Dynamic Diagnosis and Integrated Intelligent Management of Urban Drainage Ecological Risks

Rapid advances in artificial intelligence (AI) and high-throughput environmental sensing technologies are transforming conventional approaches to urban drainage ecological risk management. Urban drainage systems (UDSs) are exposed to complex environmental contaminant mixtures and dynamic environmental conditions, making traditional static and chemistry-based assessments insufficient. This perspective emphasizes a shift toward biologically informed, system-oriented, and AI-assisted dynamic diagnosis. We summarize recent progress in AI-assisted approaches based on biological imaging, metabolites, and functional gene big data. High-throughput imaging systems enable real-time quantitative extraction of behavioral and morphological-response indicators. Metabolic sensing across in situ, remote, and underwater platforms captures dynamic physiological and biochemical responses. Functional gene biomarkers provide mechanistic, process-level signals that link contaminant transformation to ecological health effects. These complementary bioindicators support earlier, more mechanistically interpretable detection of ecological stress. We also discuss stability-based and interpretable analytical strategies that integrate biological and chemical information under multicontaminant exposure scenarios, particularly in data-limited real-world scenarios. Building on these foundations, we outline an integrated management platform for UDSs that connects monitoring-network optimization, ecological risk diagnosis, source tracing and early warning, plant–network–river coordination, and source-control strategies. This platform supports closed-loop intelligent management and ecosystem resilience in urban aquatic ecosystems. This perspective provides a scientific foundation, strategic direction, and methodological guidance for advancing AI-enabled ecological risk diagnosis and intelligent management of UDSs.