So you have a set of positions for some services (coffee shops, groceries), a set of positions for cameras and a set of positions for homes.

But how you model people actually driving each day and their paths? Because the accuracy of your conclusions seems to heavily depend on the accuracy of modelling.

I don't model daily paths as in 'coffee before work, then groceries on the way home'; I do straight shots from residences to each of these amenities. I don't know of a better way to do it than this; any more complicated model that tries to model 'daily routines' risks losing simplicity, as well as straying too far from actual driving behavior, and my main goal is extremely simple statistics.