AI tech provides new insights on food-grade polypropylene recycling streams

A new study has been published by Closed Loop Partners that reveals high volumes of recycled food-grade polypropylene (PP) are successfully captured at material recovery facilities (MRFs). 

The study was led by the Closed Loop Foundation and Closed Loop Partners' Center for the Circular Economy, collaborating with technology company Greyparrot AI and four U.S. MRFs. The groups leveraged AI-powered vision systems to characterize the PP recycling stream with unprecedented detail. These results fill data gaps on the availability of food-grade PP, which can create new opportunities to return material to food service packaging supply chains.

The findings have been released amidst growing demand for recycled food-grade PP, driven by policy shifts — including recycled content mandates and Extended Producer Responsibility laws — and commitments from brands to incorporate more recycled materials in their in-store and shipped packaging. Despite increased market demand, there has been a significant lack of data on the available volumes of recycled food-grade PP in the recycling system. With no easy way to track, differentiate, and separate food-grade and non-food-grade PP, these materials typically blend together at MRFs, making it challenging to amass the appropriate quantities of food-grade PP to meet growing market demand.

With funding from the NextGen Consortium and the Closed Loop Foundation, the Center for the Circular Economy teamed up with Greyparrot and four MRFs to reveal what was in the PP recycling stream, including the volume of food-grade and non-food-grade items, as well as colour, format, and other critical identifying features. Nearly 45 million individual PP and non-PP objects were characterized over the course of the study, revealing newfound granular details on the PP comprising the stream.

The study revealed 3 key findings:

• Clear and white food-grade PP is the most common: On average, more than 75 percent of the PP captured in the study was white or clear, most of which was also presumed to be food-grade. Also, over 30 percent of clear PP packaging identified were beverage cups. This finding has important implications for meeting growing food-grade PP demand and how to retain more value in the system.
• AI-enabled technologies can reliably quantify and classify recyclables with granularity, at scale: AI systems, such as the Greyparrot Analyzer, proved reliable in providing effective material characterization data at previously unavailable scales. This solution indicates the potential for AI to drive value to MRFs through increased intelligence and data granularity on material flows.
• AI can help measure and track facility and equipment performance: Upgrades to optical sortation technology at MRFs had a notable impact on improved material sortation. This was progress that the AI technology was able to track and provide critical analytics on, indicating the potential for AI to offer enhanced performance evaluation data for MRF operators.

A collaborative approach to improve recovery rates 

While the characterization study focused on uncovering better data at one stop of the PP recycling value chain, the results underscore that strong recycling data can improve recovery opportunities and create new value up and down the recycling system.

These insights lay the groundwork for additional studies that can generate new, valuable recovery opportunities further downstream. Moving forward, Closed Loop Partners' Center for the Circular Economy will continue to identify opportunities to leverage new data and innovations to advance material recovery with the recycling industry and global retail and food service brands.

"The data captured demonstrates what is possible for the future of recycling and circular materials management when powered by technology that can enhance transparency in the recycling system and increase high-quality material recovery," said Kate Daly, managing partner and head of the Center for the Circular Economy at Closed Loop Partners. "As we continue our work with many of the world's largest retailers and foodservice brands, we look forward to identifying more opportunities to pull valuable food-grade materials back into foodservice packaging supply chains — a critical step toward recycled content goals and packaging circularity."

"This work provides important data and transparency around the performance of AI technology and its capabilities within MRFs. We are proud to contribute critical data on the presence and quantity of food-grade objects within the PP stream," said Gaspard Duthilleul, COO of Greyparrot. "In just three months, Greyparrot Analyzers characterized over 45 million PP and non-PP materials — a process that would take nearly four years manually, as manually characterizing just 1,000 pounds of material can take an entire day. This scale uncovers new opportunities for data collection at recycling facilities, serving as the foundation for increased recovery of valuable materials."