TOWARDS DATA-DRIVEN SIMULATIONS IN URBAN MOBILITY ANALYTICS
Frédéric Roulland, Cesar De Souza, Luis Ulloa, Arturo Mondragon, Michael Niemaz, Victor Ciriza
In this work, we present recent advances on creating data-driven models to address transportation planners’ needs to better understand mobility and simulate and predict the impact of changes. This new modelling approach leverages the massive amount of data collected in the field from the daily users transactions and sensors output, and proposes to use in a more extensive way the machine learning techniques that have emerged over the last decade. We first present how this new approach of modelling transportation differs from the traditional practices. We then illustrate it through some specific use cases where we have applied it and the preliminary results we obtained. We finally end up with a discussion highlighting the main advantages and the high potential of the adoption of such approach in transportation planning domain but also the main obstacles to be overcome before a large adoption can happen.
14th ITS Asia Pacific Forum, Nanjing, China, April 27-29, 2015.
2015-011.pdf (756.41 kB)