Groundbreaking Machine Learning Approach to Remove PFAS from Drinking Water
Imagine a future where your drinking water is free from harmful chemicals that stick around forever. Researchers at Georgia Tech are making this possible with cutting-edge technology aimed at removing PFAS (per- and polyfluoroalkyl substances) from our water supply. These toxic chemicals are known for their ability to persist in the environment—and even the human body—for years, causing a range of health problems like immune system suppression and an increased risk of cancer.
As more communities across the United States struggle with water contamination, a team of engineers, led by Professor Yongsheng Chen, has set out to find a practical, scalable solution. Their research could be a game-changer in addressing the PFAS crisis, helping millions of Americans get access to cleaner, safer drinking water.
The Problem with PFAS: A Growing Crisis
PFAS, often called “forever chemicals,” are found in common items such as makeup, nonstick cookware, and food packaging. The problem is, once they get into the environment, they don’t break down easily. They’re in our water, soil, food, and even the air we breathe. For years, these substances have been linked to serious health issues, making them a major public concern.
In the U.S., over 200 million people are exposed to PFAS through their drinking water. Alarmingly, more than 1,400 communities have water with PFAS levels that exceed health experts’ recommended safety thresholds. Tackling this issue head-on, Professor Chen and his team are pushing the boundaries of environmental engineering to provide a solution that could help mitigate the damage these chemicals cause.
The Georgia Tech Team’s Innovative Solution
Professor Chen, an expert in environmental engineering, leads a multi-university team focused on developing a better membrane to filter out PFAS from drinking water. Unlike traditional water treatment methods that often produce harmful byproducts, the new system would offer a more efficient and sustainable way to remove these chemicals.
“This project is really about finding a scalable solution,” Chen said. “The goal is to create a membrane system that can be used in communities across the country to reduce exposure to PFAS.”
To make this happen, the team turned to machine learning. Machine learning is revolutionizing various industries, and now, it’s being applied to environmental engineering. By using advanced algorithms, the team has been able to model and design a membrane capable of trapping PFAS molecules effectively.
How Machine Learning is Changing the Game
Machine learning plays a crucial role in this research. Chen’s team, supported by over $10 million in funding from the U.S. Department of Agriculture (USDA), the National Science Foundation, and the Environmental Protection Agency (EPA), relies on machine learning models to predict which materials would best separate PFAS from water. The team is working to identify and develop the most effective membranes for this task, marking a breakthrough in water purification techniques.
Rather than simply applying existing solutions, the team is using ML to design a targeted solution. Instead of treating the whole body of water, the team isolates PFAS from the water stream. This focus on PFAS isolation and the development of efficient filtration systems could pave the way for a new generation of water treatment technologies that specifically address this chemical contamination problem.
The Collaborative Effort: A Multi-University Approach
Georgia Tech is not tackling this challenge alone. Collaborating with the University of Wisconsin-Madison (UWM) and Arizona State University (ASU), Chen’s team has built a powerful research network. UWM helped validate the machine learning model with molecular simulations, while ASU contributed valuable data and insights from their lab work. Together, the three universities form an unstoppable force dedicated to solving the PFAS problem.
Tiezheng Tong, an associate professor at ASU, emphasizes the importance of this collaboration. “Applying machine learning to membrane separation represents an exciting frontier for environmental engineering,” he said. With each school contributing unique expertise, this project has the potential to change the way we think about water treatment.
The Path Forward: Real-World Impact
The impact of this research could be transformative. If successful, the development of a PFAS-removing membrane could be adopted in communities nationwide. With more than 200 million Americans impacted by PFAS in drinking water, improving water purification methods is critical. But the road ahead is not without challenges. Scaling up the technology and making it cost-effective for widespread use will require further development.
However, with ongoing support from the USDA, NSF, and EPA, and the innovation driven by Georgia Tech, UWM, and ASU, this team is taking major steps toward making safer drinking water a reality for millions.
