Skip to main content

In the Loop

The technological evolution of materials recycling

It’s time to update how we sort valuable materials (a.k.a. waste).

Sims materials recovery facility in Brooklyn, New York
Recycling heaps at the Sims Municipal Recycling Center in Sunset Park, Brooklyn, New York. Image via Shutterstock/Felix Lipov

Have you been to a materials recovery facility (MRF)? I’ve had the pleasure of touring a couple, and the thing that sticks out about my experiences thus far is not the high-tech sorting machinery.

In fact, because I haven’t spent enough time in a MRF to fully comprehend all the material movements, I would best describe it as something akin to a Rube Goldberg machine. Materials seem to be moving in every direction, crisscrossing, dropping off cliffs and moving up belts. When you break it all down, I know that it is nothing like a Rube Goldberg machine, but that’s the first thought that still comes to mind for me.









Because of the massive amount of material that flows through the average MRF and the very real implications of missing valuable materials in the sorting process, a whole industry of innovators has popped up over the last decade to help MRFs operate more efficiently. 

You’ve probably heard about some of these companies. There’s Amp Robotics and Tomra, both working to better sort materials with their technologies (and both covered in this GreenBiz piece from 2020). There is MachineX, a manufacturer of a broad suite of sorting equipment for MRFs. ZenRobotics makes robots for sorting. There are also a number of companies that develop equipment to optically sort waste, including Blue Green Vision and Recycleeye. In other words, this is a space growing quickly with new solutions and technologies coming to market seemingly every day.

To try to understand this universe a bit better, I sat down with JD Ambati, founder and CEO of EverestLabs, to talk about the company’s solution and what is really needed to unlock the promise of extracting valuable materials from the vast amount of waste we create. 

A quick aside here: Ambati, who has a background in artificial intelligence and chemical engineering, built a successful 17-year career commercializing technology products for Fortune 1,000 companies. When I asked him about the hard left he took from that world to founding a company dedicated to MRF efficiency, he said that in many ways, these things are the same:

"My previous roles have always been working with companies, listening to them, listening to their problems, listening to their goals and then providing solutions. I applied that to the world of recycling by talking to MRF operators, brands, talking to municipalities and such. In a way, they are really the same, it is just understanding the problems and issues, and providing guidance."   

With that in mind, let’s dive into the challenges that MRFs currently face in sorting valuable materials from waste, and how technologies such as those developed by EverestLabs can help.

Addressing the MRF data gap

There are a couple of critical data points that measure the efficiency of any materials recovery facility. Specifically, these are the quantity of materials in and the sorted quantity of material out.

Historically, MRFs have operated mostly with only these two metrics, leaving a number of important data points within the plant, as well as those upstream and downstream from it, unknown. What haven’t we known enough about? Information about what type of recyclables are slipping through the cracks in sortation, how efficient current sorting equipment is operating versus expectations, percentages of each recyclable material type that are being sorted properly, and even what form factors (and from what brands) are not able to be sorted. These data gaps leave space for new technologies such as machine learning to step in and not only help a facility run more efficiently but also to be able to provide valuable feedback to producers to improve their packaging for recycling. 

Ambati suggested that major innovations in waste management (drum feeders, density separation and eddy current separation) really stopped in the 1990s, even amid all the innovation happening in computing. Because of that, MRFs and other waste management facilities are overdue for an influx in new technology. In general, it is true that MRF operators don’t wake up in the morning with the goal of sending more waste to landfill. They want to do the right thing and save as much material as possible. The issue is that they need the tech that will solve the efficiency problems they face, with an acceptable return on investment, that’s easy to use and that fits within their current footprint. 

Here’s the crux of the matter: In order to capture more material, MRF operators need actionable insight about the data gaps mentioned above. In other words, as Ambati put it, MRF operators must "have clarity between the major data points they already have."

The mission of EverestLabs

Self-described as the first AI-enabled operating system for recycling, EverestLabs raised $16.1 million in Series A funding last summer led by Translink Capital. According to its press release, "the funding enables the company to invest in its scaling and go-to-market capabilities."

Ambati said the EverestLabs technology can enable MRF operators to fill data gaps, increase sorting efficiency and ultimately get more value from the material running through their facilities. EverestLabs’ secret sauce is the software engine that the company has built from the ground up. Another potential advantage is that the EverestLabs software is material-agnostic and can be applied to organic materials sorting, construction and demolition waste facilities and other activities, according to Ambati. 

No silver bullets

While no one solution will unlock circularity, it is no secret that better sortation and recycling will be a key part of the future solution set. I am excited about all the new technology flowing into the material handling sector and that some of the largest facilities in the world are using these technologies to better handle valuable materials. If the largest sorting facilities can effectively implement machine learning and robotics technologies to increase efficiency, it can create a blueprint for other facilities to follow suit. 

This could be a very active space in the coming years as both machine learning and applied technologies such as robotics come to scale. These downstream sortation improvements can empower MRF operators to increase valuable material recovery while lowering operating costs to improve recycling rates. This, coupled with upstream interventions to reduce single-use materials, are both crucial pieces of the circular economy transition.

[Interested in learning more about the circular economy? Subscribe to our free Circularity Weekly newsletter.]

More on this topic