A prototype city generation framework for simulating future mobility scenarios across global urban typologies

Abstract

The goal of this project is to develop prototype cities that represent urban typologies worldwide, for the purpose of simulating future mobility scenarios. In ongoing efforts, we have discovered nine driving factors based on data from 331 cities across the world. Using these, thirteen distinct urban typologies resulted, each representing a unique mobility outcome. In order to assess the impacts of future vehicle technologies and environmental policies in these typologies, simulation-ready prototypes are required as test-beds in our state-of-the-art urban simulator, SimMobility. In my thesis, I outline the data and methods harnessed in building a pipeline for the generation of these prototype cities. As a realization of the proposed pipeline, I synthesize the Auto-sprawl prototype city, which represents the urban typology where cars are the dominant modeshare across a large metropolitan area, and public transit availability is limited. The candidate real-world city used for generating this prototype is Baltimore, Maryland. I show consistency of the generated results by comparing the generated data with the real statistical data. Finally, I demonstrate that this work is being utilized for running simulations in SimMobility and generating additional simulatable prototype cities.