Center for Digital Agriculture



The four primary research themes of the Center for Digital Agriculture are:


Technologies that reduce and augment human labor and achieve more effective agricultural outcomes through improved scalability, accuracy, and precision


All aspects of data collection, security, privacy, storage, analysis, transfer and decision making, related to agriculture and food


Research to improve animal and crop genetics, production practices, and sustainability using digital technologies


Use of digital technologies in the social and economic aspects of agriculture, food, nutrition and rural communities


Large agricultural equipment used in production today are becoming increasingly automated and instrumented. Nevertheless, the potential benefits of automated sensing and of robotics in agriculture have been explored to a very limited extent. For example, sensing and digital imaging could provide data on plant disease and drought stress susceptibility, chlorophyll content, nutrient concentrations, growth rates, and yield potential as well as food production challenges such as uninterrupted monitoring of temperature and pH.

Likewise, sensing and imaging can help improve animal welfare, health, and productivity. Computer vision can aid in phenotyping, animal monitoring, disease detection, sowing, weeding, picking, sorting, grading, and packaging. Low-energy edge computing combined with wireless technologies can make sensor use more effective through analysis and decision making in near-real-time.

The potential of robotics in agriculture has been explored to a very limited extent. For example, teams of dexterous robots could enable entirely new kinds of production agriculture systems ranging from high-throughput phenotyping and monitoring to production farm management to addressing labor shortages.

Collaborative efforts with researchers in the data, animals and crops, and people focus areas will identify and investigate suitable applications for technology. New digital capabilities can be achieved through orders-of-magnitude improvements in the energy, compute, and memory requirements of the underlying systems.


Many of today’s production practices in the United States and other advanced societies use extensive instrumentation on large farm equipment, and can provide timely and actionable information via broadband wireless or satellite connections. Today, these data already provide an opportunity to improve crop yields and reduce the excessive use of chemicals and nutrient runoff.

Digital agriculture practices can provide the data to create high-resolution models of farm-environment interaction and can inform growers on precise prescriptive chemical application.

In the future, data-driven approaches will integrate real-time data from farmers, service providers, financial providers, satellites, mobile devices, drones, weather stations, and other digital sensors to support productivity and provide access to services.

Using big data analytics, economic modeling, and forecasting, we’ll investigate the relationship between local and global food security issues and resource management. We’ll employ blockchain technology to track detailed information about inputs and production processes through many intermediaries in complex food distribution systems from farm to table.

A key obstacle to realizing this vision is that many farmers and commercial growers are reluctant to share their data because of competitive concerns. We will work with farmers, farm equipment companies, and data analytics companies to develop novel privacy-preserving analysis techniques that will be needed to reduce these risks. We will coordinate with industry groups and policy makers to develop data-sharing best practices and regulatory approaches that foster mutually beneficial data sharing and large-scale analytics beyond individual farms to span broad geographic regions.

Increased reliance on computer control and digital communications make agricultural and food systems vulnerable to remote attacks and compromises. We will explore novel algorithmic and computing systems approaches to address cybersecurity concerns.


New sensor technologies, spatial datasets, and analytics platforms are also necessary for agricultural research and outreach in animal agriculture.

Critical agricultural microbiome research furthers understanding of livestock nutrition. Research into the gastrointestinal microbiome of animals could have substantial benefits in nutrition, production, animal welfare, and targeted antibiotic use.

Precision animal management integrates applied livestock management (e.g., housing, animal behavior, animal health, stress physiology) with cutting-edge techniques in data science (e.g., high throughput phenotyping, data visualization, predictive analytics). This integrative approach will improve welfare and production efficiency of livestock species and will maximize environmental stewardship of agricultural operations.

Collaborative efforts in CDA will result in new technologies to gather data, such as drones to monitor animal movements, remote sensors for monitoring vital signs, digital video of animal behavior, photo analysis to detect conformation abnormalities, and infrastructure for high throughput phenotyping.

This emerging area can provide new means and methods for optimizing resource management, improving animal health and well-being, improving meat quality, and remaining productive in a changing climate.

Another area in which vast quantities of data are produced is in host-microbiome research. Quantitative and computational tools to analyze, visualize, and interpret large and complex datasets from host-microbiome research are inadequate or do not exist.

To address this issue, research is focusing on development and deployment of multiple capabilities, including novel bioinformatics software for understanding of the role and function of the microbiome in the host and new visualization tools for host-microbe interactions.


Continued genetic improvement of crops is a focus for many researchers around the world.

Plant biotechnologists use bioinformatics and modeling to guide crop genome modification for creating new variations while crop breeders use statistical models to guide and predict genetic crop improvement. To increase yields, these models must be sensitive and accurate.

Digital methods can greatly enhance the speed and accuracy of these approaches, leading to more rapid gains in genetic improvement in both biotechnology and breeding.

For example, automated crop phenotyping can greatly increase the quantity and quality of data on plant growth and development, allowing more accurate and narrow analysis of specific desirable traits, and the screening of larger populations to increase breeding gains per cycle. Real-time on-farm data can facilitate research on field performance and improved cultivation practices.

The soil is also an increasing focus of cross-disciplinary research. Understanding the association of plant roots with endosymbiotic and soil microorganisms and other soil parameters is an emerging research field entirely dependent on digital information from sensors, sequencers, and other data sources.

Sustainable farmland management will be based on collecting and analyzing dense data of soil characteristics along with weather and market variables. Research that advances better decision-making at the producer level, such as use of fertilizers and pesticides, will have environmental and productivity benefits.


People are a crucial part of the agricultural system as both stewards and beneficiaries of the multitude of outputs produced from food, fiber, and fuel products as well as environmental services. Understanding the interaction between agriculture and people requires modeling the interplay between human and natural systems.

For example, predicting how a new seed technology will alter land use and environmental outcomes requires researchers to understand how farmers might change their behavior. Similarly, researchers might want to predict how climate change will affect trade patterns and change farmer profits, food accessibility, composition, and nutritional outcomes.

Answering these questions requires detailed data on farm production, environment, human behavior, markets and nutrition, sensors, remote sensing, and even retailers and cellular phones. It also requires the ability to combine the data to understand the agricultural and food system as a whole.

Machine learning approaches afford the potential to build complex coupled models of social and natural systems that are flexible and scalable. These approaches can be used to understand the agricultural supply chain by predicting supply shortfalls and resulting price spikes, climate vulnerabilities, or transportation bottlenecks.

Advances in the analysis of high-frequency trading data combined with text analytics will help analysis of market information and infrastructure on price outcomes. Together, these models will allow us to identify vulnerabilities and target areas for efficiency gains.

More research is needed in agricultural environments to increase the effectiveness of agricultural practices, food processing and food safety, nutritional quality of products, and product shelf life while minimizing food/animal waste, carbon emissions, and air/water pollution.