Design engineer
- Sees
- The revision built, the inspection report and the design envelope
- Signs
- Proposes the disposition as a non-author
One case, two days before shipping
A patient-specific implant held at 08:54 and released at 11:58
A right mandible reconstruction plate: three segments, two bends, eleven screw holes. Its last records closed at 08:40. It ships in two days.
demoThe case, its plate, the surgeon, her hospital, the staff and each measurement shown are sample data.
- 0108:52
All records have closed, and no one has looked at a number yet
Design engineerOn the schematic of the patient's mandible sits the plan the surgeon signed, holes outlined and not yet read. Nothing is checked against the latest revision. The panel says what will happen: seven agents read the case file, lock the revision the surgeon signed, and compare each screw hole and bend with it.
- 0208:53
13 records read, and three of them used a revision nobody approved
Design engineerThe case file reader indexes the order, imaging, plan, three revisions, the approval, build package, build record, inspection, sterilization, label and packing list. Next, the revision lock finds the surgeon signed rev 2. The build package, the inspection drawing and the label all used rev 3, which has no engineering approval.
- 0308:54
Bend B2 is 6.1° off the signed plan, so four holes sit as far as 3.39 mm away
Design engineerAgainst rev 2, holes H8 to H11 are 0.83 to 3.39 mm from the plan, outside the sample tolerance of 0.30 mm. The other seven holes are within 0.04 mm and B1 within 0.2°. 8 of 13 features pass. The case holder placed the hold and opened a nonconformance at 08:54.
- 0408:55
The inspection report passed, measured against the rev 3 drawing
Design engineerShe opens the coordinate measuring report herself. All values are there, and each passes against the drawing it was given: B2 measured at 24.1°, within 0.1° of rev 3. That is why nothing upstream raised a flag. A check that passes against the wrong revision does not count here.
- 0508:56
Rev 3 moved B2 from 18.0° to 24.0°, as an internal draft
Design engineerThe revision itself says why: segment 3 follows the planned bone segment more closely, and the last four holes move with it. It was never reviewed by engineering and never sent to the surgeon, yet the build package rule took the latest revision. She did not author it, so she can review it inside the nonconformance.
- 0609:20
Her proposal: use as is, only if the surgeon approves rev 3
Design engineerRev 3 sits inside the design envelope, which she has opened too. Her signature states that she reviewed rev 3 as a design change within the nonconformance, did not author it, and proposes a disposition QA must agree. She types her own reason. The other choice was a remake to rev 2, after the surgery date.
- 0709:22
QA release opens the nonconformance before agreeing
QA release specialistAs the second person on the disposition, she opens the inspection report and the nonconformance record, which notes the extent check: none of the other 25 case files read this week was built from a revision other than the signed one. She agrees in her own words at 09:41, so rev 3 goes to the surgeon.
- 0811:32
The surgeon reads rev 3 on her patient's model and approves it
SurgeonIn her own guest view she sees only her case: rev 2 dashed and rev 3 solid on the model, with the bend and segment 3 moved. She opens rev 3 in full, picks her reason and signs. Declining would have sent the case to a remake. No agent offers her a recommendation.
- 0911:58
Released at 11:58, after a fresh read found 13 of 13 within tolerance
QA release specialistThe agents re-read against rev 3 at 11:34, and the label's design version now matches. She opens the re-read results, the surgeon's decision and the carton label proof, each for three seconds, then releases the case with her own words. Released this week moves to 10 of 31. The nonconformance closes as use as is.