AI-Driven Network Serviceability Mapping Using eGIS and Streaming Data Pipeline

In the face of growing urban complexity and critical infrastructure demands, ensuring resilient and real-time network serviceability is more vital than ever.This review explores how the convergence of Artificial Intelligence (AI), electronic Geographic Information Systems (eGIS), and streaming data pipelines can revolutionize network monitoring and fault prediction.The paper systematically analyzes various AI models-such as Random Forests, LSTM, and Graph Neural Networks (GNNs)applied within geospatial frameworks and assesses their effectiveness using real-world datasets and simulations.GNNs emerged as the most capable models in capturing the spatial topology of urban networks.Furthermore, the review examines how platforms like Apache Kafka and ArcGIS enable scalable, near-real-time solutions for service degradation detection.Key challenges such as data heterogeneity, model explainability, and privacy concerns are identified, with future directions pointing toward Edge AI, federated learning, and autonomous multi-agent systems.The findings suggest that AI-eGIS architectures are not only effective but essential in shaping the next generation of smart and resilient infrastructures.