From Urban Data to City-Scale Models: A Review of Traffic Simulation Case Studies

Depending on the availability of urban data and the scope of the study, various approaches has been proposed for large-scale modelling and simulation of vehicular mobility. However, studies ending up with real-world applications often adopt less efficient and simpler solutions for distinct parts of the simulation workflow (e.g., grid search for calibration instead of meta-heuristic approach) creating the discrepancy between state-of-the-art methods and the choices made by practitioners. This paper provides a systematised review on traffic simulation case studies based on a consolidated workflow for the creation and use of data-driven large-scale traffic simulations. We analyse and discuss the implementation of the various steps in the workflow, namely, data preparation, model implementation, model evaluation, refinement, and application. By reviewing more than 60 case studies from 23 countries, we identify trends and best practices in the design and development of city-wide traffic simulations, as well as formulate the current challenges and gaps that need to be addressed. As a result, this study summarises the current state-of-the-art techniques for implementing and applying large-scale traffic models and serves as a practical reference for urban researchers and practitioners who aim to develop new data-driven models for large-scale urban areas.