La transizione verde tra energia, trasporti e ripristino della biodiversità

Technology&Climate

Exploiting satellite data to anticipate climate risks

Artificial intelligence helps predict extreme weather events in the name of the UN 2030 Agenda. Startups (such as Eoliann) pioneer cleantech

by Claudia La Via

9 July 2024

4' min read

Key points

Eoliann's activity
Global Actions
Green intelligence

4' min read

Mapping deforestation, detecting changes in the structure of icebergs, enhancing intelligent and optimised waste management and, above all, predicting and curbing life-threatening environmental disasters. Today, the power of Artificial Intelligence (AI) in processing huge amounts of data and supporting human decisions is transforming many industries, but it promises to do even more: safeguard the environment and our very survival.

This is the intuition behind Eoliann, a Turin-based startup born from a simple but revolutionary idea: exploiting satellite data and artificial intelligence to estimate the probability and impact of natural disasters. A strategic and supportive possibility in climate risk management for financial institutions, infrastructure companies and insurance companies, but also a valuable opportunity to fight climate change and protect the planet.

The topic is hotter than ever. Indeed, the recent launch of a UN-led Advisory Council on AI has promoted the use of machine learning systems to find solutions to common challenges and to accelerate and expand efforts to achieve the ambitious Sustainable Development Goals of the 2030 Agenda.

The activity of Eoliann

The Turin-based start-up bases its activity on the strategic and timely analysis of satellite data. "We collect studies from all over Europe and train our models on satellite data that can more accurately reflect the current state of the land and climate from month to month," explains Chiara Mugnai, co-founder and chief data scientist at Eoliann, pointing out that in the past, satellites were mainly used only to collect short-term data for military and meteorological purposes. "Over time, space agencies and research centres have accumulated a very substantial and homogeneous amount of data, and today this perspective allows us to apply artificial intelligence algorithms that assess the slow but continuous evolution of the Earth system," says Chiara Mugnai. Recently, the volume of such data has exploded, with more than 100 terabytes of satellite images collected daily, and from this perspective, AI is crucial for organising and scaling this vast amount of data.

Il team di Eoliann

Global Actions

Today, satellites are also equipped with more advanced sensors that allow for higher resolution, and are able to capture much more informative signals on the state of water, land and atmosphere. According to the World Meteorological Organisation (WMO) of the United Nations, these technologies offer unprecedented capabilities to process huge volumes of data, extract meaningful knowledge, and improve predictive models that can help communities and authorities develop effective adaptation and mitigation strategies. In the global arena, several UN agencies can already support vulnerable communities in underdeveloped countries with predictive analytics, including for early action to be integrated into humanitarian programmes. Many global-scale initiatives based on AI are also targeting high-risk areas and feeding into local and national response plans. For landslide-prone areas, for example, mapping can help local authorities plan and implement sustainable development measures, reduce risks and ensure the safety of residents in vulnerable communities.

In this context, the UN launched the AI4Climate Action initiative, which aims to explore the role of artificial intelligence to promote and extend transformative climate solutions for mitigation and adaptation actions especially in developing countries. Harnessing the benefits of AI is also part of the 'Early Warning for All' initiative launched by the UN Secretary-General. Launched earlier this year, the action plan aims to ensure that everyone on the planet is protected from dangerous weather, water or climate events through early warning systems by the end of 2027.

The issue is not only predicting a climatic phenomenon that could turn out to be disastrous, but also the timing with which it can be known and what the scope for action might be. 'In the long term, we can decide how to act: if we know that an area is at high risk, we can protect ourselves with insurance products or provide infrastructures that lower the risk, as was done with the lamination basins in Veneto, which prevented even greater damage this spring,' points out Roberto Carnicelli, co-founder and CEO of Eoliann. 'Although the climate is changing very fast, we are trying to move even faster. Our product is in continuous development, with increasingly advanced and accurate algorithms,' he concludes, pointing out that at the moment Eoliann is, among other things, working on a Pnrr project that involves the development of an advanced model for landslide risk and its impact on the road system.

Green intelligence

Artificial intelligence, however, is not just a forecasting tool. The Economist has coined the term 'Green Intelligence' to describe the use of AI applied to the environment by also taking a sustainable technological approach. Whether it is improving recycling, accelerating renewable energy or monitoring ecosystems, the applications are endless. However, HIA is not a panacea. In addition to the hitherto unimaginable technological potential, comprehensive policies and a renewed environmental commitment are needed to make 'green intelligence' a sustainable revolution. Changing, for the better, our environmental future.