Precision Livestock Farming Potential use of an accelerometer for predicting parturition in dairy cows

Many farmers need to face the conflict between economic pressure and higher requirements in farming. High stillbirth rates and perinatal mortality are an occurring problem. Monitoring calving is time-consuming and labor intensive, and calving prediction via direct observation is not very accurate. The usage of small and lightweight sensors potentially allows automated monitoring of animals on a noninvasive or minimally invasive basis. The aim of our study was to evaluate the possibility of monitoring physiological changes of dairy cows in the prepartum period by using a triaxial accelerometer. Data collected in this study should be used to develop an algorithm to predict parturitions. For the preliminary study, an accelerometer was mounted on the upper part of the tail of 5 cows, approximately one week before the estimated calving. At the same time the dams were observed permanently by 24h-video recordings. The results showed that it was possible to sense the movement of the cow's tails and measure the raising of the tail by use of the sensor to predict the onset of calving. We used the data to develop a decision function that gave a birth alarm when exceeding a certain threshold. In the main study, the accelerometer was used in form of the SMARTBOW eartag and attached to the animals' ears. More than 1,000 calvings were monitored starting from a 12-day period preceding birth, until the expulsion of the calf. The development of the algorithm has not yet been completed. Preliminary results have shown that it is possible to predict parturitions using sensor-recorded information on lying time, rumination time and activity. Based on our current knowledge, triaxial accelerometers are suitable to monitor dairy cows and predict calvings. Sensor systems must be understood as additional tools in herd management. Their purpose is to support farmer's and veterinarians, not to replace them as decision-makers.