Measuring stellar proper motions has been possible for amateurs for a long time now, though we will not reach Gaia levels of accuracy.
In particular, I really must update my track of Barnard’s Star with images from this year. Ok, that’s an extreme case but proper motions greater than, say, 100mas per annum should be fairly easy to measure over the course of a year or few.
The pipeline I use is MaximDL for taking images; local astrometry.net using the Gaia catalogue to put a WCS on each sub; SWarp for stacking if required; and then APT for photometry (my primary interest) and initial astrometry (if desired). The source list fed into APT contains Gaia positions and its output includes substantially sub-pixel precision on the stars’ centroids, in both pixel and sky co-ordinates. That’s good enough for me (and I repeat that my primary interest is in photometry rather than astrometry). Perhaps I could, and should, try harder to achieve ultimate accuracy in astrometrical positions.
Paul
