Putting sensor technology to the acid test

Validation of two different accelerometers for assessing the lying behavior of dairy cows on pasture

Author: Barbara Pichlbauer

Like many other aspects of modern life, digitalization and state-of-the-art technical equipment have found their way into livestock agriculture. Huge farms could not face daily business without support from modern digital technology. At the same time, it is essential to continue working according to the “Good Agricultural Practice” and to meet steadily increasing quality standards (both legal and social), even if conditions and structures in agriculture are constantly changing.
Animals’ health, welfare, and proper husbandry no longer only concern farmers but also the public. Consumers are concerned with these topics and want to know where their food is coming from. In organic dairy farming, pasture access is mandatory. However, various quality programs also demand a certain degree of pasturing from their members since it is seen as one of the husbandry’s most natural ways.
Some farmers already implemented digital sensor technology into herd management for continuously monitoring their animals. Facing the rising demand for pasturing, those farmers need sensor systems that fit to be used inside and outdoors. To meet these requirements, they have to be robust and weather-proof, as well as reliable and accurate.
Already there are systems available that cover both bases, while many are only suitable for indoor use.
The project “Using digital sensors to monitor dairy cows on pasture” evaluates if and how sensor-based monitoring systems can be used on pasture and if differences between indoor housing and cows kept on pasture can be displayed.

Our goal for 2020’s pasture season was to test two different accelerometers that have already been proven to be reliable when applied in the stable. We used the following two sensors, focussing on continuously collecting data on the cows’ lying behavior:

  • A 3-D-accelerometer that is built into an ear-tag (SMARTBOW ear-tag, Smartbow corp., Zoetis, Parsippany, NJ)
  • A data logger with an accelerometer and a gyroscope that is fixed on the leg of an animal (HOBO Pendant G Logger, Onset Computer Corporation, Bourne, MA)

Figure 1: SMARTBOW ear-tag

Figure 2: HOBO data logger

When working with new sensors (or known ones in a new environment), the first step is to validate the systems to ensure a continuous and reliable stream of data. It is essential to make sure the conditions while evaluating are as similar as possible to the working conditions.
The validation of sensor technologies for monitoring animals’ behavior, direct or indirect animal observation, serves as a gold-standard. Direct animal observation is carried out by a trained person, who classifies the behavior immediately and keeps accurate records. When working with indirect animal observation, at first, animals’ behavior has to be video recorded. Then, the video footage can be classified and taken as a reference. The advantage here is that once the video footage is available, one can go through it as often as necessary or even in slow motion if needed. For indirect animal observation on pasture, drone-mounted cameras are a promising approach compared to fixed cameras, mainly because of their flexibility.
Once the new technology has proven itself in the validation process, the sensor data can be used for further analysis.
The reason for the lying behavior of dairy cows to be interesting is that it is a high priority behavior and can be influenced by many endogens and exogen factors. Deviations of lying time from the animal’s individual “standard” can be a sign of disease or compromised animal welfare. This is why this study aims to describe the lying times of dairy cows under pasture conditions.

There are also a few more sensor systems that are going to be used in the project “Using digital sensors to monitor dairy cows on pasture”:

  • Drone with an integrated camera
  • RumiWatch-halter
  • RumiWatch-pedometer
  • Polar heart rate belts

Figure 3: Drone with an integrated camera (producer: DJI)

Figure 4: On the left: The pressure sensor is integrated into the noseband, and the accelerometer is situated at the side. Ruminating, feeding in the barn and on pasture, and the water intake can be detected with the RumiWatch-halter. On the right: The pedometer can classify an animal’s behavior into “standing” and “lying” and detect the steps.

Figure 5: Polar heart rate belts for measuring the heart rate of the animals

Citation: B. Pichlbauer, „Sensorsysteme auf Herz und Nieren prüfen: Validierung zweier verschiedener Beschleunigungssensoren zur Beurteilung des Liegeverhaltens von Milchkühen auf der Weide“. In: DiLaAg Innovationsplattform [Webblog]. Online-Publication:, 2020