Shanghai Institute of Disaster Prevention and Relief

The integrity of urban water supply pipelines, an essential element of municipal infrastructure, is frequently undermined by internal defects such as corrosion, tuberculation, and foreign matter. Traditional inspection methods relying on CCTV are time-consuming, labor-intensive, and prone to subjective interpretation, which hinders the timely and accurate assessment of pipeline conditions. This study proposes YOLOv8-VSW, a systematically optimized and lightweight model based on YOLOv8 for automated defect detection in in-service pipelines. The framework is twofold: First, to overcome data limitations, a specialized defect dataset was constructed and augmented using photometric transformation, affine transformation, and noise injection. Second, the model architecture was improved on three levels: a VanillaNet backbone was adopted for lightweighting, a C2f-Star module was introduced to enhance multi-scale feature fusion, and the WIoUv3 dynamic loss function was employed to improve robustness under complex imaging conditions. Experimental results demonstrate the superior performance of the proposed YOLOv8-VSW model. This study validates the framework on a curated, real-world image dataset, where YOLOv8-VSW achieved mAP@50 of 83.5%, a 4.0% improvement over the baseline. Concurrently, GFLOPs were reduced by approximately 38.9%, while the inference speed was increased to 603.8 FPS. The findings validate the effectiveness of the proposed method, delivering a solution that effectively balances detection accuracy, computational efficiency, and model size. The results establish a strong technical basis for the intelligent and automated control of safety in urban water supply systems.

The reliability of urban water distribution networks (WDNs) is critical for ensuring sustainable infrastructure management. However, traditional failure prediction models often overlook the complex interdependencies between water pipelines and road networks, leading to suboptimal predictive accuracy. This study introduces a novel pipeline failure prediction framework that leverages Graph Neural Networks (GNNs) to incorporate coupled road–pipeline network features. By integrating traffic-related indicators, such as intersection proximity, pipeline–road angles, and network topology, this approach systematically assesses their impact on failure risk. A comparative evaluation of various GNN architectures, including Graph Convolutional Networks (GCNs), Graph Attention Networks (GATs), and GraphSAGE, demonstrates that GraphSAGE achieves the highest predictive performance, significantly surpassing traditional machine learning methods. The findings underscore the necessity of incorporating network topology into predictive models, validating the role of spatial dependencies in accurately assessing pipeline failure risks. This study contributes to advancing infrastructure resilience modeling by providing a robust predictive framework that supports proactive maintenance strategies and enhances risk mitigation in urban water distribution systems.