Mobility data is powerful, but it is also sensitive. Cities and platforms increasingly want to learn how people travel while protecting privacy. This project will develop methods that allow us to obtain meaningful behavioral insights via structural spatiotemporal choice models, meanwhile ensuring privacies are protected.

What you might work on.

You will build and test models that connect mobility decisions (such as route, mode, or timing) to context (time, network conditions, and constraints). You will then study how privacy-preserving data changes what can be reliably learned, and develop estimation strategies that remain stable and interpretable under these constraints.

Why it matters.

Mobility data was historically difficult to collect at scale, and privacy has been one of the main barriers. By developing privacy-preserving modeling and inference methods, we aim to extract meaningful behavioral insights that can inform better services and policies while protecting individuals. This helps move transportation planning into a true large-data regime, where more advanced and data-hungry methods can be applied responsibly.

Who we are looking for.

Students who enjoy a mix of modeling and data work. A strong fit if you like statistics, discrete choice, optimization, and reproducible computational experiments. Programming experience (Python/R) is helpful, but motivation to learn matters most.