简体中文界面。帖子、指南及政策正文保留原文;未对原文作自动翻译。

Which physical cover should our FR10 scratch label describe?

LeoBennett0710 · 2026年8月14日 14:38 UTC

回复讨论
LE
LeoBennett0710
I am preparing FR10 presentation of polished covers to a fixed camera. The bright streak shifts when presentation changes slightly, yet our examples label it a scratch. How should I establish the physical references before anyone tunes thresholds or treats the proposed inspection as commissioned?

18 条回复

IM
ImranBarnes0566

Identify the retained covers and have the acceptance owner examine the surfaces under an agreed viewing method. The same pictures used to invent a label cannot independently confirm it.

17
CA
CallumChen1143

What actually changed between pictures, and can each image still be linked to the same physical cover?

5
AN
AnilBennett0723

Keep the disputed examples visible as disputed. Quietly deleting them makes the proposed inspection appear less sensitive to presentation than it is.

4
LE
LeoBennett0710

The covers are identified and retained. They were reseated for the example pictures, with small differences visible. The scratch labels were assigned from the screen rather than a separate physical review.

6
LU
LucaChan1048

Can quality examine an accepted bright surface as well as the suspected scratch?

12
IM
ImranBarnes0566

Yes, include accepted finish variation. A system that rejects every bright line could find the scratched example and still be useless. The initial reference group should help distinguish those behaviours before a broader evaluation is planned.

16
CA
CallumChen1143

Will the seating review establish which variation is intended, rather than assume every pose in the old pictures belongs in the final process?

7
LE
LeoBennett0710

That is not defined yet. I have asked tooling to establish the intended seat while quality reviews the covers. I do not want a fixture fault disguised as a camera setting.

9
AN
AnilBennett0723

Nor the reverse. A sound seat does not establish that the lighting reveals the surface reliably. Keep both questions in the comparison.

17
LU
LucaChan1048

Agreed. Improving one photograph isn't the acceptance test.

11
IM
ImranBarnes0566

Use repeat captures of identified pieces with deliberate changes recorded separately. Otherwise a new cover, a new seat and a new light position can all change at once, and the clean-looking result tells you little about which condition mattered.

-2
LE
LeoBennett0710

Quality accepted the bright feature on one cover and confirmed a physical scratch on another. Tooling also found that one of our photographed seats was outside the intended locating condition. Both findings are recorded separately.

11
CA
CallumChen1143

Does the scratch remain visible in the intended seat, not just in the old awkward presentation?

24
AN
AnilBennett0723

And can another person reproduce that seat from the instruction? Otherwise the comparison may depend on the demonstrator's careful placement.

5
IM
ImranBarnes0566

Those are the two useful next checks. Keep the original misseated image as an explained example, not a normal-condition test result and not a reason to erase the fact that the former setup allowed it.

3
LE
LeoBennett0710

The intended-seat comparison retained the confirmed scratch and reduced the moving bright feature on the accepted cover. A second loader reproduced that seating. This is only the initial two-cover comparison; the wider finish range remains untested.

7
LE
LeoBennett0710

Thanks, Luca. Including the accepted bright cover stopped us judging the setup only by whether it found our obvious reject. The reference records now contain the physical decisions and both capture conditions.

12
IM
ImranBarnes0566

That gives the wider study a defensible starting point. Have quality and vision select evaluation pieces outside the tuning group, with the intended finish variation recorded, before claiming inspection performance for production.

17

参与讨论

欢迎来到 Application Robot

所有人都可以阅读论坛。登录或注册后即可发起讨论、回复或上传照片。

忘记密码?

注册账号即表示你同意我们的 使用条款 社区准则。请阅读我们的 隐私政策 ,了解我们如何处理你的信息。