In-space assembly robotics

The largest structures in space will be built there.

Thiral is developing autonomous modular robots that assemble, extend and repair large structures in orbit and on the Moon. Launch compact units. Build what no rocket could carry.

Request the deckSee the roadmap

01 The problem

Everything in space is still sized by the rocket.

5 mFairing width

Most of today's launch vehicles carry payloads about five metres wide. Anything larger has to fold, or fly in pieces.

42Assembly flights

The International Space Station took 42 assembly flights and more than a decade of astronaut work to build.

344Single points of failure

The James Webb Space Telescope flew with 344 of them, most in the mechanisms that unfolded it.

The structures the coming decades call for are larger than any fairing: wide antennas, power arrays, stations, lunar bases. Folding them adds mechanisms that can fail. Assembly by astronauts does not scale.

And once a structure is in place, it usually cannot be repaired or extended. Space agencies and private companies building beyond Earth all meet the same limit.

Sources: NASA, SpaceX, Arianespace.

02 Why now

Launch became twenty times cheaper. Building in space has not changed.

20xCheaper launch

From about $54,500 per kg on the Space Shuttle to about $2,720 on Falcon 9.

2026Programmes are flying

DARPA's NOM4D programme flies in-orbit assembly demonstrations.

2030A new station era

The ISS retires as soon as 2030, with commercial stations to follow.

$1.8TGlobal space economy

The projected size of the space economy in 2035, up from $630B in 2023.

Mass to orbit is no longer the scarce thing. Volume and assembly are. Redwire, Northrop Grumman, GITAI and others are working on it, with DARPA and NASA programmes behind them.

Earlier efforts such as NASA's OSAM-1, cancelled in 2024, and Archinaut, ended in 2023, show how hard the problem is, and that it is still open.

Sources: NASA, DARPA, The Planetary Society, World Economic Forum.

03 Approach

Launch small. Build big. Keep growing.

Launch compact

Small units pack tightly for launch, so the size of the structure is no longer limited by the rocket.

Build in place

Robots assemble the units into a load-bearing structure where it will be used, with no crane and no crew.

Grow and repair

Add units to extend the structure, or swap a damaged one to fix it. Nothing is built only once.

The step that is missing

Robotic assembly has been shown in the laboratory, most notably by NASA's ARMADAS project in 2024. It has not yet become a system an operator can buy and fly. Thiral is working on that step.

  • Modular

    Structures are made from a small set of standard units that can be produced in volume.

  • Autonomous

    Robots do the assembly. The build does not depend on astronauts at the site.

  • Reconfigurable

    Units can be added, replaced or moved, so a structure can change after it is built.

04 What it unlocks

Infrastructure with no fixed size.

Large antennas

Apertures far wider than any fairing allows.

Solar arrays

Power for stations, missions and surface bases.

Station modules

Commercial stations and habitats that can grow.

Lunar structures

Towers, shelters and platforms assembled on the surface.

Repair and extension

New life for structures that are already in orbit.

Service hubs

Refuelling and support points for other spacecraft.

05 Progress

Engineered before it is built.

The prototype design is closed in computer-aided engineering and its manufacturing files are released. The hardware build comes next.

470+Design iterations

CAD and CAE iterations, each with a stated hypothesis and a recorded result.

100+Structural analyses

Iterations backed by finite-element analysis of stress, stiffness and contact.

40Research studies

Prior-art and research studies drawing on more than 220 cited sources.

2xLoad margin

Joint capacity against the reference case: a five-bay bridge carrying a ten-unit payload.

All figures are from simulation and analysis: parametric CAD, finite-element analysis, multibody contact dynamics and continuous collision checking. Hardware validation follows with the prototype.

06 Roadmap

From first prototype to orbit.

Now

Design and first prototype

Engineering closed in simulation, then a prototype of 10 to 15 units that shows the core mechanism on the ground.

H1 2027

Working system

Scale up to a minimum viable system of 100 units, building real structures on the ground.

Q4 2028

Space testing

Vacuum, thermal and zero-gravity validation, once the system is proven on the ground.

Beyond

Infrastructure in place

Antennas, power and habitats assembled, extended and repaired where they operate.

The aim

Structures built where they are needed, by machines that work as one.

Request the deck

For investors, grant programmes and partners.