Preparing the published article.
How Repeatability Is Measured
Prove it with data, not with a demonstration.
A practical method
- Choose a fixed measurement point with a dial gauge or dedicated fixture.
- Run the same program at least 30 times and record every deviation.
- Report the maximum and the distribution, not the best single result.
- Repeat with and without the payload attached.
What disturbs the result
| Factor | Effect |
|---|---|
| Temperature change | Thermal distortion of arm and workpiece |
| Loose base | Whole-machine shift |
| Load and speed | Larger dynamic error |
| Pose | Error grows near singularities |