Application of sensor technologies in animal breeding

Large groups

Livestock is often kept in large groups under commercial circumstances. However, in animal breeding, individually collected data is used to improve health, welfare and performance of animals. Collecting individual data of animals kept in large groups is a challenge, especially monitoring changes in health and behavior. With the upcoming use of sensor technologies and computer vision, automatic identification and monitoring of animals is possible, which provides opportunities for animal breeding. Breed4Food (B4F) is a consortium established by Wageningen University & Research and four international animal breeding companies. One of the projects of B4F is to develop methods to track and monitor individual animals kept in groups under commercial situations using sensor technologies. Furthermore, the information will be used to measure traits related to animal behavior, health and efficiency.

Novel sensor technologies

The aim of the symposium was to bring together animal breeders and researchers using state-of-the art technologies to track and monitor animals in groups. During the symposium, different novel sensor technologies were presented. Esther Ellen from Animal Breeding and Genomics Wageningen University & Research (ABG-WUR) started the symposium with giving an overview of the challenges within animal breeding. Malou van der Sluis (ABG-WUR) gave a nice overview of the different radio frequency identification (RFID) methods and the application in laying hens. In the presentation of Michael Toscano (University Bern) laying hens kept in a large aviary system were followed over time, focusing on movement patterns using light sensors. They found large variation in movement between birds, whereas individual birds were very consistent over time. Janice Siegford (Michigan State University) used accelerometers on laying hens to track directional activity of individual hens kept in an aviary system. They were able to record vertical and horizontal movement and the falling rate, distance and force of an individual hen. Deep learning and computer vision algorithms were used by Oleksiy Guzhva (SLU Sweden) for continuous behavioral monitoring of dairy cows under commercial situations (with dust, spider webs etcetera). They were able to track individual cows under these challenging conditions. Finally, Tomas Norton (KU Leuven) gave a nice overview of all the work that is going on at KU Leuven using sensor technologies, focusing on computer vision. The presentations were followed by a plenary discussion to place the presented research in a broader perspective and identify future research directions.