Industry field guide
Nursery plant lot data for testing AI traceability
Two benches can hold the same cultivar from different propagation lots.
- 2022 NAICS
- 111421
- Practice pack
- Example and answer key (JSON)
- Updated
What this guide tests
Which batches retain verified L1 ancestry, and can P's cultivar and current bench establish the same lot?
See the specimen and fields · Work through the fictional records
Task and source context
Two benches can hold the same cultivar from different propagation lots. An assistant that reads only the plant tag may merge stock with different origins. APHIS greenhouse certification resources describe a program with traceability and program-specific requirements. A nursery can design a generic lot-matching exercise without sharing a compliance record. APHIS greenhouse program technical requirements · micro1 workplace recording listing
Make a mock pot tag, a propagation lot card, and a short move record. Include one batch divided across benches, one later relocation, and one pot with a label showing the cultivar only. Ask the model to trace the records that actually link to the source lot and to flag the pot it cannot place. Keep the readable plant name separate from verified lineage.
Protect plant breeders’ material and customer contracts. Replace real treatment records, lot volumes, and greenhouse layouts with invented values. Have the operator clear filming and plant handling in advance.
Industry research
The U.S. Census Bureau’s industry definitions establish which activities these codes cover. The workflow sources in this guide describe the task or record type. Fiund’s pilot below uses fictional material to test a specific decision within that scope.
111421 · Nursery and Tree Production
This U.S. industry comprises establishments primarily engaged in (1) growing nursery products, nursery stock, shrubbery, bulbs, fruit stock, sod, and so forth, under cover or in open fields and/or (2) growing short rotation woody trees with a growth and harvest cycle of 10 years or less for pulp or tree stock.
Build a pilot sample
Set up this pilot
Trace mock nursery lots through a split and a bench move. Place a similar cultivar on the same bench and grade ancestry from verified tags.
Prepare the test packets
- Start with the worked case. Use the fictional record types below as a guide to the source files you need. Give each packet a stable ID and save its source versions with it.
- Change one fact. Make one packet with enough evidence to answer the question and one where the evidence points to a different answer. Keep the other fields fixed so a reviewer can tell what changed.
- Remove the decisive evidence. Give the reviewer a third packet with a plausible near-match in its place. Have them write what can still be established and what evidence would resolve the rest.
Fields the reviewer needs
- Cultivar and propagation lot
- Separates variety from origin.
- Batch split and move references
- Preserves lineage through relocation.
- Pot-tag completeness and uncertainty
- Flags unsupported assignments.
Use Plan a farm, forest or fishery AI data sample to turn the specimen into a brief with a defined task, source and review owner.
Fictional exercise
A cultivar label does not identify the propagation lot
An imaginary nursery has original pot tags and abstract bench records. No breeder material, treatment history, actual volume or greenhouse layout is included.
| Record ID | Source | What it records |
|---|---|---|
| NL1 | Propagation split | Lot L1, cultivar C, split into batches A and B. Reviewed parent links supplied. |
| NL2 | Move history | Batch A moves from bench X to Z, retaining L1. Batch B remains on Y. Another lot L2 also uses cultivar C on bench Z. |
| NL3 | Loose pot tag | Pot P says cultivar C only, observed on Z. No batch, lot or move reference supplied. |
Question for the model. Which batches retain verified L1 ancestry, and can P's cultivar and current bench establish the same lot?
Show the reference answer
NL1 links both A and B to L1, and NL2 preserves A's parent through its move to Z. NL3's P lacks a batch reference, while NL2 explicitly places both L1-derived A and separate L2 material on Z. Cultivar C and current bench therefore do not resolve P's propagation lot. Preserve the readable cultivar as known and lineage as unresolved instead of merging stock with different origins.
How to check the result
Ask for the answer, the record IDs that support it and any fact the evidence leaves unresolved. Have a reviewer compare the response against the reference before deciding whether the sample is ready to expand.
Errors to flag in this case
- Treating a cultivar name as a unique propagation-lot identifier.
- Inferring ancestry from current location when different lots share the bench.
Change the inputs and run it again
Give P a batch A tag but no reviewed link between the physical pot and tag. Require the attachment evidence to be checked before clearing provenance.
Score the decision and the evidence separately
Save the reviewer’s expected answer, supporting record IDs and unresolved fields in a separate answer key. Score the model on the decision, its citations and its handling of the deliberate gap. A correct answer citing a different case counts as an evidence error.
Before expanding collection, have a second reviewer run the same checks on the held-out packet. Keep related versions, views and exports of the same case together so they stay on one side of the training and evaluation split.
Check permission and scope
Use invented tags and approved empty props. Exclude breeder rights, customer terms, treatments and layouts; clear any later filming or handling with the operator.
Remove precise farm coordinates and operator identifiers from a review copy. Keep the linkage between lot, sample and outcome intact.
Describe an authorized archive through Fiund’s existing-data assessment, or describe a new capture capability through the collection brief. The contributor agreement covers the terms for participation.
Sources and further reading
Resources for this workflow
Company and program reviews
Use these reviews to compare sourcing, capture or evaluation approaches. Confirm the applicable program, accepted sources and terms with the provider.
Compare the approach
truelabel review for teams collecting physical AI data
Compare how a capture team specifies devices, processing and quality review for an outdoor recording pilot.
Compare the approach
Graviti review for workplace video and multimodal training data
Compare a catalog with commissioned multimodal collection when a particular crop, setting or viewpoint matters.
Related industry examples
Industry guide · NAICS 111422
Cut flower bunch and bloom-stage images for AI review
A photo of a flower bunch can help count stems and classify visible bloom stages.
Industry guide · NAICS 113210
Forest nursery seed-source and seedling-lot data for AI
A seedling’s current bench tag does not necessarily reveal where its seed originated.
Industry guide · NAICS 112512
Farmed shellfish tag and harvest-lot data for AI
Two harvest tags can look alike and still describe different lots.