How can our bezel reference teach the UR5e what finish to tolerate?

HassanAdams0095 · 25 Aug 2026, 13:38 UTC

Reply to discussion
HA
HassanAdams0095
I need an accepted bright-streak reference as well as a genuine defect for our UR5e bezel camera plan; the present images label moving highlights as scratches, and I want physical decisions and recorded presentation before commissioning or threshold tuning.

10 replies

NA
NathanChen1156
Replying to HassanAdams0095

Give quality the identified bezels and disputed regions, then preserve its physical judgement independently of each image. A whole-bezel pass label may not explain which bright patch was acceptable or which other region carried a defect.

18 points
HA
HarishBarnes0576
Replying to HassanAdams0095

Could the first repeated captures keep the existing light fixed? That gives you a baseline before a holder change and a lighting change become inseparable.

13 points
HA
HassanAdams0095
Replying to HarishBarnes0576

That is the request now; quality has the retained parts, and the capture plan separates ordinary reseating with fixed settings from any deliberately changed lighting.

14 points
PA
PavelBarnes0527
Replying to HassanAdams0095

What is the intended seat, though? I've seen people repeat a part carefully in an undefined holder and call it a controlled comparison

-5 points
HA
HassanAdams0095
Replying to PavelBarnes0527

Tooling supplied the intended contacts and checked the holder; no mismatch found in that review, so the normal reseating comparison will use that identified configuration rather than a guessed pose.

12 points
NA
NathanChen1156
Replying to HassanAdams0095

That is a basis for the comparison, not yet its result. Include how the loader confirms the intended seat, so the operator does not have to reproduce the designer's private technique.

7 points
HA
HassanAdams0095
Replying to NathanChen1156

Quality returned an acceptable bright region and a separate real rim scuff; the loading method is recorded, and the fixed-setting comparison shows the accepted region changes brightness while the scuff remains visible in the checked views.

10 points
HA
HarishBarnes0576
Replying to HassanAdams0095

Does that mean you can leave the lighting alone for now and test candidate thresholds against those views?

2 points
NA
NathanChen1156
Replying to HarishBarnes0576

For this bounded development set, perhaps. Reserve independently judged examples for later evaluation, and keep the presentation range explicit. These useful views do not establish performance across all incoming bezels.

6 points
HA
HassanAdams0095
Replying to NathanChen1156

Engineer will use the recorded condition for initial threshold development and keep separate judged pieces for evaluation; nothing is being called a commissioned inspection, but we finally have a reason to retain the bright region rather than crop it out.

10 points

Add to the discussion

Welcome to Application Robot

Everyone can read the forum. Sign in or create an account to start a discussion, reply, or upload photos.

Forgot your password?

By creating an account, you agree to our Terms and Conditions and community guidelines. Read our Privacy Policy for how your information is handled.