The first industry off Earth will mostly supply itself and the people and machines nearby: oxygen, water, propellant, building material and power hardware. Very few things are worth shipping back to Earth. The exceptions are small, high-value products that come out better when made in microgravity, such as certain crystals and semiconductor materials.
Why make things off Earth at all
The case comes down to gravity. To leave Earth, a rocket has to reach about 11.2 km/s. To leave the Moon, it needs about 2.4 km/s. That makes material lifted from the lunar surface far cheaper, in energy terms, to deliver to lunar orbit or elsewhere in space than material launched from Earth, once the equipment to make it is in place.
The biggest prize is the most ordinary one. Oxygen makes up most of the mass of a typical rocket's propellant, and people need it to breathe. Lunar regolith is roughly 40% to 45% oxygen by mass, locked in minerals. If oxygen can be made on the Moon, the largest share of what a lunar base would otherwise import from Earth can be produced on site.
The products, and how far each has got
| Product | From | Where it stands, October 2026 |
|---|---|---|
| Oxygen (Mars) | Carbon dioxide in the Martian air | Shown on Mars. NASA's MOXIE, aboard Perseverance, made 122 g over 16 runs from 2021 to 2023, at up to 12 g per hour and 98% purity or better. |
| Oxygen (Moon) | Minerals in regolith | Ground tests. ESA's lab plant has made oxygen from simulated regolith since 2020. In February 2026, a NASA team used concentrated sunlight to extract oxygen from simulated soil. It's a NASA LEIA funding topic. |
| Water | Ice in permanently shadowed polar craters | Confirmed, not yet mined. NASA's LCROSS impactor detected water in a polar crater in 2009. How much ice there is, and how easily it can be reached, is still uncertain. |
| Metals | Iron, aluminum, titanium and silicon in regolith | Lab scale. The process ESA's oxygen plant uses, adapted from UK firm Metalysis, leaves behind a mixed metal alloy powder that could feed 3D printers. |
| Building material | Regolith, sintered or bonded | Ground tests. ICON's Olympus system for NASA is in development, and China plans to test brick printing on Chang'e 8 in 2028–29. |
| Power hardware | Local silicon and metals, eventually | Imported for now. NASA LEIA funds vertical solar arrays and radioisotope generators, and NASA wants a 100 kW fission reactor ready to launch around the end of 2029. Making solar cells on the Moon is still a concept. |
The pattern is that every link of the chain has been shown somewhere, mostly in laboratories on Earth with simulated soil. None has run at useful scale on the Moon, and none is yet joined up with the next link. Turning these separate demonstrations into a working chain is the core of what we call extraterrestrial industrialization.
What's worth sending back to Earth
Shipping goods down to Earth is expensive, so only light, high-value products make sense. Three kinds are being tried:
- Pharmaceutical crystals. In microgravity, some drug molecules form crystals that are more uniform. Varda Space Industries' first capsule grew crystals of the HIV drug ritonavir in orbit and returned them to Earth in 2024.
- Semiconductor materials. Space Forge aims to grow crystals for chips, such as gallium nitride and silicon carbide, in orbit and bring them back as seed material for factories on Earth.
- Specialty optical fiber. Fiber made from ZBLAN glass may form fewer defects when drawn in microgravity, and it has been tested on the ISS.
All three use material launched from Earth. They are space manufacturing, but not yet resource use. They matter for Exoforge because they are building the business of flying and returning factories, which a later industry would rely on.
Mining asteroids for platinum-group metals to sell on Earth gets the most attention, and it is the furthest from working. AstroForge, the best-known startup in the field, launched its Odin probe toward a near-Earth asteroid in February 2025. It was lost within days of launch. The company plans a follow-up mission, Vestri, designed to land on the asteroid.
Who owns what's mined?
The legal picture is more settled than it looks:
- The Outer Space Treaty (1967) says no country can claim the Moon or other celestial bodies as its territory.
- The US Commercial Space Launch Competitiveness Act (2015) says US citizens and companies own the resources they extract. Luxembourg passed a similar law in 2017, and a few other countries have followed.
- The Artemis Accords (2020), led by the US and signed by a growing list of countries, say extracting resources doesn't count as claiming territory.
So the working rule among most spacefaring countries is that nobody can own the ground, but you can own what you dig up. China and Russia haven't signed the Accords, and how disputes over the same ice-rich crater would be settled is an open question. Our planned explainer "Who owns the Moon?" will cover this in more detail.
Where superintelligence comes in
Industry off Earth will be run mostly by machines, because people are expensive to keep alive there. That makes AI central rather than a bonus. The jobs that most need it are operating processes on variable raw material, coordinating many machines across a site, and redesigning equipment so more of it can be made locally. We cover the first in Can factories in space run themselves? and the last in Self-replicating factories.
An industry beyond Earth will start with supply, not export. Oxygen, water, building material and power, made and used locally, are the first products. Each has been demonstrated in pieces, almost all on Earth.
The milestone to watch is the first link that runs on the Moon. Producing oxygen from real regolith, on the surface, at more than a lab-sample scale, would mark the point where extraterrestrial industry becomes a practical engineering project.
What to watch
- NASA LEIA awards, especially in lunar oxygen production and in-space manufacturing.
- Chang'e 7 (2027) and Chang'e 8 (2028–29): mapping south pole ice, then the first attempt to make building material from lunar soil.
- Lunar cargo landers such as Blue Origin's Blue Moon MK1: industry needs regular, heavy deliveries.
- Fission Surface Power: mining and processing need steady power, including through the two-week lunar night.
- AstroForge's Vestri and other asteroid missions: a test of whether asteroid mining can get past its first steps.
Sources and further reading
- NASA, "NASA Calls for Proposals to Accelerate Lunar Surface Technologies", September 2026. nasa.gov
- NASA, "Sunlight Extracts Oxygen From Regolith Using Solar Chemistry", 2026. nasa.gov
- ESA, "ESA opens oxygen plant: making air out of moondust", January 2020. esa.int
- NASA JPL, "NASA's Perseverance Mars rover extracts first oxygen from Red Planet", 2021, and NASA's 2023 MOXIE wrap-up. jpl.nasa.gov
- Space.com, "Hope is all but lost for private asteroid probe in deep space", March 2025. space.com
- Wikipedia, "Space Forge" and "AstroForge". wikipedia.org