| Known standard deviations | |||
|
Item |
Repeatability s(r) |
Reproducibility s(R) |
Reliability |
|
|
|
|
|
Error Sources

|
Known possibilities for error |
||
|
Error |
Frequency |
Error contribution expressed as standard deviation |
| Installation of the probe and the functionality of the work place | ||
| Condition of the probe concerning | ||
| - The size of the knife | ||
| - The sharpness of the knife | ||
| - The measuring guide | ||
| - Linear measures against standard blocks |
|
|
| Probing sites, place and angels |
|
|
|
- Insertion place |
||
|
- Insertion angle |
||
It should be noted / documented which instrument that are used, and for each instrument the program version or EPROM number should be recorded to make it possible to track the used algorithm.
|
Daily control |
Periodic control |
|
Control of instrument |
Control of instrument |
|
|
|
Control of data |
Control of data |
|
|
|
|
|
|
|
|
|
Control of reflectance with test block |
|
|
|
Control with standard block |
|
|
|
Check of performance of
classification |
It should also be a requirement that
the abattoir checks the performance of classification every day.
F.ex. that one person - on behalf of the management - must check 10 pigs
(which have been classified) per day per classifier to ensure that the
measurements have been taken in the right position (and right angle). As
it can be seen in the following pictures, the use of simple tools is
enough: |
|
|
|
|
|
|
|
|
|
| Control with steel square |
|
|
|
Check of classification results |
It is important that the abattoir several times a day checks that the average meat percentage
(and the average fat and meat depths) are as expected compared with other
days or other reference data. If it is not the case, the abattoir should make a check of the performance of classification and a check of
the equipment. |
|