Exploring the potential for AI to delve into animal cognition

In a groundbreaking development, researchers at Dublin City University's Insight Centre for Data Analytics, funded by Research Ireland, are working on a project that aims to train an AI model to recognise instinctive dog behaviours such as pawing, nudging, or barking. This is just one aspect of a broader movement in the scientific community, where artificial intelligence is being harnessed to study the emotions of various animal species.

A significant stride has been made by Milan-based researcher Stavros Ntalampiras, who has developed an AI model capable of detecting emotional tones in the calls of seven species of hoofed animals. The model has shown that high-pitched calls are especially informative in pigs, while in sheep and horses, the mid-range carries more weight.

Similar work is underway with whales, where the New York-based research organization Project Ceti is analysing patterned click sequences called codas. Long believed to encode social meaning, these are now being mapped at scale using machine learning, revealing patterns that may correspond to each whale's identity, affiliation, or emotional state.

One study has shown that subtle shifts in canine facial muscles correspond to fear or excitement, while another found that tail-wag direction varies depending on whether a dog encounters a familiar friend or a potential threat. In dogs, researchers are linking facial expressions, vocalisations, and tail-wagging patterns with emotional states.

The deep-learning model can recognise shared features of their calls, such as pitch, frequency range, and tonal quality. This discovery of emotional traits across species is a significant leap forward in a field that is being transformed by AI.

Farmers could receive earlier warnings of livestock stress, conservationists might monitor the emotional health of wild populations remotely, and zookeepers could respond more quickly to subtle welfare changes. At Dublin City University, a detection collar is being developed for assistance dogs trained to recognise the onset of a seizure in people who suffer from epilepsy.

However, it's important to remember that AI models are most reliable when interpreted in context, alongside the knowledge of someone experienced with the species. The real test is not how well we listen, but what we're prepared to do with what we hear, to ensure that AI serves animal welfare rather than simply satisfying human curiosity.

This work is part of a growing movement in institutions like the HTWK Leipzig and TU Bergakademie Freiberg, where scientists are using AI to improve animal welfare and behaviour understanding in livestock such as dairy cattle. As we continue to unravel the complexities of animal emotions, we open up new possibilities for animal conservation, welfare, and our relationship with the natural world.