INDUSTRY FIELD GUIDE
Gasket and seal data for AI placement-view evidence
NIST publishes research on deformable-object assembly benchmarks.
NAICS 339991 · Gasket, Packing, and Sealing Device Manufacturing
NIST publishes research on deformable-object assembly benchmarks. An original gasket-placement exercise can examine what a view reveals about a flexible component and its source identity. NIST assembly-benchmark context
Design a fictional flange-shaped paper fixture and two soft-paper ring variants with different reference notches. Supply an instruction and component ticket. Ask which staged placement matches the issued notch orientation. Include a closed assembly that hides the ring and a pre-closure view whose component identifier is missing. The key should preserve that identity gap.
Annotate ring variant, notch, reference side, assembly stage and evidence view. Hold out a new ring outline for evaluation. Score a hidden placement inferred from the closed exterior, a similar outline used as complete identity and notch reversal independently. A complete case should connect the pre-closure view to its component record.
Use original designs and harmless materials. Customer joint designs, proprietary compounds, licensed specifications and laboratory reports require authority and qualified review. A licensing brief should define the placement-evidence task and checked examples, with no sealing-performance conclusion inferred from the props. Confirm buyer requirements before commercial preparation or capture.
Discuss this original manufacturing sample with Fiund
Original Fiund exercise
A visible notch does not prove the ring sits flat
An original paper flange and ring represent placement on a mock fixture. The case does not test sealing.
| Record | Source | What it records |
|---|---|---|
| 1 | Layout | Ring G2 notch aligned to square; ring should appear flat in side view. |
| 2 | Top image | G2 notch aligns to square; circular outline centered. |
| 3 | Side image | One edge visibly lifts away from the fixture. |
Question for the model
How should the two placement criteria be scored?
Show the reference answer and checks
Alignment is supported by the top view under records 1 and 2. The side view in record 3 shows a lift, so flat placement is not supported. Keep both labels; a correct top outline does not establish seating or sealing performance.
- Calling placement complete from a top view
- Turning an observed lift into a real leak diagnosis
Fields to preserve
- notch_alignment
- Records in-plane orientation
- side_view_gap
- Captures out-of-plane evidence
- fixture_reference
- Aligns the views to one specimen
Test another case
Remove the side view and require an unknown seating label while keeping the alignment result.
No actual gasket specifications, torque, pressure, installation guidance or containment 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.