ESA AI4EO Accelerator
- Start date
- 21 October, 2020
- End date
- 20 April, 2022
The AI4EO Accelerator is a collaboration between the European Space Agency (ESA) and the UKRI Centre for Doctoral Training (CDT) in the Application of Artificial Intelligence to the study of Environmental Risks (AI4ER), hosted jointly by University of Cambridge and British Antarctic Survey.
Over the last decade, rapid developments in digital technologies and in our capability to monitor our home planet from space with Earth Observation (EO) satellites resulted in large amounts of data. The rate at which data is generated is ever increasing for example by the new generation of satellites coming online, including the Copernicus system and New Space.
To gain insight from this data, the European Space Agency (ESA) has therefore developed a Research and Innovation Agenda for European AI for Earth Observations (AI4EO). Within this framework, the AI4EO Accelerator facilitates high impact innovative research sprints. To select end-user relevant topics, AI4ER partners who come from industry, the third sector as well as national and international institutions, help setting several environmental themes. Themes are focussed into specific challenge questions which are tackled by the AI4ER master students in two teams from December to March.
Artificial Intelligence provides insight into and makes sense of data in many different areas of science and society. Nevertheless, the entrance barriers are still proving too high for some organisations and researchers. The AI4EO Accelerator aims to bridge this gap by connecting partners in industry, national and international organisations and the third sector with students on short innovative projects. Themes are selected for their value, impact, feasibility and relevance to earth observation research.
At the end of the projects, the results are presented to interested parties. The approach to the problem is also presented as an online showcase to inspire what results using Artificial Intelligence can be obtained from which type of data.