Control charts applied to simulated sow herd datasets

J. Krieter, J. Engler, K.-H. Tölle, H.H. Timm, E. Hohls

(view pdf)

This paper presents an approach for computer-based farm analysis, to support the decision-making of the farmer. The aim was to develop a method enabling the farm manager to quickly detect with statistical certainty emerging process changes in swine production. To gain more insight into the optimal design of the CUSUM as well as the EWMA (exponentially weighted moving average) chart in agriculture production processes without any side-effects, a simulation study was performed.

It was found that both the EWMA and CUSUM charts enabled a clear graphical presentation of the results. The classification performance was comparable. The presentation of the EWMA chart may be easier to understand, because actual values are assigned in contrast to CUSUM, which works with deviations. An advantage of CUSUM is the possibility to differentiate between calculating the positive and negative deviations.

Modern computer technology enables the implementation of control charts in online production monitoring. Further, analysis of the relevant (out-of-control) deviations should examine the underlying causes to improve farm performance.