Data Spaces and Sovereignty by Design: A Proof of Concept in Agriculture

Rosa Gil and Roberto Garcia are associate professors at Universitat de Lleida, Spain. Rosa is also the coordinator of the Digital Design and Creative Technologies Degree, and Roberto is the deputy vice-rector for technology transfer. Their research focuses on facilitating data sharing while considering human values, especially sovereignty, through initiatives like Data Spaces. Among other projects, they have recently participated in the EU project AgriDataSpace to define the roadmap toward the Common European Agriculture Data Space. Currently, they are participating in the EU project AgrifoodTEF, a €60M project involving 36 partners from 9 countries, which aims to establish a European network of testing and experimentation facilities for agrifood innovation using artificial intelligence and robotics.

The session will present the Data Spaces initiative, its links to data sovereignty, and its implications for the agri-food sector. In addition to a general introduction, a practical experience running a proof-of-concept data space in the pig sector will also be presented. The Data Spaces initiative aims to lower data sharing barriers by guaranteeing that data originators remain in control of their data and receive fair compensation for it. This data sovereignty is especially important in sectors like agrifood, which have many small players in the process of digitalizing their agricultural holdings. Data availability in the agri-food sector is key for its digitalization and is a prerequisite for innovation based on applications of Artificial Intelligence. However, it is understandable that many agrifood stakeholders are reluctant to share their data if they cannot control how it is used and receive fair compensation for the data they share. The Data Spaces initiative aims to address these concerns by guaranteeing that data originators remain in control of their data, providing data sovereignty by design. We apply a Value Sensitive Design approach to create a proof-of-concept data space that balances all stakeholders’ values. It provides data sovereignty by design to guarantee data originators that their data can be used (for instance, to train AI models or make predictions) without being copied elsewhere and getting out of their control. Real cases in the pig sector, such as well-being assessments without farm image leakage, are currently available through the pig sector data space and will be presented.