Anne-Val�rie Preto

Talks [1]

2025

Preto A., Danalet A., Joschka B., Rezvany N., Torres F., Bierlaire M. Differentiation of Modal Preferences in Public Transportation May 14, 2025. 25rd Swiss Transport Research Conference (STRC), Swiss Federal Railway, Ascona, Switzerland.


@Misc{2025Preto_STRC,
  author = {Anne-Val\'erie Preto and Antonin Danalet and Bischoff Joschka and Negar Rezvany and {Fabian Alejandro} Torres and Michel Bierlaire},
  title = {Differentiation of Modal Preferences in Public Transportation},
  howpublished = {\url{https://transp-or.epfl.ch/documents/talks/2025Preto_STRC.pdf}},
  year = {2025},
  note = {Presented at 25rd Swiss Transport Research Conference (STRC), Swiss Federal Railway, Ascona, Switzerland}}}

Research projects [1]

2025

Extracting Activity Schedules from the Continuous Mobility Panel Data

SBB is facing changes in mobility behavior that challenge traditional planning models. This project uses the Continuous Mobility Panel data and EPFL�s expertise in activity-based modeling to provide new insights into travel demand. We will first develop a typology of travelers, capturing the heterogeneity of weekly mobility patterns. Then, we will extend the activity-based model to predict activities and trips over a full week. These outcomes will help SBB design services that reflect new behaviors, improve efficiency, and increase customer satisfaction.
- Sep 10, 2025 - Oct 31, 2026
- Sponsor: Swiss Federal Railways (SBB)
- Principal investigator: Michel Bierlaire
- Project manager: Anne-Val�rie Preto

Semester projects [1]

2025-2026

Theory and implementation of connectors in public transport demand models, Emilien Ulrich

Supervision: Evangelos Paschalidis, Anne-Valérie Preto, Michel Bierlaire
Transport demand models play a role in evaluating public transport policies. They allow stakeholders to analyze different scenarios: how schedule changes and infrastructure projects affect passenger behavior and accessibility. A component of these models is the representation of connectors, i.e., the virtual links connecting demand zones to boarding stations or stops. These connectors determine access and egress possibilities and directly influence the perceived resistance of an origin�destination (OD) relation. In high- density urban networks, the choice of connectors strongly impacts the realism of resistance calculations, as passengers may flexibly choose between several nearby stations depending on service frequency, connec- tivity, or travel time. For now, the SMA demand model treated connectors without considering different alternatives for origin and destination stops: it connects a zone to the closest public transport stop, which is a reasonable assump- tion for high-scale analysis like a national railway system, as train stations are very often important hubs, either when they are main stations in cities or the only stop in a town. However, this assumption lacks precision for the analysis of dense urban networks, as more public transport stops are candidates to be the origin or destination of a zone. Integrating connectors into a utility-based framework would allow demand models to better reflect pas- sengers� trade-offs when selecting boarding and alighting stations. This project will investigate the theory behind connectors. It will analyze their treatment in the literature and in Swiss practice, and develop a framework for integrating connectors into a utility-based demand model. The ultimate goal is to test this framework in a realistic public transport context and assess whether it improves the accuracy and behavioral plausibility of resistance calculations.