Skip to content

6.5.2.9 ObservationSelection

6.5.2.9.1 Purpose

Chooses one preferred copy among observations that are proved to be copies of the same original measurement, for example a reading received both directly and through the management software. The source evidence is never changed.

6.5.2.9.2 Questions it answers

  • Which copy of a duplicated measurement do consumers read?
  • Is the selection disputed?
  • Which rule set made the choice?

6.5.2.9.3 Fields

Field Type Cardinality Description
id string (uuid) 1..1 Key.
candidates CanonicalObservation 1..*
preferred CanonicalObservation 0..1
status enum 1..1 Values: selected, disputed, unresolved, cancelled.
identity_basis CodedValue 1..1 Versioned matching rule plus common upstream/sample/session evidence in the decision audit.
rule_set ReconciliationRuleSet 1..1
revision RecordRevision 1..1
issues DataIssue 0..*
created_at datetime 1..1

6.5.2.9.4 Relationships

References

Uses types

CodedValue, RecordRevision

6.5.2.9.5 Rules & constraints

  • Versioned selection among proved copies of one original measurement; no change to source evidence.
  • Needed when a farm reaches us through two channels.
  • Preferred belongs to candidates and was available/non-cancelled at the decision time.
  • Compatible profile, qualifiers, method, body site and temporal support are necessary, but similarity alone is insufficient.
  • Lab versus independent sensor measurement never becomes a duplicate merely because animal/day/type match.
  • A cancellation/revision triggers a new selection; an alternative still-valid copy may become preferred.
  • At a query cutoff, a revision participates in at most one active duplicate-selection group.
  • Unselected independent observations remain visible. Consumers retain disputed status rather than silently treating it as truth.