THE FARM OF THE FUTURE: HARNESSING DATA-DRIVEN TECHNOLOGY FOR A SUSTAINABLE AND RESILIENT US AGRICULTURAL SYSTEM

<b>Project Methods</b><br> CAST will include the state-of-the-art Cornell University Ruminant Center, Cornell Teaching Dairy Barn, and Musgrave Research Farm, which comprise a large (~2,550 acres), diverse land base in NY State. Through systematic integration of data, coordinated technology testing and demonstration, and exchanges of physical materials, these rural farm units will form an advanced hub for research, extension, and education that helps the FotF to fulfill its promise. CAST will focus on field crops and dairy production as models of the US ag economy, these being among the largest sectors in volume and value and offering some of the greatest challenges and opportunities for mitigating climate change. CAST's multi-site nature is a major strength: more can be learned about technologies, practices, and intelligent systems if they are applied across operations of varying size, type, and management. CAST's two crop production units (~2,550 acres available) and two dairy herds (~825 adult cows, 500 youngstock) will generate enough data to realistically model key challenges of integration and analytics.RESEARCH. CAST's four research thrusts will support innovation through cycles of development, deployment, and evaluation of technological and data-driven breakthroughs and test and demonstrate existing and emerging technologies and practices under commercial-farm-like conditions. Research on innovation in technology and farm practices will develop, deploy, test, and demonstrate innovative technologies and management practices under working farm conditions.Under research thrust 1,Innovation in Technology and Farm Practices, Specific Objectives forTechnology-Enhanced Field Crop Productionwill develop, refine, test, and demonstrate technologies and management practices in (1) Precision management of crop inputs, (2)Cover cropping and (3) Soil amendments such as rock dust and biochar.Under Specific ObjectiveSmart Automation and Data-Driven Precision Animal Managementthis project willevolvea suite of technologies in support of precision management and automation in animal systems that can enhance animal and human health, well-being, and performance while improving farm profitability and sustainability. Specifically, research will develop and demonstrate data-driven technologies for (1) Precision feeding and nutritional managementby data-driven ration formulation and automated monitoring and management, (2) Precision health management throughdata-analytic tools using integrated sensor and non-sensor data to predict automatically and in real time health outcomes of cattle, and (3) Precision reproductive monitoring and managementthroughautomated estrous detection tools, data-driven decision support methods using integrated data,automated fertility control,and point-of-care diagnostic devices.Under research thrust 2, Data Integration,the Cornell-developed Software Defined Farm, will convert raw, uninformative data from multiple sources and in diverse formats into high-quality data streams that enable efficient, accurate data analytics, and reporting. Data generated by thehardware and software infrastructure at CAST will be used to develop fully automated processes for real-time data capture and initial processing at the source, standardization of data and metadata, transfer to intermediate and centralized data streams, automated identification of technical failures, and automated discrimination of technical failures versus biological variation.Under research thrust3,Data Analytics and Decision Support,the wealth of data generated and captured by technologies and intensive management practices at the CAST will be interpreted by machine learning algorithms (MLAs) and data analytics to be maximally useful to farmers. We will create, refine, and test several algorithms for improved decision making for soils and crops and for animals at CAST. For cattle outcomes algorithms willautomatically predict health and reproductive outcomes in real time. For field crops outcomes, we will develop AI tools for site-specific management for adjusting inputs rate on a per-zone basis and understand the main drivers of productivity for each zone.Adata integration and fusion infrastructure integrating historical and real-time sensor and non-sensor data will enable the exploration of multiple interactions between the numerous sources of variation driving outcomes of interest.Under research thrust 4,Farm, Food, and Social Systems Impact Assessment, we will build on the knowledge gained to evaluate the expected economic, animal health, environmental, and social/socioeconomic outcomes of adoption of novel data-driven and farm management practices. Specific Objectives include evaluation of (1)Farm Financial Feasibility, (2) Whole Farm Animal and Environmental Health, and (3) Social and Socioeconomic Impacts.EXTENSION.By means of an actively involved stakeholder network, knowledge transfer, and education and training, CAST will promote development, marketization, acceptance, and adoption of the technologies and methods that it develops, tests, and demonstrates. Under Specific Objective 1, we will develop the CAST Network for Extension and Teaching (CAST-NET)in which stakeholders will participate in technology development, refinement, evaluation, and demonstration.Under Specific Objective 2,Knowledge Transfer,extension activities will target knowledge transfer and communication beyond CAST-NET. TheCAST vision and activities will be communicated through virtual platforms, on-site activities, and contributions to existing PRO-Dairy and Cornell Cooperative Extension programming. A website willaggregate and link to virtual content including live and recorded video streamed that will provide virtual CAST demonstrations for stakeholders. Two virtual courses for precision crop and animal management will be built on knowledge generated at CAST.Under Specific Objective 3,Demonstration and Training, industry engagement will include access to demonstrations, hands-on testing, and evaluations for stakeholders interested in technology and management practices employed at CAST. Industry organizations thatpartner with CAST will have access to arich data collection and integration platform for testing, awhole farm impact assessment framework for evaluation of financial and environmental feasibility, and a communications and outreach program for demonstration, on-site training, and visibility.EDUCATION. Educational efforts will focus on experiential learning–creating, touching, doing, interacting, evaluating, and reflecting, with CAST as a living classroom. Under Specific Objective1,a minor in digital agriculture (DA) will be developed to help undergraduates gain new perspectives, network across the university, do research pertinent to DA, and engage with CAST and CIDA.UnderSpecific Objective 2, aVirtual Course in DA for URM Studentswill disseminate knowledge from CAST and provide URM students from outside Cornell with an opportunity to learn from cutting-edge activities in real time.Cornell and the University of Arkansas at Pine Bluff (UAPB) will collaborate on developing a DA introductory course to a virtual format.UnderSpecific Objective 3, we will develop aResearch and Extension Summer Internship Program at CAST.Thisinternship program complements classroom learning with opportunities for extended, hands-on engagement in the development and evaluation of data-driven technologies and management practices at CAST.UnderSpecific Objective 4,Engage CAST in the CIDA Hackathon, CAST will belinkedto the annual CIDA Hackathon which gathers students, faculty, and community members to develop innovative tools and analyses driving DA research.