On-line measurements in pig carcass classification: Repeatability and variation caused by the operator and the copy of instrument.

For nearly all pigs slaughtered in the EU, the lean meat content is assessed on-line at the slaughter line. The assessment is made indirectly by an instrument performing a number of informative measurements including the thickness of back fat as one of the most important and common measurements. Several types of instruments are used for making the measurements. The quality of the calibration (the prediction ability) has to be approved by the EU Commission. However, the maintenance of instruments, training of operators, working conditions and other factors influencing the routine are quite as important for the accuracy as the calibration. As a part of an EU funded project, partners representing thirteen European countries have investigated the instruments used in their countries focusing on the precision of indirect measurements. The preconditions have differed considerably between the countries resulting in a wide range of estimates of the repeatability and the reproducibility (precision) of fat and muscle thickness. Totally, there have been three different types of manual instruments - invasive probe instruments from three manufacturers, non-invasive ultrasound and callipers. Furthermore, the precision of two automatic instruments with respect to lean meat content has partly been estimated. Even though neither the aim nor the design of the experiments was set for a direct comparison between different instruments, none of them seemed to deviate notably from the others with respect to the precision of fat thickness. In this study, the only investigated influencing factors were the variations in operators and copies of instruments. Generally, the variations between operators were more important than the variation between copies of the same type of instrument.