NC State University researchers will use a grant from Schmidt Sciences to develop an AI-enabled, open-source platform that can analyze the potential to convert agricultural leftovers — biomass such as misshapen sweet potatoes or their discarded green tops — into biofuels or other bioproducts, creating new business opportunities. 

The four-year project, Resource Engineering Framework for Responsible, Augmented Modeling and Engagement (REFRAME), supports the path to a circular bioeconomy, where biomass is repurposed to create new materials. This approach reduces waste, creates opportunities for entrepreneurs, and provides new sources of income for farmers and agribusinesses.

To find the best ways of converting sweet potatoes and green tops into biofuel, for example, a sustainable aviation fuel company will be able to use the new platform to analyze the current yield in an area and determine where to locate a collection facility or biorefinery. The platform will be designed to assist farmers, processing plants, researchers and policymakers in making decisions. 

NC State University’s Daniela Jones will lead the project. She is an assistant professor in the Department of Biological and Agricultural Engineering, joint appointment faculty member with Idaho National Laboratory, and N.C. Plant Sciences Initiative Education and Workforce Development Platform director. REFRAME co-leads are NC State University’s Debjani Sihi from the Department of Plant and Microbial Biology, Joe Sagues from the Department of Biological and Agricultural Engineering, and Cranos Williams from the Department of Electrical and Computer Engineering. Additional collaborators are from Idaho National Laboratory, the University of Illinois Urbana-Champaign, ICF International and McCarl & Associates.

We asked principal investigator Jones to explain the team’s plan for REFRAME.

How will you approach the REFRAME project?

My expertise is in supply chain analytics, so our efforts cover the full supply chain. So in terms of availability, where is this agricultural biomass? In terms of accessibility, how do we move it from point A to point B? How can we characterize it? What are its qualities? What can it become through conversion? What would be the cost? What would be the impact or the shock from introducing a new product into the market? How would farmers respond? How would biomanufacturing affect land use? These are examples of the questions we want to be able to answer much more rapidly with an AI-enabled platform.

Will you incorporate any existing digital models or tools?

For all of these supply chain components, we have identified some legacy models, models that have been developed for years that are open source and that have been tackling one aspect of the supply chain — for example, agronomic models that analyze availability of biomass. However, these legacy models have always worked in silos. A lot of people have worked very diligently on one model, but strong connections between models don’t currently exist. The pipeline, in waiting for the delivery of one model’s output so that it can go into another as an input, takes months and years.

How will you apply AI to help with these issues?

What we are creating is first a digital backbone, a digital environment where we can store and have the computing capacity to be able to run a model multiple times so that we can create AI surrogates that are much faster. By creating simplified versions of these models that run faster, we have the potential to predict things that were not previously possible with the legacy models. So that’s the gist of it. We will create this pipeline of models and it will become accessible to anyone out there as an AI-enhanced, open-source platform.

The REFRAME project is part of Schmidt Sciences’ Virtual Institute on Feedstocks of the Future (VIFF).

About Schmidt Sciences 
Schmidt Sciences is a nonprofit organization founded in 2024 by Eric and Wendy Schmidt that works to accelerate scientific knowledge and breakthroughs with the most promising, advanced tools to support a thriving planet. The organization prioritizes research in areas poised for impact, including AI and advanced computing, astrophysics, biosciences, climate and space — and supports researchers in a variety of disciplines through its science systems program.

This post was originally published in College of Agriculture and Life Sciences News.