ALFTECH / DEEPTECH LAB

Luxembourg · 2026

FOUR INTERACTIVE STUDIES / TWO RESEARCH DIRECTIONS

A laboratory for possible futures.

Keep a lunar habitat powered. Test a controller before a field trial. Compare a cooler site plan. Choose a materials experiment. Four hands-on prototypes, with their assumptions in view.

01

MATERIALS · SPACE RESOURCES

MatterLoop

Which materials experiment should you run next? Try different processing recipes, compare recovered oxygen with energy use, and follow the results in a synthetic lab.

  • Active experiment selection
  • Interpolated surface with exploration scoring
  • Auditable recipe log
Open the experiment console
02

ARCHITECTURE · CLIMATE

UrbanProof

Could trees, porous ground and shade make a site cooler and reduce rainwater runoff? Change a synthetic site plan and compare it with the reference case.

  • Counterfactual design comparison
  • Uncertainty and provenance
  • Machine-readable evidence bundle
Open the design evidence room
03

PHYSICAL AI · SAFETY

TwinSafe

What happens when a sensor disappears or a command arrives late? Run two controllers through the same simulated problems, then replay a failure to understand it.

  • Monte Carlo fault injection
  • Safety invariant evaluation
  • Deterministic failure replay
Open the scenario test range

THE MEETING KIT / FIVE SLIDES EACH

A short introduction.
A concrete pilot question.

Read a pitch in your browser or download an editable PowerPoint. Each explains the idea, what works today and what needs validation. English first.

Tecton

Can a lunar outpost get through a power failure?

Read slides ↗PowerPoint ↓

TwinSafe

Which controller handles difficult conditions better?

Read slides ↗PowerPoint ↓

UrbanProof

Which site changes are worth studying further?

Read slides ↗PowerPoint ↓

MatterLoop

Which processing recipe should a lab test next?

Read slides ↗PowerPoint ↓

Have a reference case, a dataset or a question? a@alftech.space

THE RULES OF THE LAB

Credible by construction.

01 / Local first.
No upload, account or hidden external service is needed for these demonstrations.

02 / Claims carry evidence.
Every result exposes assumptions, uncertainty or a reproducible scenario seed.

03 / Prototype honestly.
Synthetic data is labelled. The computational workflow is real; production connectors are the next layer.

NEXT RESEARCHED BUILDS / 2026–27

Where the laboratory goes next.

Two possible next directions: circular construction and orbital operations. These are research concepts, not working demos yet.

04 / CIRCULAR CONSTRUCTION

ReForm

A material-passport and salvage decision engine for buildings.

Connect BIM components to product passports, compare demolition with selective reuse, and produce a traceable embodied-carbon and residual-value case.

DECISION
Reuse, remanufacture or replace?
PROOF
Building section + component ledger + DPP export
European DPP programme ↗
05 / SPACE TRAFFIC

Apsis

A conjunction decision room for satellite operators.

Propagate uncertainty, test maneuver windows and turn collision probability, mission impact and fuel cost into an auditable operator recommendation.

DECISION
Maneuver, monitor or accept risk?
PROOF
Monte Carlo encounter + covariance view + decision log
Luxembourg Space Strategy ↗