Infra-Red Sensor-Based Technology for Collecting Speed and Headway Data on Highways Under Mixed Traffic Conditions

Abstract

Traffic data is the new oil for designing highway facilities. Collecting reliable and precise traffic data is a challenging task at locations where a higher degree of traffic heterogeneity prevails. To address the traffic issues effectively, accurate and real-time vehicle recording is essential. Conventional techniques such as videography may be used for collecting traffic data under lane-disciplined and homogeneous traffic conditions, but these techniques are no longer considered suitable for Indian traffic conditions because of plying of the wide variety of vehicle types, which follow weak lane-discipline. One of the promising and robust techniques for collecting traffic data on highways is the Infra-Red (IR) sensor-based technique. Due to the non-intrusive nature of the IR sensor-based devices, they are used for collecting continuous and massive traffic data. In this study, using a traffic detector device based on IR sensors, named Transportable InfraRed Traffic Logger (TIRTL), a large amount of accurate speed and headway data were obtained on Indian highways that exhibits heterogeneity in its traffic composition. Using this data, the probability distribution functions of speed and headway for various traffic flow and density levels have been studied. Later, the goodness-of-fit test (Kolmogorov-Smirnov test) is used to find the best fitted probability distribution function to assess the statistical validity of each fitted probability distribution function. The results of the study show that speed and headway data follow different probability distribution functions under different traffic flow and traffic density levels.