This paper describes methods to identify trip details, including the mode of transport for each trip, from smartphone app data and from mobile network data. Use cases include travel demand surveys, travel behaviour gamification, mobility-as-a-service and automated ticketing. In the context of transport planning, the paper examines solutions to protect privacy and to enhance the representativeness of mobile phone data samples. It makes recommendations to overcome the many obstacles involved, in particular the scarcity of annotated training data.
Inferring Modal Split from Mobile Phones
Principles, Issues and Policy Recommendations
