City logistics routing requires time-dependent travel times for each network link. We rely on the concept of Floating Car Data (FCD) to develop and provide such travel times. Different levels of aggregation in the determination of time-dependent travel times from a database of historical FCD are presented and evaluated with regard to routing quality. Furthermore, a Data Mining approach is introduced, allowing for a substantial reduction of the volume of input data required for city logistics routing. The different approaches are investigated and evaluated by a huge amount of FCD collected for the urban area of Stuttgart, Germany. The results show that the Data Mining approach enables efficient provision of time-dependent travel times without a significant loss of routing quality for city logistics applications. © 2011 Elsevier Ltd.
https://doi.org/10.1016/j.trc.2011.11.004Cite as:
@article{Ehmke_2012,
doi = {10.1016/j.trc.2011.11.004},
url = {https://doi.org/10.1016%2Fj.trc.2011.11.004},
year = 2012,
month = {apr},
publisher = {Elsevier {BV}},
volume = {21},
number = {1},
pages = {338--352},
author = {Jan Fabian Ehmke and Stephan Meisel and Dirk Christian Mattfeld},
title = {Floating car based travel times for city logistics},
journal = {Transportation Research Part C: Emerging Technologies}
}