Today’s herd management undergoes a major transformation triggered by the penetration of cheap sensor solutions into cattle farms, and the promise of predictive analytics to detect animal health issues and product-related problems before they occur. The latter is particularly important to prevent disease spread, ensure animal health, animal welfare and product quality.

Sensor businesses entering the market tend to build their solutions as end-to-end pipelines spanning sensors, proprietary algorithms, cloud services, and mobile apps. Since data privacy is an important issue in this industry, as a result, disconnected data silos, heterogeneity of APIs, and lack of common standards limit the value the sensor technologies could provide for herd management.

In the last few years, researchers and communities proposed a number of data integration architectures to enable exchange between streams of sensor data. This paper surveys the existing efforts and outlines the opportunities they fail to address by treating sensor data as a black box. We discuss alternative solutions to the problem based on privacy-preserving collaborative learning, and provide a set of scenarios to show their benefits for both farmers and businesses.

F. Papst, O. Saukh, K. Römer, et al., Embracing Opportunities of Livestock Big Data Integration with Privacy Constraints, in: Proceedings of the 9th Internat Conf on the Internet of Things, Bilbao/Spain (2019) Art No 27