The Space Development Agency wanted cheap satellite buses. What it got was a lesson in what "commodity" actually means in a supply chain that does not have enough suppliers.
In 2024, the first deployment of PWSA satellites four different manufacturers built: L3Harris, Lockheed Martin, York Space, and SpaceX. Every bus had problems. Maj. Gen. Stephen Purdy, the SDA's director, said it plainly at a recent conference: "In the first deployment of PWSA satellites in 2024, the biggest challenge we had was the buses, which were supposed to be a commodity, and none of them were. They all had challenges. Every single one of them." according to SpaceNews.
That is not a budget problem. That is not a design problem. That is a hardware problem, and it is representative of a structural constraint the US space industrial base is running into right now — not somewhere over the horizon, but today, on programs already in production.
The numbers behind the story: US launch activity has grown nearly tenfold in six years, going from roughly 340 objects in 2019 to more than 3,400 in 2025. The industrial base that supplies the components inside those vehicles has not kept pace. A study released this spring by the Aerospace Industries Association and PwC identified nine specific bottleneck components where domestic supplier capacity is three qualified sources or fewer. Optical inter-satellite links. Actuators. Rocket motor nozzles. Composite overwrapped pressure vessels. Valves. FPGAs. Connectors. Switchgears and transformers. Post-processing. For some of these, there is effectively one qualified domestic supplier.
A Breaking Defense report found this capacity gap for specialized components gnarls the supply chain.
The supply chain is not abstract. Purdy called it what it is: a domain of war. "The supply chain is a domain of war," he said. That framing — putting logistics alongside sensors and weapons as a theater of competition — reflects a recognition that the US learned painfully in other programs and is now re-learning in space.
Some of the pressure is self-inflicted. Xenon, used as a propellant in electric propulsion systems, saw supply disruption after Russia and Ukraine together accounted for a large share of global output. Russia invaded Ukraine in 2022, and the propellant market restructured. For a defense program betting on electric propulsion at scale, that is a materials risk, not just a schedule risk.
The $613 billion global space economy in 2024 is real, and growing. But the infrastructure behind it is aging. Aerospace industrial facilities now average nearly 26 years in age. You cannot surge a supply chain that has consolidation at the component level, an aging manufacturing base, and demand that has outrun planning assumptions by a factor of ten.
Congress is starting to respond. The NDAA for fiscal 2027 is shaping up to focus explicitly on industrial base and supplier base growth, with a recognition that the foundational framework for any surge capacity is the supplier layer underneath the prime contractors. Wittman said the committee intends to make growing that base a priority.
The SDA's answer to bus problems has been to work with the suppliers it has, iterate, and push forward. That is the right operational call. But the structural fix — actually developing multiple qualified suppliers for optical inter-satellite links, for actuators, for COPVs — takes years, not quarters. The PWSA is projected to cost nearly $35 billion through fiscal 2029. You cannot budget your way around the physics of building a supplier base from scratch.
Voyager Space has been building toward this with acquisitions: ExoTerra Resource for $93 million, Estes Energetics for $64 million, ElectroMagnetic Systems for $33 million, Optical Physics for $9.9 million. These are bets that vertical integration inside a satellite builder solves the supply concentration problem. Whether that works depends on whether the suppliers being absorbed were the bottleneck or merely another link in a fragile chain.
The lede is Purdy's quote, and it should be. Not because supply chain is a sexy topic, but because it is the one that will determine whether the 1,000-satellite constellation that is supposed to replace lost spacecraft in conflict actually exists when it is needed. Every component — laser crosslinks, valves, connectors — runs through a supplier base that cannot currently scale. That is not a software problem. It does not ship on time because someone typed faster.
