Smart Traffic Signal Management System: An AI-Powered Real-Time Traffic Monitoring and Adaptive Signal Control System

Abstract

Traffic congestion is a major challenge in urban areas, particularly at road intersections where a large number of vehicles compete for limited road space. Conventional traffic signal systems generally operate using predefined time intervals, which do not always reflect the actual traffic conditions on the road. As a result, vehicles may wait unnecessarily on less congested roads while heavily congested roads continue to face long queues. To address this issue, this paper presents a Smart Traffic Signal Management System that uses artificial intelligence and computer vision techniques for real-time traffic monitoring and adaptive signal control. The proposed system uses camera-based video input to continuously observe traffic at a four-way intersection. A YOLO-based object detection model is used to identify and count vehicles from the captured video, while the detected vehicle information is analyzed to estimate the traffic density and queue level on each road. Based on the calculated traffic condition, the system dynamically determines the priority and duration of the green signal for different directions. This approach allows the signal timing to respond to changing traffic conditions instead of relying only on fixed schedules. The system also includes a monitoring interface through which authorized traffic personnel can observe vehicle counts, traffic density, signal status, and other relevant information. A manual override facility is provided to allow human intervention in situations where the automated detection system is affected by poor visibility, unusual traffic conditions, or other operational issues. The collected traffic information can be stored in a centralized database and used for generating periodic reports and analyzing traffic patterns. The proposed system was developed and tested in a controlled environment using traffic video data to evaluate its ability to detect vehicles and determine traffic density. The results demonstrate that an AI-assisted adaptive signal system can provide more responsive traffic management compared with conventional fixed-time signal operation. The system can potentially reduce unnecessary vehicle waiting time, improve the utilization of green signal periods, and provide traffic authorities with useful real-time information for intersection management.

Introduction

The research proposes a Smart Traffic Signal Management System using Artificial Intelligence (AI) and Computer Vision to reduce traffic congestion and improve vehicle movement at four-way intersections. Unlike traditional traffic signals that operate using fixed timing schedules, the proposed system uses a YOLO-based object detection model to detect and count vehicles, estimate traffic density, and dynamically allocate green-light durations according to real-time traffic conditions.

1. Introduction and Problem Statement

The rapid growth of urban populations and vehicle numbers has increased traffic congestion, waiting times, fuel consumption, and air pollution. Conventional traffic signals use predefined timings without considering the actual number of vehicles waiting on each road.

This can result in heavily congested roads receiving insufficient green-light time while less busy roads receive unnecessary signal time. Manual traffic monitoring also requires considerable human effort and makes consistent data collection difficult.

The proposed system addresses these challenges by automatically analyzing traffic conditions and adjusting signal timings accordingly.

2. Proposed System

The system uses cameras to capture live or recorded traffic video at a four-way intersection. A YOLO-based object detection model identifies different types of vehicles, including cars, buses, trucks, and motorcycles.

The detected vehicles are counted separately for each direction, and their numbers are used to estimate traffic density. The system then prioritizes roads with heavier traffic and allocates suitable green-light durations while maintaining the required signal sequence.

Its main objectives are to:

• Reduce unnecessary vehicle waiting time and congestion.

• Improve the utilization of green-light durations.

• Automate vehicle detection and counting.

• Provide real-time traffic monitoring.

• Allow authorized personnel to override automatic signal control.

• Store traffic data for reporting and future analysis.

3. System Architecture and Methodology

The system follows a continuous processing pipeline:

1

Traffic Video Acquisition

Capture live or recorded intersection footage.

2

YOLO-Based Vehicle Detection

Identify vehicles in video frames.

3

Vehicle Counting

Count vehicles separately for each road.

4

Traffic Density Estimation

Determine which directions have heavier traffic.

5

Adaptive Signal Timing

Allocate green-light time according to traffic conditions.

6

Monitoring and Manual Override

Allow traffic officers to monitor or take control.

7

Database and Reporting

Store traffic records for analysis and reporting.

The process continuously updates vehicle counts and traffic-density estimates. Signal decisions are adjusted as conditions change, with yellow and red phases incorporated into the signal sequence.

4. AI-Assisted Features

The proposed system incorporates several intelligent functions:

• Vehicle detection: YOLO identifies vehicles from camera footage using bounding boxes.

• Automatic vehicle counting: Counts detected vehicles in each direction without manual observation.

• Traffic-density estimation: Classifies traffic conditions and identifies potentially congested roads.

• Adaptive signal control: Assigns longer green-light durations to directions with greater traffic demand.

• Real-time analysis: Updates traffic information as new video frames are processed.

• Human-assisted control: Allows authorized traffic officers to intervene during emergencies or detection failures.

• Traffic analytics: Maintains records that can support congestion analysis and planning.

5. Administrative Structure

The system provides three levels of access to support security and operational control.

• System Administrator: Manages users, system configuration, database records, and overall application operation.

• Traffic Control Officer: Monitors vehicle counts, traffic density, and signal status, and operates manual controls when necessary.

• Higher-Level Officials: Review stored records, reports, and traffic patterns for administrative decisions and future planning.

This role-based structure helps restrict access to critical controls and maintains human supervision of automated traffic management.

6. Database and Reporting

A centralized database can store vehicle counts, traffic-density measurements, signal activities, user information, and system events. These records can be used to generate periodic reports, identify frequently congested directions, and support data-driven traffic planning.

Conclusion

The proposed Smart Traffic Signal Management System provides an AI-based approach for monitoring and managing traffic at road intersections. The system uses camera-based video input and a YOLO-based object detection model to identify and count vehicles on different roads. The detected vehicle information is used to estimate traffic density and support the selection of suitable signal timings. [1] Unlike conventional fixed-time traffic signals, the proposed system considers the current traffic condition while deciding signal priority. This can help reduce unnecessary waiting time on roads with higher vehicle density and make better use of the available green-light duration. The real-time monitoring interface also allows authorized traffic personnel to observe vehicle counts, traffic density, and signal status from a single platform. [10] The inclusion of a manual override facility provides an important layer of human control when the automated system is affected by camera limitations, unusual traffic conditions, emergencies, or other unexpected situations. The ability to store traffic information in a centralized database also provides useful data for monitoring, reporting, and future traffic analysis. Overall, the system demonstrates the practical use of Artificial Intelligence, Computer Vision, and adaptive signal control for improving traffic management. Although the system is primarily developed and evaluated in a controlled environment, the approach provides a foundation for developing more advanced intelligent traffic management solutions.

References

[1] A. Bochkovskiy, C. Y. Wang, and H. Y. M. Liao, “YOLOv4: Optimal Speed and Accuracy of Object Detection,” arXiv preprint arXiv:2004.10934, 2020. [2] A. Agrahari, M. M. Dhabu, P. S. Deshpande, A. Tiwari, M. A. Baig, and A. D. Sawarkar, “Artificial Intelligence-Based Adaptive Traffic Signal Control System: A Comprehensive Review,” Electronics, vol. 13, no. 19, 3875, 2024. [3] Y. Wang and X. Yang, “A Review of the Self-Adaptive Traffic Signal Control System Based on Future Traffic Environment,” Journal of Advanced Transportation, 2018. [4] M. Mileti?, E. Ivanjko, M. Greguri?, and K. Kuši?, “A Review of Reinforcement Learning Applications in Adaptive Traffic Signal Control,” IET Intelligent Transport Systems, vol. 16, no. 10, pp. 1269–1285, 2022. [5] S. Gupte, O. Masoud, R. F. K. Martin, and N. P. Papanikolopoulos, “Detection and Classification of Vehicles,” IEEE Transactions on Intelligent Transportation Systems, vol. 3, no. 1, pp. 37–47, 2002. [6] M. A. Haque, M. M. Rahman, and M. A. Hossain, “Vehicle Detection and Tracking in Traffic Surveillance Systems,” International Journal of Computer Applications, 2014. [7] G. Jocher, A. Chaurasia, and J. Qiu, “Ultralytics YOLOv8,” Ultralytics, 2023. [8] J. Guo, L. Li, and X. Ban, “Urban Traffic Signal Control with Connected and Automated Vehicles: A Survey,” Transportation Research Part C: Emerging Technologies, vol. 101, pp. 313–334, 2019. [9] A. M. Shams, A. M. T. Emtenan, and C. M. Day, “Review and Taxonomy of Advanced Traffic Signal Control,” Journal of Transportation Engineering, Part A: Systems, vol. 152, no. 8, 2026. [10] J. Li, L. Peng, K. Hou, Y. Tian, Y. Ma, S. Xu, and T. Z. Qiu, “Adaptive Signal Control and Coordination for Urban Traffic Control in a Connected Vehicle Environment: A Review,” Digital Transportation and Safety, vol. 2, no. 2, pp. 89–111, 2023. [11] R. Szeliski, “Computer Vision: Algorithms and Applications,” Springer International Publishing, 2nd Edition, 2022. [12] R. C. Gonzalez and R. E. Woods, “Digital Image Processing,” Pearson Education, 4th Edition, 2018. [13] I. Goodfellow, Y. Bengio, and A. Courville, “Deep Learning,” MIT Press, Cambridge, Massachusetts, 2016. [14] M. Abadi et al., “TensorFlow: A System for Large-Scale Machine Learning,” 12th USENIX Symposium on Operating Systems Design and Implementation (OSDI), pp. 265–283, 2016. [15] M. Treiber, A. Hennecke, and D. Helbing, “Congested Traffic States in Empirical Observations and Microscopic Simulations,” Physical Review E, Vol. 62, No. 2, pp. 1805–1824, 2000. [16] D. C. Gazis, “Traffic Theory,” Kluwer Academic Publishers, Boston, Massachusetts, 2002. [17] Google, “Cloud Firestore Documentation: Cloud-Based NoSQL Database for Application Development,” Firebase Documentation, Google, 2026.

Copyright

Copyright © 2026 Mr. A. Alagar, Mrs. M. Nithya, M. B. Suriya Vijay, M. Durgesh Rajan. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.