This new grouping is using mathematical and computational approaches to develop novel measures of animal welfare and address animal welfare and health issues.

Increasing computational power, the development of machine learning and AI algorithms, and the availability of ‘big data’ provide new opportunities for employing mathematical and computational approaches in animal welfare science. We are using these approaches in a number of animal welfare contexts including traditional epidemiological studies of risk factors for animal health and welfare problems, computational modelling to reveal processes underlying behaviour, affect and decision-making, the use of computer vision to develop automated methods for welfare assessment, and implementation of deep-learning algorithms to detect data patterns (e.g. variation in thermal images) predictive of poor welfare and disease.

Our work involves collaboration with BVS researchers Data Analytics, and Platform Technologies. We also have collaborations with researchers within and outside the wider University as detailed below.

Ongoing work and collaborators include

• Development of cognitive measures of affective state and welfare using parallel studies of animals and humans, and computational modelling to tease apart links between affect and decision-making (Vikki Neville, Prof Mike Mendl, Dr Liz Paul (BVS); Collaborators: Prof Peter Dayan (Max Planck Institute for Biological Cybernetics, Tübingen, Germany); Prof Iain Gilchrist (Psychology))
• Using thermography and deep learning to automatically detect risk of disease in calves (Dr John Fennell, Dr Laszlo Talas (BVS))
• Developing computer vision approaches to animal welfare assessment: the defence cascade response in pigs (Dr Poppy Statham, Prof Mike Mendl (BVS); Collaborators: Dr Neill Campbell and Dr Sion Hannuna (Computer Science))