Wearable sensors such as activity monitors have the potential to indicate when an animal's activity changes or is different from others. During my post-doc, my group used the Actiwatch Mini®. to automatically record the behavioural activity levels of sheep. We were able to demonstrate its ability to detect differences between low, medium, and high activity (McLennan et al., 2015). 

Now, I am working with my colleagues in Cambridge, Marwa Mahmoud, and some summer interns to develop a visual computer based system that would be able to detect lameness in a sheep using images of the sheep whilst walking or even standing. We are using posture and behavioural expression to train the system to detect changes that are expressed by lame sheep, such as limping, head nodding and toe pointing.