Police GIS

A recent piece from Miller-McCune looks at changes in how police are using GIS as a crime fighting tool. It includes a look at a model that sounds like an integrated form of agent-based and cellular automata:

One simulation that Liu and Eck supervised involved a section of Cincinnati and two types of agents — store robbers and managers. Robbers traveled randomly within the neighborhood, deciding to steal from a store if their skill exceeded its level of protection, while managers increased store protection depending on the frequency of robberies. The end result? The pattern of simulated robberies began to look familiar — in fact, it was startlingly similar to the real pattern of thefts in the actual area of the city.