Technical Report 337, c4e-Preprint Series, Cambridge

Authors: Jiying Chen, Jethro Akroyd, Sebastian Mosbach, Feroz Farazi, Xinhong Deng, Kok Foong Lee, Mei Qi Lim, George Brownbridge, Tom Bishop, Will Cuningham, Tomas Gonzales, Thomas Burgoine, Jenna Panter, Louise Foley, Søren Brage, Nicholas J. Wareham, and Markus Kraft*

Reference: Technical Report 337, c4e-Preprint Series, Cambridge, 2025

Associated Themes:
 

• Dynamic individualised environmental exposures calculated in The World Avatar
• Ontologies create interoperability across environmental data
• Distinguishes historic features of the environment and time-specific availability
• Enables scalable and transferable mobility-aware assessment
• Demonstrated using location tracking data in the UK and Singapore

The environment in which people spend their lives influences their health behaviours and health outcomes. Traditionally, attempts to study this have used fixed descriptions of the environment and ad hoc analyses to characterise the environmental exposures of people at static locations, analysing the relationship between one environmental factor and one aspect of health at a time. However, the environment and the exposures of people who move through it are dynamic. We introduce a digital twinning approach that leverages The World Avatar to allow the interoperable calculation of individualised dynamic environmental exposures. The approach uses a computational agent to calculate exposures based on a semantic representation of features in time and space. The time-specific part of the exposure calculation considers both the historical context (i.e., whether features of the environment had been built) and current context (i.e., whether something was open or accessible) of the exposure. The scalability of the approach is illustrated using GPS smartphone data describing the movement of people in the UK and Singapore. The resulting exposures capture historic year-specific changes in greenspace and the time-specific accessibility based on opening hours of the food retail environment. This approach could facilitate large-scale digital studies, yielding critical insights that help create healthy cities.

PDF (6.6 MB)