The effectiveness of smart urban solutions depends on the properties of road networks, population distribution and transportation flows. Proper assessment demand unlimited set of road networks with the superimposed transit lines, and population patterns that cover the entire spectrum of urban variability;

The examples we use should be both complete and scalable. On the one hand, we need the full and coherent set of urban layers in order to feed the greedy algorithms that aim at highly informative city of the future; on the other, to test solution’s performance, the datasets must be of all scales;

Collection of real-world examples is a cumbersome process and too often results in the understanding that important patterns and parameters are unavailable.

​

We thus propose a novel framework for establishing broad, scalable and controllable spectrum of urban patterns and cities: instead of collecting real data let us generate synthetic cities that repeat the characteristic of the real ones.

We consider synthetic cities modeling as a new domain of the urban science which importance will grow in time.