5. Research into virtual fencing

Independent research on Halter

Internally, our team does continuous extensive research based on millions of days of cow behaviour in order to advance our system and improve animal welfare outcomes. We’re also conducting important independent research studies into Halter’s system with the Tasmanian Institute of Agriculture (TIA). These studies are led by Dr Megan Verdon, Research Fellow and an expert in animal behaviour with a focus in the development of livestock systems that achieve continual improvement in animal welfare and productivity. These studies will be the most comprehensive research conducted into pastoral dairy cows and virtual fencing. We have partnered with TIA because of their team’s invaluable experience from previous research into virtual fencing technology and animal welfare, and because their dairy research facility and technologies are world-class. The full results from this research will be published in 2024 after completing a peer-review process. In the meantime, we have released preliminary findings of the first research study. Click here to read a summary fact-sheet and read below for an overview of the findings.

Overview of the research study with the Tasmanian Institute of Technology (TIA)

TIA assessed the effectiveness of Halter’s technology to manage lactating dairy cows. This represents the longest study of virtual fencing on lactating dairy cows, and the first to study the application of Halter technology. Eighty mid-lactation dairy cows were split into two groups and managed with the Halter system for a 10-day training and 4-week management period. In tandem, an additional eighty mid-lactation dairy cows were split into two groups as control groups and were managed with conventional methods; cows were contained with electric fences and were shifted with farm workers. Cue data was monitored across two scenarios: cows being held in a zone, and cows transitioning to the dairy shed.

Training period and management period

TIA analysed extensive data from a 10-day training period and a 4-week management period following training. In total, this was equivalent to 2,950 days of collars on cows. The objective of the below cue data is to show the cows' level of understanding of the primary and secondary cues and the frequency of them receiving the secondary cue (pulse).

  • During the 10-day training period, a typical cow received less than three low-energy pulses per day.
  • During the management period, most cows received less than three pulses a week, and that number continued to decline each week.
  • Once trained, the guidance cues that a typical cow receives each day are almost entirely sound and vibration.

Cows are intelligent and typically learn to stay within a virtual boundary within a matter of hours of beginning training. While cows generally show significant progress in their understanding within the first 24 hours, Halter prescribes a full training programme of seven days, with specific daily monitoring and training modules assigned to farm staff on each of these days.

Cue ratios (pulse-to-sound ratio)

A low ratio of pulse-to-sound cues demonstrates that a cow is consistently responding to the sound cues by changing her direction to avoid a pulse. By the fourth week of TIA’s management period, the pulse-to-sound ratio was two pulses per 100 sound cues. This shows that Halter-trained cows had a strong understanding of the system’s primary cues. Halter monitors cue data and analyses scenarios where cows receive a high number of pulses. Halter is continually improving the ratio of pulse-to-sound cues through improvements in cow training, farmer training, and product upgrades.

For various reasons, a small minority of cows receive more pulses. For example, in the TIA study, there was a small subset of cows (5% of the herd) that received over seven pulses per week, which was over triple the typical number of pulses. Our observations are that the most common reason is that these are the more confident cows choosing to ignore the primary cues and cross the virtual fences. Verdon et al. (2020) have demonstrated that young cattle that were assessed as being less fearful, using a startle test, tested a virtual fence more frequently and received more pulses than cows assessed as being more fearful. There is also an evolutionary explanation; herd animals like cattle that have evolved from prey species generally contain individuals who explore the boundaries, looking for new resources for the group. This is supported by external evidence that the animals who tend to push boundaries are the more dominant animals (Campbell et al., 2019). There is also some evidence that animals receiving more than the average number of pulses are higher producers (AWC, 2022).

Further research study with the Tasmanian Institute of Agriculture focused on Halter’s technology

TIA will lead a longitudinal (multi-year) study that quantifies the benefits of Halter for pasture production, labour requirements and animal performance. This study has been established to build an in-depth and long-term understanding of the impacts of Halter’s technology. The study will be undertaken in partnership between TIA, Halter and Tasmanian farmers, with support from the Tasmanian Government’s Agricultural Development and Agricultural Innovation Funds. This study is due to commence in 2024.

General research on animal welfare and virtual fencing technology

Halter implements best practice from the scientific literature on animal welfare and virtual fencing technology, specifically from:

  • Published academic research on virtual fencing technology, for example, by Dr Megan Verdon, Research Fellow at the Tasmanian Institute of Agriculture (TIA), and by Dr Caroline Lee, who leads the Animal Behaviour and Welfare Team at CSIRO (Commonwealth Scientific and Industrial Research Organisation) - both leading researchers in the fields of animal welfare and virtual fencing technology.
  • The Five Domains Model - developed by Professor David Mellor and Dr Cam Reid. This is the leading standard for categorising the assessment of animal welfare into the domains of nutrition, physical environment, health, behavioural interactions, and mental domain.
  • Published opinions of regulators based on detailed literature reviews:
    • The New Zealand Animal Welfare Advisory Committee’s Code of Welfare Evaluation Report Dairy Cattle, Ministry of Primary Industries (2022).
    • The UK’s Animal Welfare Committee opinion on the welfare implications of using virtual fencing systems to contain, move and monitor livestock (2022).
    • The Opinion of the Panel on Animal Health and Welfare of the Norwegian Scientific Committee for Food and Environment: A review on the use of electric devices to modify animal behaviour and the impact on animal welfare. Mejdell CM, Basic D and Bøe KE. (2017).