Efficient Map-Matching for Urban Freight Applications

We present a framework to simplify the process of offline map-matching, applied to real GPS data from vans in Oslo and Akershus in Norway. Filters are established to exclude GPS points associated with stops and minor movements. We also highlight challenges related to sparse and noisy GPS data and discuss advantages and disadvantages using different open-source map-matching tools. Map-matching resulted in a driving distance 11% longer, compared to straight-line GPS measurements. Furthermore, we review existing research on Vehicle Kilometres Travelled (VKT) and ‘cruising for parking’ for commercial vehicles and provide practical examples of how such analysis may be implemented using map-matching. Regarding ‘cruising for parking’ analyses for commercial vehicles, we find little or no research applying map-matching techniques. Therefore, we see a potential to supplement ‘cruising for parking’ research on commercial vehicles, by utilizing advanced map-matching techniques.