Application domains
Turning Earth Observation into Effective Action
Addressing today's evolving environmental challenges requires a comprehensive approach: analyzing historical trends, monitoring real-time conditions, and forecasting future states.
Observing the Earth
Remote sensing is at the core of this strategy, enabling access to vast, heterogeneous, and continuously evolving data sources.
From Data to Insight
By merging scientific advancements with advanced analytical models, we transform raw information into cross-domain solutions that support decision-making, risk mitigation, and effective action across multiple application areas.
Remote Sensing for Environmental Intelligence
Environmental monitoring
We harness geospatial data to generate high-level products for environmental monitoring across the Earth system's atmosphere, land and water domains.
By developing advanced algorithms for biogeophysical parameter estimation, we analyze and assess:
atmospheric composition including air quality parameters and greenhouse gases;
land surface dynamics such as soil moisture, surface temperature, and biomass;
water quality indicators like chlorophyll and suspended matter concentration.
Advanced image processing
Beyond parameter estimation, our work focuses on advanced image processing techniques that enable a broad range of environmental monitoring applications.
These include:
classification tasks, such as land cover mapping and vegetation categorization;
object detection, with particular attention to maritime scenarios, such as ship and oil spill detection;
anomaly and change detection, including flood mapping, deforestation monitoring, and urbanization analysis.
Data compression & feature extraction
We also investigate compression and feature extraction strategies aimed at reducing data dimensionality and identifying the most informative components of large-scale Earth Observation data.
To this end, we exploit both AI-based approaches, such as autoencoder models and pruning techniques, and traditional methods, such as principal component analysis (PCA).
Extreme Weather & Climate Events Monitoring
A changing climate
The increase in the frequency and intensity of extreme weather events poses significant environmental and social challenges.
Today, prevention is one of the key actions to counter the growing effects of climate change and build climate resilience.
Earth Observation & AI
Earth Observation technologies provide a systemic view of our planet, offering a broader perspective than traditional survey methods.
To transform the large volume of satellite data into strategic information, we develop advanced solutions that integrate geospatial data with Artificial Intelligence.
By combining deep learning and multi-decadal data records from leading EO programs, such as Copernicus, we build predictive models that monitor evolving climate trends, map territorial vulnerabilities, and identify potential threats.
Climate resilience in practice
These AI-driven capabilities empower organizations to improve the protection of communities and infrastructure.
Urban planning & public administration: guiding the development of resilient infrastructure and sustainable cities to mitigate climate risks.
Disaster management: equipping agencies with early warning systems and critical data to coordinate effective emergency responses.
NAT CAT Insurance
The insurance industry is facing unprecedented challenges as climate change drives the frequency and severity of Natural Catastrophes (NAT CAT).
To move beyond the limitations of conventional models, accurately pricing risk and managing post-disaster claims now require advanced geospatial intelligence.
To support this digital transition, we propose U-Climat (Urban-CLimat Impact Mitigation Adaptation Tool), a dedicated tool for climate risk and impact assessment.
By combining satellite Earth Observation (EO) data with Artificial Intelligence (AI), U-Climat provides the insurance sector with reliable information to evaluate physical risks, allowing insurers to upgrade their NAT CAT policies and transition to highly accurate, data-driven models.
Key applications for property and casualty insurers include:
risk assessment & pricing: evaluating the exposure of specific areas to NAT CAT (e.g. floods, wildfires, severe storms) to support accurate policy pricing;
damage assessment: mapping affected areas and estimating the extent of structural impacts following an event, to support the claims evaluation process.
Multi-risk & climate impact assessment.
Image: example of post-flood mapping for the 2023 event in Emilia-Romagna, Italy. Courtesy of ESA
