Motoyasu Natsuga

Online Real-time Monitoring of Milk Quality during Milking by Near-infrared Spectroscopy

Recently, there has been a need by dairy farmers for a method to assess milk quality in real time during milking. We have constructed an on-line near-infrared (NIR) spectroscopic sensing system on an experimental basis. The NIR system can be used for real-time monitoring of milk quality items such as fat, protein, lactose, somatic cell count, and milk urea nitrogen during milking with sufficient precision and accuracy. We tried to improve the robustness of calibration (CAL) models for measurement of milk quality items using NIR spectrum data obtained from two dairy herds. When CAL models developed from data obtained from one herd were used for validation of data obtained from the same herd, the milk quality items could be measured with high levels of accuracy. On the other hand, when the CAL models were used for validation of data obtained from the other herd, the levels of accuracy in measurements of all milk quality items except fat were low. The low levels of accuracy may be caused by factors such as differences in cow individuality, lactation stage, calving times, feeding stage and experimental period (temperatures in the dairy barn). To develop robust calibration models for measurement of milk quality items, therefore, data acquisition from various milk spectra caused by these factors is necessary.

On-line Near Infrared Spectroscopic Sensing Technique for Assessing Milk Quality during Milking

There has been a need in recent years for a method that will enable dairy farmers to assess milk quality of individual cows during milking. We constructed on-line near-infrared (NIR) spectroscopic sensing system on an experimental basis. This system enables NIR spectra of unhomogenized milk to be obtained during milking over a wavelength range of 600 nm to 1050 nm. We also developed calibration models for predicting three major milk constituents (fat, protein and lactose), somatic cell count (SCC) and milk urea nitrogen (MUN) of unhomogenized milk, and we validated the precision and accuracy of the models. The coefficient of determination (r2) and standard error of prediction (SEP) of the validation set for fat were 0.95 and 0.42%, respectively. The values of r2 and SEP for protein were 0.91 and 0.09%, respectively; the values of r2 and SEP for lactose were 0.94 and 0.05%, respectively; the values of r2 and SEP for SCC were 0.82 and 0.27 log SCC/mL, respectively; and the values of r2 and SEP for MUN were 0.90 and 1.33 mg/dL, respectively. These results indicated that the NIR sensing system developed in this study could be used to assess milk quality in real time during milking. The system can provide dairy farmers with information on milk quality and physiological condition of individual cows and therefore give them feedback control for optimizing dairy farm management.