FAIRiCUBE Logo

Welcome to the FAIRiCUBE Knowledge Base Digital Library#

Explore and benefit from the experience, the know-how and the services of the FAIRICUBE project.

This digital library aims to share know-how on extracting insights from large/complex datasets using Machine Learning (ML) techniques, enabling actors beyond the traditional Earth Observation domain to access, process and share gridded data and algorithms.

Browse through the different sections of the menu on the left to find user guides for the FAIRICUBE services, as well as recommendations, technical and implementation expertise on data analysis and processing, based on experience and understanding of project use cases. Following instructions in the dedicated part of the User Guide, it is also possible for the community of ML and GeoDatacubes implementers to contribute to documentation in this Knowledge Base.

Thanks to the interactive FAIRICUBE Query tool, both ML and data processing experts and non-experts can easily discover and analyse the project's data analysis and processing resources (pipelines, pre-processing, ML models and algorithms...).

By going through the Tips & Tricks, users will discover the challenges faced by the use cases and the associated solutions, workarounds, failures and lessons learned.

Sections of this Knowledge Base:

  • Overview

provides general information about the FAIRiCUBE project, the FAIRiCUBE Hub and the two pillars of the projects’ activities: the GeoDataCubes and the Machine Learning. Go through this section to learn about the core mission of FAIRiCUBE, what the Hub is with its architecture and platform, and the basics of GeoDataCubes and Machine Learning, which in the project were combined to extract meaningful insights from complex and voluminous Earth Observation data.

  • Use Cases

introduces to the project’s use cases, all focusing on urban and regional scale. The FAIRiCUBE use cases address EU green deal action items, like climate change, circular economy, energy and biodiversity. Specifically, they investigate the adaptation to climate change, the nexus between biodiversity and agriculture, the environmental adaptation genomics in drosophila, the spatial and temporal assessment of neighbourhood building stock, the validation of Phytosociological Methods through Occurrence Cubes.

provides access to the FAIRiCUBE Services: the FAIRICUBE Metadata Catalog describing the project’s datasets (used and produced by the use cases) and processing resources (pipelines, pre-processing, ML models and algorithms), the Query Tool to query over the resources, the EOX Lab, the rasdaman Lab and the FAIRICUBE storage to let the user try the environments and the algorithms. For each service detailed descriptions, examples and instructions for use are available.

provides insights into the basic concepts of data science as applied/applicable to FAIRiCUBE, together with a list of useful links to external resources, instructions for use and examples. Those new to ML and geodatabases can use the topics in this section to acquire the basic concepts/skills needed to understand and benefit from the work done in the use cases.

introduces the most used tools and frameworks in the AI field, with description and examples of use. In the documents of this section one can find link to practical resources and code snippets that demonstrate how to apply AI methods across a range of domains, from machine learning and deep learning to natural language processing and computer vision.

introduces the most used tools and frameworks in the GeoDatacubes field, with links to description page and examples of use. In the documents of this section one can find link to tools enabling use of the GeoDatacubes in a wide range of applications that require large-scale geospatial analysis, like the environmental monitoring, climate change research, agriculture, biodiversity and conservation, urban planning, disaster management and much more.

documents the use cases challenges, their successes & failures, solutions & workarounds. Reading through the contents of this section will prevent you from wasting time trying to solve problems that have already been faced by use cases and for which they have already found a solution. You will also learn which approaches have been successful, which have failed, and which points are still open.

  • External Resources

contains useful links to external resources used in FAIRiCUBE, like the Sentinel Hub, the Copernicus Hub providing comprehensive and accessible access to a wealth of Earth observation data gathered by the Sentinel satellites, the EOxHub / Euro Data Cube providing the FAIRiCUBE EOX Lab, the rasdaman platform the integrated solution for managing and providing access via standardized API to spatio-temporal datacubes.