In my last post I discussed the progress to date on the Farm Data Code of Practice, a New Zealand initiative that is receiving a level of interest internationally. The Code of Practice is intended to support the development of trust between primary producers (farmers) and organisations that collect and manage data – and between organisations as well.
Once the necessary level of trust and the business needs for data exchange are in place, the next hurdle is the technical details of data exchange. The critical starting point is a common vocabulary that allows organisations to use the same terms to describe the same sort of data (and which helps us to know when we should use different terms because the data is different).
A number of organisations around the world have started work on gathering and standardising agricultural data, including some substantial work in Europe and the United States. In New Zealand, we have built on some of the earlier work, and have defined a number of standardised vocabularies for different types of data. These are published as “Data Dictionaries” – tables that name and describe different pieces of data, and you can find a number of them at www.farmdatastandards.org.nz.
An example is the Animal Data Standard, which was based on work carried out in Germany on data standards, the work of the International Committee for Animal Recording in its Guidelines, and the New Zealand Herd Testing Standards, as well as the databases of a number of recording organisations. The standard covers:
- Ways of defining or standardising animal identifiers and other identifiers (such as farms and cohorts or herds);
- Attributes of animals (such as Birth Date and Sex);
- Statuses of animals that may change over time; and
- Observations made about animals – everything from weights and milk records to movements, treatments, and test results.
There are similar documents about applications of fertilisers and sprays, effluent and irrigation systems, and feeds and grazing.
Of course, such vocabularies won’t stand still. Just like spoken and written languages, vocabularies for describing data must evolve over time. Sometimes naming conventions become more generic as they are applied to other species (for instance, Parturition rather than Calving). Sometimes new observations will be required to match new measurement technologies or areas of interest – and sometimes adjustments and corrections will be needed.
We intend to version the data dictionaries, and support a regular process of seeking input and applying revisions. Right now the industry groups that have funded this work are considering how this process might be managed in the future. We’d appreciate your input – a discussion in the NZ Farm Data Standards LinkedIn group might be a good place to start, or drop Don Wilson or me a note.
If you are considering importing, exporting, or interchanging farm data in your business, take a look at the farm data standards web site. We can also provide some advice and assistance with adoption, so please get in touch.