Energy-optimizing machine learning-driven smart traffic control system for urban mobility and the implications for insurance and risk management
Abstract
Heavy traffic during peak hours, such as early mornings and late evenings, is a significant cause of delays for commuters. To address this issue, the prototype of a dual smart traffic light control system is constructed, capable of dynamically adjusting traffic signal duration based on real-time vehicle density at intersections, as well as the brightness of the streetlights. The system uses a pre-trained Haar Cascade machine learning classifier model to detect and count vehicles through a live video feed. Detected cars are highlighted with red squares, and their count is extracted. The vehicle data is then transmitted to an Arduino microcontroller via serial communication, facilitated by the pySerial library. The Arduino processes this information and adjusts the timing of the traffic lights accordingly, optimizing traffic flow based on current road conditions. A novel approach involves optimizing energy usage through real-time data integration with the power grid. Street lighting is then dynamically adjusted at night times—brightening during high-traffic periods and dimming during low-traffic times. The brightness levels are set at 30%, 50%, 75%, and 100% based on the number of cars detected, with above 50% indicating the presence of cars. This adaptive control enhances energy efficiency by reducing energy consumption while maintaining road safety. The simulated and experimental results are provided. The former demonstrated a lower accuracy compared to the latter, particularly during the transition to the green light, across all traffic density levels. Additionally, the simulation was only capable of representing discrete lamp brightness levels of 0%, 50%, and 100%, in contrast to the experimental results, which showed a clear differentiation between 50%, 75%, and 100% brightness levels. Details of the model limitations are outlined with proposed solutions. The implications of the optimized system for auto insurance, liability coverage, and risk management are explored. These are areas that are rarely addressed in current research.
References
[1]Thomas P, Welsh R, Mavromatis S, et al. Survey results: Road safety data, data collection systems and definitions. Safer Africa Project; 2017.
[2]Adeloye D, Thompson JY, Akanbi MA, et al. The burden of road traffic crashes, injuries and deaths in Africa: A systematic review and meta-analysis. Bulletin of the World Health Organization. 2016; 94(7): 510–521A. doi: 10.2471/BLT.15.163121
[3]Kazeem Y. Death rates from traffic accidents are higher in Africa than anywhere else. Available online: https://www.weforum.org/agenda/2019/02/death-rates-from-traffic-accidents-are-higher-in-africa-than-anywhere-else/ (accessed on 5 December 2024).
[4]Hartenstein H, Laberteaux LP. A tutorial survey on vehicular ad hoc networks. IEEE Communications Magazine. 2008; 46(6): 164–171.
[5]Yan G, Rawat DB. Vehicle-to-Vehicle Connectivity Analysis for Vehicular Ad-Hoc Networks. Ad Hoc Networks, 2016; 58 (C), 25 - 35. doi: 10.1016/j.adhoc.2016.11.017.
[6]Raya M, Hubaux JP. Securing vehicular adhoc networks. Journal of Computer Security. 2007; 15: 39–68.
[7]Taleb T, Samdanis K, Mada B, et al. On Multi-Access Edge Computing: A Survey of the Emerging 5G Network Edge Cloud Architecture and Orchestration. IEEE Communications Surveys & Tutorials, 2017; 19(3), 1657-1681. doi: 10.1109/COMST.2017.2705720.
[8]Veres M, Moussa M. Deep Learning for Intelligent Transportation Systems: A Survey of Emerging Trends. IEEE Transactions on Intelligent Transportation Systems, 2020; 21(8), 3152-3168. doi: 10.1109/TITS.2019.2929020
[9]Arifin S, Razi SA, Haque A, et al. A Microcontroller Based Intelligent Traffic Control System. American Journal of Embedded Systems and Applications. 2019; 7(1): 21–25. doi: 10.11648/j.ajesa.20190701.13
[10]Nyiekaa EA, Irokwe JN, Shaapera TM. Design and Implementation of a 4-Way Traffic Control and Detection System. International Journal of Innovations in Engineering and Science. 2022; 10(4): 16–27.
[11]Fadoro JO. Design and Implementation of Electronic Traffic Light. Villanova Journal of Science, Technology and Management. 2019; 1(1).
[12]Bisen K, Shahare R, Wasnik K, et al. Machine Learning-Based Intelligent Traffic System. International Research Journal of Modernization in Engineering Technology and Science. 2023; 5(4). doi: 10.56726/irjmets35340
[13]Balasubramanian SB, Balaji P, Munshi A, et al. Machine learning based IoT system for secure traffic management and accident detection in smart cities. PeerJ Computer Science. 2023; 9: e1259–e1259. doi: 0.7717/peerj-cs.1259
[14]Tiwari P. The machine learning framework for traffic management in smart cities. Management of Environmental Quality. 2024; 35(2): 445–462. doi: 10.1108/MEQ-08-2022-0242
[15]Khan H, Kushwah KK, Maurya MR, et al. Machine learning driven intelligent and self adaptive system for traffic management in smart cities. Computing. 2022; 104(5): 1203–1217. doi: 10.1007/s00607-021-01038-1
[16]Lanke N, Koul S. Smart Traffic Management System. International Journal of Computer Applications. 2013; 75(7): 19–22. doi: 10.5120/13123-0473
[17]Al-Abaid SAF. A Smart Traffic Control System Using Image Processing: A Review. Journal of Southwest Jiaotong University. 2020; 55(1). doi: 10.35741/issn.0258-2724.55.1.31
[18]Sobral AC. Vehicle Detection with Haar Cascades. Available online: https://github.com/andrewssobral/vehicle_detection_haarcascades?tab=readme-ov-file (accessed on 5 December 2024)
[19]Aulia U, Hasanuddin I, Dirhamsyah M, et al. A new CNN-BASED object detection system for autonomous mobile robots based on real-world vehicle datasets. Heliyon. 2024; 10(15). doi: 10.1016/j.heliyon.2024.e35247
[20]Khan D, Waqas M, Tahir M, et al. Revolutionizing Real-Time Object Detection: YOLO and MobileNet SSD Integration. Journal of Computing & Biomedical Informatics. 2023; 6(1). doi: 10.56979/601/2023
[21]Balasubramanian R, Libarikian A, McElhaney D. Insurance 2030—The impact of AI on the future of insurance. Available online: https://www.mckinsey.com/industries/financial-services/our-insights/insurance-2030-the-impact-of-ai-on-the-future-of-insurance (accessed on 5 December 2024).
[22]MAPFRE. The growing potential of technology in traffic accident reconstruction. Available online: https://www.mapfre.com/en/insights/innovation/technology-traffic-accident-reconstruction/ (accessed on 5 December 2024).
[23]Chen Y, Zhang Q, Yu F. Transforming traffic accident investigations: A virtual-real-fusion framework for intelligent 3D traffic accident reconstruction. Complex Intelligent Systems. 2025; 11. doi: 10.1007/s40747-024-01693-9
[24]Pádua L, Sousa J, Vanko J, et al. Digital Reconstitution of Road Traffic Accidents: A Flexible Methodology Relying on UAV Surveying and Complementary Strategies to Support Multiple Scenarios. International Journal of Environmental Research and Public Health. 2020; 17(6): 1868. doi: 10.3390/ijerph17061868
[25]Li HJ, Luo XG, Zhang ZL, et al. Driving risk prevention in usage-based insurance services based on interpretable machine learning and telematics data. Decision Support Systems. 2023; 172: 113985. doi: 10.1016/j.dss.2023.113985
[26]Ziakopoulos A, Petraki V, Kontaxi A, et al. The transformation of the insurance industry and road safety by driver safety behaviour telematics. Case Studies on Transport Policy. 2022; 10(4): 2271–2279. doi: 10.1016/j.cstp.2022.10.011
[27]Aldegheishem A, Yasmeen H, Maryam H, et al. Smart Road Traffic Accidents Reduction Strategy Based on Intelligent Transportation Systems (TARS). Sensors. 2018; 18(7): 1983. doi: 10.3390/s18071983
[28]Bhavana S, Praveen P. IoT-Enabled Intelligent Traffic Navigation with Accident Management for Critical Emergency Response in Heavy Congestion Using Machine Learning. Available online: https://advance.sagepub.com/users/810526/articles/1212008-iot-enabled-intelligent-traffic-navigationwith-accident-managementfor-critical-emergency-response-in-heavy-congestion-using-machine-learning (accessed on 5 December 2024).
Copyright (c) 2025 Author(s)
This work is licensed under a Creative Commons Attribution 4.0 International License.
Authors contributing to this journal agree to publish their articles under the Creative Commons Attribution 4.0 International License, allowing third parties to share their work (copy, distribute, transmit) and to adapt it for any purpose, even commercially, under the condition that the authors are given credit. With this license, authors hold the copyright.
