Graphic Tools to Simulate the Evolution of Cities: A Review of Recent Experiences from Cellular Automata to Artificial Intelligence

In the past decades, models such as cellular automata and agent-based systems, as a discretisation of morphogenetic models coming from the fields of biology and chemistry, have been applied to the study of urban form, in order to describe, especially a posteriori, the stochastic geometry underlying cities in different steps of their development. These techniques are at the basis of more recent studies, also involving artificial intelligence in the process. This turns out to have, at least to some extent, a predictive nature: we cannot foretell the exact shape of a city in the future, but we can observe how the variation of some parameters might affect its development, giving rise to a range of different scenarios.