Artificial-Intelligence-Based Structural Health Monitoring of Prestressed Concrete Metro Viaducts: A Framework Applied to the Nagpur Metro Elevated Corridor

Vollständiger Abstract

Worum geht es in dieser Arbeit?

ABSTRACT Urban metro networks across Indian cities, including Nagpur, rely heavily on precast, segmental, prestressed-concrete girders to carry their elevated corridors. Once in service, these girders face repeated dynamic loads from passing trains, continuous exposure to weather, and a slow, ongoing loss of prestress from creep, shrinkage, and tendon relaxation — deterioration that routine visual inspection is poorly suited to catch before it becomes serious. This paper sets out a multi-tier framework for monitoring the structural health of prestressed-concrete metro viaducts using artificial intelligence (AI), combining a mixed sensor network, an edge-to-cloud data pipeline, deep-learning-based damage identification, and digital-twin visualisation. Rather than testing this framework in a laboratory, it is grounded in a real, in-service structure: the elevated viaduct of the Nagpur Metro (Maha-Metro), built mainly from precast segmental box girders and I-girders. The paper traces how SHM practice has developed, draws together AI methods reported in recent literature, and identifies gaps that matter specifically for prestressed-concrete construction, limited explainability of deep-learning models, a shortage of verified in-service datasets, and the difficulty of separating genuine structural change from environmental noise in strain and vibration signals. It then discusses how the proposed framework could map onto a corridor such as Nagpur Metro to support condition-based rather than fixed-schedule maintenance, closing with implementation challenges and directions for future work, including physics-informed learning and edge-based inference for sites with limited connectivity. Keywords: Structural Health Monitoring; Artificial Intelligence; Prestressed Concrete; Metro Viaduct; Digital Twin; Deep Learning; Nagpur Metro