Rethinking Virtual Fence

Winter 2023-24: An experiment was conducted in a Y-maze composed of two phases. Phase 1 compared audio, visual, and combination treatments where the exclusion zone was preassigned to Y arms randomly for all treatments. Phase 2 re-examined audio and visual treatments where visual cues were preassigned randomly and audio cues were triggered in the first arm the animal entered.

Phase 1-Visual cues and the combination of visual+audio cues were highly effective at reducing the number of times animals entered the exclusion zone. Both treatments demonstrated three successes for every one attempt, with very few failures. There was no difference in how cattle respond to visual cues alone or visual+audio cues. Phase 2- Animals with visual cues were four times more likely to avoid the exclusion zone than those receiving audio cues only. Animals in the audio treatment were six times more likely to fail, though almost half of their runs with classified as attempts.

This research was conducted as Hope de Avila's Master's thesis:  Exploring Cue Salience in the Cue-Consequence Relationship of Virtual Fence for Cattle - ProQuest (click link to view) 

Fall 2023: A hay bale was set in their paddock and surrounded with temporary white fence posts with flagging tape as a visual cue to delineate the exclusion zone. Active devices were put on the 4 cows' ears.

At the end of a week, there was no evidence of use within the exclusion zone. After a few initial interactions with the boundary, the cows had learned to avoid the area.

In less than 4 hours after the virtual boundary had been removed, the cows entered the exclusion zone and started eating the bale.

Spring 2021: This was a test conducted on 15 young bulls in test groups of three. These animals were brought into a test pen with devices attached to the ear. At the opposite end of the pen was an open gate with an exclusion zone of 4 meters in which animals would receive a sound-shock sequence upon entering the zone. This sequence would repeat until the animal moved out of the exclusion zone or 20 seconds of the sound-shock sequence was met. Each trial lasted 20 minutes.

This experiment showed that after four days of testing, 100% of animals were contained and did not cross into or over the exclusion zone that acted as a gate. Even after the second day, the number of animals that stayed in the pen almost doubled, exhibiting the learnability of this technology.

Spring 2020: We examined which types of cues (audio or visual) were most effective for animals to learn a virtual boundary.

Animals were released into a pen with an established boundary. The visual cue was a bright yellow rope laid on the ground, the audio cue was emitted from a device on the animal's halter, and there was a control that received neither of these cues. Animals received a shock if the animal ignored the boundary. The number of times the animals crossed the first line of the boundary and time spent in the exclusion zone were recorded.

Animals presented with a visual cue tended to cross into the exclusion zone fewer times, while animals with an audio cue spent less time overall in the exclusion zone. Animals without a cue had more line crosses and spent more time in the exclusion zone.

Winter 2020: We explored the potential effectiveness of an e-stimulus to the ear to deter an animal's movement toward a location.

In this experiment, Sportdog collars were wrapped in ear as close to the head as possible. The animals were released into a trial pen where a yellow rope, representing a boundary the animal should avoid. If the animal attempted to cross the rope, it would receive a shock.

The majority of those tested displayed progress in learning to associate the rope, a visual cue on the ground, with the shock, and decided against proceeding across. By the third trial, two of the animals stopped before the rope and turned around of their own accord without any electrical stimulus.