Urban mobility analysis and prediction for non-routine scenarios using digital footprints
In this project we propose to study individual’s mobility for mining non-routine (leisure, social, etc.) mobility patterns from multiple data sources. The following mobility patterns are of great interest: locations of significance, modes of transport, trajectory patterns and location-based activities for destination choice modelling. Data collected via ubiquitous devices and smart metering combined with data from social media platforms provides a range of new close-to-real-time information that can be combined with the data from more traditional sources (surveys, transport system records and static data) for urban efficient mobility planning and management. When considered in isolation, each of these data sources has gaps/missing observations, so the matching of multiple data sources can facilitate transport analysis, and enable operators to better tune – even on the fly – public transportation within cities with the aim of travelling at lower costs, faster and producing a smaller carbon footprint.