How artificial intelligence can help reduce traffic congestion in Tashkent.
- 1Independent researcher
Abstract
This article discusses the growing problem of traffic congestion in Tashkent, its environmental and economic consequences, and the limitations of traditional approaches to addressing this issue. It examines measures taken to tackle the problem, including government policy amendments, the "Uzbekistan-2030" national strategy, and the installation of 4,000 smart traffic lights. The study also reviews international experience in integrating artificial intelligence (AI) into intelligent transport systems (ITS). As a potential solution, a multi-agent traffic control system based on a deep Q-learning algorithm is proposed. Simulations were conducted using the SUMO (Simulation of Urban MObility) platform, modelling an 11-intersection corridor along Amir Temur Avenue during peak hours (800–1,000 vehicles per hour). The simulation results indicate that the proposed "Predictive MADQL" algorithm reduced the average queue length to 16.0 vehicles - an improvement in efficiency of 40% compared to static-timer traffic lights (26.8 vehicles) and 24% compared to traditional sensor-based traffic lights (21.2 vehicles). The paper concludes with practical recommendations for the Tashkent Traffic Management Centre regarding the integration of artificial intelligence into intelligent transport systems.
Related
- Explainable artificial intelligence of things integrated spatio-temporal feature modeling for traffic congestion framework on intelligent urban mobility system
- An operational spatiotemporal prediction framework for sector congestion in air traffic management
- An optimal path-finding algorithm for African cities: a case study of Nairobi, Kenya, considering traffic congestion and air pollution
- Clinical features of COVID-19 patients hospitalized at the Tashkent State Medical University and risk factors for intensive care unit admission: a cross-sectional study from Uzbekistan, Central Asia
- Population-based registry analysis of cervical cancer trends in Tashkent, Uzbekistan 2014-2024.
BibTeX
@misc{maksudi2026artificial,
title = {{How artificial intelligence can help reduce traffic congestion in Tashkent.}},
author = {Musa Maksudi},
howpublished = {Zenodo},
year = {2026},
month = oct,
doi = {10.5281/zenodo.23623583},
url = {https://zenodo.org/records/23623583},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - How artificial intelligence can help reduce traffic congestion in Tashkent.
AU - Musa Maksudi
PY - 2026
DA - 2026/10/24
PB - Zenodo
DO - 10.5281/zenodo.23623583
UR - https://zenodo.org/records/23623583
AB - This article discusses the growing problem of traffic congestion in Tashkent, its environmental and economic consequences, and the limitations of traditional approaches to addressing this issue. It examines measures taken to tackle the problem, including government policy amendments, the "Uzbekistan-2030" national strategy, and the installation of 4,000 smart traffic lights. The study also reviews international experience in integrating artificial intelligence (AI) into intelligent transport systems (ITS). As a potential solution, a multi-agent traffic control system based on a deep Q-learning algorithm is proposed. Simulations were conducted using the SUMO (Simulation of Urban MObility) platform, modelling an 11-intersection corridor along Amir Temur Avenue during peak hours (800–1,000 vehicles per hour). The simulation results indicate that the proposed "Predictive MADQL" algorithm reduced the average queue length to 16.0 vehicles - an improvement in efficiency of 40% compared to static-timer traffic lights (26.8 vehicles) and 24% compared to traditional sensor-based traffic lights (21.2 vehicles). The paper concludes with practical recommendations for the Tashkent Traffic Management Centre regarding the integration of artificial intelligence into intelligent transport systems.
ER -