Converting scientific research into a practical tool co-designed with the stakeholders in R Shiny: a web-based application for managing the main reservoir of the drinking water supply system in the Romagna region, Italy

<p>Complex water optimisation problems represent one of the biggest challenges of the near future due to human and climate impacts. On the one hand, stakeholders in the water supply sector require high-level knowledge of the whole water cycle process at different scales, with the aim to either assess the risk for uncertain future water availability or rely on more analytic approaches for decision making. On the other hand, scientific research produces high quality models, algorithms and schemes capable of solving the water problems, but scientists often struggle when it comes to deploy tools that deliver their research outcomes to stakeholders and decision makers that ultimately will use them. The principal goal of this project is to fill the gap between the development of innovative research methodologies and their practical usability in the real world. We present “RApp”, a web-based application written purely in R within the Shiny framework and developed in collaboration with the water supply company Romagna Acque SpA. RApp simulates and visualizes the behavior of the reservoir that sustains the drinking water supply system of the Romagna region, Italy, in order to support its optimal management. Reservoir simulations are obtained connecting, through a unique and site-specific modelling chain, the inflows from the upstream catchments, the functioning of the reservoir, the potential of the treatment plant and the water demand. The optimized monthly-based management rules were obtained off-line, through a multi-objective optimization algorithm by maximizing the water yields and, at the same time, minimizing the occurrence of water outages during drought periods. The RApp user can produce quick reports of the past and expected reservoir yields and stored volumes, in terms of either graphical or table outputs, as a function of different initial and boundary conditions provided by the users, such as the initial stored volume, the expected inflows, the adoption of optimized or user-defined management rules, the occurrence of an abrupt change in the water demand, thus, allowing stakeholders to explore the impact of different scenarios and management options. For developing the tool, a very close interaction between the research group and the stakeholders was required, and is still ongoing, in order to define and then expand the functionalities of the software that are most needed for its practical use.</p>