Synthetic Agents
Data mined activity patterns for an activity-based travel demand model

Master ︎ Thesis
Supervisors ︎ Prof. Dr Axel Häusler, MA Andrea Kondziela
Collaborator ︎ Maximilian Müh
Tools ︎ Python, Gama
Place ︎ Cologne, Germany
Year ︎ 2020

Page under construction ︎

Read the thesis here ︎︎︎

When designing a city, one of the most important aspects to consider is the mobility patterns of its citizens. Who they are, where they go and when they go will determine how a city should be planned to attend those needs. Agent-Based Models (ABMs) have been long used by planners to simulate these patterns using data from surveys, which usually covers a small sample of the population and are too time and money consuming to update on a regular basis. Social media with its constant stream of data may offer a great opportunity for urban planners to have constantly updated data about citizens. Location-Based Social Networks (LBSN) like Twitter and Instagram allow its users to share their location when posting. By following a user’s posts, it is then possible to track their activity patterns and even determine socio-economic aspects. The aim of this master thesis is to use social media data to create a synthetic population for an activity-based travel demand model for the future Deutzer Hafen district in Cologne, Germany. The ABM is a joint work with Maximilian Müh, who also built a Tangible User Interface (TUI) table for visualization of the model and interaction with stakeholders.