AI implementations for Dutch water management: a literature study

Abstract

This thesis consists of a literature review on how three types of Artificial Intelligence (AI) method applications may be used for improving data-driven decision making in the context of Dutch governmental water management, for the Directorate of Water and Soil. Water management officials may utilize the possibilities brought by recent developments in the field of AI. The literature review distinguishes three AI method applications in relevant works: (1) network planning and optimization, (2) monitoring and anomaly detection, and (3) simulation and prediction. These method applications are individually described according to a state-of-the-art benchmark, and then combined with the domains of seven governmental tasks of the Directorate of Water and Soil. Each intersection is assessed based on research maturity, expected research growth, and level of policymaking impact. As such, this literature review identified three key intersection fields for the Directorate of Water and Soil to further investigate: firstly, monitoring and anomaly detection intersected with pipeline construction and maintenance. Secondly, network planning and optimization intersected with waterway safety and flooding. Thirdly, simulation and prediction intersected with waterway safety and flooding. Further research on many intersection fields is suggested, but especially regarding the aforementioned three.