**2. From Theory to Practice**

Water systems and services are highly complex [9] as they are tasked to balance water resources with demands through complex interconnected infrastructure. As such, decision making about these systems and services (at strategic, tactical and operational scales) need to be taken within a continuously changing landscape where water quality and quantity are uncertain [10]. These systems are also influenced by climatic changes and human practices water demand patterns are shifting as urbanization continues [11], influencing demands [12] as standards of living rise [13]. Lastly" environmental legislation and customer expectations are also shifting and with them [14,15], the thresholds against which the water sector's performance is measured also change. This dynamic decision landscape is further complicated by aging infrastructure [16] and the advent of new (disruptive) technologies and concepts.

Figure 2 presents an overview of some of the main technologies and concepts that have emerged in the past few years and are influencing both research and practice in the urban water managemen<sup>t</sup> field and hydroinformatics specifically. In this necessarily brief and elliptical sketch, new real-time information coming from smart sensors, including smart meters, also in the context of IoT developments, stored and managed through (often cloud-based) information platforms [17,18], allow for the remote monitoring and control of new more distributed interventions in the urban water cycle integrated into (and extending the useful life of) existing centralised systems and networks. This is possible due to, also, new analytics that are developed to exploit and extract value from this new information in view of design, tactical and operational decisions (from locating new technologies, to rehabilitating piped networks to understanding and managing water demands [19]). Part of the value in this improved understanding of subsystem functions is in being able to develop and calibrate whole cycle (socio-technical) system models. They are now increasingly being applied to improve the understanding of the interplays between centralised and decentralised systems as well as the interaction between infrastructure and the end users. These new, more inclusive modelling approaches underpin a more engaging approach to decision support in the form of serious games (SG), and augmented/virtual reality (AR/VR) environments, challenging and disrupting the very way decisions are made in the water sector [20]. The latest developments in artificial intelligence (AI) and machine learning (ML) have already shown that AI/ML enabled software systems can beat human players in complex games, such as chess or Go [21]. Through reinforcement learning, these systems can learn by playing games, which can be a guiding light to developing decision-support systems capable of assisting human water system operators in performing complex operational, tactical or strategic tasks. Similarly, robotic technologies

and AI, which have been making grea<sup>t</sup> strides in the manufacturing and consumer industries, are starting to find their way to water management, e.g., underground asset inspection [22]. Lastly, the authors argue that with these data, tools and models at hand, the sector is now developing more sophisticated ways of stress-testing new and existing infrastructure, developing new methodological approaches around resilience [23]. In the remaining part of this section, a brief overview of some key literature on the subjects highlighted above is provided and an outline of their current state of art is discussed.

**Figure 2.** A shifting landscape for hydroinformatics research and practice.
