3. Animal Welfare Charter

Halter follows these standards to ensure animal welfare is protected when training, containing and guiding animals with virtual fencing technology.

Halter has trained more dairy cows with virtual fencing technology than any other wearable in the market. Over seven years we have developed and refined our technology and gained deep expertise in training and guiding animals while protecting their welfare. We take very seriously the high standards around responsible animal learning and training experiences. Below are Halter’s animal welfare standards:

The System

(1) The system should support the animal’s natural behaviour and should minimise disruption to their natural behaviour. The system should support animal agency by working with the animal, not against it.

(2) Trained animals should be confident with the system and trusting of the guidance cues. Animal learning must ensure that the aversive cue (pulse) becomes predictable and controllable for every animal (Kearton et al., 2020):

  • Predictable: the animal knows when the consistent development of primary cues will escalate to a secondary aversive cue (pulse); and
  • Controllable: the animal can choose to change its direction to avoid the aversive cue, or can ignore the primary cues knowing the consequence of doing so.

Cues must be precise and consistent to avoid confusing or stressing the animals and to minimise the welfare impact (Weiss, 1970; Lee et al., 2018).

(3) Whilst humans determine the destination of animal movements in the Halter app (i.e. a ‘pasture break’ or dairy shed),

  • Humans must never have direct control over the guidance cues sent to animals, and the only inputs to determine the cues should be the animal’s behaviour and location.

Safeguards

(4) If an animal fails to respond to the guidance cues, the system should automatically disable. The collar should not reactivate until the animal has demonstrated it is able to move freely.

(5) If an animal crosses a virtual boundary at high speed (for example, if spooked), the system should automatically deactivate.

(6) Halter must monitor cue data of animals and, if required, take action to protect animal welfare.

(7) Continuous monitoring should be in place to disable guidance cues for an entire mob if a subset of the mob fails to respond to the guidance cues (for example, if a mob is blocked from moving down a race by a fallen tree).

(8) The system should have in-built monitoring to detect key component failures and deactivate the system if required. The system must not attempt to guide animals if the collar is not operating according to its functional design.

(9) The system should alert users if they inadvertently set pasture allocations with too small of an angle (i.e. <45°) or stocking densities that are too high (m2/cow).

(10) If a communication network failure occurs, the system should put the farm into emergency mode and alert the farmer; the system should hold animals in their current zone for a set time period, allowing the farmer adequate time to respond. In this scenario, a farmer should have the ability to disable the system and take over manual control to shift the animals.

Aversive cues

(11) The system should use the minimum aversion necessary per individual animal. This should be automatically customised based on each animal’s tolerance to the aversive cue.

(12) Aversive cues must have time-bound lockouts to avoid the risk of animals experiencing aversive cues for continuous periods. These lockouts should be customised per individual animal, based on previous behaviour.

Developing system

It’s taken many years of research and development to develop the Halter system to reliably teach and guide animals. We are continuously refining safeguards and standards. At times there have been instances where we’ve seen unexpected issues, which has triggered rapid action and improvement.

No system for managing animals is perfect - whether modern technology or conventional methods. Even the most experienced farm staff using conventional methods for managing animals cannot effectively monitor all their cows 24/7. The level and speed of animal oversight that Halter provides allows for the early detection of issues impacting individual animals as well as impending emergencies affecting the whole herd. Halter provides distinct technology advantages that enable us to optimise for animal welfare outcomes including:

  • Continuous monitoring of the behaviour of all animals which enables deep insight about animal behaviour and health;
  • Vast data (42 million days) of animal behaviour, which enables deep research, learning, and innovation, and;
  • The system is updateable in the field, so as we unlock innovations and improvements we can then send them to all collars in the field (which update usually within 24 hours), as opposed to waiting years for the next generation of hardware product.

These characteristics allow us to research, learn, iterate and improve rapidly. We see endless potential to keep improving and delivering new value that protects and advances animal welfare.