I remember reading that some inventors could form mental images accurate / with enough fidelity to visualize complex machinery interacting. In architecture scho

I remember reading that some inventors could form mental images accurate / with enough fidelity to visualize complex machinery interacting. In architecture school there was gradient in how well people could mentally visualize spatial relationships including measurements. I remember a project on a fairly narrow site and during initial sketch exercises, people were coming up with (to me), obviously unworkable solutions because the programming (i.e. rooms) they'd need to fit into certain volumes, couldn't. Spectrum in spatial visualization becomes very obvious when you see tacticle / visual learners iterate through a bunch of designs, whether with modelling or with CAD, and after a few lines, I (and some others) knew immediately the proportions / scale was wrong. Became very obvious when projects started incorporating building codes and designs needed to account for required # egress (fire exists etc) XYZ meters from any point in floor plan. Some people can mentally iterate through designs/measurements in their heads, some people have to draw things out to figure out.