TimeTravel – Automated Image Recognition to Improve Long-term Urban Planning in Response to Pandemics


Crossing socio-cultural datasets and the Covid-19 pandemic in space over time to improve long-term urban planning in response to pandemics

The COVID-19 pandemic exemplifies that the spread, the intensity, and the control of the disease vary in relation to spatially and temporally uneven distribution of socio-cultural, economic, and environmental resources, qualities, and threats within the built environment – i.e., the access to health facilities, population density, or exposure to pollution.

Existing research on corona-related data tends to focus on studies on individual health risk factors and their identification.  Such studies explore the collection and overlapping of already pre-determined sets of data (epidemiological and socio-economic), defined through previous research or experts´ knowledge.    

time travel

To overcome such bias and to help provide new (planning) strategies for healthier urban environments now and in the future, the Time Travel project proposes to collect, display, and cross diverse, non-pre-determined spatial, socio-cultural, and Corona-related datasets (Covid-19 cases, vaccinations) in space and time. City Planning, Architecture and Planning History, Computer Science disciplines are involved in this project. The project embraces the idea of “Digital Humanities” as interdisciplinary scholarly activity shared between digital/computer science and humanities/social science.  

Using a supervised machine learning methodology allows crossing some hundred non-predefined, multi-source, fine-scaled and historical (“time travel”) socio-cultural, economic, and environmental datasets with different COVID-19 factors – i.e., COVID-19 disease spread, hospitalization, or vaccination rates. Such an explorative and computer-based approach can help reveal unexpected relationships between spatial forms, social determinants, and COVID-19 over time (complementing the results from the traditional methods).

By identifying unexpected relationships and patterns between linear or even non-linear combinations of geospatial and corona factors, new research trajectories open up. These findings can serve as foundation for further analysis to validate potential correlations or even causalities between corona-related data and the built environment.    

Drawing conclusions from overlapping large multi-source datasets at diverse spatial and temporal scales within a transferrable framework/methodology can help planners, decision-makers, and civil society with evidence-based planning strategies to promote adaptive and liveable urban environments.   


Volkswagen Foundation 

Overall budget: 
€ 120.000 

Role TU Delft:  
Lead partner

Project duration: 
May 2021 – August 2022 

HCU researchers:  
Prof. Dr. Dirk Schubert  
Prof. Dr.  Jörg Pohlan 
M.Sc. Hülya Lasch 

TU Delft researchers: 
Prof. Dr. Ing. Carola Hein  
M.Sc. Lukas Höller 
Dr. Christoph Lofi 
Dr. Lixia Chu  

Project partners 
HafenCity University Hamburg, Statistic Office Netherlands (CBS), PortCityAtlas Project TUDelft 

Prof.dr.ing. Carola Hein 
+31 641141071