Data-based design of an intelligent control chart for the daily monitoring of the average egg weight

Both animal dependent and non-dependent factors and their interactions have an effect on the quality of consumption eggs. With the recent development of fast, objective and non-destructive measurement methods for egg quality, extensive information on all input and output aspects of the production process of consumption eggs are available. This enables the application of the concepts of statistical process control (SPC), such as control charts, in order to detect anomalies. In this paper a quality control chart is presented for the daily monitoring of the average egg weight. In a first step for the design of the control chart, a non-linear model was developed to detect the natural increasing trend of the egg weight with the increasing hen age. This trend was then subtracted from the measured egg weight and the residual values were inserted into a cusum control chart. The data originating for two flocks of laying hens, a small flock (72 hens, deliberately subjected to challenges), and a large-scale experimental flock (500 hens) in an aviary housing, were used to evaluate the developed control charts. The results show that the quality control chart enables to quickly detect a decrease of the average egg weight. Similar algorithms for all variables on both input and output will make it possible to monitor the whole egg laying process to detect or to prevent a decreasing egg quality and hence increase profits for the poultry farmer.