This column brings you information regarding some of the research being done around the world and published in the Journal of Dairy Science. The objective is to bring to light areas of research that may have an immediate practical application on a dairy farm, as well as research that, even though it may not have a practical impact now, could be interesting for its future potential application. The idea is to give a brief overview of select research studies but not go into detail on each topic. Those interested in further in-depth reading can use the citations to find each study.
'International Symposium on Ruminant Physiology: Leveraging computer vision, large language models and multimodal machine learning for optimal decision-making in dairy farming'
Journal of Dairy Science Vol. 108, No. 7, 2025. This article, written by researchers from the University of Wisconsin, explores various applications of artificial intelligence (AI) technologies in dairy farming, including the use of computer vision systems (CVS) for animal identification, body condition score (BCS) and body shape analysis, and potential uses of large language models (LLM) in the dairy industry.
The authors explain that in recent years, precision livestock farming technologies have emerged as powerful tools for improving the efficiency and productivity of livestock farms. These technologies enable faster and more informed farm management decisions and enhance genetic selection capabilities through high-throughput phenotyping. Among precision livestock farming technologies, computer vision systems have received great attention due to their potential for monitoring animals in a highly scalable and nonintrusive way, with few devices being able to collect phenotypes from multiple individuals at a time. While RFID involves attaching eartags to each animal and using antennas placed at key locations (e.g., near cameras collecting phenotype data) to link the collected phenotypes to specific animals, computer vision algorithms can identify animals directly from the phenotyping images, removing the need for external tags or systems. Additionally, it addresses animal welfare concerns associated with the use of such equipment.
In terms of using computer vision for BCS, the authors indicate that assessing BCS in commercial farms is time-consuming, requires trained evaluators and can lead to inconsistent results due to its subjective nature. Due to this, computer vision systems (CVS) have been developed to perform BCS evaluation in a more automated and systematic way. There are several models being developed for this purpose, some of them with great accuracy. Rather than relying on subjective BCS and locomotion scoring, CVS can compute quantitative body shape characteristics (e.g., volume, area and angularity) and mobility metrics (e.g., head bobbing, stride length and stride duration) for a more objective health assessment in dairy herds.
The potential of LLM in the dairy industry is vast. These advanced tools can facilitate the seamless integration of structured and unstructured data from diverse sources, for example, images, text and tabular data at the same time, and function as virtual farm assistants. By leveraging these capabilities, LLM could revolutionize technical consultancy and enhance decision-making, operational efficiency and animal health on dairy farms.
The authors say that, despite significant progress in recent years, agriculture continues to lag other industries in adopting AI and digital technologies. Technology adoption in agriculture is more challenging than in other industries due to limited connectivity, harsh environmental conditions and lower financial incentives. Within agriculture, livestock systems lag further behind crop systems, primarily due to lower financial investment and the inherently greater challenges of monitoring animals compared with crops, as animals move both within farms and across farm boundaries. The development of scalable and impactful technologies in livestock systems relies on several key factors. First, education is crucial, as it enhances knowledge transfer through academic research and ensures its translation into commercially viable solutions. Additionally, financial investment is essential to increase funding for research, entrepreneurship and education in digital agriculture. Although there are many potential challenges ahead, these technologies offer significant opportunities for advancing animal health monitoring, farm management and individual phenotyping.
'Long-term effects of preweaning social housing on growth and reproductive development of dairy heifers'
Journal of Dairy Science Vol. 108, No. 9, 2025. This article, from researchers from the universities of Florida, Helsinki and Missouri, assessed the effects of preweaning social housing on growth until 1 year old, mounting activity and age at onset of cyclicity of heifers. For this, the authors enrolled dairy heifer calves to one of three treatments: (1) individual housing (n = 55 heifers), (2) pair housing (n = 55 focal heifers; one focal heifer per pair) or (3) group housing (n = 53 focal heifers; selected from nine groups of 10 calves per group). Calves assigned to group housing were managed identically to pair-housed calves until 15 plus or minus 3 days old, at which point they were introduced to a group of 10 calves (less than 10 days age range) and fed via an automated milk feeder. All calves received 8 litres per day of milk replacer and were weaned over 10 days beginning at 7 weeks old. At 9 weeks old, calves were mingled between previous housing treatments and housed on pasture. Growth outcomes were recorded weekly from 6 to 12 months old. The main objective of the study was to address gaps in knowledge of the long-term effects of different early life social housing systems on growth performance and reproductive development of dairy heifers.
The authors provide some background for the study, indicating that raising dairy calves in social groups with same-age peers is emerging as a common alternative to the historically widespread practice of individual housing following maternal separation at birth. Social housing provides well-established performance benefits for dairy calves, including increased solid feed intake and weight gain, and behavioural effects with benefits for animal welfare, including earlier expression of positive social interactions (e.g., social grooming) and reduced fear and avoidance of novelty. They say that despite scientific consensus on short-term benefits of social housing, individual housing remains common in some regions, and there is a lack of knowledge of long-term effects of social housing on heifer developmental outcomes that are relevant for production.
Evidence of growth advantages for calves housed in pairs compared with individually are apparent following weaning, and some recent findings suggest pair-housed calves may maintain a bodyweight advantage into pregnancy and following calving. Long-term performance effects of social housing may arise from improved preweaning weight gain, which is known to translate to longer-term growth advantages when enhanced by nutritional factors. Further, longer-term effects may be mediated by behavioural responses shaped by early experience, as early life social isolation can have long-term detrimental effects on brain function and social behaviour. Alongside potential long-term effects on growth outcomes, social housing may have implications for reproductive development. In gregarious mammals, social isolation can delay puberty attainment and disrupt the development of normal sexual behaviour.
The study results showed heifers raised in pair-housed and group-housed during the preweaning period had growth advantages relative to individually housed heifers that were apparent postweaning. However, more group-housed calves per week had a fever compared with both individual and pair-housed. Seasonal effects suggested benefits for calves born in cooler months (Q1 and Q4 versus Q2 and Q3) with respect to health (9.7% versus 18.3% of calves per week with a fever) and marginally for bodyweight at 60 days (+4.1 kg). Preweaning social housing treatments had long-term effects on growth from 6 to 12 months of age. Bodyweight over the six-month data collection period was subject to an interaction between treatment and month, where bodyweight at 6 months did not differ significantly between previous housing treatments, but final bodyweight was greater for pair-housed compared with individual-housed heifers, with group-housed intermediate. Hip height was also subject to an interaction between treatment and month and main effect of treatment, where pair-housed heifers had greater final hip height than both individual-housed and group-housed heifers. In terms of effects of preweaning social housing treatments on outcomes related to mounting activity and reproductive cyclicity, mounting activity was indicated at a younger age in pair-housed and group-housed heifers than individually housed heifers. However, the age at onset of cyclicity did not differ between previous housing treatments.
Whereas growth advantages were clearly attributed to heifers reared in pair-housed versus individual-housed, group-housed heifers had outcomes that were largely intermediate or resembled outcomes in individual-housed calves. The authors say these findings suggest that long-term developmental effects of preweaning social housing are dependent on the nature of the housing environment and management. Whereas pair-housed heifers experienced otherwise identical management to individual-housed heifers, the group-housed treatment introduced factors that often vary with management of larger social groups on-farm: a stressor of introduction to the larger social group, learning to use an automated milk feeder (can disrupt rest and can reduce milk intake while calves learn to use the feeder) and exposure to a larger age range of animals, in addition to the larger group size. Further, social housing within larger groups may present health challenges, whereas pair-housing does not appear to increase disease risk.
The authors conclude that the study supports long-term growth benefits attributed to social housing, particularly for heifers raised in pairs compared with individually. Long-term growth benefits were lessened in heifers raised in larger groups, an effect which may be attributed to management stressors or increased disease risk.
