Artificial Intelligence and Big Data Techniques for Copernicus Data: the ExtremeEarth Project
Earth observation (EO) data is arguably one of the most important digital resources available to mankind today. Lots of this data is open and freely available these days around the world. Europe is a pioneer in this area with its flagship Earth observation programme Copernicus.
Copernicus EO data is a paradigmatic case of big data giving rise to all the relevant challenges, the so called 5 Vs: volume, velocity, variety, veracity and value. Copernicus data is processed using Artificial Intelligence and Big Data technologies to extract information and knowledge which are subsequently used to develop applications of environmental, societal and economic value.
ExtremeEarth (http://earthanalytics.eu/) is a H2020 RIA with three objectives:
(i) extracting information and knowledge from big Copernicus data using scalable algorithms,
(ii) managing this information and knowledge efficiently, and
(iii) integrating it with other data sources to develop demo applications of economic, environmental and societal value.
ExtremeEarth is currently in its final year.
Its main achievements so far are the following:
(i) two implemented use cases focusing on Food Security and the Polar Regions
(ii) new deep learning architectures for crop type mapping in the context of the Food Security use case
(iii) new deep learning architectures for sea ice mapping in the context of the Polar use case
(iv) the development and open publication of very large datasets for training the deep architectures
(v) scalable semantic technologies for managing, as big linked geospatial data, the information and knowledge extracted from Copernicus data
(vi) the ExtremeEarth platform which brings all the above technologies together and is used to implement the two use cases.
The session will present the two use cases of the ExtremeEarth project and the technologies that have been used for their implementation.