摘要：AbstractUrbanization growth, together with the limited capacity of the road network, has worsened traffic congestion inside the cities. To improve the urban traffic conditions, it is essential to better understand and measure urban-traffic behavior not only on different period time of a day (e.g., morning or evening peaks) but also on different days (i.e., Monday to Sunday). Using real and recent data from the Google Maps API, this paper proposes a new approach to estimate the speeds within defined geographical areas (i.e. zip codes) per daytime and per weekday. Using this input a statistical analysis including k-means clustering is adopted to classify and define different urban-congestion levels according to the estimated speeds the number of inhabitants the zone types and the type of roads in each zip code. In order to validate our approach, we conduct an experimental analysis in Boston, US. Our results provide managerial insights for key stakeholders (i.e., Carriers, Consumers, and Government) to improve the efficiency of the road network and reduce traffic congestion in cities.
关键词：KeywordsUrban logisticslast-mile deliveryGoogle APIstatistical analysiscase study
