Extension Logo
Extension Logo
University of Minnesota Extension

The future of precision dairy farming

Precision dairy farming involves the use of technologies to measure physiological, behavioral and production indicators on individual animals.

Two people standing in a long aisle in a modern dairy barn

Technologies include

  • Wearable sensors (neck, ear, legs or tail)

  • Rumen boluses

  • Subcutaneous implants

  • Inline or online milk sensors

  • Cow-side tests

  • Video analysis and facial recognition

  • Automated systems for feeding and milking cattle


Sensors that can measure various parameters in dairy cows have been developed since the 1970s. They initially included pedometers, milk conductivity and cow identification.

A study conducted at the Reproduction Laboratory in Beltsville, MD (published in the Journal of Dairy Science, 1977) developed an inexpensive, waterproof case to fit human pedometers on cows. They demonstrated consistent changes in activity when cows were in heat and concluded that activity monitoring could be used to help in heat detection.

Close up of cow lower leg with watch-like pedometer strapped to it
Cow pedometer, 1977

Technology has really improved since then.

The application of technology to biological processes has become increasingly more feasible.

  • Wireless data transmission is less expensive and more reliable.

  • Sensor and sensing (camera, microphone) technologies needed to develop precision dairy products are small and can withstand the harsh environment of a farm.

  • The cost of devices such as mobile phones has decreased.

  • Some mobile phone technologies (gyroscope, accelerometer) can be used for on-farm applications.

  • The use of cloud-based connectivity to integrate and network sensors for data collection and analysis is becoming commonplace.

It is important that sensors are validated by third-party research to establish whether the sensor measures the behavior it is supposed to measure (such as rumination, lying, standing, feeding, activity, steps). Validation studies are conducted around the world.

  • At the University of Minnesota, we evaluated the use of an ear-tag accelerometer in grazing dairy cows. The visual observations were highly to weakly correlated to the data collected by the sensors. We concluded that, in a grazing system, this particular technology would be a good indicator of eating behavior.

  • The University of Kentucky found that visually recorded feeding behaviors in cows housed in freestalls were highly correlated with an ear-tag accelerometer and a leg-mounted sensor.

    • Visually recorded rumination behaviors were strongly correlated with one brand of ear-tag accelerometer, but more weakly correlated with another ear-tag brand.

    • Visually recorded lying behaviors were strongly correlated with three brands of leg-mounted pedometers or accelerometers.

In addition, research needs to be conducted to investigate whether the sensor’s data can accurately detect the animal condition that it is supposed to detect, for example, heat or metabolic disease.

Studies have shown promise on using rumination, activity, feeding or standing behavior for early detection of transition cow disorders, such as ketosis, metritis or retained placenta.

In one of our studies, 296 cows were fitted with rumination sensors (neck collars) and housed in a conventional freestall barn. 

  • We found that cows diagnosed with metritis had reduced daily rumination time right after calving compared to cows not diagnosed with metritis.

  • In addition, retained placenta tended to reduce activity prior to calving and was associated with reduced activity after calving. 

  • Cows diagnosed with metritis and subclinical ketosis had reduced activity after calving.

On the farm, the data collected by the sensors should result in taking action with that animal to improve performance and health of the herd in an economically sustainable manner.

The integration of technologies and interpretation of data are critical to make improvements possible. More work is yet needed to improve integration of data.

Automated feeding and milking

Feeding calves

Computerized automated feeders can feed whole milk, milk replacer, or combinations of the two to individual calves in a controlled manner according to a predefined feeding plan. This technology provides individual animal data that can be used to more precisely manage the herd.

Automated milk feeders for raising calves in groups give producers labor flexibility. They also satisfy consumers who want animals to have a more natural life.

Feeding calves in groups allows calves to express some natural behaviors that cannot be expressed when housed individually, but offers some challenges in relation to maintaining good health. Good health and low mortality rate rely on appropriate management and maintenance of the feeding equipment and cleanliness of the calf environment along with daily visual observation of the calves by the caretaker.

Feeding and milking cows

Feeding cows in single box robotic milking systems requires adjustments on ration formulation to address the need to entice cows to the milking box. Factors related to cow health, barn design and herd management may affect attendance to the milking box and influence milk production.

Join the Extension Dairy Team at the second annual Precision Dairy Farming Conference in Rochester, Minnesota, June 18–20, 2019. 

Marcia Endres is an Extension diary scientist in the CFANS Department of Animal Science.

Page survey

© 2023 Regents of the University of Minnesota. All rights reserved. The University of Minnesota is an equal opportunity educator and employer.