Science In The Spotlight: What Really Drives Farmers' Decisions In AMS Mastitis Detection?

At DeLaval, we don’t just create technology – we go into the field to ensure that what we develop truly meets farmers’ needs. This hands-on approach provides insights that shape our solutions.

One example of this commitment is a study co-authored by Dorota Anglart, one of our udder health specialists, in collaboration with the Swedish University of Agricultural Sciences and Växa Sverige.

The study, published in the Journal of Dairy Science, shines a light on mastitis detection in Automatic Milking Systems (AMS) and asks an important question: “Which data do farmers actually rely on – and what drives their decisions?”

Why is this study important?
Mastitis remains one of the most costly and challenging health issues in dairy farming, impacting animal welfare, farm profitability, and even the environmental footprint of milk production. AMS offers a wealth of data and detection tools, but with so much information, the risk is that farmers either miss key signals or become overwhelmed. Understanding what really drives decision-making is essential for designing solutions that deliver value, not just features.

How was the research conducted?
The study used a mixed-methods approach:

• National survey: 246 Swedish dairy farmers using AMS completed a detailed questionnaire covering herd size, AMS brand, technology features, and their mastitis detection routines.
• In-depth interviews: Nine farmers were interviewed to explore their daily practices, challenges, and the reasoning behind their choices.

This combination of methods allows researchers to capture broad trends while adding rich, real-world context.

And what did they find?
The survey and interviews revealed that mastitis detection isn’t just about having the right technology – it’s about how farmers interact with it. Key findings include:

1. The AMS brand sets the stageThe strongest factor influencing how farmers detect mastitis was the AMS brand itself. Each system presents data differently, shaping not just what information is available, but how it is used. For example, Lely and DeLaval systems differ in how they measure and display somatic cell count (SCC) data and generate alerts, leading to distinct routines on different farms.
2. Observation and intuition still come firstDespite advanced detection tools, many farmers rely first on practical observation – such as checking cows that are late for milking (‘red cows’) – before turning to specialised data like SCC or mastitis detection indexes (MDi). “It is no high-tech solutions we’re using. She doesn’t show up for milking, that’s how we see it.” (Farmer 5). This approach, while rooted in experience, can delay detection and treatment, increasing costs and animal welfare risks.
3. Somatic cell count is the trusted benchmarkWhen available, SCC data from cell counters (e.g., DeLaval OCC) becomes the main decision tool. Farmers find SCC easier to interpret than electrical conductivity, as it’s a standard measure with clear benchmarks set by dairy plants. “We use OCC very frequently and that’s how we find out if they are sick or not. This means that the other functions are actually a bit superfluous.” (Farmer 2). However, there’s a risk of “SCC tunnel vision”, where other valuable data is overlooked.
4. Information overload is a real challengeAMS provides a wealth of data, but too much information can be overwhelming. Most farmers adapt by creating their own workflows, often ignoring features they find confusing or unnecessary. “It’s a bit like not seeing the forest for all the trees.” (Farmer 5). Farmers want simple, actionable decision support – early warnings, not just more data.
5. Integration with herd health management is limitedAMS data is rarely integrated with other herd health systems or used by veterinarians and advisors. This is a missed opportunity for more holistic, data-driven herd management. “They [the veterinarians] look at the animal only, it is an examination of the individual for their part.” (Farmer 1). Some farmers even consider leaving national milk recording schemes (like SOMRS) because they feel their AMS provides enough data – though others value the sense of community and additional insights these schemes offer.

So why does this study matter? Practical implications for DeLaval and our customers
Success depends on designing systems that fit real-world behaviour and decision-making. This means we should:

1. Design for clarity, not just capability: Farmers value tools that provide clear, actionable insights. There’s a need to balance the richness of AMS data with simplicity in presentation and interpretation.
2. Support integration: Making it easier to combine AMS data with other herd health records and advisory systems could unlock new value for both farmers and their advisors.
3. Invest in training and support: Many farmers stick with default AMS settings, missing out on customisation that could improve their workflow. Training and peer learning can help bridge this gap.
4. Co-create with end users: The study highlights the importance of involving farmers in the development and testing of new features, ensuring solutions fit real-world routines and needs.

What’s next?
To unlock the full potential of AMS data for mastitis detection and herd health:

• Focus on integration – with herd health programmes, advisory systems, and veterinary workflows.
• Develop training and support that empowers farmers to customise and use their AMS to its full potential.
• Continue research that puts the farmer’s perspective at the centre of innovation.