ROAD TRAFFIC PLANNING IN THE CONTEXT OF THE SUSTAINABLE URBAN TRANSPORT SYSTEM

Shepelev, Vladimir; Almetova, Zlata; Korzan, Mikhail; Charbadze, Irakli · 2020 · Crossref

DOI: 10.14529/em200218

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study addresses the challenge of optimizing urban road traffic capacity, specifically focusing on the negative impact of heavy freight vehicles on intersection throughput. Traditional methods for assessing traffic flow rely on static statistical data, which often fail to capture real-time dynamics. The authors propose a solution within the context of a "Sustainable Urban Transport System" by leveraging Big Data and computer vision to monitor traffic in real time. The research aims to improve decision-making for traffic organization by identifying how freight vehicle presence correlates with congestion and reduced capacity at street-road network nodes. The methodology employs an Artificial Intelligence Monitoring System (AIMS) utilizing neural networks to process video data from street surveillance cameras. The researchers selected a high-traffic intersection in Chelyabinsk, Russia (Pobedy Avenue and Molodogvardeytsev Street) for analysis. To train the system, they created a dataset of 1,000 labeled images, categorizing vehicles into five types: cars, trucks, trams, minibuses, and buses. The technical infrastructure included the Darknet framework, OpenCV library, and NVIDIA CUDA for GPU-accelerated computation, using the YOLOv3 architecture for object detection. This setup allowed for the real-time classification and counting of vehicles, enabling the analysis of daily traffic intensity patterns and the specific influence of freight vehicles on lane capacity. The results demonstrate a significant correlation between the presence of freight vehicles and reduced intersection capacity. Data analysis revealed that freight traffic peaks during morning and afternoon hours, with the highest concentrations occurring in the western (51%) and eastern (29%) directions. The study found that freight vehicles in the eastern direction reduced passenger car intensity by 9.5% per hour during peak times. More severely, freight presence in the western direction reduced node capacity by 30% on the second lane and by 36% on the third lane. Calculations indicated that the actual saturation flow was approximately 20% lower than the ideal theoretical value due to these disruptions. The authors modeled two mitigation strategies: constructing additional lanes, which would cost roughly 9 million rubles and reduce congestion depth by 20–25%, and restricting freight vehicle access during peak hours. The latter was identified as the more efficient solution, potentially increasing capacity by up to 50% in certain lanes and reducing congestion depth by 25–30% with minimal financial and environmental impact. The significance of this work lies in its validation of real-time computer vision as a viable tool for urban traffic management. By establishing a quantitative link between freight vehicle density and intersection capacity, the study provides a data-driven basis for implementing dynamic traffic restrictions. The proposed AIMS framework not only enhances throughput but also supports sustainable urban planning by reducing fuel consumption and emissions associated with idling traffic. The findings suggest that integrating AI-driven monitoring into traffic infrastructure offers a cost-effective alternative to physical road expansion, contributing to more efficient and environmentally friendly urban transport systems.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• traffic density