This project develops a novel activity-based modelling approach based on optimization of individual daily mobility plans. Implemented within SBB’s existing nationwide model for Switzerland, it will be used for decisions on investment and service planning.
Planning an efficient transport service needs taking into account the factors affecting people’s decisions on travel in the future. Simulation models which closely represent mobility behavior are needed to predict the impact on travel demand of factors such as demographic or policy changes as well as new technologies (like self-driving cars and other emerging forms of transport service).
This new approach targets a superior predictive performance of the model and incorporates multiple individual choices (activity choice, trip scheduling, mode choice, destination choice) into a single optimization model. Modelling these choices together allows for full feedback between each decision and the representation of complex behavioral decision interactions, laying the ground for future transport simulation.
This eighteen-month project is a collaboration between EPFL and the implementation partner SBB funded by Innosuisse. The scientific part is carried out by the Transport and Mobility Laboratory (TRANSP-OR), led by prof. Michel Bierlaire. TRANSP-OR has vast competence in modeling, optimization and simulation of transportation systems.
|Prof. Michel Bierlaire
|Tim Hillel (EPFL), Patrick Manser (SBB)