Header

Search

Defense Changyu Han

Modeling Restorative Places from Daily Mobility Data - Changyu Han has successfully defended her PhD thesis.

Autor:
Modeling Restorative Places from Daily Mobility Data

Environments are planned and designed with particular purposes in mind, but how are they actually experienced in everyday life?

The PhD project, conducted in collaboration with the UZH Healthy Longevity Center, investigated how restorative places can be identified and modeled from people’s everyday mobility and lived experiences. It brings together three key elements—mobility trajectories representing individuals’ movements, places as locations of potential restoration, and lived restorative experiences—within a coherent conceptual framework that views psychological restoration as emerging from human–environment interactions in everyday life.

Building on this framework, the project developed a novel method for trajectory-based detection of place locations—or place footprints—that integrates both stop and move episodes. Conventional trajectory-based approaches typically define places from stay points alone, which can miss or only partially represent places experienced largely through movement, such as parks and recreational areas. The method offers a more complete representation of the places people encounter, providing a stronger foundation for downstream place-based analyses.

Using this mobility-inclusive perspective on places together with a tabular foundation model, the project then developed a computational framework for predicting whether mobility-derived places are experienced as restorative or non-restorative. This goes beyond existing research that primarily focuses on estimating the general restorative potential or perceived restoration of environmental scenes. By linking restoration to ordinary places as they are experienced in daily life, including during movement and across different situations, the framework extends place-based analysis beyond predefined environmental settings and supports more context-aware place recommendation and interventions.

Subpages