Animal-Computer Interaction
Animal-Computer Interaction (ACI) at the University of Glasgow investigates the intersection between computer and animal science. We examine animals and computing from two perspectives: first, how animals engage with computer systems by developing novel interactive technologies that give them agency over their environments and digital lives; and second, how computers using machine learning and artificial intelligence can detect animals' emotional states and pain.
Our work spans the full pipeline, from designing and building interactive systems that animals can control to developing AI tools that help us understand what animals are experiencing. We are committed to open, ethical, and animal-centred research, publishing open-access software, data, and systems to support reproducibility and collaboration across the global research community. Together, the two sides cover both the design and interaction aspects of animal technology and the AI-driven behavioural understanding, making Glasgow a unique hub for animal-computing research.
Project Highlights
Parrot Video Calling System
The Parrot Video Calling System (Royal Society of Edinburgh Fellowship, 2021-2022) produced an agency-based video-calling system enabling parrots to connect with other parrots, exploring whether animals can meaningfully engage in internet-based social communication. The project examines animal autonomy and the design of species-specific interfaces, with findings published at ACM CHI.
DogPhone
A device allowing dogs to initiate video calls with their owners, putting the dog in control of when social contact happens rather than the human. The project challenges conventional assumptions about which species "uses" technology, demonstrating that dogs can meaningfully and intentionally trigger digital communication.
Automated Animal Pain & Behaviour Detection
Using computer vision and multimodal machine learning to automatically detect pain, emotion, and behaviour in animals — including sheep facial expression analysis for pain estimation. This work has direct applications in farm animal welfare and veterinary care, where timely and accurate pain detection can significantly improve animal outcomes.