lesions and α=0.55 for tail lesions when the observer assessed theindicators directly in the abattoir (SOD). These lower values for thedirect assessment were most probably related to the practical constraintsat the conveyor belt. Most restricting, the time to examine thesingle carcasses was very limited, with approximately 8 s per pig at anoperation speed of up to 450 pigs per hour. For the evaluation of thetaken pictures, no time limit was set, and the “specifications” withexamples pictures could be consulted in edge cases. Also, the intraobserverreliability for SO assessing based on pictures and directly atthe abattoir (ear lesions: α=0.89; tail lesions: α=0.71) revealed thedifficulties that go along with the assessment directly at the conveyorbelt. Here, another limitation of the direct assessment became obvious:In a standing position, it is easier to look at the ears than at the tails ofthe hanging carcasses, depending on the evaluator's body size. This factcan also partly explain differences between the single observers, whovaried in height. Due to place restrictions, it was not possible to install agangway for better inspection. The camera-based system overcame this