New EPFL model improves prediction of commuter behavior
Published: Wednesday, Apr 10th 2024, 11:31
Volver a Live Feed
A new forecasting model developed by a Lausanne engineer predicts everyday travel decisions more realistically than previous models. It integrates people's reactions to unforeseen events, such as train cancellations.
City officials could use the model for their long-term planning to determine what kind of transportation infrastructure should be developed, the Swiss Federal Institute of Technology in Lausanne (EPFL) announced on Wednesday.
With more and more people working from home, car-sharing schemes being introduced and infrastructure being improved to allow employees to live further away from their employer, commuting patterns have changed significantly, according to EPFL. EPFL engineer Janody Puogala wanted to respond to these changes with her new model.
Hypotheses from sociology and statistics
In contrast to conventional traffic forecasting models, the new model does not consider a person's individual travel decisions separately, but plans activities over the course of an entire day. This makes it possible to take into account interactions between different modes of transportation, according to EPFL. For example, if a train is canceled, a person could switch to the bus instead of waiting for a replacement train.
This is a new way of modeling trade-offs, according to the EPFL. "My model can predict how different people would react under these circumstances and how long they would tolerate situations that they don't actually like," Pougala was quoted as saying in the press release.
Pougala took behavioral hypotheses from sociology and translated them into mathematical equations. She then combined the equations with data from commuter surveys and statistics so that the model could produce forecasts that were as realistic as possible.
©Keystone/SDA
Spanish 
English 
German 
French 
Arabic