INDUSTRY FIELD GUIDE
Cotton gin bale lineage data for AI traceability
A gin is in a different position from a grower.
NAICS 115111 · Cotton Ginning
A gin is in a different position from a grower. It can show the relationships among incoming modules, processed bales, and classing results, where authorized. USDA describes cotton classing services and publishes cotton data resources. The useful AI task is to preserve which result belongs to which output, particularly when a module produces several bales. USDA cotton classing · USDA national cotton database
Create two mock gin runs and several bale IDs. Map the input module to its authorized outputs. Give one bale an individual result, another a result with a missing association, and a third an explicit group-level record. Ask the model which facts can be applied at bale level. A reviewer can check whether the answer preserves method, unit, and ID when it cites a result.
Keep grower settlements, actual tag codes, customer labels, and proprietary process parameters private. Use invented result fields and ask the gin, warehouse, or classing organization who controls any real data.
Describe a bale-lineage exercise
Original Fiund exercise
A module's group result does not become an individual bale measurement
Two imaginary gin runs use original module and bale aliases with invented quality fields. No real tags, classing values, settlement information or process parameters appear.
| Record | Source | What it records |
|---|---|---|
| GB1 | Run lineage | Module M1 produces bales B1 and B2. Module M2 produces B3. Explicit mock output map supplied. |
| GB2 | Result cards | Individual card R1: B1, demonstration score 7. Group card RG: M1 outputs B1/B2, demonstration average 6; method labeled group average. |
| GB3 | Unmatched card | R-U says score 8 but lacks bale or module reference. B3 has no linked result in this packet. |
Question for the model
Which results can be cited as individual measurements for B1, B2 and B3, and how should the group average be represented?
Show the reference answer and checks
GB2 supplies an individual score of seven for B1 through R1. RG describes the B1/B2 group average of six, not a direct measurement of each bale, so B2 has only that stated group context. GB3's unlinked eight cannot be assigned to B3 or another output. Preserve result level and method alongside GB1's lineage rather than filling missing individual results from a group average or an unmatched card.
- Writing a group-average result into every bale's individual measurement field.
- Attaching an unreferenced result to the only bale lacking a value.
Fields to preserve
- Module-to-bale output map
- Defines the processing lineage.
- Result target and aggregation level
- Distinguishes individual from group evidence.
- Method, unit and unmatched state
- Preserves the meaning of each result.
Test another case
Add a later B2 individual result explicitly linked to its tag. Require both the individual measurement and earlier group average to retain their separate meanings.
Use original dummy tags and invented fields. Exclude settlements, real classing files and customer labels; gin, warehouse and result-provider rights need separate review.
Discuss a contribution
Describe the sources you control, the task they could support and what has already been checked. Use the existing-data assessment for an archive or the collection brief for a new recording project. The contributor agreement defines Fiund’s licensing authority and compensation. A guide or fictional exercise does not establish an accepted collection or a sale.