What information counts when detecting mastitis in automatic milking systems? A mixed methods approach from a Swedish...

The increasing adoption of automatic milking systems (AMS) in modern dairy farming has shifted mastitis detection from traditional human-animal interactions to technologically mediated processes. This study used a mixed methods design, combining a quantitative survey of Swedish dairy farmers (n = 246) with in-depth qualitative interviews (n = 9). The survey explored the use of AMS data for mastitis detection across herds varying in size, AMS brand, and technological features. The interviews provided rich insights into farmers' practices, challenges, and decision-making processes regarding udder health management. Our findings revealed that AMS brands and tools create distinct working environments, influencing farmers' behaviors around mastitis detection. A common practice used to detect cows with udder health problems was to monitor the behavior of animals, for example, examine cows that are late for milking, rather than following the more direct udder health parameters, such as SCC or electrical conductivity. Farmers emphasized SCC as the key indicator of udder health. Integration of AMS data into broader herd health strategies, including collaboration with veterinarians, remains underused. Enhanced training in AMS customization and closer integration with advisory systems could optimize the use of available data. These insights offer a foundation for refining mastitis management and improving udder health in AMS-managed herds.