Somebody Finally Built a Satellite to Watch Satellites

Astranis unveiled Perceptor on August 13, 2026 — its first spacecraft built specifically to inspect, characterize, and monitor other satellites in geostationary orbit.

Space domain awareness is the military term for knowing what is in orbit and what it is doing. In low Earth orbit, that job is largely handled from the ground: radar and optical telescopes track objects a few hundred kilometers up with reasonable fidelity.

Geostationary orbit is 35,786 kilometers — roughly 22,236 miles — above the equator. From the ground, a satellite at that distance is a dot. You can tell it is there. You cannot tell what it is carrying, where its antennas are pointed, or whether it just quietly maneuvered toward one of yours.

Perceptor is designed to fly up and look.

The heritage argument

The technically interesting part of the announcement is not the sensor package. It is the bus.

Perceptor is built on MicroGEO, the small geostationary platform Astranis already operates commercially. Five MicroGEO satellites are currently flying in GEO carrying communications payloads for paying customers.

That matters more than it sounds. Geostationary orbit is a punishing environment — higher radiation exposure than LEO, no realistic servicing, and mission lifetimes measured in years where a single component failure ends the spacecraft. A design that has survived up there is qualified by the only test that counts.

Astranis's specific claim is that no other company selected under the Space Force's Andromeda program has a commercial satellite bus with on-orbit heritage at that altitude. Everyone else building a GEO inspector is building a first-flight vehicle for a first-flight mission. Astranis is putting a new payload on a platform that has already proven it works.

In a procurement environment where schedule risk is the thing program offices fear most, that is the entire pitch.

Andromeda, and the number that moved

Andromeda is a ten-year indefinite-delivery, indefinite-quantity contract awarded in April 2026 to 14 companies. It launched with a ceiling of $1.84 billion.

One month later, the Space Force raised that ceiling to $6.24 billion.

A 3.4x increase to a contract vehicle thirty days after award is not a routine administrative adjustment. The stated justification was "the escalating threat environment projected for calendar year 2030 and beyond."

Read plainly: between the award and the amendment, the Space Force's assessment of what it needs to watch in orbit changed enough to more than triple the money it wanted available to spend on watching. Whatever drove that reassessment is classified. The budget line is not.

Maneuverability is the actual feature

Each Perceptor uses electric propulsion and is designed to be highly maneuverable — capable of traversing the full GEO arc and repositioning dozens of times across its mission life.

That specification is the design.

A fixed observation post in geostationary orbit sees whatever passes its slot. An inspector that can relocate repeatedly across the arc can be tasked: go look at that object, then that one, then return. Astranis describes Perceptor conducting flybys of other spacecraft — approaching closely enough to characterize what a target actually is, rather than confirming that something is present.

Electric propulsion is what makes this affordable. It delivers very high efficiency per kilogram of propellant at the cost of low thrust, which is a bad trade when you need to move fast and an excellent one when your maneuvers can take weeks and your mission runs for years. Chemical propulsion would give a GEO inspector a handful of repositions. Electric gives it dozens.

Constellation logic

Astranis also notes that multiple Perceptors could operate together, letting operators observe a target from several vantage points at once — and making it substantially harder for that target to evade observation.

This is the point where the concept stops being a telescope and becomes a network.

A single inspector can be avoided. Its position is knowable, its approach is slow by physics, and a satellite with a decent propulsion budget can simply not be where the inspector is heading. Several inspectors distributed across the arc, each able to reposition independently, create a geometry problem with no clean solution: evading one moves you toward another, and evasive maneuvers themselves are observable behavior that reveals something about the spacecraft doing them.

That is a meaningfully different capability from "we can take a picture."

Why a broadband company is doing this

Astranis built its business on MicroGEO satellites providing broadband to underserved regions — small, cheap, dedicated spacecraft as an alternative to the enormous multi-hundred-million-dollar GEO comsats that dominated the category for decades.

The commercial version of that market has become hard. Starlink and its competitors are eating regional broadband from low Earth orbit with an economics story that GEO cannot match on latency or unit cost.

Government space domain awareness is a different market with different physics. It is inherently a GEO problem — you cannot inspect a geostationary satellite from LEO — and Astranis's core competence, building small capable spacecraft that survive at that altitude cheaply, is exactly the right competence for it. The company has also been selected for the Space Force's Resilient GPS program, which points the same direction.

This is a company pivoting the same hardware into the customer that needs it most, which is the healthiest kind of pivot there is.

The read

Space domain awareness in geostationary orbit has been an acknowledged gap for years, discussed in defense circles and largely unaddressed commercially because the market was theoretical and the engineering was hard.

The Space Force just tripled the money available. Astranis just announced a vehicle built on a platform that is already flying up there.

The gap is closing, and the thing worth noticing is that a commercial broadband company got there first with a satellite bus it built for a completely different reason.