Using Map Matching for an Online Traffic Congestion Application

Abstract

Optimising the flow of traffic on road networks can be achieved through sharing information about congestion events to road users. Using GPS sensors to track vehicle movements can provide information to monitor speeds and give an indication of where traffic delays are occurring. While GPS can give location estimate at a fine granularity, GPS readings also have a certain amount of positional error. This research successfully implemented a map matching algorithm to increase the accuracy of GPS data and position vehicles directly onto correct road links. These positions were used to estimate average speed of vehicles which was used as a proxy for road network performance. Areas that were experiencing speeds under the speed limit were considered to be congested. This information was presented on a user friendly and highly accessible web map. A small area in London and sample vehicle trajectories provided a way of testing and evaluating these concepts.