Worm-like robots swimming in soil to measure crop underworld

Crop scientists over the years have learned a lot about how plants grow above the ground, but much less is known about roots and their interactions with the soil.

Now, a Cornell project funded by two separate three-year grants will develop worm-like soil swimming robots to detect and record the properties of soil, water, soil microbiome, and how roots grow.

A $ 2 million grant from the National Science Foundation (NSF) led by principal investigator (PI) Taryn Bauerle, associate professor in the horticulture section of the School of Integrative Plant Science (SIPS) at the College of Agriculture and Life Sciences, will focus on plants and soil perspective.

Meanwhile, a $ 750,000 grant from the National Robotics Initiative NSF to PI Robert Shepherd, an associate professor at the Sibley School of Mechanical and Aerospace Engineering at the College of Engineering, will develop the soil monitoring robots.

The project will focus on maize, with the ultimate goal of incorporating factors related to root growth to improve breeding efforts and soil management that directly affect productivity and food security.

“We plan to develop new tools so that we can tap into the underground environment of plants and soil in a way that allows us to shine light in a black box of plant-soil interactions,” Bauerle said.

“This is really the next frontier in plant biology,” said project co-PI Michael Gore, Liberty professor Hyde Bailey and professor of molecular breeding and genetics in the SIPS section of genetics and plant breeding. By quantifying the characteristics of the subsoil, researchers can then identify relationships to the characteristics of the soil, Gore said.

To capture these measurements, the team will develop 1- to 2-foot worm-like robots that emulate the way a hole punches into the soil, combined with a peristaltic motion that mimics the way worms move through the soil.

“The front loosens the dirt and the back pushes that dirt forward and pushes that dirt into a tunnel wall,” Shepherd said. They expect a robot to collect continuous data up and down an entire row of corn.

The team will experiment with a range of sensors and strategies. A robot’s ability to push through soil can reveal properties such as soil density and firmness. The robots will also be equipped with small temperature and humidity sensors.

Fiber optic cables could provide a range of measurements, including direct root imaging to measure growth and angles. The team plans to employ “AquaDust” developed in the laboratory of the co-PI project Abraham Stroock, Professor Gordon L. Dibble ’50 of the Smith School of Chemical and Biomolecular Engineering of the College of Engineering. AquaDust fluoresces in different wavelengths based on the amount of water in the soil.

Optical fibers could also allow measurements of the excitation and emission wavelengths of soil microorganisms and root chemistry, including carbon compounds exuded from plant roots. “We should be able to roughly determine which chemicals and organisms are prevalent on the root surface and in the surrounding soil,” Shepherd said.

By quantifying root characteristics, soil properties, compounds, microorganisms and water, researchers can then use predictive models to combine soil and soil characteristics to predict things like grain yield and stress tolerance. , Gore said.

Another goal of the project will be to evaluate how plants might respond to the effects of climate change, such as water availability. Root growth measurements, taken into account with environmental data, can provide insight into how roots grow based on external conditions, such as drought.

Since soil is not a good medium for wireless transmission, the researchers will test prototypes that record data in memory, to be retrieved later. They may also be able to experience acoustic communication through soil and wires running along a row of corn plants. At the end of the project, the researchers hope to show live demonstrations of prototypes in a wheat field.

The preliminary work was made possible by the initial funding of a grant from the Cornell Initiative for Digital Agriculture.