ABOUT
Embed2Scale is an innovative project aimed at transforming the accessibility, efficiency, and scalability of Earth Observation (EO) data utilisation. By leveraging state-of-the-art Artificial Intelligence (AI) compressors trained on High-Performance Computing (HPC) systems, Embed2Scale seeks to distil valuable embeddings from vast quantities of raw data, enabling decentralised consumption across various modalities.
Funded under the European Commission’s Horizon Europe programme and supported by the European Union Agency for the Space Programme (EUSPA), Embed2Scale marks a pioneering step towards revolutionising geospatial data utilisation within the Copernicus Programme, the Earth Observation component of the European Union’s Space Programme.
Our mission is centred around the development and deployment of cutting-edge technologies to significantly reduce the computational demands for utilising Copernicus data. The project is dedicated to addressing the critical challenges posed by the sheer volume of geospatial data through innovative AI-based data compression techniques. By generating reusable data embeddings, Embed2Scale aims to facilitate efficient data exchange, develop decentralised applications and enable real-time analytics across various sectors Technical developments of Embed2Scale will be tested on real use cases in diverse domains such as air pollution, vegetation studies, maritime awareness, and agriculture, aiming to improve the current practices in these targeted fields.
Embed2Scale focuses on a broad range of aspects varying from data integration, AI-compressor research, ML-Ops and workflows, to validation activities in downstream applications, impact creation, and project management. Through this comprehensive approach, Embed2Scale aims to deliver transformative solutions for near-real-time EO services, contributing to a sustainable and accessible digital Earth observation ecosystem.