Open Earth Monitor — Global Workshop 2023

GeoData-Based Artificial Intelligence Architectures in the Green Deal Data Space Development
2023-10-05, 16:40–16:45, Poster presentation

Nowadays, global warming is an indisputable fact; the temperature and other natural phenomena are being modified because of human action. These changes are visible in Europe’s biodiversity, where there are animal migrations out of time; the air quality indicators in the centre of Europe are the worst in the last century; and Berlin lakes are undergoing changes in their parameters, affecting its wildlife dramatically.

For this reason, it is vital to develop tools and measurement systems to evaluate and predict disorders caused by global climate change. This knowledge can make a difference to warn the involved stakeholders, such as governments or supranational organizations, helping them for designing the most adequate policies and making decisions; private companies, guiding them in a long-term sustainable way of conducting their business; particulars, raising awareness about the magnitude of the global warming issue based on facts and promoting sustainable habits; etc.

In this context, AD4GD (All Data 4 Green Deal) European project is trying to give an answer to the mentioned above issues through the development of a Green Deal Data Space, a common European space for sharing, analyzing and processing data from nature. Information acquired by Copernicus and Landsat satellites will be processed by applying the most suitable and modern AI-based technologies. Specifically, a Machine Learning Operations (MLOps) architecture implementing commonly used open source tools (Kubernetes, Kubeflow, Jupyter Notebooks, Docker) is being developed to acquire and create knowledge from the measurements obtained in the following environments: biodiversity corridors in Catalunya, lake water quality in Berlin, and air quality in some parts of Europe.

The project aims at creating and designing a data platform based on the latest initiatives that are designing the concept of Common European Data Spaces (e.g., IDSA, GAIA-X) to enable diverse information collected by multiple sources to be shared and managed in a secure and fair manner. The idea is to develop models which allow extracting knowledge from remote-sensing data and using an architecture based on Convolutional Neural Networks (CNNs), segment the information and mix it with data collected by IoT sensors in order to obtain an accurate solution. In order to allow the processing of large amounts of information and to ensure the scalability of the solution, a hybrid approach combining High-Performance Computing (HPC) resources and services deployed in the cloud is envisioned.

The resulting platform will help European stakeholders in making better decisions related to the climate change field and implement actions aligned with the Green Deal strategy. The results obtained from the processing of data coming from IoT devices and remote sensing using the previously mentioned artificial intelligence techniques will contribute to reducing environmental impacts and will be one of the first steps in the construction of a common architecture for both researchers and ecologists.

Do you accept that a video-recording of your talk is published under CC-BY license via – yes What is your current associations to EU Horizon projects?

AD4GD Grant agreement ID: 101061001

Engineer in telecom systems, specialised in artificial intelligence oriented to computer vision and time series systems. Additionally, I’m doing my PhD in ML, AI, and CV. Also, I love DeVop systems, and I can work with K8s, containers, Rancher, etc.

I am Artificial Intelligence Researcher in Atos (Madrid), where I work in different Horizon Europe Projects (FERMI, AD4GD and DIH4AI) applying the latest technologies in the field. Previously I was a postdoctoral researcher in the Signal Theory and Communications Department, at University of Alcalá. I received the PhD in Information and Communications Technology (2021), with a thesis about sound event detection in smart cities.