Project description
Explainable AI pipelines for big Copernicus data
The EU-funded DeepCube project leverages advances in the fields of artificial intelligence (AI) and semantic web to unlock the potential of big data from Copernicus. DeepCube aims to address problems of high socio-environmental impact and enhance our understanding of Earth’s processes correlated with climate change. The project employs ICT technologies such as the Earth System Data Cube, the Semantic Cube, the Hopsworks platform, and a state-of-the-art visualisation tool, integrating them into an open interoperable platform that can be deployed in cloud infrastructures and high-performance computing. DeepCube will develop deep learning (DL) architectures that extend to non-conventional data, apply hybrid modelling for data-driven AI models that respect physical laws, and open up the DL black box with explainable AI and causality.
Objective
DeepCube leverages advances in the fields of AI and semantic web to unlock the potential of big Copernicus data. DeepCube is impact driven; our objective is to address new and ambitious problems that imply high environmental and societal impact, enhance our understanding of Earth’s processes, correlated with Climate Change, and feasibly generate high business value.
To achieve this we bring mature and new ICT technologies, such as the Earth System Data Cube, the Semantic Cube, the Hopsworks platform for distributed DL, and a state-of-the-art visualisation tool tailored for linked Copernicus data, and integrate them to deliver an open and interoperable platform that can be deployed in several cloud infrastructures and HPC, including DIAS environments.
We then use these tools to develop novel DL pipelines to extract value from big Copernicus data. We implement a shift in the use of AI pipelines. DeepCube 1) develops novel DL architectures that extend to non-conventional data and problems settings, such as interferometric SAR, social network data, and industrial data, 2) introduces a novel hybrid modeling paradigm for data-driven AI models that respect physical laws, and 3) opens-up the DL black box through Explainable AI and Causality. We showcase these in five Use Cases (UC), two business, two on earth system sciences, and one for humanitarian aid. These are:
UC1: Forecasting localized extreme drought and heat impacts in Africa,
UC2: Climate induced migration in Africa,
UC3: Fire hazard short-term forecasting in the Mediterranean,
UC4a: Automatic volcanic deformation detection and alerting and UC4b: Deformation trend change detection on PSI time-series for critical infrastructure monitoring,
UC5: Copernicus services for sustainable and environmentally-friendly tourism.
Fields of science
Programme(s)
Funding Scheme
RIA - Research and Innovation actionCoordinator
11810 Athina
Greece