Space

Caltech ASTRA: a robot starts building an antenna in orbit

Key takeaways

  • ASTRA carried out assembly of a third truss bay in orbit without astronauts, which Caltech's team calls a first for autonomous construction in space
  • The demo will not finish all 11 bays before the mission ends in December, and the team plans to simplify the design
  • Ground control windows of about 15 minutes, seven or eight times a day, forced the team to debug on a twin flight spare

A robotic truss builder designed and built at Caltech has assembled part of a structure in orbit without astronaut help, which team leader Sergio Pellegrino says is the first time a structure has been autonomously constructed in space. The machine is the Autonomous In-Space Tensioned Reflector Assembly, or ASTRA, funded by the Defense Advanced Research Projects Agency (DARPA) under its Novel Orbital and Moon Manufacturing, Materials, and Mass-efficient Design (NOM4D) programme.

ASTRA is not a product you can buy. It is a proof-of-concept payload, and this assessment treats it as a tool: what it does, where it struggles, and who should care.

What it does well

ASTRA's core idea is simple to state. An autonomous manipulator, an armlike robot, picks up struts and joints from nearby motorised dispensers and repeatedly assembles identical sections, called bays, of a perimeter truss. The target is a 1.4-metre-diameter reflector antenna built from composite fibre tubes, aluminium joints and a Kapton reflective surface.

The payload is compact, roughly 1 cubic metre. It launched on 30 March 2026 from Vandenberg Space Force Base as one of 10 payloads on Momentus Inc.'s Vigoride 7 orbital vehicle, which rode the SpaceX Transporter-16 mission. It carried two prebuilt bays plus every part needed for the rest of the truss.

In orbit, the team released the launch restraints, pushed out the two prebuilt bays, then had the robot assemble the third. Nominally, the manipulator builds each bay on an assembly plate, then attaches a netlike perforated film to form two reflective cable nets, prestressed against each other by springs. Each finished bay is pushed out and the cycle repeats.

The reflective material is a notable piece of the design. Caltech co-invented it with University of Pennsylvania collaborators led by Professor Igor Bargatin. Compared with the knitted metallic mesh normally used in furlable reflectors, the perforated film needs less prestress, which lets engineers design lighter structures. Weight is the currency of launch, so that matters.

The programme timeline also deserves credit. The space payload was completed, tested and delivered to host Momentus in less than 12 months, by what Sergio Pellegrino, the team leader, called a small team.

Where it falls short

The headline limitation is completeness. A finished ASTRA truss needs 11 bays. The team hit time limits and built only the initial section, and with the mission ending in December, the demonstration appears unlikely to complete the full assembly.

The team also hit what Jaden Suh, a former Caltech postdoctoral scholar on the team, called "little surprises". Commands were sent through the Vigoride vehicle, results were checked against photos downloaded from space, and operations were repeated to overcome unforeseen problems, of which Caltech says there were many. Communication is tight: only about seven or eight windows of about 15 minutes each day. That makes every retry expensive.

The team's workaround was a flight spare, a twin of the payload kept in Pellegrino's Caltech lab. Former Caltech systems engineer Charles Sommer said: "Every time we confirm there is an issue in space, we come to the laboratory and try to mimic the problem on the ground." It works, but it means a second full-spec build is part of the cost of doing this at all.

The team says the lesson is that the design needs simplifying and refining. That is an honest admission that the architecture is viable but not yet efficient.

How it compares

Until now there have been two routes to big structures in orbit: fold them up and unfurl them, or piece sections together manually across several launches. NOM4D adds a third. The other phase 3 project, led by the University of Illinois Urbana-Champaign, takes a different approach: it will use a module attached to the International Space Station to test hardening carbon tubes in space.

ApproachHow it worksMain constraint
Fold and unfurlStructure is packed for launch and deployed in orbitLimited by rocket fairing size
Manual assemblySections sent up over several launches and pieced togetherNeeds a series of launches and manual work
Caltech ASTRARobot assembles truss bays from struts and joints in dispensersOnly the first section built so far; design needs simplifying
Illinois Urbana-ChampaignHardens carbon tubes in space, tested on an ISS moduleMethod still being tested

The scale ambition sets this apart. DARPA programme manager Andrew Detor said "very large" means hundreds of metres or kilometres, and that nothing approaching that can be achieved today.

Pricing and value

The source gives no cost figures for ASTRA, the launch or the NOM4D programme, so there is no price to judge. What can be said is that DARPA started NOM4D in 2022, aiming ultimately to "ferry raw materials from Earth for in-orbit manufacturing", and that only two projects reached phase 3. Value here is measured in retired risk: the concept now has a space result rather than only laboratory evidence from the earlier phases.

For UK readers, the source names no British partner or funding, so there is no direct UK effect to report. Anyone following orbital infrastructure and its economics can read it alongside our look at Google Suncatcher's orbital TPU, where the cost of building in space is the central question.

Who should use it

Nobody can use ASTRA yet. The people who should watch it are:

  • Space telescope and astronomy planners. Pellegrino notes large reflective surfaces could act as enormous mirrors, letting astronomers observe ever more distant objects. Those thinking about servicing and upgrades may also follow NASA's Hubble rescue call for industry.
  • Communications engineers. The same structures could form the aperture of large radio frequency antennas.
  • Space solar power developers. Pellegrino co-directs the Space Solar Power Project, which aims to harvest solar power in space and wirelessly transmit it to Earth's surface, and he points to very large solar arrays as a use.

Those who need a flight-proven, complete assembly system today should wait.

ASTRA showed the feasibility of a robot building structural bays in orbit from stored parts, and the payload was delivered on a short timeline. It has not proved the full 11-bay build, and the team expects to simplify the design. The twin-spare debugging method, with its tight 15-minute communication windows, is a practical template for anyone planning remote robotic work.

Verdict

ASTRA is a credible first step, not a finished capability. It is worth following for anyone whose plans depend on apertures or arrays larger than a fairing, because it shows the assembly concept is feasible. Watch for the December mission wrap-up, the team's simplified redesign, and the results from the Illinois project, which will show which route to kilometre-scale structures looks more practical.

More from Future Technology HQ