In fact, MRF workers are more than twice as likely to be injured on the job as the average worker. Seventeen of them were killed at work between 2011 and 2013 alone, making recycling one of our country’s most dangerous occupations.
But that could be changing. With the rise of artificial intelligence, sophisticated robotics are now turning up at MRF’s across the nation. That’s where cameras and computers guide and train these trash-sorting robots to recognize specific objects.
Here’s how they work: The robots’ arms glide over a moving conveyor belt searching for their target. Then they use their suction cups or over-sized tongs to snag cans, glass, plastic containers, cardboard and other recyclables and toss them into nearby bins. They do this at a rate of about once every second. That’s up to twice as fast as their human counterparts, and with greater accuracy.
Three manufacturers are predominately responsible for the new technology:
AMP Robotics, based in Denver; Bulk Handling Systems (BHS) of Eugene, Ore.; and ZenRobotics of Helsinki, Finland.
And not a moment too soon.
Last January, China (the world’s leading importer of recyclable materials) stopped accepting our trash due to contamination concerns. (See related article, “Exporting Our Trash: Who Will Take It?“) This sent recycling companies scrambling for solutions.
With robotics technology, recyclables are cleaner and better sorted. Less risk of contamination means more viable markets for the recycled materials. As explained in Waste Management World, “Intelligent robotic systems can process almost any given waste stream and sorting capabilities can be redefined for every new market situation — even on a daily basis.”
Combine this with the aforementioned worker safety concerns, and robotics appear to be the perfect answer.
The computer algorithms embedded in this robotics technology utilize what’s known as “deep learning” to improve garbage sorting. Unlike standard robotics used in manufacturing (which are programmed to perform mundane, repetitive tasks), the recycling robots employ a unique “visioning system” in their software.
For instance, take AMP Robotics and its trash-sorting robot called Cortex. According to company CEO Matanya Horowitz, each Cortex is trained by showing it thousands of examples of bottles, cans, packages, and other items. “It learns to identify all of these materials on its own,” Horowitz said. It does this by looking for “logos, shapes, and textures.”
While most optical sorters used at MRF’s can identify one material, Cortex can pick multiple materials. The robot detects a material’s location on the conveyor belt and intercepts it using grippers, and deposits it into the appropriate bin.
AMP Robotics has received grants from the National Science Foundation, the Closed Loop Foundation, and the Carton Council. The company’s Cortex systems are now in use at three U.S. recycling plants.
Finland-based ZenRobotics has been around more than 10 years. Its systems incorporate robotic sorters with sensors and arms for specific applications, such as construction and demolition (C&D) waste. (According to the EPA, the vast majority of U.S. waste is comprised of C&D debris.)
ZenRobotics’ Recycler creates a real-time analysis of the waste stream using multiple sensors. Then, based on the analysis, the Recycler makes autonomous decisions on which objects to pick and where to place them.
With C&D waste, for instance, the robot sorts heavy metal, wood and stone fragments. In fact, the company’s largest robot, called Heavy Picker, can lift objects weighing up to 60 lbs. And its gripper can accommodate materials up to 20 inches wide.
Since it began, ZenRobotics has installed its AI recycling system at two U.S. locations and in 10 other countries, including Australia and Japan.
Bulk Handling Systems (BHS) makes a garbage-sorting robot called Max-AI. The spidery-armed robot uses a suction cup as a grabber. And it’s able to make multiple sorting decisions autonomously.
For example, it can separate thermoform trays, aluminum, and paper while also removing residue from a stream of PET bottles. All at superhuman speed.
According to Thomas Brooks, BHS director of technology development, “Max is more than just a robotic sorter. Max-AI technology will soon become the active brain of our MRF’s, controlling various robotic, optical, and other sorting equipment, providing real-time material composition analysis, and making autonomous decisions.”
It can also integrate seamlessly with an MRF’s existing optical sorters. Company officials anticipate that this technology will lead to a decrease in plant operating costs and downtime, and an increase in throughput, recovery and purity.
So far, three sites in the U.S. and three sites in Europe are using Max-AI, with more robots on order.
But what about the workers? Should they be concerned about machines replacing them?
Probably not. Recycling robotics are currently appropriating only the dirtiest and most dangerous jobs on the sorting lines. These are the positions that are the hardest to fill and keep filled.
“Labor is a significant challenge for MRF operators,” explained BHS CEO Steve Miller. He believes robotics technology “will be very beneficial in helping our customers manage that aspect of their business.”
And ZenRobotics CEO Timo Taalas even goes a step further. He speculates that the increased efficiency at MRF’s may actually create more jobs at paper mills, plastic recyclers, and other firms that reuse raw materials.