Vehicle Speed Monitoring System Based on Edge Computing

Traffic surveillance and speed monitoring are important to reduce accidents and save lives. Several systems have been proposed to monitor vehicle speed and determine violating ones. However, the accuracy of these systems still need improvement because they do not cover the entire road network, depend on complex estimation method, and can be largely affected by weather conditions and obstacles. Therefore, this paper proposes a vehicle speed monitoring system that utilizes the edge and cloud computing technologies. The system uses GPS to determine vehicle coordinates and the Haversine algorithm to estimate the speed. The system was evaluated using smartphones of different computation powers and equipped GPS, where each smartphone represents an edge vehicle. The results show that the system accurately estimates vehicle speed with an accuracy 96.7%. Also, edges with high computing performance outperform other edges.