Purchasing Guide for the Dairy: Precision Technology System

Introduction

The integration of precision technologies for dairy farm management has emerged as a vital tool for optimizing farm management practices to improve labor efficiencies and animal monitoring abilities. Among these advancements, Precision Livestock Farming (PLF), or the ability to use a sensor and software to make inferences about livestock for decision-making, stands out as a modern approach in the dairy industry. By processing real-time data and incorporating analytical software, dairy producers can use data to make new decisions and insights into their operations, enabling them to enhance dairy cow productivity [1]. Furthermore, PLF only works if the farm receives high-quality data [2]. Dairy farmers must plan and research the system they are interested in buying.

Below is a planning checklist for dairy producers to use before making a PLF system purchase.

Is it validated? Search for online publications that were conducted at independent research institutions to make sure the technology is validated for what you are buying it for. For best results, use simple search terms such as "accelerometer" rather than the commercial name of the system. Also, check that the system is validated and tested on dairy farms, some systems may only be tested in controlled lab settings. Note that all milking parlors and voluntary milking systems are legally required to be validated and this applies to wearable systems such as ear collars and neck tags.

Precision: The PLF system must be precise, meaning that it consistently measures variability day-to-day for breeding or health alerts to be reliable [8]. Before purchasing, research a system that is > 90% precise for the intended outcome of interest.

Accuracy: How well the PLF system measures a true value is known as accuracy [9]. This is not required for screening tools, but it is required if a system intends to be used as a diagnostic tool. For example, it is not necessary for a health alert system that screens transition cows to correctly measure how many times a cow ruminates in an hour, but it must measure when the cow deviates from her normal pattern (precision). Similarly, to capture heifers in estrus we need to know that she increased her step count, but we do not need to know how many steps she took. However, if we care about the true value, such as using a rumen bolus to identify a cow with a fever, then accuracy becomes extremely important. Check publications for 80% accuracy if using the system as a diagnostic tool.

Infrastructure and Connectivity: Adequate infrastructure for device communication is important when integrating precision technology on a farm. Producers must check with their internet provider to ensure that they have enough streaming capability before making a purchase. Most sensors also have limited memory before they need to be "read" to a base station. Plan for the "hub" or base station to go where cow traffic is common, such as en route to the milking parlor [3]. Some companies sell solar panels and other remote-based options for pastured cattle [4]. Prioritize the location of the base station for communication between precision technologies (i.e. sensors) and a cloud storage interface to maximize the efficiency of the software.

Designated personnel: Someone on the farm must perform routine maintenance such as checking the internet connection, troubleshooting if the internet fails, tracking battery life on sensors, and replacing parts [5]. It is also fundamental that a person is assigned the task of routinely managing cows who enter and leave the herd. For example, most PLF systems require an acclimation period of at least 8 days of behavioral information on a cow before reliable predictions are made. If a manager is using a system to monitor transition cow health, the protocol be written to enroll cows with collars 14 days before their scheduled calving date. This ensures that the system knows the cow's behavioral baseline before she is sick.

Training: Personnel must receive proper training from the technology company on how to effectively use and interpret the PLF data. This enables that person with proper knowledge on how to best adjust the settings of the PLF system's alerts to meet that farm's goal. However, there is often a learning curve at first. It is key to request training and education for farm staff from the technology company before purchasing to best understand how to overcome challenges.

Cost: The integration of PLF technology has great value, however, there is a cost [6]. Some precision technologies, such as automated feeders, voluntary milking systems, and electronic wearable systems, can be a large upfront investment. The rule of thumb is that the system must last long enough to pay back labor savings and offer a return on investment. Many PLF companies offer leasing options if you are on the fence and not sure if you want to invest in the technology permanently. This can be an advantageous option if a farm wants to purchase the system for some of the herd (i.e., only heifers to get them bred on time, or only transition cattle). Remember that not all PLF systems are expensive. For example, farms can install automation sensors that regulate curtains and fans based on weather conditions, which has great payback by reducing the heat stress of the herd [7]. Ongoing expenses for maintenance, upgrades, and training also contribute to the overall cost. The biggest takeaway is to ensure that the farm has an actionable plan to use the data to improve herd performance to have the system pay for itself.

Sensitivity (skilled labor) vs. specificity (save labor): Finally, select a PLF system based on the farm’s primary purpose of interest. It is important to select a system for either sensitivity, or screening to find the animals of interest, or specificity, eliminating animals from the checklist [10].  Sensitivity refers to how well the system classifies cattle as true positives [11], with the end goal of using the PLF system to screen for cattle. Specificity refers to how well the PLF system classifies cattle as true negatives [11] with the end goal of using the PLF system to remove animals that require an exam.

Conclusion

In summary, precision livestock farming has the potential to revolutionize the dairy industry, offering innovative solutions to many challenges. However, producers should make sure the technology is validated for the farm’s goals. If a producer is using the sensor to screen for sick cows or catch estrus, aim for systems with high sensitivity, and high precision. If a producer is using the sensor to remove healthy animals from a standard exam, aim for systems with high specificity, and high precision. Accuracy is only required if the system is diagnostic. Ensure the system has a productive life that offers a return on investment. Precision livestock farming may be a strategic investment for the future and an option for better farm management.

References

[1] Kleen, J. L., & Guatteo, R. 2023. Precision livestock farming: What does it contain and what are the perspectives? Animals. 13: 779. doi.org/10.3390/ani13050779.

[2] Bahlo, C., Dahlhaus, P., Thompson, H., & Trotter, M. 2019. The role of interoperable data standards in precision livestock farming in extensive livestock systems: A Review. Comput. Electron. Agric. 156:459–466. doi.org/10.1016/j.compag.2018.12.007.

[3] Symeonaki, E., Arvanitis, K. G., Loukatos, D., & Piromalis, D. 2021. Enabling IOT wireless technologies in sustainable livestock farming toward agriculture 4.0. IoT-Based Intel. Model. Env. Eco. Eng. 67:213–232. doi.org/10.1007/978-3-030-71172-6_9.

[4] Aquilani, C., Confessore, A., Bozzi, R., Sirtori, F., and Pugliese, C. 2022. Review: Precision livestock farming technologies in pasture-based livestock systems. Animal. 16. doi.org/10.1016/j.animal.2021.100429.

[5] Munz, J., and Schuele, H. 2022. Influencing the success of precision farming technology adoption—a model-based investigation of economic success factors in small-scale agriculture. Agriculture. 12:1773. doi.org/10.3390/agriculture12111773.

[6] Lovarelli, D., Bacenetti, J., and Guarino, M. 2020. A review on dairy cattle farming: Is precision livestock farming the compromise for an environmental, economic, and social sustainable production? J. Cleaner Prod. 262. doi.org/10.1016/j.jclepro.2020.121409.

[7] Chen, C.-S., and Chen, W.-C. 2019. Research and development of automatic monitoring system for Livestock Farms. Appl. Sci. 9:1132. doi.org/10.3390/app9061132.

[8] Michaud, E. J., Liu, Z., and Tegmark, M. 2023. Precision machine learning. Entropy. 25:175. doi.org/10.3390/e25010175.

[9] Rojo-Gimeno, C., van der Voort, M., Niemi, J. K., Lauwers, L., Kristensen, A. R., and Wauters, E. 2019. Assessment of the value of information of Precision Livestock Farming: A conceptual framework. NJAS: Wageningen J. Life Sci. 90–91. doi.org/10.1016/j.njas.2019.100311.

[10] Bausewein, M., R. Mansfeld, M.G. Doherr, J. Harms, and U.S. Sorge. 2022. Sensitivity and Specificity for the Detection of Clinical Mastitis by Automatic Milking Systems in Bavarian Dairy Herds. Animals. 12:2131. doi.org/10.3390/ani12162131.

[11] Fernandes, A. F., Dórea, J. R., and Rosa, G. J. 2020. Image Analysis and Computer Vision Applications in Animal Sciences: An overview. Front. Vet. Sci. 7. doi.org/10.3389/fvets.2020.551269.