Research, testing and product development · Lesson 4 of 4

Research testing and product development: defects, bad evidence and case studies

Ways to verify a tested edge, a lifted coping cover, a heavier cleat with a lower published figure, a data sheet that is not a lab report, and how AFS declines to overstate.

You will be able to: Work a failed edge, a size or cleat substitution and a data sheet versus a test report to a go or no-go, with the evidence to request, and recognise honest research.

10 min · 11 slides · 6 questions

All lessons in this module

Slide 1 of 11

Four ways to verify a tested edge

SPRI's seminar lists four ways to verify an edge was ES-1 tested: a manufacturer certificate, a manufacturer test report, an FM Approval listing, and NRCA sub-listing.

It says a manufacturer with a certificate should also hold the test report, and for NRCA sub-listing you check that both the detail and the fabricator are listed.

Sources

SPRI Wind Design Seminar: Perimeter Edge Metal Systems for Low-Slope Roofing (slides, uploaded October 2025) - SPRI

Glossary

Manufacturer certificate: One of four ways SPRI's seminar lists to verify an ES-1 tested edge: a certificate supplied by the manufacturer or fabricator, who should hold the test report behind it.
Golden run: In AFS's takeoff gate, a test drawing with fixed expected answers. The AI model is not deterministic, so the drawing passes only when 3 runs out of 3 match.
Failure mode: A way a tested edge loses securement. ES-1's commentary lists full fastener pull-out, fastener pull-through, collapse of a cleat, fascia or cover, and disengagement of the cover from a cleat or clip.