During the 2024 Emilia-Romagna floods, the regional Civil Protection Agency used SaferPlaces to map flood extent and water depth from Earth Observation data and high-resolution spatial analysis, during the response. That experience shaped what we are building next.

We are extending SaferPlaces Agentic AI from a conversational way to reach flood-risk information into a layer that helps coordinate operational decisions before, during and after an event.

The problem it addresses is familiar to anyone who has managed a flood emergency. The data usually exists: forecasts, hydrodynamic models, satellite observations, exposure datasets, but it sits across separate tools and modelling environments. Pulling it together under time pressure is slow, and it takes specialist skills that an emergency room rarely has to spare.

Take a single question an operator might face: which access roads to this hospital become unsafe if the river rises?

Answering it today means sourcing terrain data, running a simulation, overlaying the road network and the building footprint, then reading the result. The Agentic AI is designed to sequence those steps from the question itself: activating the relevant SaferPlaces capabilities, high-resolution digital twin, flood models, forecasts and exposure layers, and returning maps and a clear explanation rather than raw output.

The same logic carries across the emergency cycle. Ahead of an event, it supports anticipating where impacts could fall and what is exposed. During one, it helps track how conditions are shifting and where to prioritise attention. Afterwards, it makes post-event satellite mapping easier to interrogate: which buildings were affected, how observed flooding compares with the modelled scenarios.

One thing does not change: the science underneath. Hydrodynamic simulations, EO analysis and exposure calculations stay grounded in dedicated, validated methods. The Agent coordinates the workflow and makes the result usable: it does not generate predictions of its own, and it does not make the call. Professional judgement and official procedures stay where they belong.

Further demonstrations and early-access opportunities will follow in the coming months. If you represent a Civil Protection authority, emergency agency or humanitarian organisation and would like to know more, get in touch with the SaferPlaces team.

 

Watch a guided, narrated walkthrough of SaferPlaces Agentic AI on the city of Bologna, Italy.

Follow what the Agent is doing: from understanding the question to retrieving data, running simulations and presenting results.

Privacy Preference Center