First steps
- Run an attribute agreement study with known good and bad parts.
- Improve limit samples and lighting.
- Move detection closer to the cause.
How to measure success
Escape rate and kappa of the inspection. Record the value before you start, so the effect of every action can be shown.
When these methods are the wrong choice
- Attribute agreement analysis: Measured characteristics.
- MSA / Gage R&R: Nothing: do it before you trust any data.
- Poka-yoke: Causes in the material or the supplier’s process.
- PFMEA / DFMEA: Fixing a problem that has already happened; update it after the 8D instead.
- 5 Why: Problems with many interacting causes; use data and DOE there.
- Control plan: Nothing in series production: every IATF process needs one.
The methods in one line
| Method | What it is |
|---|---|
| Attribute agreement analysis | Checks visual and go / no-go inspectors against each other and a standard (kappa). |
| MSA / Gage R&R | Checks that the measurement system can see the variation you want to control. |
| Poka-yoke | Makes the error impossible or detects it immediately, so it does not depend on attention. |
| PFMEA / DFMEA | Analyses how a process or design can fail, rates the risk (AP) and defines prevention and detection. |
| 5 Why | Ask “why” until you reach a cause you can act on, with one chain for occurrence and one for escape. |
| Control plan | Lists every characteristic, how it is checked, how often and what the reaction plan is. |