It's what it sounds like: I want to know whether sending a probe builder to orbit a moon or a planet (for this question I'm thinking of Titan or another smallish body in the outer system.), to build probes there instead of here. I can think of issues with this already: RTG manufacturing, powering the builder etc. However, the main advantage I see is that the journey would take a while with today's technology, and thus being able to manufacture the latest tech for the probe once it gets there instead of when it leaves (or years before, in some cases) would seem to be very beneficial, especially as we might have come up with a better probe design by the time it gets there. This would give the engineering and science team a greater freedom with design too: the probes, which have empty space in them, could be manufactured from bulk resource instead of to fit inside a payload bay. They also wouldn't have their electronics degrade on the way there. Can someone please tell me this is wrong in some significant way? I feel like something is, but I can't work out what... Thanks guys
In addition to the point raise by notovny even once we do have some degree of on site fabrication we probably end up with worse tech than flying an earth built version.
This is because many tech advances in hardware are not 'recipes' but 'things'. An obvious example is silicon fabrication, where current chips and transistors sit on the end of a long series of machines that humans used to build a slightly better machine. For ore smelting and machining this is less of problem, but if a clever process requires an instrument or sensor using a rare and complex to extract element/compound then things get tricky for our factory in a box.
So it is likely that any near future version of this will fly a bunch of CPUs and other complex parts, and the 'made' parts will be boring things like structural elements and wheels. Which also benefit from in situ iteration as you find out what works far more than making a better CPU does.
Where this process makes sense is when you can setup a decent number of humans somewhere. They have much better chances of using the machines they have to make better machines, and it quite possibly makes more sense to have a team building and repairing rovers controlled by scientists from earth than trying to do all the science themselves.
It's not possible at this time. Humanity has no experience in building a dollar-store-quality hammer from raw materials on any non-Earth body, much less a space probe.
If it were possible to autonomously build a space probe from local raw materials gathered on-site in a package less massive than sending an assembled space probe, that level of autonomous manufacturing capability would already be in use everywhere on Earth for everyday household items.