INDUSTRY FIELD GUIDE
Concrete placement and specimen records for AI review
A cylinder result may carry the same date as a pour without being its specimen.
NAICS 238110 · Poured Concrete Foundation and Structure Contractors
A cylinder result may carry the same date as a pour without being its specimen. ACI publishes field-testing and certification context. A concrete contractor can test whether an assistant links a practice placement record, specimen label and review before reporting anything about a result. ACI field testing certification context · FHWA e-ticketing context
Invent two placements and three training specimen labels. Pair one placement with a verified test record, and give the other a result referencing the wrong specimen. Add a duplicate label photo that does not mean a second test occurred. Ask which evidence can be attributed and which status remains unresolved.
Keep real mix designs, client tests, jobsite information, engineering tolerances, and people outside the sample. Use dummy tags and avoid active concrete handling.
Describe a concrete specimen records sample
Original Fiund exercise
Match a result to its specimen before matching its placement
An invented concrete packet uses dummy tags and arbitrary result units. It has two placements, three specimen labels and a duplicate image of one tag.
| Record | Source | What it records |
|---|---|---|
| CP1 | Practice placement register | Placement P-A links specimen C1; placement P-B links specimen C2. |
| CP2 | Fictional test report | Reviewed result 14 practice units applies to C1. A second result of 16 applies to C3, with no placement link. |
| CP3 | Tag images | Two photos both show C1. A file caption says P-B, but no authorized correction changes the register. |
Question for the model
Which result can be attributed to P-A or P-B, and how many independent C1 tests do the two photographs establish?
Show the reference answer and checks
CP1 and CP2 support attributing C1's reviewed 14-unit result to P-A. P-B is linked to C2, for which no result is supplied. The C3 result cannot fill that gap. CP3 shows two images of the same tag, not two independent tests. Its caption conflicts with the register and should be flagged rather than silently overriding the placement link.
- Assigning C3's result to P-B because it is the remaining unmatched report.
- Counting repeated photos of C1 as separate physical specimens or tests.
Fields to preserve
- Placement-to-specimen relation
- Defines the source of attribution.
- Test event identifier
- Separates observations from duplicate media.
- Conflicting caption and register
- Preserves disagreement for review.
- Result unit
- Keeps arbitrary exercise values separate from engineering measures.
Test another case
Add a valid correction linking C3 to P-B and explicitly superseding C2's original tag entry. Ask the model to resolve attribution while retaining the correction trail.
Use paper tags and invented reports. Exclude actual mix designs, engineering tolerances, client test results and active concrete handling.
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.