From the model already operational on the Lamone to the Senio: a project within the national program “Regioni per l’Intelligenza Artificiale” (Regions for Artificial Intelligence) that uses artificial intelligence to improve flood forecasting and make the Region’s data assets queryable.
The Program
REG4IA (Regioni per l’Intelligenza Artificiale – Regions for Artificial Intelligence) is the national program promoted by the Department for Digital Transformation of the Presidency of the Council of Ministers alongside all Italian Regions and Autonomous Provinces to test artificial intelligence (AI) solutions across public services. Funded by the Technological Innovation and Digitization Fund, it is structured into four interregional hubs. Ex Machina was selected as a strategic AI-specialized technology partner within the hub dedicated to territorial resilience and safety, led by the Tuscany Region in partnership with Friuli-Venezia Giulia, Emilia-Romagna, and Sardinia. This specific project is part of PoC (Proof of Concept) 3.7, “Simulation models for river and watercourse flooding.”
The Project in Emilia-Romagna
It is a project promoted by the Emilia-Romagna Region through the Directorate General for Resources, Europe, Innovation, and Institutions, the Regional Agency for Territorial Safety and Civil Protection (ARSTPC), and ARPAE (the Regional Agency for Prevention, Environment, and Energy). Working alongside Ex Machina are the University of Padua, the University of Parma, the Polytechnic University of Milan, and the CIMA Research Foundation (International Centre on Environmental Monitoring). The initiative focuses on testing a solution prior to its structural adoption. Activities officially began in spring 2026.
From the Lamone to the Senio
The starting point is an application that is already operational: the real-time simulation of flooding scenarios resulting from levee breaches—the failure of an embankment—developed for the Lamone River following the May 2023 floods. The system couples a hydrological model, RHYME, developed by the University of Padua, with a two-dimensional hydrodynamic model, PARFLOOD, from the University of Parma, integrated into the FEWS (Flood Early Warning System) interface. Leveraging the computing power of the MarghERita regional supercomputer, it can simulate a scenario involving three simultaneous levee breaches in real time.
The objective is to extend this approach to the Senio River, the final right-bank tributary of the Reno: a watercourse stretching approximately 101 kilometers, continuously embanked downstream of Castel Bolognese and therefore particularly vulnerable to breach hazards. Here, AI serves to improve forecasting and enable timely warning and evacuation procedures, as well as broader land and hazard management.
The Role of AIRAS
Ex Machina contributes the expertise gained with AIRAS (Artificial Intelligence Response on Alert System), the decision-support platform engineered to safeguard communities and territories against extreme weather events. The mission within REG4IA is to transform the Emilia-Romagna Region’s unstructured data assets and the technical model outputs developed by the universities regarding the Senio River and its levees into actionable insights.
The contribution focuses on the most innovative links in the operational workflow. Ex Machina is responsible for developing an AI-powered conversational agent capable of integrating the model outputs from partner universities with the documentary knowledge base, making them accessible to operators on the ground. Alongside the Polytechnic University of Milan and CIMA Foundation, it is also working on developing an Impact-Based Forecast approach to assess flood impacts in real time and generate “what-if” scenarios critical for emergency planning.
A Methodology, Not an Isolated Solution
The overarching goal of the project is to establish a methodology that is scalable and replicable, creating an AI-driven knowledge base that is both accessible and queryable to optimize hydrogeological risk forecasting and management.
