Q&A With Environmental Specialist Cliff Walls: AI Recycling Program Wraps Up
In September 2022, East Lansing announced a test program that would use cameras in recycling trucks to identify contaminants in curbside recycling and help educate residents about good and bad recycling habits.
The City partnered with The Recycling Partnership, the Michigan Department of Environment, Great Lakes and Energy (EGLE) and Prairie Robotics to launch the first pilot municipal campaign in the United States, using artificial intelligence (AI) to try to cut down on recycling contamination.
Cameras were installed in recycling trucks to look into curbside bins for 24 weeks, monitoring residents recycling and sending them a postcard if a contaminant was identified. The postcard also included information about recyclable materials and links to the City’s education resources.
With the program wrapping up and Ohio State University finalizing their findings of the study, comparing the AI approach to traditional audits, ELi followed up in an email with East Lansing Environmental Specialist Cliff Walls about the results of the program and the city’s next steps.

Why did the City choose to participate in this study?
“The yearly influx of new students and residents brings with it a diversity of recycling habits and know-how, often habits that aren’t consistent with the nuances of our program and our processor’s rules. We had already been exploring ways to better engage and educate residents at the household level, as opposed to broad education campaigns, when a chance to partner with EGLE, The Recycling Partnership, and Prairie Robotics on the pilot emerged. Having already witnessed how AI has streamlined processing facilities, we knew curbside innovation was next. The chance to test the technology here, at no cost to taxpayers, was an obvious decision.”
How does the technology work?
“The technology uses Artificial Intelligence that has been trained on millions of unique data points to be able to identify contaminants – items that aren’t accepted for curbside recycling, such as plastic bags, yard waste, and Styrofoam – in the recycling cart. This occurs “on-truck” once the carts are emptied into the hopper. GPS technology determined the address of the cart, and if contamination was found, a postcard was generated and mailed to the address with information about acceptable materials and links to the City’s educational resources, like the Recycle Coach app. While City sanitation have long been equipped with cameras to alert drivers of dangerous or damaging materials, this new, on-truck technology enables the City to provide answers to common recycling questions as they occurred and educational material where they had the most impact.
“The project was a modified version of The Recycling Partnership’s ‘Feet on the Street’ cart tagging recycling program – an initiative to improve the quality of recycling in curbside recycling carts by providing residents personalized and real-time education and feedback. Traditionally, this is done by someone reviewing the contents of a curbside cart and tagging it with educational stickers if contaminants are found and repeating manual inspections over the course of weeks or months. Many recycling programs across the nation utilize these manual audits, and AI presents an opportunity to replace these labor intensive, time-limited, and weather-dependent operations with on-board tech.”

How frequent were mistakes?
“Mistakes/misidentifications are seen less than 0.5% of the time. Also, if the computer wasn’t certain of the source or address of the contaminants, no postcard was mailed.”
What are some of the findings from the Ohio State University study?
“Ohio State University was engaged to study the efficacy of the AI approach compared to traditional, manual audits. While the findings have yet to be formally published, they found that the approach utilized by East Lansing was just as effective in reducing contamination manual efforts. East Lansing households that received a postcard with personalized feedback contaminated 23% less than the control group. The project also resulted in increased participation, with households receiving postcards increasing their cart set-out rates by 45%. The educational feedback improved both the quality of the recycling stream and household participation in the recycling program.”
What are the takeaways for the City from these findings?
“The findings support what I anticipated: That personalized, targeted feedback is more impactful than broad education campaigns. Recycling can be confusing. Every city and processor has its own rules and accepted materials. Manufacturers ‘green wash’ products by claiming they are universally recyclable when they aren’t. We see news reports that recycling is ‘broken’. Addressing the curbside confusion needs to happen at the curb.”

Prior to the study, what was the proportion of bins that had contaminated contents? After the program?
“Contamination was quantified in the recycling stream as a whole and as instances of contamination found. We saw a 25% reduction in contamination following the project.”
What were the most common contaminants?
“The most common contaminants were bagged material, polystyrene (Styrofoam) and plastic film.”
How many households were notified about the presence of contaminants?
“Over 5,000 postcards were sent to households with targeted feedback.”
With this study completed, what are the City’s plans moving forward?
“The City is exploring how we can keep the momentum going following this project and considering continuing our use of the technology to both monitor the quality of our recycling stream and educate residents. The project has certainly helped highlight the materials residents need the most help with, which informs our outreach strategies moving forward.”
As a result of the project, the City of East Lansing was named a top three finalist for National Program of the Year by Resource Recycling, Inc. at the nation’s largest municipal recycling conference.