Research

Synthetic populations are detailed digital representations (digital twins) of a territory’s population, including households and persons with sociodemographic attributes such as income, age, and gender, as well as spatial attributes such as their place of residence. Furthermore, synthetic populations provide representative activity chains which describes what type of activity, at what time, and at which place a synthetic persons is likely to to on an average day. I am the author of the eqasim-synpop framework that allows generating synthetic population data sets based entirely on open data for France and a couple of other use cases.

Agent-based simulations of the transport system take synthetic populations as input and simulate the persons’ movements between their daily activities. This way, traffic patterns of the real world can be replicated. In a second step, the simulations are used to assess the impact of new technologies, transport policies, and societal changes - on the system or individual (by place of residence, age, …) level. I am an active maintainer of the agent-based transport simulation framework MATSim and its eqasim-java extension.