It seems like every month there’s a new precision dairy monitoring technology introduced to the dairy market. Each introduction coincides with a lot of hype about how this new technology is going to change the dairy industry, with articles in mainstream media about how these new “Fitbits for cows” can be used to identify sick cows and cows in heat.

Bewley jeffrey
Innovation & Analytics Scientist / Holstein Association USA
Jeffrey Bewley was formerly Alltech’s dairy housing and analytics specialist. He has a MS from th...

Some of these new ideas are repackaged or improved ideas from older technologies that have been available for a decade now. Luckily, this also means we, as an industry, have quite a few years of experience with these technologies in both research and commercial settings. We have seen some technologies come and go. More importantly, we have learned valuable lessons from both successes and failures with precision dairy technology adoption and implementation.

As a refresher, precision dairy monitoring is the use of technologies to measure physiological, behavioral and production variables on individual animals to improve management strategies and farm performance. Precision dairy monitoring technologies provide tremendous opportunities for improvements in individual animal management on dairy farms. These technologies are changing how dairy producers manage reproduction and health.

Precision dairy monitoring technologies include wearable technologies to monitor animal behavior, image-based technologies to measure body condition score or lameness, real-time location systems to track animal movement and milk-based technologies to measure biomarkers for animal health and reproduction.

Wearable technologies incorporating accelerometers, which measure movement, have been applied to the leg, neck and ear to measure variables like lying time, number of steps, rumination, feeding and panting behavior. These technologies have been shown to be particularly effective for monitoring estrus behavior. Additionally, many dairy producers have had success using these technologies for disease detection.


So if you are excited by the many options available to you at this point, what criteria should you use to select the right technology for your farm? The following are some key areas to examine with any technology you are considering.

1. How well does the technology actually measure what the manufacturer says it does? This might seem like a simple question, but not all technologies are created equal. Some are better than others at measuring the variable they are marketed to measure. Just because one technology accurately measures something like rumination or feeding behavior doesn’t necessarily mean all of them do.

The best technologies have had third-party validations demonstrating how accurate they are at measuring whatever they are reported to measure. Thus, a simple question to ask of technology manufacturers is: “Can you provide a third-party reference demonstrating how well your technology works?”

2. How well does the system stand up to the harsh conditions of a dairy farm? A dairy farm is a difficult environment to keep technologies running in. Wet, dirty areas with large, clumsy animals create challenges for wearable and stand-alone technologies. It’s not easy to manufacture technologies to survive in these conditions. Rodents and weather conditions can also have a negative impact.

3. How often do the devices fail and need to be replaced? With all technologies, devices will fail periodically. A good goal should be to have less than 3 percent of devices fail per year. The best technologies on the market today meet this standard. It is truly amazing how reliable some of these technologies are. On the other hand, other technologies may need to be replaced much more often.

This becomes a costly, logistical nightmare for dairy management. Eventually, all devices will reach the end of their useful life. And, at some point, they are technically out-of-date anyway. Few of us have the same cellphone we did five or 10 years ago. It’s not realistic or necessary that devices last forever, but a low failure rate during the first few years should be expected.

4. Which device location fits your farm situation best? I have heard compelling, passionate arguments from dairy producers for why leg, neck, ear, tail or rumen device locations are the best location for a recording device, which usually include reasoning for why the alternative locations won’t work. The reality is: There are pros and cons to each device location.

The key is to decide which device location fits your farm situation best. What works best for you might not be the same as what works best for your neighbor. Talk to dairy producers using different devices and make the best educated guess for your situation. Remember, also: The “location” might be in the form of an image or milk-based technology rather than having devices on or in every animal.

5. Is the information presented in a logical, easy-to-use format? Information overload can become a problem quickly with these technologies. Knowing the exact location of every cow in your herd every second of the day is overkill. Having this information summarized to show how much time the cow spends at the feedbunk or in freestalls is likely more useful.

But, still, you don’t necessarily have time to sort through this information for every cow every day. Rather, this information needs to be summarized to show which cows might be deviating from their normal behavior. Algorithms that help point to the most important information in user-friendly software or phone apps make a huge difference in practical application of these technologies.

6. How well does the data integrate with other information on your dairy? Precision dairy technologies provide novel, exciting information. However, this information has less value if it cannot be combined with basic cow demographic information like days in milk, lactation number and pregnancy status.

For optimal use, precision dairy technologies need to interface well with your herd management software. Ideally, they also interface well with other sources of data. Integration of data from multiple sources is both a challenge and an opportunity in this new data era within the dairy industry.

7. Can this information be turned into a meaningful action? Once we know the technology is technically accurate, then we need to ask how we will use the information provided to undertake a meaningful action. For example, a large increase in activity telling us a cow is in heat can be turned into the meaningful action of insemination.

Many of the variables measured by these technologies provide us with information that can help us treat a sick cow. However, with some information, we might have to ask ourselves, “OK. I can measure that, but so what? I can’t really do anything meaningful with that information.”

8. Does the technology address a real need, problem or opportunity for your farm situation? Technology investment decisions are very farm-specific. Just because other farms are having success with a particular technology doesn’t necessarily mean it’s a good investment for your farm. If you are a producer who does an excellent job managing body condition score, maybe an automated body condition score system isn’t the best investment for your herd.

If you have a high pregnancy rate through successful synchronization, maybe an automated estrus detection isn’t the best fit for your situation. On the other hand, if those areas are challenges, they may be the right fit for you. It’s important to ensure the technology meets a real need or opportunity for your particular farm situation.

9. What are the false-positive and false-negative rates for the technology? If we are using a technology for management by exception, the accuracy of alerts provided by the system is essential. In an ideal world, each technology provides alerts identifying all events of interest without ever providing false alerts. In reality, this isn’t easy to achieve.

A false negative occurs when an animal has a condition (i.e., estrus or disease) but the system does not provide an alert. A false positive occurs when an alert is provided but the animal is not experiencing the condition. We want to minimize both false-positive and false-negative rates. If the false-negative rate is too high, we miss the condition we were looking for.

If the false-positive rate is too high, we end up examining, treating or breeding animals that don’t need action. The right balance of false positives and false negatives depends on the condition, the farm, economics and personal preference.

10. Will you observe an economic return from investing in the technology? This is a key question, no matter what the milk price is. If we are going to spend money on a technology, we need to understand the economic returns from investing in the system.

Improving reproductive efficiency, lowering disease costs and increasing cow longevity can all help pay for these systems. But we need to calculate the return on investment for a technology for each herd situation. Some of these systems have a hefty price tag. We just need to make sure we see a strong return for that investment. end mark

Jeffrey Bewley has a Master of Science from the University of Wisconsin – Madison and a Ph.D. from Purdue University.

Jeffrey Bewley is with BoviSync CowFocused Housing. Email Jeffrey Bewley.