Author(s): Lameea Khan; Wenyan Wu; Waheb A Jabbar; Essa Shahra
Linked Author(s):
Keywords: Combined sewer overflow (CSO); Sensor optimisation; SWMM; Artificial intelligence
Abstract: Combined Sewer Overflow (CSO) events are becoming more frequent under urbanisation and climate-driven rainfall variability, posing significant environmental and regulatory challenges. Accurate monitoring requires the integration of hydraulic modelling with intelligent data-driven techniques. This paper presents a systematic review of recent advances (2021–2025) in CSO monitoring, focusing on dynamic sensor deployment optimisation and the integration of EPA Storm Water Management Model (SWMM) with artificial intelligence (AI). We adopted a structured review methodology based on PRISMA to identify, screen, and analyse relevant studies. Selected studies are classified using a three-layer framework comprising: (i) SWMM-based modelling approaches for hybrid hydraulic simulation, (ii) AI-driven tasks including CSO prediction, and optimization, (iii) sensor deployment strategies categorised as static, heuristic, and dynamic. Comparative analysis reveals that most existing approaches rely on static sensor placement and show limited integration between physics-based and data-driven models. Finally, a conceptual roadmap is proposed for dynamic sensor deployment optimisation to support resilient, interpretable, and intelligent CSO monitoring systems.
Year: 2026