Construction in space, remarkably and wonderfully, is something we know that we can do, because we have done it. And it is worth saying something about, if only as a change from the distressingly popular subject of blowing things up in space.
To be sure, what we do now is only the final stage of assembly, and from kits. We all know how smoothly that can go. Like rendezvous and docking it might easily have been impractical, but it works about as well in real life as it did in Heinlein stories.
Thus we can build our ships in space, as big as we need them to be, instead of having to launch them all up from Earth. In this as in so much else, the ISS amounts to a training mission for deep space travel.
Actual fabrication of structures is another matter, but we are not likely to do it (other than as an experiment), or need to do it, for quite a long time to come. The reason lies in the odd economics of space travel, and Cobb's Law. If orbit lift is cheap, we can bring up the things we need. If it isn't cheap, we aren't up there either.
Cheap is a relative term when it comes to space travel, but I'll resort to my standard rule of thumb and say that with technical and economic maturity the major component parts of a spaceship - hab pods, tanks, drive units, radiator panels; some assembly required - might cost roughly a million dollars a ton, here on Earth at the factory gate. Launching them into space, along with assembly crews, might cost another $1 million a ton, making the on orbit flyaway cost $2 million per ton.
The good news is that this places deep space craft in the same broad range of cost, size for size and mass for mass, as jetliners, of which thousands are in regular service. The bad news is that the lunar spacecraft industry is strangled in the crib. It has no market unless its production cost at the factory cargo airlock can be brought down to twice Earth production cost at the factory gate. That is an awfully tough standard for a new industry on a new planet.
(If lift cost to orbit can be brought down to $100,000 per ton - which calls for the magic of heavy traffic demand, supporting regular flights by production vehicles - even food production is cast in doubt. $450/lb is a lot to pay for cornflakes. But cheap local produce from the ecohab is only cheap after someone has built the hab, and mastered space gardening techniques.)
More complete fabrication in space will grow partly out of special requirements for structures too big to lift and not suited to snap-together construction. But more than that it will evolve from the duct tape side of the equation. People will start modifying pods, authorized or otherwise. Over decades a boneyard of old structures and equipment will accumulate, begging to be repurposed. Orbital (or lunar, etc.) machine shops will appear to do this work.
And the machine shop is pretty much the basis for all industrial technology, a production line being simply a series of machine tools lined up and set up to run a standard job repeatedly.
Once space facilities can fabricate their own structures, the 'supply chain' economics changes. Now the lunar industry doesn't have to compete in the market for expensive prefabricated pods, but for much cheaper rolled aluminum and the like. Raw aluminum currently costs on order of $3000 per ton - even if the lunar cost were ten times greater, it would be cheaper than 'cheap' orbit lift.
A note of caution. As always in space economics there are some devils in the details. Inherently the trip from the lunar surface to lunar orbit and back, a round trip delta v of 3.2 km/s, is much cheaper than from Earth to LEO. A sensible, single stage vehicle can do it. But it will only be cheap when there is enough lunar traffic for frequent service, so that economies of scale can start to kick in.
So a lunar industry may have to start big - perhaps meaning for political reasons - to permit economies of scale in transportation, and likely in production as well. In the pristine world of economics there is no particular reason for this to happen, suggesting that politics might rear its ugly head.
Therein, presumably, would lie a tale.
Image from Atomic Rockets.
Related post: A Solar System For This Century.