INDUSTRY FIELD GUIDE
Cotton harvest module records for AI review
A cotton grower can know which field produced a module without knowing how the gin later grouped its outputs.
NAICS 111920 · Cotton Farming
A cotton grower can know which field produced a module without knowing how the gin later grouped its outputs. USDA publishes cotton classification resources and public discussion of module averaging. An AI evaluation can focus on the handoff from field creation to transport without making claims about the gin’s processing or grade. USDA cotton classing services · USDA module averaging pilot FAQ
Invent two module tags and show a created time, a planned pickup, an actual pickup, and a receiving confirmation. Make the planned truck assignment differ from the actual one for a module. Ask the model to identify what happened and which record confirms it. A useful extra case is a field tag visible in a photo with no receiving record. Keep any eventual classing label separately identified.
Do not export true field boundaries, fleet identifiers, ginner schedules, or private grade and payment details. Check ownership of machine-generated module logs with the operator and service provider.
Describe the field-to-hauler evidence
Original Fiund exercise
The planned truck and the confirmed receiving record can differ
An invented cotton-module timeline contains abstract field and truck aliases. No actual route, fleet ID, field boundary, schedule or commercial grade data is supplied.
| Record | Source | What it records |
|---|---|---|
| CM1 | Module creation and plan | Module M-A created from field F-A day 1. Planned pickup truck T1 day 2. |
| CM2 | Actual handoff | Day 2 pickup log names M-A and truck T2. Receiver G acknowledges M-A from T2 on day 3 with matching inert transfer reference H1. |
| CM3 | Other module | M-B field tag visible in a practice photo description. No pickup or receiving acknowledgment supplied; a classing-label field is blank. |
Question for the model
Which transport and receiving events are supported for M-A, and what can be said about M-B's handoff or eventual classing?
Show the reference answer and checks
CM2 documents M-A's actual pickup by T2 and later receiving acknowledgment at G under H1. CM1's T1 assignment was a plan, not the documented carrier. CM3 supports only M-B's field-tag observation, with no confirmed handoff or classing result. Keep field origin, planned movement, actual transport and downstream classing as distinct stages rather than treating a visible tag or receiving record as proof of every later event.
- Reporting the planned truck instead of the actual confirmed one.
- Inferring a receiving or classing outcome from a field tag alone.
Fields to preserve
- Module and field-origin alias
- Preserves the creation link.
- Planned versus actual transport
- Separates intended and observed events.
- Receiver acknowledgment and downstream state
- Limits what the handoff establishes.
Test another case
Add a receiving card for M-B with a conflicting transfer ID. Require reconciliation rather than acceptance based only on a matching module label.
Use synthetic logs and original labels. Exclude real boundaries, fleets, gin schedules and private grades; machine-generated export rights require 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.