Preliminary outcomes of a low-power cow oestrus detection system in dairy farms

In livestock management, accurate and timely detection of oestrus is a priority aspect to improve production systems. In a previous study, a moving mean-based algorithm for dairy cow’s oestrus detection from uniaxial-accelerometer data acquired in a free- stall barn was developed. In this study, the algorithm was implemented in a stand- alone smart pedometer (SASP) which is a customised electronic device designed to be connected to Low-power wide-area networks (LPWAN). In detail, the SASP was specif- ically developed to provide farmers with a real-time tool able to detect the ‘standing to be mounted’ behaviour by computing a specifically oestrus index. The SASP was equipped with both a triaxial accelerometer, which acquired data at 4 Hz, and a micro- controller which calculated the moving-means. Then the computed means were sent to a cloud server at 15-min intervals. A WebApp was specifically developed to monitor the oestrus status by producing a graph of the oestrus index. The SASP was tested in a free-stall barn for dairy cows during summer. The farmer selected six cows among those at thirty days distance on average from calving and six SAPSs were attached to the cow forelegs. All cow oestrus onsets were detected through the WebApp and then validated by farmer. Moreover, the graphs of the moving mean highlighted the occurrence of other atypical conditions related to cow locomotion activity. This study makes a new step forward to development of cow’s oestrus monitoring systems based on low-power wide-area networks (LPWAN)