Periodic Reporting for period 1 - dAIry 4.0 (Advanced, trustworthy AI and data solutions for individualised automated milking & feeding of dairy cows)
Reporting period: 2023-10-01 to 2024-09-30
Producing more with fewer resources and less impact on society, animal welfare, climate and biodiversity is essential for modern dairy farming. Automated Milking Systems (AMS) have greatly enhanced farm profitability by reducing manual labour and costs while enhancing animal welfare, milk quality and farm management. The dAIry 4.0 project advances AMS by integrating AI-driven tools that allow individualised and optimised milking, feeding, and gating control to improve system performance. New sensing systems in the AMS enable general health inspection of the animal even at the single teat level and support milk quality evaluation, partitioning milk into separate tanks. Additionally, a novel Decision Support System (DSS) leverages multimodal learning techniques to generate recommendations for managing each cow, allowing farmers to adopt more precise and proactive management practices.
Initially, work focused on identifying specific requirements for each stakeholder. Surveys have been conducted across the EU to gather these requirements. In parallel, six farms that already use Automated Milking Systems (AMS) were selected to serve as the primary use cases and real-world demonstrators. Progress was also made in the identification of the data that will serve as the basis for further research regarding the AMS adjustments. Historical data has been identified, in parallel with actual data from the farms. Several technologies for the dAIry 4.0 project are now under development, as next project phase has started. New camera-based sensing systems are being designed for integration into milking robots, enabling real-time health monitoring at the single-teat level. These cameras, along with novel sensors for milk quality assessments, will allow milk to be partitioned into different tanks based on its quality. Additionally, data from all sensors will feed into a so called "Animal Status Classifier" module, which will monitor and assess animal health and support related decision-making at the single animal level. All gathered data will be integrated into the new dAIry 4.0 management platform, under development. Supported by the Decision Support System, a Simulator and a farm Digital Twin, this platform will generate real-time recommendations and notifications for farmers.
dAIry 4.0 project aims to go beyond the current state of the art by focusing on the individualised, AI-driven optimization of milking and feeding processes at the single animal and single teat level. By leveraging advanced AI and data analytics, the system will implement adaptive milking settings for improved udder health and reduced CO2 emissions. Additionally, AI-powered individualised feeding strategies will optimize nutritional balance and cost efficiency, enhancing both animal welfare and farm sustainability. Early progress will also include the development of robust animal health and welfare monitoring systems that provide reliable, actionable insights, integrating novel data techniques and AI models for real-time analysis.