Automatic On-line Monitoring of animal Health and welfare by Precision livestock farming

Livestock production today has become a very complex process since several requirements have to be combined such as: food safety, animal welfare, animal health, environmental impact and sustainability in a wider sense. The consequence is a growing need to balance many of these variables during the production process. In the past farmers were monitoring their animals in their daily work by normal audio-visual observation like ethologists still do in their research. Today however the number of animals per farm has increased so much that this has become impossible. Another problem is that visual observation never can be done continuously during 24 hours a day. One of the objectives of Precision Livestock Farming (PLF) is to develop the technology and the tools for the on-line monitoring of farm animals and this continuously during their life and in a fully automatic way. This technology will never replace the farmer but can support him as a tool that automatically and continuously delivers him quantitative information about the status of his animals. Like other living organisms farm animals are responding to their environment with several behavioural and physiological variables. Many sensors and sensing techniques are under development to measure such behavioural and biological responses of farm animals. This can be done by new sensors or by sound analysis, image analysis etc. A major problem to monitor animals is the fact that animals themselves are complex systems that are individually different and that are so called time varying dynamic systems since their behaviour and health status can change at any time. Another challenge for PLF is to develop reliable monitoring tools for such Complex Individual Time varying Dynamic systems (“CITD” systems). In this paper we will talk about what is PLF and what is the importance of PLF. Next we will explain the basic principles. Further we will show examples of monitoring tools by PLF such as on-line monitor for health status by analysing continuously the sound produced by pigs. Another example shows the on-line automatic identification of the behaviour of individual laying hens by continuous analysis of 2D images from a top view camera. Next we will demonstrate the potential of PLF for more efficient controlling of biological processes. Finally we will discuss how implementation might be realised and what risk and problems are. The technology that is already available and that is under development today can be used for efficient and continuous monitoring if an engineering approach is combined with the expertise of ethologists, physiologist, veterinarians who are familiar with the animal as a living organism.