Sustainable Edge AI and IoT-Based Smart Transportation System for Real-Time Road Condition Analysis

The advancements of intelligent transportation systems and sustainable urban infrastructure are discussed, focusing on the need for energy-efficient road monitoring solutions in real-time. Anomalies that may occur in the roads, such as potholes, cracks, bumps, and rough road surfaces can significantly impact the overall safety and effectiveness of the transportations and fuel consumption rates. Conventional approaches used for inspecting anomalies in roads are mainly manual processes, costly, time-consuming, and not designed for real-time transportation monitoring systems. Recent innovations and progressions in the fields of AI and IoT have improved the accuracy in anomaly detection but many existing solutions are based on cloud-centric designs, which increase the computational costs and energy usage.

The paper introduces a Sustainable AI-driven IoT framework for real-time detection of anomalies on roads and smart transportation systems. This framework uses multi-sensors IoT technology for gathering data, lightweight edge-based AI algorithm to analyse the collected data in real-time, federated learning model for training models and connected vehicle communication for transmitting data. The core principle behind this framework is green intelligence in transportation, where the goal of the system is to minimize communication costs and dependency on clouds while optimizing computing performance. Moreover, the framework allows for mobility and maintenance management and sustainable infrastructure design. Finally, important research gaps and potential research directions for smart transportation systems are identified