A new perspective for precision evaluation of large-scale traffic sensor data measurement

Abstract

ABSTRACT Use of sensor data has been increasingly common in recent years, yet there is still a knowledge gap in evaluating the precision of traffic sensor data being used in traffic analyses for developing a transportation management plan. This paper fills this gap by exploring a new approach to evaluating the level of precision of large-scale traffic sensor data. The proposed analytical framework incorporates a spatiotemporal domain for the purpose of projecting spatiotemporal characteristics of the data into a repeatability and reproducibility (R&R) study. The main finding of this study is that the proposed framework is effective in examining the precision level of large-scale data spatiotemporally. The proposed framework would be useful for researchers and practitioners to benchmark precision measurements of traffic sensor data in a way to gather quality data and avoid any potential biased result of deeper traffic analyses.