6. Virtual fencing background
Research on electric pulses
There is no evidence that receiving pulses distresses any cows, as long as pulses are predictable and controllable. Kearton et al. (2020) showed that sheep wearing collars similar to those that Halter uses on cows did not show signs of distress (behavioural changes or elevated plasma cortisol levels) if they had been trained with warning cues before receiving pulses. This is consistent with the classical studies conducted on rats by Weiss (1970) that showed predictability and controllability both reduced the stress caused by electrical shocks. Rollin (2006) has pointed out that the perception of discomfort varies widely between individual animals depending on how well they are managed to cope with it. Halter continues to work towards minimising the number of pulses received in set time windows by any cow and sets variable maxima of pulses based on the response of each cow.
Most researchers report a wide variability in the number of pulses required to train animals to a virtual fence, with some cows experiencing higher numbers of both primary sound cues and electrical pulses before they became fully trained. This coincides with Halter’s experience. Grumett & Butterworth (2022) described this phenomenon as slow learning. In fact, the literature on bovine learning indicates that cows generally learn new tasks rapidly (Marino & Allen, 2017). There is no evidence that it is a result of impaired cognition and it could equally be the result of greater stoicism or less fearful behaviour. There is no evidence of ‘slow learning’ occurring on Halter farms.
Benefits of virtual fencing
Virtual fencing allows farmers to set and adjust boundaries in real-time very easily. This advancement in herd and pasture management enables farmers to shift mobs virtually via the app as required. Cows can be moved to ensure nutritional intake is optimised and can be moved to safety during a weather event e.g. heavy rainfall or flooding. Virtual fences free farmers from the labour intensive management and maintenance of physical fences allowing them to commit more time on cow nutrition, safety and welfare. Halter increases herd productivity while reducing the need for electric fences, dogs and motorbikes for mustering cows, potentially leading to less stress and fewer injuries on farm.
GPS accuracy
Every Halter collar has a GPS tracker to continuously track the location of each cow. The guidance cues are informed by the location and behaviour of each cow. “GPS drift” is a minor inaccuracy between the location detected on a device and the device’s true location. It is common to all GPS-enabled technologies. For Halter, this inaccuracy occurs infrequently and is caused by external factors like the number and position of the satellites, as well as occasional interference from buildings, trees or atmospheric conditions.
Currently, the level of accuracy with Halter’s collars is approximately 1.5 metres. This can result in the infrequent occurrence of either overallocation or underallocation of feed using virtual fencing by this margin. To mitigate this, Halter’s collars compare GPS readings with what it expects to see from internal accelerometers, to discard obviously incorrect GPS data. Halter also recommends to farmers to not ‘feed out’ along a virtual boundary, and to not graze mobs in long, thin breaks.
Nationally, there is continual investment and advancement of GPS accuracy. In 2023, the New Zealand government announced funding to invest in an innovative satellite-based augmentation system (SBAS) that will improve the accuracy of GPS for a range of enabling technologies. This investment will improve the accuracy to less than 1 metre.
