Not Your Parents' Trash: How AI is Revolutionizing the Solid Waste Sector

Raj opened the discussion, explaining that “artificial intelligence is the field of science concerned with building computers and machines that can reason, learn, and act in a way that would normally require human intelligence.” AI is already part of our daily lives, from text predictions on our phones to Netflix recommendations and facial recognition. The focus now is how AI can make recycling more efficient, increase recycling rates, and enhance safety for workers. 

Recycling has its obstacles, with contamination alone costing the US approximately $300 million in operational expenses annually. One promising AI-driven solution to tackle this is a smart container known as the “trash bot.” This container scans and weighs items to determine their appropriate handling based on composition. According to Musetti, the trash bot learns over time: “If a water bottle that is full of liquid is entered (into the machine), it’s able to ask the question, how should this be disposed? So you’re able to kind of program it to keep up with you and how things are emerging.” During a thirteen-week trial at a US airport, the trash bot sorted 2,500 items with approximately 96% accuracy—three times higher than the national average recycling rate. 

Fleming highlighted safety as another crucial area for improvement. “In 2022, recycling facilities alone had an injury and illness rate of 4.4 for every 100 full-time employees. AMP states that their technology, the Vac Tech, can remove film more efficiently and can replace five human sorters.” By automating sorting with AI using a robotic arm, facilities can reduce workers' exposure to potentially dangerous materials. 

AI is also making waste collection more efficient through route-tracking, a practice that optimizes route schedules for the day and ensures they are not repeated. For instance, “the City of Glendale, Arizona, did this and…saved about 155 miles per day, or 24,000 miles annually” by balancing their workload and standardizing route schedules with RubiconSmartCity. Additionally, existing containers don’t need to be replaced to utilize AI. Sensors can be added to containers to monitor fill levels, optimizing collection schedules and avoiding unnecessary pickups. “Compactors and open-tops are typically charged by weight, but front loads are typically charged per haul. By putting a sensor into this container, you’re able to use it as kind of compactor gauge,” Musetti noted.  

Another advantage of AI is its flexibility across platforms. AI can “identify anomalies or inefficiencies,” aid in scheduling preventative maintenance, and adjust settings based on variables like weather or gas composition. Facilities can even use AI to create a digital twin—a virtual model that mirrors real-world operations. With a VR headset, operators can walk through digital replicas of solid waste and resource management facilities to conduct training and safety drills, all without having to physically enter the building. Ariens concluded, “Digital twins are going to be instrumental in enhancing operational efficiency and safety.” 

A common concern is that AI might lead to job loss by replacing human roles. The panelists addressed this, noting that the impact would likely be most significant in undesirable and high-risk jobs, potentially improving safety and job quality. Ultimately, the introduction of AI in waste management offers significant benefits, from cost savings and improved safety to increased recycling rates. SWANA is optimistic about AI’s potential to enhance predictive maintenance, boost recycling, and streamline operations in the years ahead. 

Disclaimer: As noted during the presentation, any and all companies mentioned are referenced for data purposes only and not as endorsements.