New AI Institute Focused on Climate-Smart Agriculture and Forestry

The National Science Foundation (NSF) has just announced a $20 million grant for a new AI Institute led by the University of Minnesota, focusing on climate-smart agriculture and forestry. Named AI-CLIMATE (for Climate-Land Interactions, Mitigation, Adaptation, Tradeoffs, and Economy), the new institute will be led by PI Prof. Shashi Shekhar, Distinguished McKnight University Professor from Computer Science and Engineering. The USDA National Institute of Food and Agriculture (NIFA) and NSF jointly funded the institute.

The Minnesota Robotics Institute (MnRI) played a crucial role in coordinating the preparation and submission of the proposal for this AI Institute. It also coordinated similar proposal submissions in previous funding cycles where the research team was invited for reverse site visits but was unsuccessful. “It is immensely gratifying to see our efforts finally succeed,” says Nikos Papanikolopoulos, Director of MnRI. Some critical participants of AI-CLIMATE are faculty members of MnRI, including Maria Gini, Nikos Papanikolopoulos, and Ju Sun.

AI-CLIMATE is a joint effort between the College of Science and Engineering (CSE), the College of Food, Agriculture, and Natural Resource Sciences (CFANS), and the Office of the Vice President for Research (OVPR) at the University of Minnesota Twin Cities. Other key universities participating in the institute include Cornell University, Colorado State University, Delaware State University, Purdue University, and North Carolina State University. The researchers will also collaborate with the American Indian Higher Education Consortium (AIHEC) and the tribal nations it represents.

“One of the driving factors of climate effects is carbon emissions,” says Shashi Shekhar, director of the institute. “By 2050, the United States aims to have net zero carbon emissions, and one of the most promising ways to do this is to use natural systems like forestry and agriculture as ‘carbon sinks.’” If a farm or forest sucks more carbon dioxide out of the atmosphere than it releases, it is a carbon sink. Farmers and foresters may be rewarded for doing this through carbon markets or systems in which property owners can sell “carbon credits”—equal to the amount of carbon dioxide their farm or forest has sequestered—to companies trying to offset their carbon emissions. With current technology, however, it’s difficult and expensive for farmers and foresters to measure how much carbon they’ve sequestered accurately. Using new AI techniques like deep learning and knowledge-guided machine learning, researchers at the AI-CLIMATE Institute will improve accuracy and lower the cost of accounting for carbon and greenhouse gases in farms and forests, ultimately making the process more accessible for more people.

“When farmers touch, smell, and look at the soil, they can tell if it’s carbon-rich,” Shekhar said. “But they can’t manually survey thousands of acres of land yearly. We’re developing easy-to-use tools to show all of these measurements and images of the soil to an AI neural network and let it calculate the amount of carbon for us, saving farmers and foresters time, energy, and money."

More information about the new AI Research Institute can be found on the NSF website.