A City-Scale Dataset of Traffic Flows, Travel Times, and Urban Context
Provides a benchmark dataset for traffic prediction and urban computing research, but is incremental as it applies existing data fusion methods to a new city.
The authors present a multi-source traffic dataset from Padua, Italy, combining traffic volumes, travel times, and urban context data, validated to capture expected patterns like rush hours.
We present a multi-source traffic dataset derived from Automatic Vehicle Identification (AVI) recordings in Padua, Italy, spanning from February 2026 to April 2026. The dataset combines traffic volume time series, aggregated at 10-minute intervals, with time-varying trajectory-based flow statistics including transition probability matrices, average travel times, and flow residuals. To enrich the traffic measurements with urban contextual information, we integrate Points Of Interests (POIs), demographic data, meteorological variables, and road infrastructure data. All components are accessible through a Python class that loads temporal and contextual data exploiting a spatio-temporal graph representation. Validation analyses confirm that the dataset captures expected traffic patterns, such as morning and evening rush hours, as well as weekdays vs. weekend days traffic routines.