Lesson learned in big data for dairy cattle: advanced analytics for heat stress detection

This report provides an overview of the strategies for data management and data analysis developed within the EU project EIT Food DairySust “Big data and advanced analytics for sustainable management of the dairy cattle sector”. The main ambition of this project is to improve sustainability and animal welfare, besides productivity, in dairy farming, through advanced data analytics for every level of stakeholders. Good data management, in terms of acquisition, processing, harmonization and imputation, is required for good modelling for early diagnosis and for the identification of optimal prevention strategies, particularly in fields where monitoring can collect very heterogeneous data, and for which agreed protocols have not yet been standardized. The project investigated the “ecosystem” of data and application strategies for sharing computer resources and information in a secure and organic manner. This research first developed an optimal computational ecosystem based on the integration and harmonization of heterogeneous data types. Classical and advanced modelling strategies were used and compared. The results are suitable to provide the stakeholders with improved decision-making process about animal welfare and sustainability of the production. This report focuses on the implementation of a numerical model for the assessment of the impact of heat stress on milk production and provides a feedback on it.