From Moon Rover, which encourages predators to hunt and feed with packs, to applying cutting-edge algorithms to try and understand the facial expressions of Sumatran orangutans, artificial intelligence and robotics are giving zoos and researchers new insights into animal welfare.
Developed by researchers at the University of Adelaide and with support from researchers at Zoos in South Australia, these high-tech approaches have been tested at Adelaide Zoo and Monato Safari Park, where keepers provide unprecedented access to animal behaviors that support animal welfare in real time.
For African painted dogs at Monarto Safari Park, this unique approach allows them to engage in a new kind of naturalistic enrichment that reflects the thrill of pursuit.
“We use semi-autonomous rovers, a highly remote controlled vehicle, to carry food corpses through painted dog habitats and simulate the movement of live prey. This new approach encourages dogs to interact, communicate and move together like they are in the wild.”
“This is a great example of how ongoing development of AI, remote sensing and robotics can continue to increase opportunities for humans to understand, preserve and protect wildlife.”
Not only does rover stimulate instinctive behavior, but also strengthens the essential pack bonding, which is an important part of the species' social structure, usually in groups of up to six, as African painted dogs rely on cooperative hunting for bonds.
“When births are delivered this way, there is a significant increase in social communication and bonding,” said Rachel Heming, Carnivore supervisor at Monato Safari Park.
“The most important relationship for painted African dogs is their relationship with the pack. Encouraging them to move, hunt and feed together helps them stay cohesive and promote natural behavior.”
The Adelaide Zoo focuses on understanding the welfare state of Sumatran orangutans. Camera systems are installed in habitats, monitoring activity around the clock, and computer vision algorithms detect and track body movements to build and analyze “skeletal” models. This tool allows researchers to classify their behavior more accurately.
A facial expression tracing is also underway to gather more information about the orangutan emotional state. The collected data will be interpreted along with the welfare knowledge of the experts at the next stage of the project.
“Orangutans, like humans, show distinct facial features and micro-expressions that convey emotions such as curiosity, pain, and relaxation. Our AI model recognizes individuals with high accuracy in facial expressions to estimate emotional states, says Georgia Mason, a third-year computer science student who is part of the project.
Students in Electrical and Electronics, Computer Science, Mathematics, Veterinary Biology, and Animal Behavior collaborate with research teams to collect data, develop models, and field test potential solutions.
“These innovative engineering solutions have already advanced our ability to monitor and enrich the lives of zoos animals and show great expectations for the future,” said Mark Smith, Conservation Manager at Zoos SA.
Researchers hope to secure more funding to expand their research into other species and potentially develop tools like mobile apps that keepers can use to access live behavioral insights.
