Cultivating FAIR principles for agri-food data

• Automated tools are urgently needed for findable and reusable data in agrifood. • We need a community-based approach for defining metadata vocabularies in agrifood. • Data cannot be shared by an open-by-default policy only. • Scientific insight is needed in how data is (re)used in scientific communities. Data generated by the global food system is crucial in the transformation towards sustainable, resilient, and high-quality food production. Although the amount of potentially useful data is growing rapidly, its (re)use is still limited. The FAIR-principles have been developed for making data findable, accessible, interoperable, and reusable both by humans and machines. This paper explores the further operationalization of the FAIR principles in agriculture and food. Experience shows that several conditions must be fulfilled before data can be effectively shared and reused. First, automated tools must be available for data providers and users. Secondly, we need a community-based approach in developing tools and vocabularies. Thirdly, data cannot be shared by an open-by-default policy only. Finally, scientific insight is needed in how data is actually (re)used in scientific communities. We conclude that bringing the FAIR-principles to full maturity requires a fair balance of efforts within the agri-food communities, supported by a proper infrastructure.