An Intelligent Edge-AI Assisted Smart Pedestrian Crossings Framework for Smart Transport System of Smart Cities

The rapid urbanization necessitates effective management, driving the development of smart cities. A key component of smart cities is a secure and efficient transportation system, which is vital for residents’ well-being. However, the growing complexity of road networks and increased vehicle volume present significant challenges in maintaining service quality. Pedestrian crossings, especially for children and the elderly, remain problematic in urban traffic worldwide. To address this, we propose an intelligent pedestrian crossing system that leverages advanced technology and real-time data processing to improve service quality and sustainability. Despite challenges such as data management, infrastructure costs, and privacy concerns, smart cities are adopting cloud and fog-based solutions, like smart traffic signals and intelligent pedestrian crossings, to enhance traffic flow and safety. This work presents an Edge-AI-assisted framework for pedestrian crossings, using edge devices as roadside units to collect real-time data. Images captured by cameras are processed with YOLOv8, an object detection algorithm, to analyze road and pedestrian crossing traffic. The edge-AI assisted smart pedestrian crossings Framework adapts the flow of objects based on the real-time data and objects over pedestrian. Simulation results demonstrate its effectiveness across various complex road transport datasets.