Towards an Automated Detection of Oestrus in Dairy Cattle

Abstract

Heat detection is a key factor in the profitability of dairy herds. However, this detection demands a significant part of the breeder's working time and is made difficult by the short duration and the discrete behavioural changes associated with oestrus in modern dairy cows. Progress has been made in monitoring cow with electronics, biosensors and computer. As a result, automated heat detection systems have been developed. Currently available tools are automated detectors of standing heat, activity-metres and automated in-line systems measuring milk progesterone. Camera-software systems and monitoring of body temperature are being developed and may also be used as heat detection tools. The heat detection rate of most systems is above 80% with a specificity of detection generally higher than 90%. The accuracy, however, may vary considerably depending on the tool and model developed. The initial investment of several thousands of euros required for these automated systems becomes a source of profit in large herds, provided the recorded data are properly managed.