Less sedimentation
Particles and growing crystals settle less. That can open a longer window for suspension and self-assembly.
01 / THE MATERIALS
Small structures.
Extraordinary
possibilities.
Microgravity changes how matter moves and assembles. The opportunity is to turn that difference into materials that do something better.
Explore the first five ↓MATTER / REIMAGINEDCRYSTAL LATTICE · CONCEPTUAL
01.1 / WHY MICROGRAVITY?
Change the environment.
Change what can form.
Different transport
Reduced buoyancy-driven flow can make diffusion more important. Surface-tension flows and forced mixing can still dominate.
New structures to test
Changes in growth can affect size distributions, crystal forms and defects. Better functional performance must be measured separately.
Microgravity is a process condition, not a guarantee of improvement. Radiation, launch loads, thermal history and return introduce additional constraints. Protein-crystal study ↗ · Colloidal-crystal study ↗
01.2 / OUR FIRST FIVE
A portfolio of possibilities.
A test for every one.
Five initial research targets—not five qualified products. Select a material to explore the opportunity, the flight evidence, and the experiment we need next.
Loading the material dossier…
01.3 / REAL SPACE DATA IN
We read the archive.
Every crystal grown in orbit.
Before choosing what to fly, we mined decades of public records on MIT's Engaging cluster: the Protein Data Bank, NASA's ISS physical-science archive and open materials databases. Each panel says what it shows and what it does not.
192 GBpublic data processed
256Mrecords analysed
64Slurm jobs on MIT Engaging
36public datasets
How we did it ↗REAL ISS FRAMES · NASA PSI / DECLIC-DSI (CNES/NASA) · PUBLIC DOMAIN
Each frame looks down on the solidification front of a transparent alloy grown on the ISS. To its right is the radial power spectrum our pipeline uses to measure cell spacing. The last two frames show the gate rejecting a planar front and an interference fringe.
01.4 / FROM CLAIM TO MEASUREMENT
One result.
Precisely stated.
A published pembrolizumab study reported a lower-viscosity crystalline suspension after growth in orbit. Here is what the comparison actually shows.
PUBLISHED COMPARISON · 2019
33.2%
lower reported viscosity
At 50 mg/ml, using the values in Table 1.
DYNAMIC VISCOSITY / cPLOWER IS EASIER TO FLOW
Whiskers reproduce the reported spread (table label: geometric SD). They are not confidence intervals. This is a formulation-specific result, not a clinical or economic benefit.
Read the primary paper ↗Measurement notes & evidence limits +
Values are taken consistently from Table 1. The paper’s prose gives slightly different values for the 50 mg/ml flight suspension. Raw independent replicates are unavailable here, so we do not calculate an effect confidence interval. Flight insights also improved Earth processing; our next experiment must compare against that stronger ground baseline. Full extraction notes and results ↗
01.5 / HOW WE FIND THE WINNERS
From image to insight.
From insight to proof.
Traceable flight data
NASA PSI images, process telemetry and ground controls. Keep sample, run and condition identities intact.
Validated features
Segment crystals and growth fronts. Measure size, defects and texture against manually reviewed labels.
Independent batches
Estimate effects and uncertainty at the batch level. Avoid counting adjacent image frames as independent experiments.
Functional outcomes
Test viscosity, selectivity, cycling or optical loss after return. Retain only the advantages the application can use.
Current baseline: 12 NASA DSI images passed access and image-quality checks. This verifies the pipeline, not a material advantage. Engaging can scale validated extraction across runs.
Inspect the data plan ↗01.6 / BIG DATA, NEXT
From archive
to flight manifest.
The archive tells us where to look. Our own flights will produce the data that tells us what works. This is how the evidence base grows.
One pipeline for every space-grown crystal.
Link each microgravity PDB entry to its raw diffraction images where they are public, then reprocess flight and ground crystals with the same software. This removes the beamline and era differences that blur archive comparisons.
Every batch becomes a data point.
Each foundry cartridge writes a provenance record: recipe, thermal history, images, residual acceleration and return timeline. Matched Earth controls use the same schema.
Choose the next flight with a model.
Join in-situ images to post-return assays (viscosity, selectivity, optical loss) and train models that predict which recipes benefit from orbit. Active learning picks the most informative next batch.
Turn the literature into records.
Extract conditions and outcomes from decades of microgravity papers into structured evidence records with links, reviewed by a scientist before use.
WHAT EACH ANALYSIS ASKS FOR NEXT
Loading the extension plan from each analysis…
01.7 / WHO WE CALL FIRST
Partners,
chosen by the evidence.
Organisations whose data, hardware or demand our materials pillar depends on. Cards marked IN OUR DATA appear in the archives we analysed. No partnership is claimed.
NEXT / 02 — THE FOUNDRY