delivery than the old cap allowed. The pipeline is being widened before the missions are announced.
Space Force's Andromeda program brought 14 vendors into a $1.8B SSA buying lane: Per DefenseScoop, the 10-year Andromeda vehicle pulls traditional primes and newer space companies into the same procurement, with the first order tied to RG-XX, the GSSAP successor. That's less a contract and more a market template — the government wants optionality on commercial space situational awareness before it picks an architecture.
Eastern Range supported five different rocket types in April — a 60-year record: Space Launch Delta 45's announcement reads like a win, but vehicle diversity stresses safety reviews, airspace coordination, and pad turnaround in ways that show up later as scheduling capacity limits. The record is also a warning.

📅 What to Watch

If the FCC grants Amazon Leo a clean deadline extension with no conditions, other operators will cite that precedent to seek similar leniency, weakening the FCC's leverage to enforce milestone-based deployment and potentially delaying consolidation and investment decisions across the industry.
If JAXA's H3 F6 flies cleanly in the June 10–30 window, MICHIBIKI No.7 reschedules and Japan's positioning constellation timeline unfreezes; if it slips, every payload riding H3 — including Mars-related campaigns — moves with it.
If Blue Origin's New Glenn investigation extends past summer, the company misses the FY27 task-order eligibility window and the Space Force's three-provider strategy stays theoretical for another full year.
If a Federal Register notice appears on launch range fees, expect immediate lobbying from small launch operators on a marginal-cost-per-flight basis — and watch whether the rule carves out cadence-based discounts that effectively favor SpaceX.
If a second Starlink satellite of the same generation suffers the same internal energetic anomaly, the conversation shifts from "anomaly" to "fleet-wide design issue," and insurers reprice the entire LEO broadband sector.
If Eutelsat's MaiaSpace contract converts into a firm flight manifest before MaiaSpace's first orbital launch, Europe just funded a reusable-launcher program through commercial demand rather than government subsidy — a structural shift in how European launch gets capitalized.

The Closer

A 6-ton satellite finally finishes a constellation a decade after groundbreaking; a reused booster lands cleanly while its customer's satellite is towed to the deorbit queue; a 1990s spectrum cap dies of old age while regulators wave through 1,618-satellite extension requests. Somewhere in low Earth orbit, a second Starlink is venting from an "internal energetic source" that SpaceX is patching with software — which is one way to describe a hardware problem you really hope is a software problem.

Stay sharp out there.

Forward this to the satellite operator in your life who's been quietly refreshing the FCC docket page.

From the Lyceum

FERC published its data center power rule and the litigation started before the ink dried — relevant if you're financing ground stations or edge compute. Read → FERC's Deadline Is Today