Orlando, Fla. — With all the talk about artificial intelligence, Tomra Recycling wants to make one thing perfectly clear: AI has been around for decades and has been used in company products that help recycled plastics sortation.
Software and algorithms have allowed companies to use optical sortation to handle and separate large amounts of plastics far faster than humans could ever hope to accomplish.
What's new is what Tomra calls is the deep learning aspect of AI that's now allowing for increased efficiencies and applications.
"The industry has seen significant improvements in traditional optical sorter sensors and sorting algorithms to provide a more granular sort of plastic feed material and improved sorting accuracy, high throughput and better yields for recovered materials," said Ty Rhoad, vice president of sales Americas for Tomra Recycling, in an email interview.
"Plus, the industry has seen the genesis of deep learning-based artificial intelligence sorting, which is helping to close the recycling loop with some plastic materials," Rhoad said.
Tomra showcased its GAINnext technology at NPE2024 in Orlando.
"GAINnext leverages AI's deep learning subset, the current game-changing technology for the plastics recycling industry, that further advances sorting accuracy and adaptability to material streams. Using thousands of images that hold a pool of object information, large datasets of trained neuronal networks enable deep learning materials to recover materials that are difficult or even impossible to distinguish using traditional technologies," Rhoad said. "The network recognizes patterns in the data and connects this information pool to the sorting task."
Rhoad said it is important to define what AI is for the recycling industry.
"AI consists of software and algorithms, not the hardware. Traditional AI has been used in optical sorters for decades in the recycling industry. It is used in combination with a variety of sensors — RGB cameras, near-infrared (NIR), laser, electromagnetic (EM) and X-ray — to automatically detect and process plastic fractions. Traditional AI is a manual process whereby a technician engineers the classifiers and deploys it into the in-house developed sorting software," Rhoad explained.
Tomra launched what the company calls five new sorting applications for GAINnext to be used with its Autosort optical sorter. Three of the applications include separating food-grade from nonfood-grade PET, polypropylene and high density polyethylene. Another GAINnext nonfood-grade sorting application is aimed to high purity PET bottle streams.
"The true power comes from combining deep learning AI of GAINnext with traditional optical sorter sensors. With this power and the ability to have specific applications, you can sort to a higher purity and differentiate materials previously only capable before by a human eye — for example, food-grade vs. nonfood-grade material," he said.
"When applied correctly, today's emerging AI technologies have transformative powers, driving automation and efficiently enabling more granular sorting of complex material fractions. But they can also be a waste of both time and financial investment if not considered as part of a holistic solution," Rhoad said. "Deep learning advancements in AI are application-specific, so the technology supplier must provide the recycler with material testing, training and optimization. The supplier cannot simply drop AI-based equipment into the sorting circuit and leave. To leverage the full potential of deep learning-based AI sorting, it must be through a well optimized and positioned machine that workers train on how to get the most out of it."
It's been six years since anyone has exhibited at NPE, and Tomra has seen some changes over time.
"The biggest change, especially in the Americas, has been the infrastructure investments in the recycling industry. With the improvements of upgrading the technology and overall expansion, it has been widely documented that companies have made significant investments in new infrastructure to support the evolving market," Rhoad said.
Tomra Recycling
S35185