16AGRITECHCAT5: Early detection of tail biting in pigs using 3D video to measure tail posture

Tail biting in pigs is a serious and unpredictable animal welfare problem for farmers worldwide. It results in losses for
farmers of £10.4M a year in the UK alone, mainly from carcass condemnation. Before damaging tail biting begins, pigs hold
their tails down. This project will develop a system to detect these tail posture changes using a 3D video system giving
farmers advance warning of tail biting in time to intervene. We will 1) Collect continuous 3D video from pigs at a high risk of
tail biting to capture the changes in tail posture pre-tail biting, 2) Provide a detailed behaviour analysis of tail posture
changes and 3) Develop software algorithms to automate this. The project partners provide expertise in pig behaviour
(SRUC), a route to market and algorithms for automated 3D video analysis (Innovent Technology Ltd), pork supply chain
knowledge (Sainsbury's) video expertise and access to a network of expertise in engineering and precision agriculture
(Agri-EPI centre).

Tail biting in pigs is a serious and unpredictable animal welfare problem for farmers worldwide. It results in losses for
farmers of £10.4M a year in the UK alone, mainly from carcass condemnation. Before damaging tail biting begins, pigs hold
their tails down. This project will develop a system to detect these tail posture changes using a 3D video system giving
farmers advance warning of tail biting in time to intervene. We will 1) Collect continuous 3D video from pigs at a high risk of
tail biting to capture the changes in tail posture pre-tail biting, 2) Provide a detailed behaviour analysis of tail posture
changes and 3) Develop software algorithms to automate this. The project partners provide expertise in pig behaviour
(SRUC), a route to market and algorithms for automated 3D video analysis (Innovent Technology Ltd), pork supply chain
knowledge (Sainsbury's) video expertise and access to a network of expertise in engineering and precision agriculture
(Agri-EPI centre).

Tail biting is an unpredictable, costly, multifactorial problem for the pig industry with no simple solution. This project aims to
develop and validate the prototype of an automated system based on 3D video to detect tail position to alert farmers to when an outbreak of damaging tail biting is imminent, so they can act to prevent it (lowered pig tails, rather than the normal raised tails are a sign that damaging tail biting will soon begin).

Innovent Technology Ltd (an SME) will benefit directly from sales of the system once fully developed. Sainsbury's Supermarkets Ltd will benefit through their involvement in the project by gaining insight into the problem of tail biting, which affects their supplier farms but impacts on the pork supply chain, and also by addressing their customer's ethical requirements for good animal welfare. Pig producers affected by tail biting will benefit financially from the system, as tail biting results in considerable costs and waste in the food chain, primarily through condemnation of carcases infected via open tail wounds. The first to benefit would be existing customers of Innovent and Sainsbury's supplier farms (particularly those involved in the further development of the prototype). The system will work in any indoor pig facility, so will also be sold to other producers, in the UK, EU and worldwide. Pigs will also benefit through improved animal welfare (reductions in both pain from tail biting and morbidity from infections).

From a policy perspective, tail docking is widely used in the UK and across the EU, despite its 'routine' use being banned by an EU directive. Farmers perceive that they have little option but to use this (partly) preventive measure, however policy makers, responding to pressure from citizens, have tried to reduce tail docking (which is an acutely painful mutilation). For example AHDB Pork in the UK has published advice on enrichment materials and agriculture ministers from Denmark, Netherlands and Germany pledged to reduce tail docking in December 2014. Our product will reduce the likelihood and severity of tail biting, increasing the confidence of producers to stop tail docking. If adopted by UK farmers first, our system could boost production and give farmers an animal welfare marketing advantage in export markets if tail docking was
stopped.

There is growing interest in the use of technology to monitor and control various aspects of livestock production ('Precision Livestock Farming'), with various academic research groups and companies working in the area. In the UK, Agri-EPI will become a centre of excellence, bringing together a network of expertise in technology, engineering and agriculture. SRUC is involved in a bid for a Centre of Excellence in Livestock (CIEL), a UK-wide consortium of academic and commercial partners enabling the sharing of expertise and knowledge exchange. Beyond farming, the use of machine vision for automated real time analysis of video has a wide range of applications in identification of human behaviours (e.g. for crowd safety, detection of crime) or animal behaviours (e.g. in lab, home, zoo and wild settings). The time of flight 3D video technology used in our project has great potential in a wide range of applications.

The requirement to protect Intellectual Property may limit knowledge transfer at first, but once IP protection is in place, we will use various channels to publicise our results. Agri-EPI network will facilitate a cross-fertilisation of ideas and best practice with Agri-tech stakeholders. Other outputs will include scientific publications, articles to promote the product in industry publications such as Pigworld (and The Pig Site), and direct communication with pig producers via AHDB Pork or QMS Knowledge Exchange channels, and via the considerable existing customer base of Innovent Technology Ltd and supplier farms to Sainsbury's Supermarkets Ltd.