Sensors show how dairy cows graze

Background and Objective

Functional traits are an important part of the breeding goal to increase the lifetime of dairy cows. The monthly animal recording provides data concerning milk yield, milk composition and - indirectly - udder health and metabolic status of a cow. These data are used in the estimation of breeding values. However, other information, e. g. lameness scores, are not available on such a regular base. The project focused on these readings. Accelerometers, which offer the opportunity to record the activity and resting behaviour directly on the animal, were evaluated with regard to other applications. We tried to monitor the grazing behaviour of cows.

Approach

In repeated studies, direct observations on pasture were compared with the signals recorded by the sensors. Accelerometers were attached either with a halter around the neck or hanging on a neckband. Algorithms were developed to differentiate between grazing and other head-neck-movements. In another experiment, the sensitivity and specificity of algorithms and sensors were evaluated.

Repe

Results

The tested accelerometers (IceTag3D, IceRobotics Ltd., UK) can - on principle – be used to record grazing behaviour of dairy cows and distinguish grazing from other head-neck-movements. However, sensitivity and specificity of sensors and algorithms have to be improved before the system might be used to automatically monitor grazing behaviour in dairy cows.

Links and Downloads

www.optiscore.eu