INDUSTRY FIELD GUIDE
Pulp mill data for AI handsheet and sample matching
TAPPI publishes material on preparing handsheets for pulp evaluation.
NAICS 322110 · Pulp Mills
TAPPI publishes material on preparing handsheets for pulp evaluation. A mill can propose a document task that keeps the laboratory sample tied to the correct pulp source and preparation record. TAPPI handsheet context
Create fictional pulp-bale tickets, a composite-sample log and a handsheet report. Combine material from two bales in one sample while keeping a third bale separate. Ask an assistant which source relationships support the requested report. Include a duplicate laboratory export with a corrected sample identifier. The key should preserve both versions and use the effective correction only for the affected report.
Annotate bale, sample, preparation reference, report version and source field. Hold out a different sample-naming scheme for evaluation. Score wrong-source matching, loss of the composite relationship and use of a superseded identifier independently. A correct result should establish which records belong together without inferring a property for untested bales.
Use invented values and self-authored sample forms first. Actual customer furnish requirements, licensed test procedures and outside laboratory reports need permission for the proposed AI use. A contributor brief should identify the association task, reviewed examples and available source history. Confirm the buyer's required metadata and acceptance method before preparing a larger mill archive for licensing.
Discuss this original manufacturing sample with Fiund
Original Fiund exercise
A handsheet from a mixed sample
Paper cards represent fictional pulp bales and a composite sample. The exercise tracks source contribution rather than pulp processing.
| Record | Source | What it records |
|---|---|---|
| P1 | Bale register | Bale B1 from lot X; B2 from lot Y. |
| P2 | Composite ticket | C9 contains 6 g from B1 and 4 g from B2. |
| P3 | Handsheets index | H3 formed from C9; reported at handsheet level. |
| P4 | Draft label | H3 tagged as lot X only. |
Question for the model
Which lineage labels belong on H3?
Show the reference answer and checks
H3 traces through C9 to both lot X and lot Y using P1–P3. The composite input mass fractions were 60% and 40%; they describe the prepared sample, not two separately measured handsheet outcomes. Replace the X-only label in P4 with composite lineage.
- Assigning an entire report to the majority input lot
- Treating input fractions as separately tested result fractions
Fields to preserve
- composite_id
- Connects handsheet to its preparation record
- input_mass_fraction
- Quantifies each source on the stated input basis
- outcome_grain
- Identifies a single composite measurement
Test another case
Record amounts by wet mass with dry content missing; do not infer dry-fiber proportions.
No chemical conditions, actual supplier lots, proprietary blends or strength claims.
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.