A new visualization approach for processing the vehicle movement data received an award as one of the best papers at the PacificVis 2018 conference.

How do vehicles move through a city? This question is interesting not only for urban planners, but also for owners of vehicle fleets, emergency and security agencies, or general companies. Today, taxis and other fleet vehicles record large quantities of movement data via GPS. One requirement for further analysis is usually that the GPS measurements are mapped to a route through the road network. But the measurement is subject to uncertainties and the sample rate might be low. Automatic procedures — called map matching — try to compensate for uncertainties and assign a specific route. However, these algorithms are complex to configure and have to be tailored to recording characteristics and local specifics of the city.

In a joint project with the University of Stuttgart, the working group of Prof. Dr. Fabian Beck developed an approach to simplify the configuration of a map matching algorithm and to make it more user-friendly. They employed various visualization techniques to reveal errors in the matching process and to suggest automatic improvements to the configuration. The paper describing the approach was accepted for publication and presentation at the international visualization conference PacificVis. As one of the best contributions of the conference, the paper has been selected for direct publication in the renowned journal IEEE Transactions on Visualization and Computer Graphics and received an "Honorable Mention Award".

Learn more about the publication.