Projects

DSS Ridracoli: decision support system

Using Predictive Models to Manage Water Resources Efficiently and Sustainably

Funding: IFAB call for members

Enabling technology: Machine Learning

Sustainable Development Goals

The DSS Ridracoli project involved the development of an innovative water resource management system for Romagna Acque. The system includes two advanced predictive models designed to monitor and forecast the levels of two key water sources: the River Po at Pontelagoscuro and the Ridracoli Dam. The models provide accurate ten-day water-level forecasts, supporting real-time water resource management. Through an interactive web-based dashboard, Romagna Acque operators can easily monitor these forecasts, significantly improving decision-making and optimising the use of available water resources.

Objectives

By using predictive models, the project aims to provide a practical response to the growing need for more efficient and sustainable natural resource management. In particular, the development of advanced forecasting models strengthens Romagna Acque’s ability to manage and optimise water resources by providing operators with advanced decision-support tools. The project also aims to deliver an initial system capable of forecasting water levels and improving resource-use planning, thereby increasing both efficiency and accuracy.

Initial Challenge

The project’s main challenge was to develop water availability forecasting models capable of supporting operators in their decision-making and enabling effective planning of water resource use.

Solution

The project uses Long Short-Term Memory machine learning models to forecast the levels of two key water sources. The system provides ten-day forecasts, allowing operators to anticipate water availability and optimise resource use. The forecasts are accessible through an interactive dashboard delivered as a web application, enabling operators to view the results easily and make more effective day-to-day decisions.

Every morning, the dashboards automatically provide updated forecasts, ensuring that operators have timely access to the latest available information. In addition, the dashboard dedicated to the Ridracoli Dam allows authorised Romagna Acque users to enter specific input values through password-protected access. This makes it possible to generate and analyse alternative scenarios in support of more informed strategic decisions.

Benefits

The DSS Ridracoli system represents an important step towards the digitalisation of water resource management, combining technological innovation with sustainability.

The system provides an effective tool for managing and forecasting water levels, contributing to the responsible use of natural resources.

More specifically, it enables: more accurate planning of water withdrawals; greater efficiency in daily resource management and allocation; reduced risk of water shortages or inefficient use of available resources; improved resilience and adaptability of the water management system; lower risk of operational errors through continuous monitoring and predictive analysis; increased overall operational efficiency.

IFAB’s Role

IFAB contributed to the technical development of the project across five main areas:

collecting and integrating heterogeneous data from historical weather records, company time series and sensor systems;
developing source-specific models to estimate available water resources, calibrated using different meteorological scales and groundwater remote-control data;
assessing water quality through Sentinel-2 satellite imagery and developing seasonal forecasts to support planning;
developing constrained multi-objective optimisation algorithms to manage the trade-off between cost and quality while complying with operational constraints;

Finally, integrating the models into the existing software infrastructure and validating them against historical events.

Participation in the project enabled Fondazione IFAB to consolidate advanced expertise at the intersection of hydrological modelling, satellite data analysis and multi-objective optimisation. It also strengthened IFAB’s ability to apply artificial intelligence tools to complex operational challenges in the management of natural resources.