The European Federation of Animal Science held its 74th annual conference in Lyon, France, between the 26th August and 1st September. The dairy use-case of AGROS was present with an oral presentation on breath analysis and a poster presentation on 3D computer vision.
In his talk, researcher István Fodor summarized the results of the collaborative efforts on using breath analysis for disease detection in dairy cows. The primary focus has been on developing a proof-of-concept for non-invasive early detection of ketosis and deviations from a healthy status. He emphasized that breath analysis as a non-invasive diagnostic tool can potentially evolve together with efforts to upscale methane measurements on dairy farms.
Researcher Marjaneh Taghavi shared insights from a pilot study on using 3D computer vision for estimating feed weight in dairy cows. She concluded that the exact type of model and the distance between the feed and the camera are important drivers of prediction accuracy, and having relatively large amounts of input data is essential for model training.
In the next steps we will continue exploring the diagnostic capabilities of breath composition and test new applications of 3D computer vision in dairy herds.