Developing travel behaviour models using mobile phone data by Andrew Bwambale

This presentation was delivered as part of the Institute for Transport Studies research seminar series. Mobile phone data has recently emerged as a very promising source of data for developing mobility models. Previous transport applications of mobile phone data have however focused primarily on extracting travel patterns and trends. This research aims to extend the application of mobile phone data to mainstream travel behaviour modelling and policy analysis by augmenting the data with information from other sources. This comes along with significant challenges arising from the anonymous and noisy nature of the data. Consequently, novel modelling frameworks have been developed in the context of trip generation, route choice and departure time choice modelling to address these challenges. These have been tested using different types of mobile phone data from Switzerland, Senegal and Bangladesh, providing us with rich insights into the strengths and weaknesses of the different data types. In all the cases, the model estimation results are consistent with the expected travel behaviour. Moreover, validation exercises with respect to hold-out samples and/or valuation metrics from other sources (such as the value of travel time) produce satisfactory results, thereby demonstrating the potential of mobile phone data for transport policy analysis. The findings are expected to be useful to transport practitioners - particularly those working in contexts where traditional data sources for transport modelling are scarce.

University of Leeds

11/12/2018 10:54:02 AM