An effective flood early warning system is one that tells you exactly what is going to happen in a specific riverbed, when, and with what consequences.
Today’s meteorological models are better than they were three decades ago. The leap is so significant that the precision once achieved four days out is now reached at seven. However, this progress has physical limits situated around 14 days, the horizon beyond which atmospheric forecasting irreversibly loses reliability.
And the fact is, we must remember that the atmosphere is a chaotic system, and at a local scale, that uncertainty is amplified.
So, forecasting the arrival of a severe storm is not enough. What a city council, a refinery, or a railway manager needs to know is exactly what is going to happen in their specific stretch of the river when that storm arrives.
That gap between regional forecasting and local impact is exactly the problem that conventional alert systems fail to solve.
From the warning that doesn’t say enough…
It is two in the morning, and the weather forecast has been warning of heavy rainfall in the area for hours.
But the duty personnel at the city council do not know whether they should close the riverside road, call emergency services, or notify the industrial estate of the threat.
The warning doesn’t tell them. Where exactly will the river overflow? Is there still time to close the south access, or is it already too late? These are the questions that matter, and no weather bulletin provides the answers.
That limitation finds its answer in a concept that is gaining ground in emergency management: impact-based forecasting. An effective forecast doesn’t just describe whether it will rain, but rather what consequences that weather will have on specific people, infrastructures, and territories.
A warning stating that “the Mañaria river will exceed the alert threshold at the Santa Ana bridge in less than 4 hours” does allow for a decision to be made.
That precision doesn’t come from a meteorological model. It comes from integrating, in real time, what is happening upstream in that specific riverbed, with that specific terrain. This is the starting point from which IntelliAlert has developed FAST.
…to FAST, the flood warning system that turns data into decisions.
FAST is a flood prediction platform powered by Artificial Intelligence, developed by IntelliAlert Technologies. It combines real-time data from hydrometeorological stations, river level sensors, and weather prediction models, all weighted through proprietary algorithms.
The difference with other solutions is not just technical, but operational.
A traditional model takes between 30 minutes and 2 hours to generate an update. FAST does it in less than 10 seconds and repeats the cycle every 10 minutes.
But prediction alone doesn’t close the loop.
FAST incorporates emergency management modules that allow users to define protocols linked to specific alert levels, centralize key contacts, and log every incident to improve future response.
Additionally, we are also developing a virtual assistant that will support operators during the early stages of an emergency, when both time and information are in short supply.
Three environments, the same problem.
In Durango, Bizkaia, FAST monitors the Mañaria and Ibaizabal rivers with predictions up to 12 hours in advance. The City Council can make preventive decisions before the water reaches the streets, rather than reacting once it’s already there.
“The tool developed by IntelliAlert provides real-time data on river levels and can predict and anticipate future levels up to 12 hours in advance. This AI-powered tool facilitates emergency management within our entity.” — Technical Office Representative, Durango City Council.
In Muskiz, the municipality where the Petronor refinery is located—the largest refinery on the peninsula—FAST protects urban areas from the swelling of the Cotorrio and Barbadún rivers.
Fuente: Simulación de escenario de inundación sobre las instalaciones de Petronor, Muskiz. Fuente: IntelliAlert / Google Earth
In Málaga, ADIF uses FAST to monitor the Campillos-Almargen section of the Bobadilla-Algeciras line, an area where intense rains can turn runoff into torrents in a matter of minutes, allowing for the management of railway traffic with sufficient advance notice.
These three contexts share the same logic: the risk is local and the warning must be too.
The context in Spain is concerning. According to a study by the Sustainability Observatory, Spain has more than 4340 flood-prone critical infrastructures. 71% of them are in a very serious hazard situation. Among the study’s primary recommendations is precisely this: that each of these infrastructures must have a flood early warning system with clear protocols and preventive compliance, a point also highlighted by the European Directive on the Resilience of Critical Entities (CER).
One final question to close
Water risk management software that allows for flood anticipation with local precision exists. It is operating today in municipalities, transport infrastructures, and industrial facilities.
The question that remains unanswered in many city councils and companies is still the same: how many of your emergency protocols are linked to a specific alert level, generated in real time, for your specific riverbed?
FAQS. Frequently Asked Questions
FAST is an artificial intelligence platform developed by IntelliAlert that predicts floods in specific riverbeds. It is used by city councils and infrastructure managers to make preventive decisions based on real impact, rather than just general meteorological forecasts.
The system achieves an accuracy of over 90% in flood forecasts up to 8 hours in advance, surpassing the 70-80% reliability of traditional models under changing conditions.
While traditional models take up to 2 hours to update, FAST generates a new prediction every 10 minutes in less than 10 seconds. Furthermore, it integrates local IoT sensors to provide hyperlocal, rather than regional, warnings.
It is ideal for municipalities, critical industrial plants, and transport infrastructures, complying with the requirements of the European Directive on Resilience (CER).
