Automated oestrus detection methods in group housed sows: Review of the current methods and perspectives for development

This paper reviews the current automation methods of oestrus detection in group housed sows and points out their limitations. Results from non-automated methods are also presented, in an effort to draw conclusions to assist in the development of new automatic methods. Automatic measurement of physical traits, e.g. body temperature, collect non-specific information but may be combined with activity measurements, where information specific to the onset of oestrus should be given more weight. The perspective for the development of new methods relies mostly upon the measure of behavioural traits; Global Positioning System or computer vision present interesting possibilities that assist in the recognition of behavioural changes related to the onset of oestrus. Sensitivity and specificity of the methods must be carefully balanced so a better detection does not mean a higher work input associated to checking sows that are not in oestrus.