Advances of early automated disease detection in calves – a review of the past decade

Young calves are highly prone to disease and incur high mortality rates. A literature review was conducted to provide an overview of advances in automatic health monitoring in calves over the past decade, to reveal the gaps between the application of available automatic tools and PLF health monitoring in calves, thus point to opportunities for future research. A systematic literature search was conducted using the Web of Science database. 127 articles were found as primary outcome based on selected search terms, and following 4-step, 44 articles were left. We found that current stages of calf disease detection are at applying sensor technology at individual level and interpreting the data to describe ongoing diseases. Accelerometers (22 articles) and automatic feeding systems (AFS) (12 articles) were most studied, which can record daily feeding behaviours (i.e. milk intake, milk drinking time, and frequency) and daily locomotor activities (i.e. lying time, lying frequency, step counts). A sick calf generally decreases milk intake, milk drinking speed, visits to the milk feeder, and duration of each visits; regarding the changes in lying time and frequency, conflicting results were shown in different studies. We concluded that changes in feeding and lying behaviours were linked to early stages of disease and enabled faster detection of disease compared to clinical evaluation by a person. More researches, however, that can look at these data in a more detailed way are needed, i.e. reveal the behavioural patterns across the day. In addition, more variables need to be investigated to find good combinations of indicators for early disease detection. Furthermore, bridges between technology, biology, and farmers have not yet been built. Multi-disciplinary research is therefore needed to make this connection, which will assist in developing an integrated system to automatically detect calf disease at an early stage.