Virtual fencing: progress slow but sure

Virtual fencing, or to give it its correct technical term, Directional Virtual Fencing, is a concept that has been around for over a decade, and significant advances have been made on it, both in Australia and the US. But is electronic ‘fencing’ going to replace barbed wire, posts and strainers on your property any time soon? The jury is out on that one.

Dr Dean Anderson is the acknowledged expert who has been developing the concept through his work with the US Department of Agriculture research section, at the Jornada Experimental Range in Las Cruces, New Mexico. Anderson spent five months in Queensland in 2005 with CSIRO researchers, demonstrating the effectiveness of DVF.

According to Anderson, current, fixed fencing to control livestock movement has limitations. “Facilitating proper animal distribution is second only in importance to proper stocking in range management,” he says. This is a global challenge for many of the world’s 4 billion domestic animals that forage on some of the world’s 13 billion hectares.

Currently, conventional fencing controls most free- ranging animals in the developed world. Although it is essential where health and safety of humans or animals are concerned, conventional fencing’ limitations include its cost to build and maintain, the fact that it is spatially fixed, thus limiting dynamic flexible management, some people consider conventional fencing aesthetically undesirable, and it may have a negative impact on wildlife movement.

“Directional Virtual Fencing (DVF) defines a novel, new methodology for the control of animals,” Anderson says. It involves controlling the animal’s location and direction of movement through the application of audio and or electric shock cues applied to either the animal’s right or left side to initiate directional movement. DVFTM combines physics, electronics, biology and ecology to determine the animal’s location on the landscape using Global Positioning System (GPS) technology, and then using
Geographic Information System (GIS) data to direct the animal to the proper location, when necessary.

DVF uses animal behavior to change the animal’s location by conditioning the animal using programmable cues that are applied autonomously in a manner that is consistent with low-stress animal handling procedures. The animals can be located in a static or moving ‘virtual paddock’ (VP) which is defined by the electronic GIS sensors embedded in it to create invisible or virtual fences.

However, Anderson stresses, that where it is important for the health or safety of either the animal or the humans around it to have “a non-leaky fence”, DVF is not the answer.

CSIRO’s Livestock Industries Research Scientist Dr David Swain, Dr Anderson and colleagues published a paper in 2006 on testing DVF in Australian conditions. The paper resulted from the testing done at the JM Rendel research centre in Rockhampton during Anderson’s stay in Australia. Virtual Fencing Applications: implementing and testing an automated cattle control system concluded that “virtual fencing has potential for controlling cattle in extensive grazing systems.”

Twenty-five Belmont Red steers with a mean live weight of 270 kg were each randomly assigned to one of five treatments, consisting of a combination of cues (audio, tactile and visual stimuli) and consequence (electrical stimulation). The treatments were electrical stimulation alone, audio plus electrical stimulation, vibration plus electrical stimulation, light plus electrical stimulation and electrified electric fence (6 kV) plus electrical stimulation. Cue stimuli were administered for three seconds followed immediately by electrical stimulation (consequence) of 1 kV for one second.

The cattle were fitted with a collar-halter device to carry the electronics, batteries and equipment providing the stimuli (including audio, vibration, light and electrical) as a prototype virtual fencing device. Cattle were allowed to travel along a 40 metre alley to a group of peers and feed while their rate of travel and response to the stimuli were recorded. The animals’ rate of travel along the alley demonstrated the large variability in behavioural response associated with tactile, visual and audible cues.

Swain’s report said “The prototype virtual fencing system was successful in modifying the behaviour of the cattle. The experiment demonstrated virtual fencing has potential for controlling cattle in extensive grazing systems.”

CSIRO has already done considerable research on wireless sensor networks (WSNs) and their potential for increasing agricultural efficiency by providing access to data such as soil moisture and dam levels. CSIRO is looking at the potential for monitoring cattle using information from WSNs, and the possibility of combining that information with the NLIS tags.

Remote sensing is already in use in some large pastoral properties in the Northern Territory and Queensland, mainly for monitoring water supplies at bore pumps and troughs, to check water levels, contamination and flow rates. However, WSN has the potential to extend to monitoring livestock, and could assist in the development of the virtual fence.

The Queensland CSIRO team was not the only research group in Australia working on virtual fencing. In Western Australia the, Department of Agriculture & Food led the conceptual development of this technology in Australia in the late 1990s. However, Dr Robert Rouda, Senior Research Officer said DAFWA never did any field trials and is “no longer involved in the development of Virtual Fencing.”

The Kondinin Farm Group has also been interested in the concept, as has the Meat & Livestock Authority. The MLA has had discussions with groups of researchers in Australia and the US to develop the concept further, according to Wayne Hall, the MLA’s Northern Production Manager, but proposals “were either not well thought out or demanded exorbitant amounts of grant funding.”

However, a research project funded by the MLA allowed Kondinin research engineer Ben White and colleagues to travel round rural Queensland, Northern Territory and NSW looking at properties that have implemented remote sensing technologies and gauging the level of interest in DVF. White says there have been several successful trials which show the potential of the virtual fencing concept to control large herds of cattle, and there is certainly interest among producers in the potential of remote sensing to cut labour costs and improve efficiency. However, he points out that to date, the “fantastic concept [of virtual fencing] has promised a lot but delivered little.”

Proof of concept

DVF has been shown to work under experimental conditions. Why isn’t it readily available for the large scale grazier to implement? Anderson’s answer is simple: “Virtual fencing could be feasible immediately if there were a commercial company that would begin building and selling devices. These initial devices would probably not be perfect, but if Microsoft’s business model works with less than perfect products still under development, why could it not be possible in the virtual fencing arena?”

White says the equipment is too big and bulky and too easy for animals to damage. “The biggest hold-up is power sourcing. The present headgear is cumbersome and it is not very hardy. It’s subject to quite a lot of damage from the animals. Ultimately we’d like to see the control equipment in an ear tag, but that’s a long way off.”

White suggests that virtual fencing will take off when batteries are sufficiently miniaturised, and recharging can be achieved either by kinetic recharging (from the animals’ head movements) or by solar power. The ideal would be to combine the virtual fencing tag with the NLIS tag. He puts it at 5-10 years away.

Hall says remote sensing is useful, but he doesn’t think that proof of concept of virtual fencing has been clearly demonstrated. “There are a lot of animal behavioural and technical issue to be overcome before it is feasible.”

He cites the same technical issues as White, and adds that animal behaviour issues need to be clarified before virtual fencing can be proved to be effective.

– How will the animals respond to the stimuli?

– What percentage of them will respond?

– Will others follow if only a few are given the stimulus?

– Under what conditions will virtual fencing operate – for example, keeping bulls away from breeding cows?


How will the cost stack up against standard barbed wire and electrified fences? Anderson says this is difficult to guestimate today since virtual fencing is still only being used in a research venue. But he says the benefits will far outweigh any disparities in cost.

“Virtual fencing offers much more than a new gadget to conduct business in the same way you would using conventional fences,” he says. “The biggest advantage with virtual fencing will be an ecologically based control of free-ranging animals in real time or near real time. Being able to dynamically and flexibly manage stocking density and foraging pressure with herders is no longer possible in most of the developed world as we move into the 21st century. Yet virtual fencing will give us the opportunity to
replace ‘physical labor’ with ‘thinking labor’ while accomplishing the same
or more that we could using herders.

“I believe virtual fencing offers resource managers a more efficient way to implement optimum stewardship of our soil, plant and animal resources.”

Is virtual fencing on the horizon?

– Hall: Ten to twenty years for commercial production “to make it a cost effective application for producers. At the moment, producers can’t see the advantage of it”.

– White: When the miniaturisation is sorted out, in five to 10 years.

– Anderson: Immediately, “if there were a commercial company that would begin building and selling devices”.

– Swain: Too early to say, CSIRO is still doing developmental work.

Find out more

Dr Dean Anderson, USDA, La Cruces, New Mexico, USA; email: deanders@msu.edu;

Dr David Swain, CSIRO – Livestock Industries, Rockhampton; email:Dave.Swain@csiro.au;

Wayne Hall, MLA, Toowoomba; email:whall@mla.com.au;

Ben White, Kondinin Group, Toowoomba; email: ben@kondinin.com.au

« Back to writing examples