Data must be findable, available and reusable
Data collected from cameras on drones, satellite images, remote sensing data, digital soil maps, farmers’ data, crop prices, and meteorological data, all feed compounds databases. A growing number of terabytes of data are now available in the agricultural sector. FACCE-JPI is considering actively advocating that data in the agri-food chain should be findable, available, interoperable and reusable, according to the so-called FAIR data principle. The Governing Board has proposed launching a joint research call on large-scale data, and organizing a joint exploratory workshop on this topic.
That FACCE-JPI should advocate FAIR data, was the outcome of the workshop ‘Identifying the potential role of Big Data for food security’ in Copenhagen (November 2017). In a background paper produced for this workshop, an international group of experts explained what FAIR data are: these data are findable, not just by humans but also by machines, which in the time of the Internet of Things (IoT) is particularly relevant. Besides, the data should be available in repositories that ideally are governed by open-access policies or otherwise have well-defined licences. Data should also be made interoperable; specifically, this refers to the availability of robust and well-documented application programming interfaces (APIs) and the use of ontologies. And the data should be reusable and readily integrated with other datasets.
Affordable for users
Only making FAIR data available to the research community is not enough, according to the participants of this workshop. Datasets should also be made affordable and easy-to-interpret for users such as policymakers, consumers or farmers. Other recommendations included preventing a skills shortage in the research community, and promoting trust among organizations so that data are shared. FACCE-JPI wants to collaborate on this issue with the US Department of Agriculture and the National Institute for Food and Agriculture in the United States.