It is my belief that within a few years time, building solar panels on the Moon could be done far easier than before, and the simple reason is
personal fabricators. Now, the holy grail is using these fab-labs to reproduce themselves, which engineers are very close to achieving.
This is mainly a thought experiment.
My proposal has a lunar base launched, which will house a hydroponics unit, and quarters for perhaps up to four people, all engineers. This would work with with just telepresence from Earth too, and without the base.
The main machines to be launched to the Moon will be a solar-powered fab-lab and a solar-powered lunar dust mining machine. From here, building the solar panels will simply be an exercise in computer programming. Once deployed, the mining machine will begin mining the lunar dust for more materials. These materials will then be shipped to the personal fabricators, which will replicate itself. These two fabricators will then replicate themselves. One of the fabricators will design more minin equipment. When there are about eight fabricators, then there are more things that can be built, such as robotic constructors. It is possible to make a rocket entirely of lunar material, using solid rocket fuel (like the boosters on the space shuttle) using metals and oxygen. The robotic constructors can take the rocket parts constructed by the fabricators, and build a rocket.
Let's say that it takes two weeks for a fab-lab to reproduce, which is far too conservative of an estimate. If we leave out the construction of other machines such as miners, robotic constructors, etc., then we can have an extremely robust manufacturing system in a year. If each fabricator reproduces itself, then in less than a year, we will have 16 million fabrication units. Of course, many of these fabricators will go to work to produce other things, but in a year, there would still be millions of fabricators. Where there are enough of these fab-labs, construction of solar panels can begin, and robots will put these together.