As the volume of global plastic waste continues to grow, maintaining the circular economy and enforcing environmental regulations is becoming an increasingly urgent issue. Researchers at the University at Buffalo (UB), part of the State University of New York (SUNY) system, have developed a revolutionary system based on artificial intelligence (AI) that can non-destructively and highly accurately determine the actual recycled content of commercial plastic products, including the critical measurement of recycled content.
The Problem: Invisible Differences Between Virgin and Recycled Plastic
Until now, authorities and industry players have faced a major technological hurdle: there was no reliable method to accurately determine the percentage of recycled plastic in a finished product after the fact.
The difficulty stems from the physical process itself. When plastic is recycled, the material is melted down, cleaned, and remolded. As a result, the end product is visually identical to virgin (new) plastic, and their chemical compositions are almost exactly the same. The differences lie in microscopic structural changes invisible to the naked eye: recycled materials contain microscopic impurities as well as broken polymer chains resulting from the reprocessing.
The Solution: Multi-Modal Physical Testing and Machine Learning
A team from the Department of Chemical and Biological Engineering at the University at Buffalo—led by Amit Goyal, a SUNY Distinguished Professor—developed a new, multi-modal, non-destructive sensing technique based on these microscopic discrepancies.
The system does not rely on a single measurement; rather, it combines multiple scientific tests:
-
It examines triboelectric properties (charge retention capacity).
-
It utilizes dielectric and impedance spectroscopy, as well as capacitance measurements.
-
It analyzes the material using mid-infrared spectroscopy.
The experiments clearly proved that as the recycled content increases, physical changes occur in the material: charge retention increases, permittivity (dielectric constant) decreases, and dielectric loss increases. The researchers analyzed this highly complex, multi-sourced dataset using a machine learning model. The AI learned to recognize hidden data patterns that correlate directly with the percentage of recycled plastic.
Quantified Results: Over 97 Percent Accuracy
During laboratory testing of the developed system, the researchers achieved outstanding quantitative results. The AI-supported model:
-
Was capable of determining the exact recycled content of PET (polyethylene terephthalate) samples with over 97 percent accuracy.
-
The recycled material ratio of the tested samples ranged between 0 percent and 50 percent, which perfectly covers the expected and legally mandated ratios for commercial plastic products.
“Our goal is to create a fast and reliable tool that can verify the recycled material content and enforce recycling regulations,” Professor Amit Goyal stated in the official university report. He added: “This is an ideal example of how cutting-edge innovation in science and engineering can be combined with artificial intelligence for societal good and potentially achieve a significant societal impact.”
The Future: Portable Devices for Real-Time Market Monitoring
The research team is not stopping at laboratory successes. According to Professor Goyal, the goal for the next phase of the project (future work) is to integrate the various sensing technologies and the machine learning model into a single, easy-to-use portable handheld device. The physical production of such a device would allow authorities and companies to conduct widespread, real-time, and on-site monitoring of recycled plastics in commercial products, drastically curbing greenwashing and supporting a true circular economy.
Official Sources and References:
-
Original News Source: New system aims to detect percentage of recycled plastic in plastic products (UBNow, University at Buffalo, March 2026)
-
(Official website of the US state university institution leading the research: UB Initiative on Plastics Recycling and Innovation, a New York State Center of Excellence – buffalo.edu/iprri)