Industries / Frontier science / Space manufacturing
Products
Products are the materials made in orbit for use on Earth, and the tests that show whether they beat the same thing made on the ground.
5
Milestones achieved, from 1985 to 2026.
1
To watch for: announced, not yet confirmed outside.
2
Open deliverables. Nobody has shipped them.
The road
AchievedWatch forOpen
Achieved
1985
Space beads go on sale
The National Bureau of Standards (NBS, now NIST) wanted polystyrene spheres 10 µm and 30 µm across to calibrate instruments that measure particle size, but on Earth gravity spoiled the growth of rigid spheres that large. John Vanderhoff of Lehigh University and Dale Kornfeld of NASA's Marshall Space Flight Center grew them on the space shuttle in the Monodisperse Latex Reactor; a run on the flight STS-6 made 10 µm beads, and in 1984 NBS found them rounder, more even in size and more rigid than beads made on Earth. In 1985 NBS began selling them as Standard Reference Material 1960, which NIST calls "the first commercial product to be manufactured in space." By NIST's later account, buyers included drug makers, chemical companies and government labs, and sales averaged about 25 vials a year.
Credit
John Vanderhoff (Lehigh University); Dale Kornfeld (NASA Marshall Space Flight Center); Tom Lettieri and colleagues (NBS)
1985
A space drug loses out
McDonnell Douglas's Electrophoresis Operations in Space (EOS) program used continuous-flow electrophoresis, which separates substances in a liquid by pulling them through an electric field, and the company said the device "could, in a weightless environment, yield products in much higher quantities and purities than are possible on earth." It ran the device on six shuttle flights for Ortho, a Johnson & Johnson company; after an August 1984 flight Ortho's hormone came back contaminated, which McDonnell Douglas said happened during processing. In September 1985 Ortho pulled out, saying the product could be made more easily on the ground. A month before, J&J had agreed with Amgen to develop genetically engineered erythropoietin, the anemia hormone that Riker, Ortho's replacement, then agreed in principle to make in space, by The Pink Sheet's reports.
LessonA space process can be overtaken by a cheaper method on Earth, here genetic engineering, before it reaches market.
pink.pharmaintelligence.informa.compink.pharmaintelligence.informa.com
Credit
McDonnell Douglas; Ortho (Johnson & Johnson)
2017
Fiber drawn, results kept
ZBLAN is a glass made of fluorides of zirconium, barium, lanthanum, aluminum and sodium that in theory carries light with 1% of the loss of the silica glass in ordinary optical fiber, but tiny crystals form in it as it cools and scatter the light. In December 2017 Made In Space, working with the photonics company Thorlabs, sent a machine to draw ZBLAN fiber on the ISS and flew it four times; FOMS Inc. showed its first ZBLAN strands drawn on the ISS in November 2019. NewSpace Economy, reviewing the field in February 2026, notes that Made In Space never publicly disclosed how much fiber it made, and IEEE Spectrum describes the early attempts as producing short lengths with mixed results.
LessonA demonstration that publishes no measurements gives buyers nothing to judge the product by.
Credit
Made In Space (with Thorlabs); FOMS Inc
2019
Orbit lessons, Earth production
Antibody drugs are hard to crystallize because the molecules are large and flexible, though crystals, held as a crystalline suspension of tiny crystals in liquid, are one way to make, store and deliver a drug. Merck researchers, with the ISS National Laboratory, crystallized Merck's cancer antibody pembrolizumab (Keytruda) on the ISS, sent up on the SpaceX CRS-10 cargo mission. In 2019 they reported that the flight samples formed crystals of one even size, about 39 µm, where ground controls gave a mix of 13 µm and 102 µm, and that the flight suspensions were less viscous. They then made uniform crystals of 1 to 5 µm on Earth, using rotating mixers to reduce settling. A 2026 review by researchers at MIT and the University of Puerto Rico lists the work among the few true bulk crystallization studies done in orbit.
LessonAn orbit experiment can pay off by showing which conditions to copy on the ground.
Credit
Paul Reichert and colleagues (Merck); ISS National Laboratory
2026
Kilometers of fiber, tested
Over about a month in early 2024 Flawless Photonics drew about 12 km of ZBLAN fiber on the ISS, including one unbroken strand of 1.1 km, in a machine that could restart the draw after a break. An international group of five organizations, among them the Technical University of Denmark, the University of Warsaw and the University of Adelaide, tested about 2.1 km of it against fiber from an identical drawing machine in Luxembourg. Their 2026 paper, led by Mustafa Arat, found lower loss and higher strength in orbit-drawn step-index fiber, which has a core of one glass inside a cladding of another; IEEE Spectrum reports it as 80% stronger, with no measurable optical difference in fiber made of a single glass. Flawless's own calculation puts the best gain from microgravity at half the loss, against the hundredfold gain once hoped for, because trace metals in the raw materials and bubbles in the glass blanks remain. John Ballato of Clemson University, who was not involved, says melting in space "buys you nothing" against impurities.
LessonWeightlessness can remove one source of loss, but the purity of the starting material sets how good the product can be.
Credit
Flawless Photonics; Heike Ebendorff-Heidepriem (University of Adelaide) and partner labs
Watch for
Semiconductor crystals from orbit
On the last day of 2025 Space Forge said the furnace on its ForgeStar-1 satellite had made plasma, a hot gas of charged particles, at more than 1,000°C in low Earth orbit, conditions the company calls essential for growing advanced crystals. The company plans to grow crystals of gallium nitride, silicon carbide, aluminum nitride and diamond, materials the electronics industry uses, and expects them to be "up to 4,000 times purer" than crystals made on Earth. ForgeStar-1, launched in June 2025 on SpaceX's Transporter-14 rideshare, only tests the furnace, and its mission was set to end in a controlled burn-up, so nothing made aboard comes back. The company plans to return crystal "seeds" on later missions and grow them further on the ground.
Achieved whencrystals grown in orbit come back to Earth and a lab outside the company measures their purity and defect count against crystals grown on the ground.
Claimed by
Space Forge, December 2025
Open
01
Check the factor-of-two claim
An open calculator that splits the measured loss of a ZBLAN fiber at each wavelength into its causes: absorption by trace metals (iron, cobalt, nickel, copper) and by water, scattering from bubbles and crystals, and the glass's own lower limit. Given a stated purity, it computes how much of the loss microgravity could remove, and it is checked against the published loss curves for the 2024 space and ground fibers and for ZBLAN fiber sold today.
The briefCredit · unclaimed02
Grade the product claims
A public ledger of every measured claim made about a product made in orbit since 2014, such as "up to 4,000 times purer" semiconductor crystals, "80 percent stronger" fiber or "larger and more highly ordered" insulin crystals. Each row gives who claimed it and when, the sample size, whether a ground control was made in identical equipment, who measured it, and a status: company only, measured by an outside lab, peer reviewed, or repeated by someone else. It is updated every quarter.
The briefCredit · unclaimed
Open 01 · the brief
Check the factor-of-two claim
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
An open calculator that splits the measured loss of a ZBLAN fiber at each wavelength into its causes: absorption by trace metals (iron, cobalt, nickel, copper) and by water, scattering from bubbles and crystals, and the glass's own lower limit. Given a stated purity, it computes how much of the loss microgravity could remove, and it is checked against the published loss curves for the 2024 space and ground fibers and for ZBLAN fiber sold today.
Why it is still open
Flawless Photonics' team calculates that microgravity gives at best a halving of loss, and John Ballato of Clemson University argues that impurities, not gravity, set the limit; both views come from interviews, not a published loss budget. NewSpace Economy's February 2026 review says commercial viability still hinges on economics, station access and quality checks. A web search found qualitative lists of ZBLAN's loss sources but no open model that splits the loss by cause and states the most microgravity could gain.
Built with what exists
The 2026 paper by Arat and co-authors in Space: Science and Technology (doi 10.34133/space.0572), whose full text DTU's repository offers, for the space and ground loss data; makers' published loss curves for commercial ZBLAN fiber; published absorption coefficients of transition metals and water in fluoride glass; WebPlotDigitizer to read curves from figures; Python with NumPy and SciPy.
First step
Download the 2026 paper and digitize its loss curves for the step-index fibers drawn in orbit and on the ground.
Done looks like
A public calculator whose fitted budget reproduces the published space and ground loss curves within 10% at five or more wavelengths between 1 µm and 4 µm, and a table giving, for three purity levels, the largest loss reduction microgravity could deliver. A second person reruns it from the published inputs and gets the same table.
Open 02 · the brief
Grade the product claims
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
A public ledger of every measured claim made about a product made in orbit since 2014, such as "up to 4,000 times purer" semiconductor crystals, "80 percent stronger" fiber or "larger and more highly ordered" insulin crystals. Each row gives who claimed it and when, the sample size, whether a ground control was made in identical equipment, who measured it, and a status: company only, measured by an outside lab, peer reviewed, or repeated by someone else. It is updated every quarter.
Why it is still open
Factories in Space, which calls itself "the largest public database of companies" in the in-space economy, lists companies and product types but does not grade their claims. The 2026 MIT review found only 19 of 189 publications that compared microgravity and Earth outcomes directly. A web search found no tracker that checks product claims from orbit against evidence.
Built with what exists
Company newsrooms (Redwire, Space Forge, Varda Space Industries, Flawless Photonics); ISS National Laboratory results pages; NASA's Space Station Research Explorer; the Crossref and Europe PMC interfaces for papers; an AI model to pull claims from releases and papers, with a person checking each row (AI review of records is covered in the Note "Seven more flagged").
First step
Copy every quantitative claim in Redwire's, Space Forge's, Varda's and Flawless Photonics' news releases from 2024 to 2026 into a spreadsheet with its date and link.
Done looks like
At least 60 claims from at least 10 organizations, each linked to its source. Two people assign the same status to at least 90% of a 30-claim sample. The first quarterly update records every change of status since launch.