Telematics data for geospatial and temporal mapping of urban mobility: New insights into travel characteristics and vehicle specific power

This paper describes a new approach for understanding urban mobility called geospatial and temporal (GeoST) mapping, which translates telematics (location) data into travel characteristics. The approach provides the speedacceleration profile of transport flow at high spatial and temporal resolution. The speed-acceleration profiles can be converted to vehicle-specific power (VSP), which can be used to estimate vehicle emissions. The underlying data used in the model is retrieved from a large telematics dataset, which was collected from GPS-connected vehicles during their journeys over the UK's West Midlands region road network for the years 2016 and 2018. Single journey telematics data were geospatially aggregated and then distributed over GeoST-segments. In total, approximately 354,000 GeoST-segments, covering over 17,700 km of roads over 35 timeslots are parameterized. GeoST mapping of the average vehicle speed (traffic flow), and VSP over different road types are analysed. The role of road slope upon VSP is estimated for every GeoST-segment through knowledge of the elevation of the start and end points of the segments. Previously, road slope and its effect upon VSP have been typically ignored in transport and urban planning studies. A series of case studies are presented that highlight the power of the new approach over differing spatial and temporal scales. For example, results show that the total vehicle fleet moved faster by an average of 3% in 2016 compared to 2018. The studied roads at weekends are shown to be less safe, compared to weekdays, because of lower adherence to speed limits. By including road slope in VSP calculations, the annually averaged VSP results differ by +12.6%, +14.3%, and + 12.7% for motorways, primary roads, and secondary roads, respectively, when compared to calculations that ignore road slope.