Knowledge CentreTechnical Resources SearchConference PapersMeasuring Congestion Using Large-Scale Smartphone-Collected GPS Data in an Urban Road Network

Measuring Congestion Using Large-Scale Smartphone-Collected GPS Data in an Urban Road Network


Although travel time is an important performance measure for transportation professionals, congestion is perhaps a more important consideration for road users. Congestion is a dynamic phenomenon with variation across both space and time making it a promising application of smartphone-collected GPS data. The purpose of this study is to utilize GPS data collected using a smartphone application and regular drivers to estimate congestion at both the macroscopic and microscopic level across an urban road network. Data is collected using the “Mon Trajet” smartphone application in Quebec City, Canada. This data consists of nearly 50,000 trips collected during 3 weeks from approximately 5000 drivers. The application allowed for the collection of a large number of trips from regular drivers using a system that minimally impacts them or their behaviour. Given the large spatio-temporal dimension, several data issues are identified and corrected using the presented methodology. First, position of the GPS traces is provided in terms of a latitude and longitude and is not linked spatially to the road network. TrackMatching is a commercially available map-matching service used to match GPS data to the OpenStreetMap road network. Speeds are smoothed, and level of congestion is computed using the Congestion Index (CI). CI is computed at the individual link level at time intervals of 1 hour, to yield a detailed picture of congestion both across time and space. Finally, the progression of congestion over time is mapped across the entire road network and congestion variability across different time scales is computed.

Conference Paper Details

Session title:
Big Data Applications for Travel Demand Management
Stipancic, J.
Miranda-Moreno, L.
Labbe, A.
Transportation planning