Industries / Frontier science / Space manufacturing
Research in microgravity
Research in microgravity is the experimental work that finds out whether weightlessness changes how materials melt, mix, solidify and crystallize, before anyone tries to sell what is made.
5
Milestones achieved, from 1969 to 2024.
1
To watch for: announced, not yet confirmed outside.
2
Open deliverables. Nobody has shipped them.
The road
AchievedWatch forOpen
Achieved
1969
First welds in orbit
On 16 October 1969 the Soyuz 6 crew, Valery Kubasov and Georgy Shonin, ran Vulkan, an automatic welding unit built by the Paton Electric Welding Institute in Kyiv. Kyiv Polytechnic Institute's aerospace institute and Anatoly Zak's RussianSpaceWeb both call it the first welding in space. The crew started it from the descent module with the hatch closed. When they went back into the habitation module, by Zak's account, they found that the machine had welded through the samples and gone on to melt the tray and the decorative covering of the cabin wall, and they retreated to the descent module for the rest of the flight.
LessonA hot process inside a crewed spacecraft needs limits that stop it before it reaches the spacecraft itself.
Credit
Paton Electric Welding Institute; Valery Kubasov and Georgy Shonin
1974
Crystals grown without stirring
On Earth, heavier liquid sinks and lighter liquid rises, and this flow, called convection, stirs a melt as a crystal grows from it and leaves uneven streaks of added elements in the crystal. NASA's Skylab space station carried the M512 materials processing facility, built by NASA's Marshall Space Flight Center for six experiments "intended to explore the advantages of manufacturing materials" in near-weightlessness. Among the crystal-growth experiments flown on Skylab, Harry Gatos and August Witt of MIT grew indium antimonide crystals laced with tellurium, and reported in 1974 that growth controlled only by diffusion, the slow spread of atoms through still liquid, had been achieved, a condition "never accomplished on earth."
Credit
Harry Gatos and August Witt (MIT); NASA Marshall Space Flight Center (facility)
2000
Space crystals called inconclusive
X-ray crystallography maps a protein's atoms from the way a crystal of it scatters X-rays, and larger, more orderly crystals give sharper maps. From 1985 space agencies in the US, Europe, Japan, Canada, Germany and China flew proteins to grow such crystals in orbit, by the account of Alexander McPherson and Lawrence DeLucas. In 2000 a National Research Council (NRC) report on biotechnology research for the International Space Station (ISS) evaluated NASA's crystal program and told NASA to focus it on specific questions. McPherson and DeLucas summarize the NRC's most recent review as finding the results inconclusive, most gains incremental and proper control experiments hard to run, and add that launch delays, slow return of samples and paperwork kept most structural biologists away. A 2024 analysis of the Butler Microgravity Protein Crystal Database by Keegan Jackson and colleagues at Butler University reads the record the other way: crystals grown in microgravity "continue to show improvement across all metrics evaluated" compared with ground experiments.
LessonA better crystal from orbit cannot be credited to weightlessness without a matched crystal grown on the ground.
Credit
the National Research Council (review); Alexander McPherson and Lawrence DeLucas (history); Keegan Jackson and colleagues (database analysis)
2014
Printer settings, not gravity
In September 2014 NASA and the company Made In Space sent the first 3D printer to the ISS. It built small plastic parts by laying down melted ABS plastic layer by layer, and made its first parts in November and December 2014. NASA Marshall tested them against parts made on the same printer before launch: the space-made tensile and bending samples came out denser and stronger, and the compression samples weaker. A ground study traced most of the gap to the distance between the nozzle and the build tray, which had been adjusted after each print in orbit. After a second round in 2016 with tighter process control, NASA reported "no engineering significant evidence of microgravity effects on material outcomes."
LessonA difference between space and ground samples has to survive identical machine settings before it is put down to weightlessness.
Credit
Made In Space (printer); NASA Marshall Space Flight Center (testing)
2024
Metal printed in orbit
ESA's Metal 3D Printer, developed by Airbus Defence and Space and AddUp under an ESA contract from 2016, was launched to the ISS on 30 January 2024 and installed in ESA's Columbus module. It printed a first metal line, shaped like an S, in June 2024 and its first complete three-dimensional sample that summer, by the accounts of IMDEA Materials Institute and 3D Printing Industry. In March 2025 the first sample reached ESA's ESTEC laboratory in the Netherlands, and a second went to the Technical University of Denmark, each to be compared with the same part printed on Earth.
Credit
Airbus Defence and Space and AddUp, for ESA
Watch for
Drug crystals as a service
Redwire sells space on PIL-BOX, small crystallization units flown to the ISS for drug makers and university labs. In June 2026 it said 54 PIL-BOXes had flown, and that its latest batch, for Bristol Myers Squibb, Aspera Biomedicines, Rowan University and Purdue University, came back to Earth on 17 June 2026. Its clearest reported result so far, insulin crystals for Eli Lilly that were "larger and more highly ordered" than crystals grown on Earth, is described as an initial finding in Redwire's own news release.
Achieved whena customer publishes a peer-reviewed comparison with crystals grown in identical hardware on the ground, or names a drug process that now uses what the flights showed.
Claimed by
Redwire, June 2026
Open
01
Test the crystal record
A matched-pair table of protein structures in the Protein Data Bank (PDB), the world archive of molecular structures, whose crystals grew in orbit. Each is paired with structures of the same protein chain in the same crystal form grown on the ground, ideally by the same group within five years, and compared on resolution (the finest detail the data show), R-free (how well the atomic model fits data held back from fitting) and the Wilson B factor (a measure of disorder in the crystal). A second list records flights whose crystals failed or were never deposited, from NASA and JAXA flight reports, to show how far the archive leans toward successes.
The briefCredit · unclaimed02
Score ground stand-ins against orbit
A table of every molecule that has been crystallized both in orbit and in a ground device meant to imitate weightlessness: a clinostat or random positioning machine, which keeps turning the sample so that gravity's direction averages out; magnetic levitation, which uses a strong magnet to offset gravity; or gels and thin capillaries that slow convection. For each pair it records whether the orbit crystals and the stand-in crystals were better, the same or worse than ordinary ground crystals, and it scores how often each kind of stand-in predicted the orbit result.
The briefCredit · unclaimed
Open 01 · the brief
Test the crystal record
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
A matched-pair table of protein structures in the Protein Data Bank (PDB), the world archive of molecular structures, whose crystals grew in orbit. Each is paired with structures of the same protein chain in the same crystal form grown on the ground, ideally by the same group within five years, and compared on resolution (the finest detail the data show), R-free (how well the atomic model fits data held back from fitting) and the Wilson B factor (a measure of disorder in the crystal). A second list records flights whose crystals failed or were never deposited, from NASA and JAXA flight reports, to show how far the archive leans toward successes.
Why it is still open
Two readings of the same record disagree. McPherson and DeLucas report the NRC's verdict of "inconclusive" and its doubts about controls, while the 2024 Butler database analysis finds improvement on all metrics. A 2026 review from MIT and the University of Puerto Rico counted 189 publications on crystallization in microgravity and found only 19 that compared microgravity and Earth outcomes directly. Earlier tallies counted flights, not matched structures: Russell Judge, Edward Snell and Mark van der Woerd of NASA Marshall built an archive of flight results in 2005 and found that success could not be predicted from any property of the molecule, and the Butler group's 2023 meta-analysis covered 507 experiments. Databases of flight crystallizations have existed since 1994, when NIST's Biological Macromolecule Crystallization Database took in NASA's protein crystal growth archive, and a web search for comparisons of space-grown structures in the PDB found only the Butler analysis, with no public table of matched pairs that others can recheck. On 19 November 2026 a search of the PDB's crystal-growth notes for "microgravity" returned 28 entries, and a full-text search 86, so the starting set is small enough to check by hand.
pmc.ncbi.nlm.nih.govapi.crossref.orgpmc.ncbi.nlm.nih.govebi.ac.ukapi.crossref.orgnist.gov
Built with what exists
The RCSB PDB search and data interfaces (free); RCSB sequence clusters to find the same protein grown on the ground; PDB-REDO, which re-refines deposited structures with one consistent method so their statistics can be compared; Python with pandas.
First step
Run the RCSB query for "microgravity" in the crystal-growth notes and, for each of the 28 entries, list every other entry with the same protein sequence.
Done looks like
Every PDB entry whose crystals grew in orbit is identified, by text and through the papers that cite it, with at least one matched ground entry where one exists. The table gives the share of pairs in which the orbit crystal has the better resolution and R-free, with 95% confidence intervals. Two people agree on at least 90% of a sample of 50 pairings, and the code and table are public.
Open 02 · the brief
Score ground stand-ins against orbit
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
A table of every molecule that has been crystallized both in orbit and in a ground device meant to imitate weightlessness: a clinostat or random positioning machine, which keeps turning the sample so that gravity's direction averages out; magnetic levitation, which uses a strong magnet to offset gravity; or gels and thin capillaries that slow convection. For each pair it records whether the orbit crystals and the stand-in crystals were better, the same or worse than ordinary ground crystals, and it scores how often each kind of stand-in predicted the orbit result.
Why it is still open
The 2026 MIT review says ground stand-ins "cannot fully reproduce orbital microgravity" but may serve for early tests before flight, and finds the field held back by "the absence of coordinated space–ground experimental frameworks". Plant biologists have done this comparison: John Kiss, Chris Wolverton and co-authors set stand-ins against spaceflight results for plant growth in 2019. A web search found no such comparison for crystallization.
Built with what exists
The Europe PMC and Crossref interfaces to find papers; the Butler Microgravity Protein Crystal Database for flight outcomes; an AI model to pull the molecule, device and outcome from each paper, with a person checking every row; Python.
First step
Start with lysozyme, the protein the MIT review found most studied: list every lysozyme crystallization paper that used a clinostat, random positioning machine, magnetic levitation or gel, and every lysozyme result from orbit.
Done looks like
At least 30 molecule and stand-in pairs that also have an orbit result, each row with its sources; an agreement score for each kind of stand-in with a confidence interval; two people give the same outcome label on at least 90% of a 40-row sample.