University of Illinois Extension

Illini DairyNet Papers

Electronic Detection of Estrus in Dairy Cattle
Eeman E. At-Taras and Sidney L. Spahr


  • Estrous detection was improved by both electronic activity tags (Heat SeekerTM) and heatmount sensors (HeatWatchTM).
  • A new algorithm based on mean activity plus two standard deviations of the cow's two hour activity counts resulted in earlier detection of estrus than did the activity ratio algorithm.
  • Hot weather decreased the total duration of standing mount activity significantly.


Sound reproductive management is the foundation of a successful dairy operation, and detection of estrus is essential in an effective dairy reproductive management program using artificial insemination. It is estimated that the dairy industry loses $300 million as a result of erroneous diagnosis of and failure to detect estrus. Significant decreases in milk production, delays in days to first service, prolonged days open, and extended calving intervals are all contributing factors to this loss. Traditionally, estrus detection was accomplished through visual observation of estrous-related signs. This method is both tedious and time-consuming. Hence, alternate and more efficient methods of estrous detection are needed.

Recently several sensors have become available as aids for detection of estrus. We compared the operational characteristics of two of the systems, HeatWatchTM and Heat SeekerTM, and tested a potential new algorithm for the Heat SeekerTM system. We also used the sensors to determine the range and variation for length of estrus under specific conditions.

Description of the sensor systems:

The HeatWatchTM system consists of a pressure sensitive transmitter (contained in a pouch and attached to the rump of the cow), a receiver (which receives the signal from the transmitter), a buffer (stores mounting data until retrieved), and HeatWatchTM software on a personal computer. Mounts are recorded as they occur and are stored in the buffer until the HeatWatchTM program is accessed on the computer. The Boumatic Heat SeekerTM system consists of a battery powered activity tag, a reader, and data acquisition software. The tag is placed on the leg of a cow and a strap is fitted around the leg to secure the tag. The tag contains a mercury switch and a memory file that records leg movements by two-hour intervals. In addition, there is also a built-in algorithm that determines an activity ratio for each two hour period by comparing the activity during the last 12 hours with activity from the same 12 hour period in the previous two days. We postulated that this algorithm could be improved by incorporating standard statistical approaches, comparing recent activity with the mean plus two standard deviations of the cows 2 hour activity during non-estrus. Using a retrospective analysis to test this hypothesis, we compared activity around estrus with activity counts two days before and two days after the day of estrus to test this hypothesis. Hence, both the increased activity ratio (> 2.0) and increased activity count were used as indicators of increased activity associated with estrus.


Sixty Holstein, Jersey, Brown Swiss, and Ayrshire cows housed in free stalls and a concrete lot were fitted with a Boumatic electronic activity tag (HeatSeekerTM) on the right foreleg and a HeatWatchTM heatmount detector on the rump. Physical and mounting activity data were collected for 120 estrous cycles (82 single estrous cycles and 38 concurrent cycles). Twenty-three of the cows were monitored during spontaneous estrus; these cows were kept on the trial until diagnosed pregnant (at 35 to 45 days post-breeding) or until 150 days on the trial. The remaining 37 cows were given 5 ml of LutylaseTM (Upjohn, Kalamazoo, MI) in groups of three or more upon confirmation of a corpora lutea via rectal palpation; these cows were removed from the trial upon breeding. All cows were observed daily for visual signs of estrous. Data were analyzed using a statistical model that included parity (primiparous vs. multiparous), group (single or multiple cows in estrous), and weather (hot or cool). Summer was considered as hot while the other seasons were considered cool.


The efficiencies of estrous detection for the HeatWatchTM, increased activity ratio, and increased activity threshold count were 80.2%, 82.3%, and 86.9% respectively; visual observation only detected 54.5% of all predicted estrous periods. Parity and single or multiple cows in estrus had no effect on mounting activity. However, weather had a substantial impact on standing mount activity; average standing mount duration was decreased from 6.76 hours under cool conditions to 2.97 hours under hot conditions (P # 0.05).

Figure 1 shows a typical graph indicating start of increased activity count and increased activity ratio. The activity count threshold algorithm detected estrus earlier (2.61 hours) than did the activity ratio algorithm when considering 33 of the 35 estrous periods (P # 0.02). The remaining two estrous periods were characterized by increased activity both before and after peak mounting, but contained periods of low activity that coincided with recorded mounts; these cows were omitted from the analysis. The duration of time above the increased activity ratio threshold (2.0) exceeded the time above the increased activity count threshold (24.0 hr vs. 9.7 hr) in the 33 estrous periods that were analyzed; in spite of this, increased activity count above the threshold overlapped with mounting activity for a significantly longer period of time than did increased activity ratio (6.13 hr vs. 4.41 hr). This supports our hypothesis that the activity count threshold algorithm may be a better indicator of estrus since it corresponds more closely with standing mount behavior and detects estrus earlier than does the activity ratio algorithm.


In summary, both sensor systems used in this study improved detection of estrus relative to visual detection. The HeatWatchTM system was particularly attractive because it recorded actual standing mounting behavior. However, keeping the sensor transmitters attached to the cows required considerable maintenance; a more secure method of attaching transmitters to the cows would improve efficiency of the system immensely.

The Heat SeekerTM system required little maintenance. However, its data acquisition system (collection of the summary of activity ratios for the past 12 hours as the cow passes through an antenna at milking time) may result in an undesirable time lag relative to the instantaneous transmission of data in the HeatWatchTM system. Our results suggested that efficiency of estrous detection by the Heat SeekerTM system could be improved by collecting the activity counts at periodic intervals throughout the day, and using an algorithm to detect the presence of estrus that was based on standard statistical procedures. This procedure would require telemetric collection of activity counts between milkings, and a personal computer to perform the test to detect increased activity. Collection of activity data between milkings would require changing the method for transmission of data from the current Heat SeekerTM system that requires the cow to enter an electromagnetic field around an antenna to one that could send its data upon command to a more distant antenna, similar to the method used with the HeatWatchTM system.

The figure may be viewed in the PDF

Attached File: