How Can We Increase the Value of Our Data?
AGRO’s strategy includes a focus on strengthening international collaboration and working with other institutes to develop sustainable solutions that address societal needs for food security and green transition. Data is at the center. It is our shared knowledge base—something that must be possible to share, retrieve, understand, and reuse across colleagues, collaborators, and ourselves, both now and five, ten, or twenty years from now. This places interoperability—both in the narrow and broader sense—high on the agenda.
In the narrow sense, interoperability means that our data are well-described, represented, and documented in a structured and standards-based manner, and are unambiguously understandable, so that they can be used in combination with other and new data, applications, and purposes — and be understood by both humans and machines (software). That is the “I” in FAIR. Put more simply, it means that data can be understood and used across systems and disciplines.
The more interoperable data is, the higher its potential for (re)use.
Interoperability is therefore not only essential for the data we will create going forward, but also for the data we already have—for example, from our long-term experiments. By investing in making these data more interoperable, we increase their potential for reuse, and thus their value.
At AGRO, we are in the middle of a generational shift—among both researchers and technical staff. This is an important opportunity to revisit how we organize and document data. While the most immediate concern is avoiding the loss of data and knowledge, it is also about making future collaboration around data easier—among other things by improving interoperability.
One example is the expansion and renovation of the VirkN database. This Access database was established in 2015 for a single long-term experiment (Virk-N), and has since—through various workarounds—been used to hold data from subsequent projects. Triggered by a generational shift among academic and technical staff, the database has now been extended to accommodate data from other projects. Going forward, we will continue working to increase its interoperability—for instance, by adding field descriptions and, where possible, referencing internationally defined terminologies (e.g., for crop varieties and treatments). In the longer term, the data will be migrated to a different database structure to ideally accommodate AGRO’s long-term experiments and serve as a shared data infrastructure.
Choosing a vocabulary that others can also understand and recognize—i.e., using common, well-defined technical terms, categories, and units that can be referenced (or linked to), and that does not disappear when the person behind the dataset leaves AGRO—is a good place to start on the FAIR journey. It is a path to better data quality, greater clarity, and higher potential for reuse—whether the data lives in a database or a spreadsheet.
In a broader sense, interoperability is also about the ability of systems and parts of the organization to work together and exchange information. Here too, better use of structure, documentation, and standards makes things easier.
In this context, shared conversations around standards and structures that promote interoperability in AGRO become essential. Where it makes sense, we should consider what can be generalized and applied across sections, units, or the entire institute—for example, by agreeing once and for all on how we name and understand core concepts (such as field, block, plot, and sub-plot), introducing authoritative lists (e.g., for equipment), or something else entirely (the opportunities are many)—so we can use them consistently and achieve semantic interoperability, at least internally in AGRO.
These are conversations we need to have in the Data Management Committee—and over coffee breaks, in research groups, in committees, at field stations, in labs, and across the institute. Conversations that can be supported by sharing and learning from concrete examples.
Diba & Jens