Urban water distribution systems evolve in time due to their expansion to new areas, addition of new structures and instrumentation, decommissioning of old infrastructure, changes in the system topology, fluctuations in the population dynamics and demographics, etc.
As the system evolves, machine learning (ML) can be used to model the underlying dynamics, and understand causal relations between actions and outcomes. By using real-time data from physical sensors and actuators, consumer behavioural data and other information, models of the changing system dynamics can be continuously updated using online learning methods, and the acquired knowledge can be transferred between different agents within and between systems.
Policies are high-level guidelines that serve as a framework within which organizations operate and make decisions. Policy making most often occurs at the level of national or local administrations, based on a broad spectrum of information that is related to the long-term economic, societal and environmental goals as well as sustainable development issues, while also addressing ethical and security issues as well as the well-being of the citizens.
Research on Subjective Well-Being, in short, “happiness”, has produced remarkable results over the last years. A central concept in the “economics of happiness” is Eudaimonia. In this project, we use the term Eudaimonia in the Aristotelian sense, characterising the highest human good,incorporating virtue and its exercise, but also external goods such as health, wealth, and beauty.
Operational (real-time and short-term) decisions need to be implemented by automated systems supervised by water operators during the everyday monitoring and controlof the system.
This includes detecting and resolving unanticipated events in the system, controlling the system parameters to minimize losses and safeguard quality, optimize resource allocation and make sure the system works in the most efficient way. Short-term monitoring and control must be aligned with the long-term management and planning decisions, as well as the general high-level policies. This includes monitoring how the risk of abnormal events changes in time and how to reconfigure the system in order to mitigate extreme events.
Transitioning relates to strategic decisions taken by the water provider’s management of long-term infrastructure investment (e.g., development strategy, water metering policy, customer engagement), pricing policy, investing in new technologies, etc.
The goal of the management decision-making process is to investigate strategic policies and to implement the most beneficial options, ensuring a high quality of service to its consumers and achieving financial sustainability of their organization, as well as to make decisions on how water is allocated in times of limited resource or for intermitted supply systems. For making informed decisions, utility management must have appropriate practical tools supported by the best theoretical knowledge, which will enable them to choose wisely among an extremely large and complex space of possible solutions.