Home / Current Issue / Paper 1710940
Traffic Control and Management System Using Deep Learning
Subject area: Science,Engineering and Technology · Area of research: Traffic Control
DOI: https://doi.org/10.64388/IREV9I3-1710940-1524
Abstract
Traffic Control and Management System using Deep Learning is an Internet-of-Things (IoT)-based solution designed for the real-time optimization of urban traffic flow. The system combines IoT vehicle detection with infrared (IR) sensors and ESP32 microcontrollers that feed real-time traffic data through MQTT to a centralized traffic optimizer. Two deep-learning models, a lane classifier and a green-time regressor both implemented using CNN, compute the dynamic decision on which lane will receive the green signal and its optimum green-time duration. We used an HTML-based sensor simulator developed in Python for the training and validation of the models with realistic traffic scenarios. We developed an interactive web- based dashboard using HTML, CSS, and JavaScript to visualize the traffic status, signal timings, and performance evaluation. Experimental results reveal an increase in traffic flow following reduced waiting times, enhanced lane fairness, and an overall traffic efficiency increase, representing an intelligent analytics solution for Smart City traffic management with the potential for large-scale deployments.
Keywords
Internet of Things (IoT), Deep learning (DL), ESP32, Message Queuing Telemetry Transport(MQTT), Sensor simulator, Lane classification, Green-time regression, Smart city.
How to cite this paper
@article{1710940,
author = {Anurag Hans, Anjana Nair, Tanvir Hussain, Jinal Patel, Bharat Tank},
title = {Traffic Control and Management System Using Deep Learning},
journal = {Iconic Research And Engineering Journals},
year = {2025},
volume = {9},
number = {3},
pages = {1968-1976},
issn = {2456-8880},
url = {https://www.irejournals.com/formatedpaper/1710940.pdf},
abstract = {Traffic Control and Management System using Deep Learning is an Internet-of-Things (IoT)-based solution designed for the real-time optimization of urban traffic flow. The system combines IoT vehicle detection with infrared (IR) sensors and ESP32 microcontrollers that feed real-time traffic data through MQTT to a centralized traffic optimizer. Two deep-learning models, a lane classifier and a green-time regressor both implemented using CNN, compute the dynamic decision on which lane will receive the green signal and its optimum green-time duration. We used an HTML-based sensor simulator developed in Python for the training and validation of the models with realistic traffic scenarios. We developed an interactive web- based dashboard using HTML, CSS, and JavaScript to visualize the traffic status, signal timings, and performance evaluation. Experimental results reveal an increase in traffic flow following reduced waiting times, enhanced lane fairness, and an overall traffic efficiency increase, representing an intelligent analytics solution for Smart City traffic management with the potential for large-scale deployments.},
keywords = {Internet of Things (IoT), Deep learning (DL), ESP32, Message Queuing Telemetry Transport(MQTT), Sensor simulator, Lane classification, Green-time regression, Smart city.},
month = {September},
doi = {https://doi.org/10.64388/IREV9I3-1710940-1524}
}