Advancing animal welfare through sensing and AI technology

We have an urgent need to respond to consumers’ demands relating to animal welfare. To address this, the industry continues to transition to welfare-friendly systems, such as free range egg farms and group housing for pigs. This change presents a challenge to understand and predict animal behavior in these relatively new environments. Advancement in vision, sensing, and AI (artificial intelligence) technologies offer a crucial opportunity to develop new methods to automatically detect and analyze animal behavior. Better understanding of social interactions between animals in social groups will result in reduced instances of harmful animal tendencies such as feather pecking in laying hens and tail biting in swine.

To address these issues, we have joined the IMAGEN (AnIMAl Group SENsor) project. By combining sensing and AI technology with animal breeding and genetics, this project aims to improve the health and welfare of livestock and reduce the ecological footprint of our food production.

IMAGEN combines a team of industry and academic specialists :

• Three universities: Eindhoven University of Technology, Wageningen University and Research, and Utrecht University
• Three organizations: The Dutch Farmers’ Association, the Royal Netherlands Veterinary Association, and the Dutch Society for the Protection of Animals
• Six companies: Hendrix Genetics, Topigs-Norsvin, Vencomatic, Noldus IT, Sorama, and FarmResult

Hendrix Genetics will be closely involved in one of the sub-projects focused on behavioral interactions in laying hens. The project will span five years and contain three main tasks:

1. developing sensing and AI technologies for automated behaviour detection in laying hens
2. behavioural research of the emergence and spread of behaviours in laying hens
3. genetic modelling of the emergence and spread of behaviours in laying hens

IMAGEN will result in a prototype hardware/software system for the automated detection of behaviours of laying hens and pigs kept in large groups. Building on the data generated by these technologies, we will develop mathematical models and data analysis methods to understand the behaviour of groups of animals, and identify the genetic and environmental factors determining the health, welfare and ecological footprint of livestock. It will allow us to breed for improved future generations of animals.

The IMAGEN project also nicely aligns with the SMART Turkey project, where we focus on the automated detection of individual health and behaviour phenotypes in turkeys.

By working together with a range of experts and stakeholders in both projects, we are proud to be a part of this ambitious venture to maximize health and welfare of the animals in our care and throughout the industry.