For centuries, humanity has looked to the sky trying to interpret the whims of nature. From the color of clouds to the behavior of animals, weather forecasts were based on a mix of empirical observation and fatalism. Today, in 2026, while climate change has transformed “extreme events” into a dramatic daily reality, we can no longer rely on intuition. Entire nations, global economic systems, and critical infrastructure are exposed to an unprecedented historical level of vulnerability. Increasingly violent storms, prolonged droughts that dry up rivers, and fires that devastate forest heritage are no longer just emergencies to be managed after the fact, but macroeconomic variables that require a new and powerful defense tool: predictive algorithmic analysis.
It is in this context of global urgency that the Italian technological ecosystem demonstrates, once again, its extraordinary capacity to innovate. In Turin, from the brilliant minds of four young entrepreneurs under 30 (Roberto Carnicelli, Chiara Mugnai, Giovanni Luddeni, and Elias Abou Rouphael), Eoliann was born, a climate tech startup incorporated in 2022 as a Benefit Corporation. After years of research and validation in the field, Eoliann has just launched Airis on the European market: a revolutionary software platform destined to radically redefine the boundaries of institutional and industrial climate resilience.
Beyond Weather Forecasting: The Science of Computational Risk
To understand the innovative scope of Airis, it is first necessary to strip away the traditional idea of “weather forecasting.” The platform doesn’t tell us if it will rain or be sunny next Sunday; its purpose is immensely wider, deeper, and strategically crucial. Airis is a climate risk analytics (climate risk analysis) system designed to assess the long-term vulnerability of specific physical assets scattered across European territory.
The beating heart of this technology is a complex fusion of Earth observation from space and artificial neural networks. The platform daily ingests frightening amounts of raw satellite data from the European Copernicus constellations (such as Sentinel-1 and Sentinel-2) and weather radars. But satellite pixels, alone, do not speak. This is where Artificial Intelligence (Machine Learning) comes into play. Eoliann’s proprietary algorithms, trained for years on vast historical archives of past calamitous events, manage to “read” the invisible radar signatures to the human eye, identifying, for example, microscopic soil moisture saturation or imperceptible alterations in plant biomass.
This extraordinary computing power allows Airis to predict the spatial and temporal probability of six distinct and devastating climate risks simultaneously:
- Floods and riverine overflows
- Landslides and hydrogeological instability
- Wildfires
- Prolonged extreme drought
- Intense precipitation (flash floods)
- Anomalous wind gusts and Mediterranean hurricanes (Medicane)
Four Time Horizons and IPCC Scenarios: The Climate Time Machine
The true power of a Digital Twin of risk lies not only in analyzing the present but in its ability to make us travel through time. Airis’s Artificial Intelligence does not limit itself to photographing the hazard level in 2026, but projects the estimates on four distinct time horizons: the present (2026), the short term (2030), the medium term (2040), and the long term (2050).
But how can an algorithm predict the floods of 2050 if we don’t know how many emissions we will produce between now and then? Eoliann solved this paradox by integrating into Airis’s mathematical heart the three main climate projection scenarios (the Shared Socioeconomic Pathways or SSP) elaborated by IPCC scientists, the United Nations panel on climate.
The platform therefore calculates risk based on three different possible futures:
- The SSP1-2.6 scenario (the most optimistic, which predicts a drastic and immediate reduction of global emissions).
- The SSP2-4.5 scenario (the intermediate scenario, where emissions begin to decline only after 2040).
- The SSP5-8.5 scenario (the so-called business-as-usual, or “catastrophic scenario”, in which the world continues to burn fossil fuels without restraints).
By cross-referencing the 6 climate risks with the 4 time horizons and the 3 IPCC scenarios, Airis’s AI returns up to 72 different mathematical dimensions of risk for every single analyzed square meter. A level of granularity and detail that, until a few years ago, required months of manual work by entire teams of geologists, environmental engineers, and statistical experts, and that today is processed in seconds on the cloud.
The Defense of Critical Infrastructure: From Transmission Tower to Pipeline
All this sophisticated algorithmic architecture is not born out of academic speculation but to solve a terribly concrete problem. Electric companies, water network managers, highways, and railway transport companies manage tens of thousands of kilometers of infrastructure exposed to the elements (the so-called “assets”).
When a landslide engulfs a railway line, or when a flood inudates an electric substation leaving an entire province in the dark, the direct damage (the repair of the physical asset) is just the tip of the iceberg. The real economic disasters stem from indirect damage: service interruption (so-called business interruption), the blockage of industrial supply chains, and the collapse of essential services for citizens, from hospitals to schools.
Airis is able to analyze the climate impact on as many as 17 specific types of critical assets. By entering into the system the GPS coordinates of high-voltage transmission towers, wind farms, railway networks, water pipelines, or gas pipelines, the managers of these large companies (including giants like Terna, which has already validated Eoliann’s technologies in the field, and important utilities like CVA from Valle d’Aosta) no longer receive generic and reassuring “weather alarms”. They obtain, on the contrary, operational dashboards and high-resolution heat maps that indicate exactly which specific transmission tower has a 78% probability of being displaced by a landslide by 2030, or which precise portion of a highway will inevitably go underwater in case of intense precipitation in the intermediate climate scenario.
From Risk to Resilience: The Impact on the Insurance and Financial Sector
The transition from mere risk perception to its rigorous algorithmic quantification is profoundly shaking not only civil engineering but also the financial and insurance sector. Historically, the insurance sector (underwriting) has always based itself on analyzing the past: premiums to insure a company against floods were calculated by observing how many floods had occurred in that area over the last fifty years.
Today, this statistical retrospective is dangerously useless because climate change has made the past an unreliable guide for the future. Insurance companies are facing “incalculable risks,” leading them to disproportionately raise premiums or, worse, withdraw from entire regions declaring them “uninsurable,” leaving citizens and businesses without any economic coverage.
Airis inserts itself into this informational void as a vital compass. By translating complex climate variability into numbers, percentage probabilities, and estimates of potential damage in euros, the platform allows banks, investment funds, and insurers to assess the real exposure of their portfolios with surgical precision.
“For a long time, translating the imperative of climate resilience into concrete, measurable, and tangible choices has been complex, and often frustrating for companies,” lucidly explains Roberto Carnicelli, CEO of Eoliann, summarizing the company’s mission (whose path has already yielded important institutional recognitions, such as the Eni Joule for Entrepreneurship special mention conferred by President Sergio Mattarella). “With Airis, our ambition is to contribute to changing this paradigm definitively. We want to transform the seemingly insurmountable challenge of climate into a rational and data-driven action: something that can be read on a screen, objectively measured, and used to make better decisions about where to allocate capital”.
Conclusion: The Return to Action
The European launch of the Airis platform reminds us of a fundamental truth about the ecological transition: technology, alone, will not save the planet if we continue to pump CO2 into the atmosphere. However, while we struggle hard to reduce emissions at the source (mitigation), we cannot afford to ignore the damage that crazy climate is already causing and will cause in the coming decades (adaptation).
Knowing exactly where the next flood will strike doesn’t stop the rain, of course. But it allows a mayor not to authorize the construction of a school in a high-risk landslide area; it allows an electrical grid manager to preemptively reinforce the foundations of transmission towers or bury crucial cables before the Medicane destroys the line; it allows the State to allocate billions of euros in structural funds (such as the PNRR) exactly where hydrogeological defense works are most necessary, and not where it’s politically convenient.
Artificial Intelligence, in this its most noble and ethical declination, stops being a tool for generating synthetic texts or images for social network consumption. It returns to being what its pioneers conceived it for: a formidable extension of the human intellect, capable of finding lifesaving mathematical models in the chaos of nature. In an age dominated by eco-climate anxiety and a sense of impotence, Italian tools like Airis give us back the most precious thing: the capacity to act rationally and preventively to defend our common home.
