Technical requirements: V, I images down to a limiting magnitude of ( ) in at least 15-20 fields. Images repeated after 5 and 10 years for proper motions would be quite important.
As above outlined, we know very little on the overall shape of the luminosity function (LF) (we cannot even decide if it is bimodal or not), on the shape of the LF at the faint end (and therefore on the expected number of brown dwarfs), and the change with time (environment) of the LF. The above information are fundamental inputs for any Galactic formation and evolution theory.
All the available studies of the faint end of the LF are affect by one of the following problems: either (a) they are on large areas but not deep enough to sample faint stars in a large enough volume or (b) they are deep, but confined to a small area in the sky.
A multicolor survey down to V=27 in at least 15-20 fields aimed at mapping the different Galactic components would allow to map the faint end (down to ) of the LF in the disk. LBT prime focus imager would be the first instrument allowing this kind of (deep) survey.
The same kind of survey would be as useful in mapping the halo LF down to . A large number of faint stars is expected even for the flattest mass function hypothesis in a LBT field. A correct interpretation of the LF shape needs the knowledge of the density distribution. This problem can be overcome by mapping the halo at different galactic latitudes and longitudes.