A Data Fusion and Time Series Framework as a Service

This brand new European Space Agency (ESA)-initiated project has officially kicked-off with the start of 2022 via online channels. We are excited to work closely on it together with VITO and University of Valencia. The goal of the project is to make use of the vast amounts of spatially and temporally rich data, which is complementary in nature.

Through the project, a data fusion and time series framework will be built and provided as an on-demand service. The framework will be based on novel algorithms & technologies, and tested in real-world practical scenarios.

About AI4FOOD

The current capacity of Earth Observation data is unprecedented in terms of the coverage, amount, and types of observations made. These highly complementary data streams offer ample opportunities for combining observations and time series into fused products.

The project aims to utilize the novel Artificial Intelligence (AI) and Machine Learning (ML) technologies (especially Deep Learning) in order to create and use the fused data streams for continuous time series analytics over land environments. Additionally, the potential to use innovative multi-scale approaches (using Sentinel-3 data) will be explored.

The project will be developed with a focus on the following real-world use cases:

• Land Cover Change Monitoring (FAO)
• Cropland Phenology Indicators (ITACyL)
• Agriculture and Land Management Activities Identification (Agency for Agricultural Markets and Rural Development, Slovenia)

The End Result

Together with VITO (as the leading partner) and University of Valencia, we aim to build a data fusion and time series framework as a service, which will be open source, modular, extensible and reusable. The framework will be implemented in a cloud-based commercial service, available via an API and a GUI/web portal.

For the actual deployment of the algorithm toolbox the consortium will benefit from the existing operational services, such as OpenEO and Euro Data Cube. This cloud-based implementation will lead to a scalable software-as-a-service approach, ensuring that a wider group of users will be able to implement data fusion and time series analytics techniques to develop EO-based services.

Project Partners

• VITO - Vlaamse Instelling voor Technologisch Onderzoek, Belgium
• University of Valencia - Laboratory for Earth Univ. Observation, Spain
• Sinergise, Laboratory for Geographical Information Systems, Slovenia