Skip to content

Technical Minutes 2021 05 06

Jump to bottom
Andrew Cooke edited this page May 11, 2021 · 1 revision

ICAR Animal Data Exchange Working Group – Technical meeting 6 May 2021

  • In attendance: Andreas S, Andrew C, Anton H, Craig V, Erwin S, Thomas D, Thomas P

Agenda:

  1. Identify the items that must be closed in order to publish release 1.2.
  2. Nullable declaration for reference types (214)
  3. Animal Health Status (148,PR149)
  4. How to remove DoNotBreed from animal (213)
  5. Feed attributes and categories (172, 173)
  6. Changes to example URL scheme
  7. Herd level milk recording statistics (184)
  8. Links between resources (from 184, also 171)

1. What is needed for the release

The participants would like to publish what we have as release 1.2. This will give organisations confidence to build on the new features.

2. Nullable declaration for reference types (#214)

We noted that we may need to change how reference types are declared to be nullable, as the specifications vary for OpenAPI 3.0 and 3.1, and some code generators seem to be having problems. Anton and Andreas will investigate.

This could be addressed as a techncial release between functional releases, as a change to allow null will not generally be a breaking change.

3. Pull Request for Animal Health Status (#148,#149)

This is not a complete implementation. We decided to delete this PR and address the health and culling issue for a future release.

4. How to remove DoNotBreed from animal (#213)

We discussed how the DoNotBreed event sets the Do Not Breed status for an animal, and how this might be reversed if required. A possibility might be to post the event without a date/time (not preferred), or do an HTTP DELETE action (if supported, but this removes history). A possibility would be to redefine a BreedControl event.

5. Feed Attributes and Categories (#172, #173)

We have nothing more to add to these categories. Some of the group are investigating how these align to what is being done in the Nordic countries.

At the moment the feed resources are defined, and the categories and properties are identifier types without fixed values. This would be ok for 1.2 release, but in time we would like a set of known properties and categories.

6. Changes to example URL scheme (to address #154)

We agreed that this could now be implemented. However, the ExampleURL scheme is quite large. We wondered if this could be broken into several examples.

  • The logical way to break this up would be to create new URL scheme files for the tags in the current file - for instance, milk, health, performance, registration, reproduction. Anton will try this with performance.
  • We would like to call this "Recommended" rather than "Example" URL schemes.

7. Milk recording information on herd level (#184)

Erwin produced a data model for herd or group level milk statistics. The diagram in issue 184 shows how milk weights and samples for individual animals are combined to produce lactation information and test day results for those individuals. Test day results could also be averaged to produce milk statistics.

  • Herd averages for a number of KPIs
  • Group averages for a variety of pre-calculated groups
  • Group averages for farmer-defined groups (Animal Sets?)

The Milk Recording Organisation (MRO) would define the (sometimes complex) method by which data was included and excluded to calculate these statistics - this is why this resource would be useful for farm software and equipment companies.

We discussed links between this statistics object and the underlying test day results and milk recording results. By definition the statistics would be based on all Test Day Results for the data and Location. The Test Day Result is an entity derived by a MRO, typically based on two Milking Visit events and a Sample.

  • This will be an area of expertise for other members of the ICAR ADE Working Group, beyond the technical group members. We agreed that Andrew should ask the ADE WG for comment.

We might want to identify the organisation that created the resource. This could use the meta data - source or creator. We discussed the importance of keeping meta data through systems - particularly source, sourceId, creator, created.

8. Links between resources

For the discussion above, the links between entities suggested might be a special case related to #171 Feed Intake in milkVisitEvent. We agreed that we need to think about a generalised approach to linking to a single resource or an array of resources (events or other resources).

[Editor's note: This could use something like icarResourceReferenceType.json]

We will discuss this in a future meeting.

Next meeting scheduled for 20 May 2021 at 8:15am CET

Clone this wiki locally