In a modern dairy herd, cows are typically housed indoor in large groups with free access to a feed mixture without the farmer knowing the individual cow’s feed intake. This means that the management decisions, which are to contribute to an increased production, are based on the assumption that all cows are equally efficient. However, this is far from reality and far from optimum.

Therefore, VikingGenetics have started a new ambitious project “Cattle Feed InTake – CFIT” in cooperation with researchers from Aarhus University, SEGES and Simherd A/S. The purpose is to create technological solutions which in practice can collect data on feed intake of each individual cow in commercial dairy herds. In the project, the researchers will use 3D cameras and artificial intelligence to identify the cows, estimate their weight and quantify how much they eat.

Existing technology for data collection at an individual level is too expensive in practice

Today, feed intake is typically measured by means of weight-based technologies (e.gt. Insentec and Growsafe). These solutions are installed in several worldwide experimental herds (also at the Danish Cattle Research Centre), but they are expensive and time-consuming to maintain. Therefore, there is a need for a cheaper system with less workload which can automatically measure the cows’ individual feed intake on conventional dairy farms.

The value of feed data at cow level is great

Professor Søren Østergaard, Department of Animal Science, Aarhus University, is responsible for the part of the new project contributing with information about how valuable access to feed data from individual cows in a herd is. “We know that more than 70% of the variable costs in dairy production are related to feed production and feeding. Therefore, even a small improvement of the feed efficiency will be of great value to the farmer. In the project, we will, among other things, develop a model for calculating the economic ranking of the cows in a herd”, says Søren Østergaard.

Feed data at an individual level provide important information at many levels

From experimental herds, one knows that the feed intake of two cows at the same production level yearly may vary with up to one ton of feed measured in dry matter. Research also shows that the methane emission from cows is reduced by a lower feed intake at the same production level. Professor Martin Riis Weisbjerg, Department of Animal Science, Aarhus University, leads the part of the project which is to generate new knowledge about how feed data of the individual cow may be used for feed optimisation in practice. “Knowledge of each cow’s feed intake will provide us with the opportunity to calculate the efficiency for each cow. In our part of the project, we will, among other things, study whether the cows during a lactation and from lactation to lactation ranks in the same way regarding feed efficiency, allowing for selection for efficient animals early during the first lactation”, explains Martin Riis Weisbjerg.

The opportunity to collect data at an individual level over herd level also makes it possible for the farmer to receive important information about each cow’s health and physiological status thus act early in case of a possible beginning of health problems. Finally, according to VikingGenetics, the project and the many data at individual feed intake recordings will open for unique opportunities and new studies in the complexity to be found in cattle breeding, resource efficiency and climate impact.

Facts about the project