Gait Anomaly Detection in Dairy Cattle

Abstract

Cow lameness is a common welfare issue in the dairy industry that causes severe health and life quality issues to cows, including pain and a reduction in their life expectancy. The earlier a lame cow is detected, the earlier and more effectively it can be treated. A change in the gait of the cow is the earliest symptom of lameness. Currently, lame cows are detected by visual inspection performed by herdsmen, which is subjective and time consuming. We present an approach to automatically detect anomalies in the walking pattern of a cow as a possible indicator of lameness. The detection is done by a wearable motion sensor attached to a cow's hind left leg. Our approach builds an individual model of the usual walking pattern of a cow during the first minutes of use and detects deviations from this model afterwards. Results from a controlled experiment we conducted indicate that our approach can detect deviations in cows' gait with an accuracy of 91.1%. This information can be used by veterinarians to keep track of changes in the walking pattern of cows and to decide whether to treat a specific cow.