On the Determination of Acceleration Thresholds for the Automatic Detection of Cow Behavioural Activities in Extensive Livestock Systems

In recent years the use of wearable sensors has proved to be valid tools in the field of livestock farming. They have played a crucial role in the automation and evolution of animal health and welfare monitoring processes, especially in the case of extensive farms where there is an infrequent farmer-to-animal contact. Accelerometers allow the measurements of movements of a body in space and are currently very popular sensors in the zootechnical field for monitoring livestock as they can be worn without being invasive for animals. Therefore, the aim of this research activity was to prove the feasibility of using acceleration thresholds for the automatic detection of grazing cow behavioural activities. In this study, a customized electronic device equipped with a triaxial accelerometer was built and placed in a collar of a grazing cow. The accelerometer data acquired at 4 Hz from the device were sent to a cloud server one time an hour. Then they were subsequently analysed to be able to establish thresholds for the discrimination of cow behaviours. As stated in previous studies, algorithms based on statistically defined acceleration thresholds could be easily implemented in a firmware for real-time detection of cow behaviour, therefore, they could be a valuable low-cost tool for monitoring livestock in extensive farms.