- August 08, 2025
- By Daniela Benites
The National Science Foundation has invested $2 million in a new artificial intelligence project led by the University of Maryland to advance sustainable biopolymer production, which will help address pollution and other consequences of the global plastics industry.
The AI investigation will focus on developing an integrated machine learning and robotics design system to create and assess biopolymers, a biodegradable alternative to petrochemical-based synthetic polymers commonly used in plastics that could reduce carbon emissions; the U.S. Environmental Protection Agency identified plastics production as the source for 3.4% of such emissions in 2019.
The research, which expands on a UMD Grand Challenges Team Project grant, will be led by Assistant Professor Po-Yen Chen in the Department of Chemical and Biomolecular Engineering in collaboration with Teng Li, a professor in the Department of Mechanical Engineering, and Sanghamitra Dutta, an assistant professor in the Department of Electrical and Computer Engineering, as well as collaborators at Iowa State University.
Beyond emissions, plastics have also proved a major source of environmental pollution, with 85% of plastic waste accumulating in landfills or oceans—disrupting ecosystems, biodiversity and food chains.
“Plastic pollution is choking our planet,” said Chen. “With the support from this grant, we bring together cutting-edge tools in AI, machine learning and robotics automation to build a public, accessible foundation for sustainable materials development.”
Although biopolymers are thought to match petrochemical plastics’ performance in mechanical resilience, strength and antimicrobial features, scientific challenges have slowed their advancement. The vast number of not-yet-studied biopolymers have shown the limits of conventional research methods, and a lack of standardized protocols for preparing biopolymers undermines the accuracy of subsequent machine learning models, which must be trained on existing data about biopolymers.
To address the scientific challenges, the research team will create a system to design sustainable biopolymer nanocomposites with programmable properties. This project will consist of a comprehensive research framework that integrates biopolymer informatics, robotics systems, machine learning methodologies and advanced simulation tools.
“Our system brings together AI, robotics and simulations to quickly predict which biodegradable and sustainable materials will work best to accelerate real-world solutions to plastic pollution,” said Chen.