RT Dissertation/Thesis T1 RumiWatch - Development and assessment of a sensor-based behavior monitoring system for ruminants A1 Zehner,Nils WP 2019/08/13 AB Sustainable and competitive milk production is highly dependent on securing the performance potential, health and fertility of dairy cows. Therefore, farmers can benefit from sensor data of animal monitoring systems to improve health management and work processes in dairy farming. The research during this PhD thesis aimed to contribute to the development and evaluation of a scientifically validated, sensor-based animal monitoring system that comprises a device for measurement of ingestive behavior and a device for measurement of movement behavior in cattle that interact as a system with system-specific software. Further aim of this thesis was to evaluate application potentials for this animal monitoring system by means of calving prediction in dairy cows and measurement of chewing activity in horses. The underlying experimental work was structured into four separate studies. The aim of the first study was to develop and validate a novel scientific monitoring device for automated measurement of rumination and eating behavior in dairy cows. Research works for this study aimed to provide a complete and detailed technical specification of the functionality of this device and to perform a validation under field conditions in stable-fed cows. The objective of the second study was to develop and validate a novel algorithm to monitor lying, standing, and walking behavior based on the output of a triaxial accelerometer collected from loose-housed dairy cows. The third study aimed to use automated measurements of ingestive behavior obtained from the developed sensor device to develop and validate a predictive model for calving in dairy cows. The aim of the fourth study was to investigate the suitability and validity of the developed sensor system for automated measurement of chewing activity in horses. In conclusion, the RumiWatch noseband sensor and pedometer that were developed and validated in the current project represent a suitable measuring instrument for automated recording of ingestive and locomotor behavior in dairy cows. The system-specific software is suitable for research purposes and shows a high performance for classification of extended parameters of rumination, eating, lying, standing, and walking behavior. The achieved validation results indicate that the measuring performance satisfies scientific requirements. Further application potentials were demonstrated by means of automated calving prediction in dairy cows and automated measurement of chewing activity in horses. The development and validation of a predictive model for calving time using measurements of the RumiWatch noseband sensor revealed a high amount of false positive alerts that was prohibitive for application of the model in farming practice. However, the analyses showed that particularly parameters of ruminating behavior have predictive value and should be taken into consideration for future research on calving prediction models. Furthermore, it was successfully demonstrated that it is feasible to apply the RumiWatch noseband sensor to horses. The results of direct observation compared with the automatic measurement showed a very high overall agreement of the observed and automatically measured data and, after minor refinements, this measuring device has the potential to become a valuable and easy-to-use tool for equine research and management. K1 Sensor K1 Wiederkäuer K1 Präzisionslandwirtschaft PP Hohenheim PB Kommunikations-, Informations- und Medienzentrum der Universität Hohenheim UL http://opus.uni-hohenheim.de/volltexte/2019/1593