@article{Badran2024, 
author = {Adham Badran and Ahmed El-Geneidy and Luis Miranda-Moreno},
title = {Extracting networkwide road segment location, direction, and turning movement rules from global positioning system vehicle trajectory data for macrosimulation},
year = {2024},
journal = {Journal of Intelligent and Connected Vehicles},
volume = {7},
number = {4},
pages = {258-265},
keywords = {global positioning system (GPS), transport model, road network, intersection control, map inference, road direction, turning movement},
url = {https://www.sciopen.com/article/10.26599/JICV.2023.9210046},
doi = {10.26599/JICV.2023.9210046},
abstract = {The emergence of road users’ global positioning system (GPS) trajectory data is attracting increasing research interest in knowledge discovery to improve transport planning-related methods and tools. In fact, the widespread use of GPS-enabled smartphones and the mobile internet has increased the availability and size of such data. With the increase in GPS data coverage and availability, some research has expanded its use to estimate state-wide vehicle-miles travelled, to classify driving maneuvers for road safety assessment, or to estimate environmental performance indicators, such as vehicular fuel consumption and pollution emissions. In computer science, research has used GPS data to infer road network maps. Although the inferred maps provide a correct topology and connectivity, they lack the essential details to be used for transport modeling. Therefore, this work proposes a method to extract network-wide road direction and turning movement rules. In addition, building a road network model under the widely used macroscopic transport modeling software serves as a proof of concept. A sensitivity analysis was carried out to determine the output quality and recommend future improvements. Road segment geometry and directionality were extracted accurately (case study accuracy of 95%); however, turning movement rules can be extracted more accurately using a larger GPS vehicle trajectory sample (case study accuracy of 68%).}
}