Localization and accelerometer sensors for the detection of oestrus in dairy cattle

The aim of this work was to combine ultra-wide band (UWB) localisation tracking, a neck-mounted accelerometer and a leg-mounted accelerometer for the detection of oestrus in dairy cows. Twelve Holstein cows with successful artificial insemination (AI) were used in this study. The sensors were attached two weeks before the expected day of oestrus and removed after AI. Different cow variables (e.g. lying time, number of steps, ruminating time, travelled distance) were extracted from the raw sensor data and used to build and test the detection models. Logistic regression models were developed for each individual sensor as well as for each combination of sensors (two or three). The performances were similar when one sensor was used only as when combining the neck- and leg-mounted accelerometer (sensitivity (Se) =75-78%, area under curve (AUC) =93-94%). The performance increased when localisation was combined with either the neck- or leg-mounted accelerometer, especially for the sensitivity (80% for leg accelerometer + localisation and 88% for neck accelerometer + localisation). The AUC were nearly the same (97%). The best performance was obtained with the combination of all three sensors (Se = 90%, AUC = 99%). Future work will consist of expanding this research to other herds with larger sample size as well as considering cows’ anomalies (e.g. mastitis, lameness) and other sensors (e.g. bolus or eartag to measure the temperature).