Accelerating AI

Like many of you in the waste industry, the Waste Today team traveled to Las Vegas to attend WasteExpo 2025.

At this year’s expo, I was awed by the rapid improvements in artificial intelligence (AI) and its evolving applications in the waste industry. Just a year ago, most AI sessions were focused on data analytics, but this year’s conference sessions tackled everything from leveraging AI to create safer work environments to automating the reporting process to help scale food recovery.

One standout session was Unlocking the Value in Municipal Solid Waste with AI-Powered Next-Gen MRF/MWPF Practices, with panelists including Matanya Horowitz, founder of Amp, and Riel Johnson, the former vice president of resource recovery for Athens Services.

In discussing the emergence of new technologies and cutting-edge practices for municipal solid waste (MSW) management, Johnson said he saw the potential for AI to perform a waste assessment as materials come into a material recovery facility (MRF) and then fine-tune the facility in real time to optimize material recovery.

Horowitz dialed in on AI’s potential uses in mixed-waste processing to help these facilities compete with landfills.

“Mixed-waste processing, as we see it, is a huge opportunity in the space,” Horowitz said. “AI can change the way you go after this problem.”

In these facilities, AI could replace traditional sorters to reduce costs and increase consistency. AI also could offset the need for entire classes of more conventional recycling equipment, Horowitz said, including those machines that are known to cause downtime.

Using screening technologies augmented with AI-powered optical sorters, Horowitz envisions systems that can go through MSW, segregating the organics and recyclables.

“Retrofitting MRFs with the technology, we can get some good, incremental improvements,” Horowitz said of Amp’s technology. “But when you build it in from the ground up with a greenfield facility, we seem to have really tremendous gains, and that’s coming from the obvious places like reduction in manual sorting and such but [also] significant increases in the uptime.”

With those increases in uptime, the economics of integrating AI really start to change, he explained. These new AI-powered recovery opportunities are meant to supplement landfills rather than compete with them, identifying regions where landfills are scarce resources and there’s value in extending their lifetime and offsetting transportation costs.

“The opportunity we see is there’s lots of good stuff still in garbage,” Horowitz said. “There’s enough recyclables in the MSW that I can make very successful programs without touching the single-stream systems.”