By Krisy Gashler for the Cornell Chronicle
April 20, 2022
Innovations in digital agriculture are enabling farmers and researchers to capture data from sources like aerial drones, ground robots, satellites and mobile apps. But making sense of all this information – the integration and analysis of big data – requires further innovation.
Michael Gore, the Liberty Hyde Bailey Professor and chair of the Plant Breeding and Genetics Section of the School of Integrative Plant Science (SIPS) in the College of Agriculture and Life Sciences, is leading a multidisciplinary project to create an open-source digital ecosystem that will integrate agricultural data from multiple streams and make it available to anyone.
Gore won a $1 million grant from the U.S. Department of Agriculture’s National Institute for Food and Agriculture to support the work.
“Our goal is to create an open-source digital ecosystem that has the computational tools and data analysis approaches that will allow users to use this dataset to ask important biological questions,” Gore said. “And once these models have been developed and pressure-tested by researchers, they could be deployed in growers’ fields, as well.”
Gore said he anticipates the digital ecosystem will be most useful to people interested in using multifaceted data sources to evaluate plants, but who may not have the technical know-how to process and make use of such data.
The project should benefit crop breeders and crop modelers, who are interested in evaluating large amounts of phenotypic data (observable traits in a plant) in a small amount of time. For example, a breeder might be interested in understanding how plant height varies over a growing season in comparison to the number of leaves a plant grows or fruits it develops. An aerial drone can give information about plant health and productivity, such as photosynthesis rate, while a ground robot can collect complementary data from below the leaf canopy that describes a plant’s size and shape. High-resolution satellite data can be used to measure plant productivity, among many things.
For growers in the future, Gore said, this system could provide earlier, more accurate predictions on crop yields or disease pressures.
One of the biggest challenges in integrating all this information is determining how to geolocate data points from different data streams in a field so researchers can effectively evaluate all these traits for each plant, Gore said. Solving that problem requires collaboration between plant breeders, computer scientists, statistical geneticists, crop modelers and all the support staff who make field trials possible, he said.
Co-investigators on the project are Kelly Robbins, assistant professor in SIPS’ Plant Breeding and Genetics Section; Ying Sun, assistant professor in SIPS’ Soil and Crop Sciences Section, Noah Snavely, associate professor of computer science at Cornell Tech; and Abe Davis, assistant professor of computer science in the Cornell Ann S. Bowers College of Computing and Information Science.
In addition, the project team is working with Joseph Gage, assistant professor in crop and soil sciences at North Carolina State University, and has convened a scientific advisory board, with input from external research institutions and the USDA’s Agricultural Research Service. Once the digital ecosystem is created, the project team plans to host virtual workshops to train users in working with it.
“This is an exciting, radical collaboration: By bringing together all these individuals with their different skill sets, we’re able to break down the boundaries that have separated our disciplines in the past and make new discoveries,” Gore said. “The students and postdocs who participate in this grant will be trained in computer science, crop modeling, breeding and genetics. It will really help educate and train the next generation of digital biologists to learn how to translate discoveries into improved crop varieties.”
Krisy Gashler is a writer for the College of Agriculture and Life Sciences.