Towards an open, automated, and reproducible synthetic population of Switzerland : A preparatory, modular pipeline for agent-based mobility simulations

Abstract: This thesis explores the development of an automated re-synthesising tool, called EasySynth, to generate a Swiss synthetic population and travel diaries from publicly available data. The aim is to overcome the limitations of restricted-access real population data while maintaining control over anonymisation and statistical accuracy. The research addresses the need for reproducibility and limited data in transportation modelling and thus may be employed in contexts outside Swiss borders where data is not as ample. The methodology involves direct data manipulation and calculation using weighted probabilistic sampling to generate synthetic travel data representative of the real Swiss behavior. The results indicate a close match between the synthetic population and the real Swiss population in terms of age, sex, canton,purpose, and mode of travel. This study aims to contribute to the Eqasim pipeline for Switzerland whereby raw synthetic data is prepared for MATSim and other activity-based microsimulators. In so doing, the project demonstrates the potential of utilising open data to generate a synthetic population and travel behaviour in an automated fashion.