Source: The Pennsylvania State University
Introduction
Dairy producers use precision dairy technologies to monitor the reproductive status, feeding behavior, and health of their herd. The precision technologies are deployed using a variety of sensors integrated with software such as triaxial accelerometers, gyroscopes, geo-positioning and triangulation networks, infrared thermography, thermometers, and microphones. These sensors are often wearable and can be attached to the cow to track their behavior or physiological status. If a cow deviates from her normal behavior, an interpretive algorithm will generate an alert with a software interface to inform the producer. Producers use these technologies for a variety of functions, but most validated sensors are used for estrus detection, and to evaluate if a cow is pre-clinically sick. Capturing estrus in cattle passively with these sensor systems saves on farm labor from chalking tails to the manual observation of estrus and improves economic efficiencies related to breeding programs if cows are bred in a timely manner (Adenuga et al., 2020). Furthermore, this technology can be used to capture the early signs of sickness behavior related to disease; thus many precision sensors are also used as screening tools to passively observe for transition cattle requiring further examination. This article introduces how precision dairy technologies function and a few of the commercial precision technologies available.
How precision dairy technologies function
Many precision dairy technologies are wearable sensors that can be attached to the cow. The behavioral information on each cow is collected within a set time clock in the device, and stored within a chip in the sensor. Most sensors store this information until a cow passes a base station. Once a connection is made with a base station, the data is automatically distributed to a server, and stored into a cloud for interpretation by an algorithm embedded in the commercial software. The concept of using a sensor, in combination with an interpretive algorithm that leads to actionable decision-making for a producer is called precision livestock farming. So how are alerts created for producers to use? Often, these sensor-based systems establish a baseline for each cow’s behavior either by using a rolling average of her behavior across several days (e.g., 7+ days), by observing for relative changes from her behavior within a short period (e.g., 48 to 72 hours), or by monitoring for deviations in the cow’s behavior from a set alert threshold (Costa et al., 2020). Using individual cow behavior to create alerts for precision technology software is imperative as many validated precision technologies were designed with precision in mind rather than accuracy to preserve the battery life of each sensor so that it lasts for several years.
A useful precision dairy technology system has precision. Precision is the ability to consistently measure values closely to one another, like clustering all of your darts within a specific part of the target, while accuracy is the ability to consistently measure the true value of a behavior, like hitting the bullseye on a target (Practices of Science: Precision vs Accuracy). Precision technologies need to have precision, but not necessarily accuracy. For example, a pedometer system that monitors for deviations from a cow’s normal activity patterns to detect estrus needs to be precise to capture when she increases her overall activeness in response to estrus. However, it does not need to be accurate, we do not need to know the accurate step count to capture the increased activeness in her behavioral patterns.
One flexibility for most commercial precision technology systems is that producers can usually change the alert thresholds within the interface of the commercial software to ensure that estrus detection, or transition cow health alerts, are specific to their farm’s environment, milking routine, and feeding schedule. This is to avoid too many false positive alerts within the software, as producers will often ignore precision technology alerts if they are not highly sensitive (Eckelkamp and Bewley, 2020). As an example, some commercial transition cow health alerts were designed using smaller herds in Europe with indoor, outdoor access. A herd of 1200 milking cattle in freestalls that houses their transition cows next to the milking parlor and feeds their cattle a specific transition cow ration should change the alert threshold because their herd walks a shorter distance, eats a different diet, and is checked on multiple times a day compared to the default alert system. Altering this alert threshold would decrease the prevalence of false positives that could cause producers to ignore the technology. Contact your local extension educator, or the representative that sold your farm the commercial precision technology for guidance related to changing an alert threshold on a precision technology software.
Common commercially available precision dairy technologies
Estrus detection
One of the most common precision dairy technologies available on farm is the wearable sensor designed to detect estrus in cattle by using an accelerometer (Adenuga et al., 2020). These sensors measure acceleration in three different axis in relation to gravity and are often used as wearable pedometers which can attach to a cow’s leg with a band to track her activity behavior like a fitness tracker. Neck collars, and ear tags are other wearable sensor types often used to track cow activity behavior to signal for when a cow is in estrus. These wearable sensors work with integrated software to detect when cows are ready for breeding by measuring when she increases her overall activeness (e.g., ear tags and neck collars) or by a pedometer which can also track increased step counts, shorter lying times, higher rates of acceleration of the leg, and increased lying bouts (Adenuga et al., 2020).
Health status
Most neck collar systems are accelerometer sensors integrated with an interpretive algorithm software; producers use them to observe a cow’s digestive health status by measuring neck movements that are highly correlated with rumination bouts (Grinter et al., 2018). Another option is prescribing a bolus to a cow which resides inside the cow’s reticulorumen; these bolus contain an RFID, a thermometer to record the internal temperature within the rumen, and these systems can track stomach movement. The thermometer in the bolus also works with integrated algorithm software which have been validated for accuracy to proxy for fever detection, and the bolus can be used to automatically adjust for water intake events, and feed intake, which are known factors that affect temperature in the rumen (Cantor et al., 2018). Some producers also visually use the rumen bolus software to monitor for heat stress in the herd (trends for increased water intake with decreased feed intakes) and ruminal pH. However, one limitation to this technology is that the sensor must be replaced for each cow and cannot be reused. Neck collars and ear tags can also record total feeding time which is used to monitor for deviations in feeding behavior patterns that are associated with pre-clinical disease status in cattle (Reynolds et al., 2019). Some older models of neck collars will record burps associated with cud regurgitation by using microphones (Reith and Hoy, 2012).
The reason tracking rumination patterns are useful is because when a cow has digestive illness, rumination slows down, and transition cattle are at a high risk for metabolic disease and digestive illness (Daros et al., 2022). In comparison, healthy cattle regurgitate, chew, and swallow their cud at consistent intervals, and rumination patterns that maintain this process are unique to the cow (Lock et al., 2006). However, if a cow has a displaced abomasum, rumination patterns change and can even stop altogether because the abomasum is filled with gas (Stangaferro et al., 2016). Some producers use rumination collars only during the transition period to ensure that cattle at-risk for disease are offered preventative drenches, or are monitored more closely. Similarly, metritis, or an infection of the uterus, is detectable by monitoring for deviations in rumination patterns, and activity patterns in a cow (Stangaferro et al., 2016). Mastitis, or an infection of the udder, has also been associated with sickness behavior that results in altered activity patterns in cattle (Stangaferro et al., 2016). Therefore, systems which passively observe for rumination patterns in transition cattle may be useful to screen cattle at-risk for metabolic disease.
Limitations to precision livestock farming concept
First and foremost, ensure that any precision technology system under consideration for your dairy has been internally, and scientifically validated. It is fundamental that the technology under consideration has been proven to measure what you are interested in monitoring, and has been validated for reliability in a commercial dairy setting.
One word of caution is that precision technology systems cannot replace health exams. Plus, many producers that use estrus detection systems still synch their cows and use the precision technology system as a tool to improve their conception rate by cherry-picking cows for breeding in-between cycles. These systems cannot replace a trained herdsman that performs routine transition cow monitoring, and for estrus, the system requires someone to inseminate the cattle in a timely manner.
Based on Dr. Cantor’s extension experience, anecdotally, the most successful herdsmen that implement a precision livestock farming concept on their dairy are mindful that this reallocates their labor, rather than provides true labor savings. For example, rather than allocating labor to chalking cow tails, producers who use precision technology systems for estrus detection re-allocate the labor to maintenance, and using the data as a decision-making tool for their herd. Thus, consider precision livestock farming concepts as a tool to re-allocate labor from performing manual tasks to spending the time using data for more informed decision-making on the dairy. These tools are meant to screen for animals that deviate from their normal routine because we cannot watch a cow 24 hours a day. This is important to keep in mind as these precision livestock farming concepts are precise, not diagnostically accurate for disease diagnostics.
Milk capture to record milk production
Milk meter systems (top photo), are manually attached to milking units to measure milk production in cattle once monthly by a certified technician. Today, producers can integrate this system into their parlor by using milk flow meters, also known as milk capture systems, to capture the daily milk production of their dairy herd. Some producers with milk capture systems will monitor milk components, and the somatic cell count on each cow monthly (bottom photo).
One very common technology on farms is the use of milk capture. Milk capture is a blanket term which refers to the milk meter, and integrated software that monitors the milk production and labor efficiency of the parlor. Milk meters are units that are placed in tandem with the hoses of each stall of a milking parlor. The in-flow milk meters record milk flow to determine when to remove the claw, conductivity to proxy for mastitis, and summarize milk flow into milk weights to record the daily pounds of milk produced per cow. The systems identify cattle using Radio Frequency Identification Devices (RFID) that are placed into the cow’s ear. When the cattle pass a panel, the RFID is read so that the milk capture system associates milk production with that particular cow. Most milk capture equipment also record metrics regarding labor efficiencies such as milking time per cow, number of turns per hour, and number of teat cup re-attachments. Milk production is an important metric for producers to monitor because cattle can decrease milk production for a number of reasons such as a late stage of lactation, estrus, lameness, days carrying a calf, and disease (Zaninelli and Tangorra, 2006). Furthermore, milk production is also monitored to track herd-related factors such as poor quality feed delivery, poor quality feed, mycotoxin in the ration, and the presence of stray voltage (Zaninelli and Tangorra, 2006). Conductivity is also sometimes used by producers because it can also be used as an indirect proxy for mastitis; researchers observed that cows with increased somatic cell counts changed conductivity rates in the milking parlor (Norberg et al., 2003). Thus, milk production can be used to monitor the health of individual cattle, or be used as a tool to indicate when there are changes in overall herd performance.
Limitations to milk capture equipment
Conductivity should only be used as a screening tool for mastitis. Conductivity is only correlated with somatic cell counts in cattle and cannot replace a good standard operation of procedure to visually assess for mastitis such as stripping teats to check milk for flakes prior to milking, California Mastitis testing, recording somatic cell counts through routine milk testing, and other routine mastitis monitoring protocols (Fan et al., 2023). Furthermore, many parlors have only one RFID reader per side of the parlor. Sometimes a cow will have her tag read by the reader before she has entered the milking parlor. This can cause other cows that enter the parlor to be displaced electronically, and the wrong cow ID will be displayed. It is important that the milking parlor staff routinely observes that the last cow in a milking stall has the ID that is displayed on the screen in the parlor. A simple correction in the display for a cow that is misidentified by the reader can save herd manager time. This is also important because some milking parlors can also use sort gates if a cow is detected with an alert for herdsman intervention. The wrong cow will be sorted if the RFID is not identified correctly by the reader. Herd managers will avoid false positive alerts for milk production on a cow that is misidentified by the reader.
A note for success for any precision livestock farming concept
It is important to delegate one individual on the farm to maintain the technology by replacing units that have poor battery life, cleaning technology removed from sold cattle, and placing new units on heifers entering the herd. Similarly, a plan is needed to determine which data is the primary reason for investing in the precision technology system. Many systems have multiple potential uses, Dr. Cantor suggests that if a farm is considering a precision technology system for integration into their operation, it is recommended that they meet with the herd veterinarian first to discuss the plan for what the technology will be used for. Ensure that someone on the farm who is familiar with the current herd management software will also routinely monitor cattle as part of the daily routine for the alert of interest, and that a plan is in place for what the type of intervention is related to the alert. For example, someone needs to monitor for estrus alerts daily and have a system in place to communicate to the breeder to ensure that the cow is inseminated at the correct time. Similarly, if a transition cow creates a feeding behavior or rumination alert, someone needs to detect the alert, and delegate to the appropriate herdsman to determine what type of health exam is needed specific to their operation. All producers who are successful with the precision technology system and report a return on their investment adopt these practices to ensure communication with the sensor, software, and people on the farm are in harmony.
Conclusion
In summary, precision dairy technology systems are precision livestock farming concepts that involve a sensor, integrated software, and decision-making to intervene on a cow that changes their behavior or physiological status. Most validated precision technologies are used to monitor for changes in a cow’s reproductive status, health status, and for deviations in milk production. However, not all technologies are equal, and alert thresholds within the precision technology software should be monitored to ensure that it avoids too many false positives (i.e., healthy cows alerted as sick). Always delegate a person on the farm to routinely enroll new cows entering the herd, and to monitor, replace, and maintain the system. Furthermore, ensure that a plan is in place for who will use the data. Specifically, who will intervene to further examine a cow that generates an alert, and what actions are needed after the alert is detected. Work with the herd veterinarian prior to adopting any precision dairy technology of interest to develop implemented protocols, and to ensure that targets are met for the expectation of the technology. Finally, ensure that the precision technology and the behaviors or physiological changes it measures have been scientifically validated before considering the investment.
Authors:
