The Influence of Location Factors on C-ITS Effectiveness: A Simulation-Based Study of GLOSA Under Varying Conditions

Abstract

Cooperative Intelligent Transport Systems (C-ITS) services, such as Green Light Optimal Speed Advisory (GLOSA), are technologies designed to address current challenges in the transport sector, particularly regarding traffic efficiency, safety, and sustainability. Beyond the technical configuration of C-ITS, the implementation context plays a crucial role in shaping the extent of their benefits. Understanding how these site-specific conditions influence C-ITS effectiveness is essential for optimizing deployments and ensuring that anticipated effects are achieved. This study proposes a methodology for analyzing and modeling the influence of site-specific factors on GLOSA performance. The effects of GLOSA on travel time, CO₂ emissions, and the number of stops were evaluated through traffic simulations conducted across multiple test sites. Key factors include (i) the number of lanes, (ii) the number of traffic signals, and (iii) the traffic flow. The simulation results show that GLOSA performance varies notably under different implementation conditions, highlighting the importance of context-dependent deployment strategies. To ensure robust and generalizable insights, multiple regression models were applied using k-fold cross-validation. Thereby, tree-based regression models such as gradient boosting and random forests demonstrated high accuracy and stable results. Additionally, cluster analyses were conducted to group locations with similar outcomes and identify patterns indicative of favorable implementation conditions. Sites characterized by higher traffic volumes, single-lanes, and several consecutive traffic lights demonstrated the greatest improvements in efficiency and emissions. Overall, this research supports developing context-dependent C-ITS implementation strategies and demonstrates how site-specific characteristics can guide deployment decisions.