Energy-optimizing machine learning-driven smart traffic control system for urban mobility and the implications for insurance and risk management

Information System and Smart City, Год журнала: 2025, Номер 5(1), С. 2253 - 2253

Опубликована: Фев. 27, 2025

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 dual smart light control system constructed, capable dynamically adjusting signal duration based on real-time vehicle density at intersections, well brightness streetlights. The uses pre-trained Haar Cascade machine learning classifier model to detect count vehicles through live video feed. Detected cars are highlighted with red squares, their extracted. data then transmitted an Arduino microcontroller via serial communication, facilitated by pySerial library. processes information adjusts timing lights accordingly, optimizing flow current road conditions. A novel approach involves energy usage integration power grid. Street lighting adjusted night times—brightening high-traffic periods dimming low-traffic times. levels set 30%, 50%, 75%, 100% number detected, above 50% indicating presence cars. This adaptive enhances efficiency reducing consumption while maintaining safety. simulated experimental results provided. former demonstrated lower accuracy compared latter, particularly transition green light, across all levels. Additionally, simulation was only representing discrete lamp 0%, 100%, in contrast results, which showed clear differentiation between Details limitations outlined proposed solutions. implications optimized auto insurance, liability coverage, risk management explored. These areas that rarely addressed research.

Язык: Английский

Digital Reconstruction Method for Low-Illumination Road Traffic Accident Scenes Using UAV and Auxiliary Equipment

World Electric Vehicle Journal, Год журнала: 2025, Номер 16(3), С. 171 - 171

Опубликована: Март 14, 2025

In low-illumination environments, traditional traffic accident survey methods struggle to obtain high-quality data. This paper proposes a reconstruction method utilizing an unmanned aerial vehicle (UAV) and auxiliary equipment. Firstly, methodological framework for investigating accidents under conditions is developed. Accidents are classified based on the presence of obstructions, corresponding investigation strategies formulated. As unobstructed scene, UAV-mounted LiDAR scans site generate comprehensive point cloud model. partially obstructed ground-based mobile laser scanner complements areas that obscured or inaccessible LiDAR. Subsequently, collected data processed with multiscale voxel iteration down-sampling determine optimal parameters. Then, improved normal distributions transform (NDT) algorithm different filtering algorithms adopted register ground air clouds, combination selected, thus, reconstruct high-precision 3D model scene. Finally, two nighttime scenarios conducted. DJI Zenmuse L1 UAV system EinScan Pro 2X selected reconstruction. both experiments, proposed achieved RMSE values 0.0427 m 0.0451 m, outperforming photogrammetry-based modeling 0.0466 0.0581 m. The results demonstrate this can efficiently accurately investigate scenes without being affected by providing valuable technical support refined management analysis. Moreover, challenges future research directions discussed.

Язык: Английский