Explorative analysis of vehicular movement patterns using RFID-based transport data: An eulerian perspective

Abstract

The advancement in technology on data capture procedures has overcome many of the challenges associated with data acquisition for transportation studies. The use of Radio Frequency Identification (RFID) technology is increasingly becoming significant in transport application domains where there is the need to track and analyze patterns of vehicles movement. In this paper, we explore the efficacy of RFID technology, a eulerian perspective on movement, to extract spatial and temporal rhythms of vehicular movements in, Nanjing, China for road traffic analysis. Data mining and geo-computation methods were used to mine and extract vehicular movement. The count data, statistical, visual analytics and Geographic Information System (GIS) methods were used to determine spatial and temporal patterns of vehicular movement. Global Moran’s I, hot spot analysis and kernel density estimations were the spatial statistical methods used to determine spatial patterns of vehicular movements. The study reveals the efficacy of the usage of massive RFID data, which uses a eulerian perspective of movement for determining spatiotemporal patterns for traffic analysis. The study revealed morning peak and evening peak vehicular movements, for weekdays with Thursdays and Fridays displaying the most vehicular movements. Spatial patterns revealed a clustering of low and high vehicular counts for weekdays, weekends, off-peak and peak hours. This explorative study using RFID technology to determine spatial and temporal patterns in vehicular counts has important application for traffic analysts. This study approach supports traffic congestion monitoring, traffic flow statistics and traffic planning as well as helps to determine low and high traffic locations to evaluate the performance of a traffic system.