How robot ranch hands could protect livestock from predators
In an effort to keep livestock safe from coyotes while not harming the predators themselves, researchers carried out a study involving a remote-controlled vehicle equipped with powerful lights. The solution worked, hinting at a future in which robot ranch hands could work night patrol.
While predation causes a relatively low number of livestock deaths each year, its effect is still significant. According to the US Department of Agriculture, in 2022 predation was responsible for the deaths of over 9,000 adult cows and calves, and over 6,000 adult sheep and lambs. Most of those deaths are caused by coyotes, so seeking to deter them while not causing even more loss of life is an important area of research.
In the current study, US researchers used a captive colony of coyotes at the USDA National Wildlife Research Center (NWRC) Utah Field Station to test out a few different automated solutions to deter them from predation. In all cases, they put commercially prepared food as bait in the middle of the animals' enclosures.
In one scenario, they mounted commercially available motion-activated deterrent lights known as a Foxlights in stationary positions around the bait. In the second scenario, they mounted these lights on a small remote-controlled vehicle that made regular forays around the bait every three minutes.
In a final trial, the researchers incorporated adaptive movement by actively steering the remote controlled vehicle at the coyotes when they approached the food. If the coyotes retreated, the vehicle returned to its base. If it didn't, they steered the vehicle at the animals until they moved outside of the bait zone.
As you might imagine, the team saw incremental improvements with each scenario. In the light-only test, of 39 meals made available, only five went uneaten. The test in which the vehicles circulated on a predetermined pattern at regular intervals led to 12 out of 21 meals remaining untouched. The clear winner though, was the adaptive movement test. When the cars were steered toward the animals, 15 out of 18 of the meals stayed uneaten.
"Our observations indicated that the movement of the deterrent vehicle toward a coyote seemed to keep coyotes 'on edge' or off-balance," write the researchers. "The profound impact of adaptive movement noticed in this experiment helps justify further development of this type of vehicle. Critical for advancing this idea is the creation of a mobile deterrent that can both identify and react to animals."
The researchers acknowledge that building autonomous weatherproof vehicles that are capable of machine learning will present a technological and economic challenge. Instead, they suggest that simpler vehicles that could roam livestock grounds in much the way robot vacuums zip around rooms should be a worthwhile field of development, as such a system effectively doubled the efficacy of the stationary system tested.
The study involved scientists from the NWRC, Colorado State University, The Society for the Preservation of Endangered Carnivores and their International Ecological Study, the University of Washington, and Krebs Livestock. It is described in a paper that was published in the journal, PeerJ.