SDA admits only 17% of its orbital constellation works

The Space Development Agency has 42 satellites in orbit and no working mesh network to show for it. Of those 42, only about 17 percent are what the agency considers operational — meaning they have reached their correct orbital positions and are communicating as expected. The rest are in orbit but not yet functioning in their intended role. That gap between what is launched and what is working is the central problem facing SDA's Proliferated Warfighter Space Architecture as of late March 2026.

Acting director Gurpartap "GP" Sandhoo said that figure at the Satellite 2026 conference March 23, using "operational" in a narrow sense that he acknowledged is not obvious from the outside. "We use the term operational, but that means 17 percent," he said. The distinction matters: 42 satellites have been launched, but far fewer are currently doing what they were sent to do.

The next launch of Tranche 1 satellites will not occur until at least May or June, marking a gap of roughly seven months since the last launch in mid-October 2025. The pause is deliberate, Sandhoo said — not a scheduling accident, but a program that stopped to fix what it found in orbit. "We saw a handful of things, and we kind of stopped to make sure we fix those for the next set of launches," according to SpaceNews.

The mesh network is the clearest example. All PWSA satellites carry optical intersatellite terminals supplied by Mynaric, Tesat Spacecom, Skyloom, and CACI. Ground testing showed those terminals work. In orbit, the network they were supposed to form does not yet exist. "We have not established the mesh network for Tranche 1 yet," Sandhoo said, a Aerospace America report found.

Tranche 1 is structured across 10 orbital planes: 126 Transport Layer satellites from Lockheed Martin, York Space Systems, and Northrop Grumman, plus 28 Tracking Layer satellites from L3Harris and Northrop Grumman. Those 154 spacecraft are the first operational layer of a distributed missile warning and data relay network — the hardware pipeline feeding into the broader Golden Dome architecture, which the FY26 defense bill funds to the tune of roughly $13.8 billion for space and missile defense programs. The SDA has awarded $3.5 billion in contracts to Lockheed Martin, Rocket Lab, Northrop Grumman, and L3Harris for 72 Tranche 3 Tracking Layer satellites, launching in fiscal year 2029.

The current tranche is where theory meets hardware. Orbit raising for the Lockheed Martin satellites is running about three months behind the original schedule, Sandhoo said. One Lockheed satellite remains uncontactable after launch; the company says 95 percent of its satellites are healthy, and the architecture is designed to close orbital planes with 16 or 17 satellites rather than the full 18, accounting for some expected loss. Northrop Grumman has a software issue that surfaced just before its satellites were slated to ship to the launch site, delaying those spacecraft pending resolution.

Once satellites reach their operational orbits, payload testing moves to classified ground stations in Huntsville, Alabama, and Grand Forks, North Dakota, where SDA evaluates integration with military systems — a handoff that adds coordination complexity in a multi-vendor architecture. York Space Systems has been demonstrating Link 16 connectivity to assets in space and air, which the agency counts as progress toward integrating the transport layer with combat networks.

A January 2026 GAO report on the PWSA found SDA had overestimated technical readiness, carried unplanned contractor work, and lacked an architecture-level schedule or reliable cost estimate for a program projected at $35 billion through fiscal year 2029. DoD concurred with five of six GAO recommendations.

The slowdown reflects a tension built into SDA's founding logic. The agency was created in 2019 to move faster than traditional defense acquisition — incremental tranches, frequent launches, rapid deployment. The theory was that getting hardware on orbit quickly mattered more than getting it perfect. Tranche 0, launched in 2023 and 2024, included 27 satellites demonstrating laser communications capabilities and served as a proof-of-concept for the proliferated architecture. Tranche 1 was supposed to launch on a near-monthly cadence. The pace is now set by something slower: the work required to turn launched satellites into a functioning network.

"We plan to start launching again towards the May, June timeframe," Sandhoo told Defense Daily. Whether the fixes hold, and whether the mesh network eventually establishes, will determine whether the next 112 satellites deploy on a revised timeline — or whether this pause becomes a feature of how proliferated constellations actually work.