A Review and Analysis of Traffic Data Sources

Abstract

Transportation is essential for economic and social development, and vehicle flow data can be used for safety monitoring, pollution analysis, and traffic flow management. Unfortunately, traffic management and control centres do not always comply with codified standards, making it difficult to obtain up-to-date data. This paper analyses open traffic datasets and Italian public traffic data sources available online, providing a knowledge base for transportation managers and researchers. Open traffic datasets are dimensionality-reduced and clustered. An event with 209,135 visitors is used to benchmark the public data sources, the time series of traffic flows are decomposed and a regression tree is used to identify different periods. The results suggest that the available Italian sensor grid is not fine enough to identify all incoming and outgoing traffic, more infrastructure investments are required or the available measurements should be coupled with other evaluation approaches capable of extending the punctual data through mathematical means.