Real time operational support in young stock rearing

The aim of the Dutch Smart Dairy Farming project is to develop decision support models for operational support in farm management. This project started in 2012 and involves collaboration between several institutions and companies in the dairy chain. The project has three focus areas: animal health, fertility and feeding. The objective of the 'animal health' focus area is to optimise young stock rearing as a basis for extending the life of cows. Lowering the annual replacement rate, which is currently about 30%, has a significant positive impact on financial results. The main objectives of this young stock rearing project were to develop state-of-the-art tools which would help the farmer to optimise growth and development of young stock. Milk feeders and water drinkers combined with weighing scales were used for data collection at individual calf level. Daily data about milk intake, milk feeder visits, water intake and body weight were used to build detection models that generate alerts when measured values deviate and indicate a possible health problem. Body weight measurements were also used to determine the growth rate of a calf and to produce an alert if this deviated too much from a desired growth rate. All alerts were translated into messages with work instructions for the farmer. Weekly feedback from the farmer indicated that almost 60% of the messages were correct. It is suggested that the large number of wrong messages (false positives) can be reduced by applying more advanced analysis techniques.