Edge-Cloud Artificial Intelligence predictive maintenance model for road infrastructure management

The deterioration of road infrastructure presents a persistent challenge to transportation efficiency, safety, and economic productivity, particularly in resource-constrained environments where maintenance is often reactive rather than preventive. This study proposes an edge-cloud artificial intelligence predictive maintenance conceptual model designed to enhance real-time monitoring, early fault detection, and proactive decision-making for road infrastructure systems using a qualitative, design science research (DSR) approach. The model integrates edge computing nodes, Internet of Things (IoT) sensor networks, and lightweight machine learning deployed at the edge to enable low-latency data processing and reduce dependence on centralized cloud infrastructures. By leveraging continuous data streams, the model will facilitate timely identification of structural anomalies including cracks, potholes, and subsurface degradation. The proposed model further incorporates adaptive learning mechanisms to improve predictive accuracy over time while optimizing bandwidth usage and operational costs. In addition, a decision-support layer is introduced to assist infrastructure managers in prioritizing maintenance interventions based on risk and severity indices. The conceptual model demonstrates the potential of decentralized AI architectures to transform traditional road maintenance practices into intelligent, predictive, and scalable systems suitable for both urban and rural contexts.