When testing material, recyclers are responsible for proper material preparation. “We recommend that customers pretreat the material with upstream processing equipment prior to sending it to TOMRA, so it will be similar to what our equipment will sort in the plant,” says Rhoad.
Plant builders like WENDT CORPORATION and VAN DYK often offer more upstream processing equipment as part of their test facilities, so less material preparation may be required. “In addition to TOMRA machines, Van Dyk’s test center includes an elliptical screen along with other mechanical 2D and 3D material separators,” says Wolf. For processing end-of-life vehicles, “our test lab includes magnets, screens and eddy currents in addition to the TOMRA FINDER, X-TRACT and COMBISENSE,” adds WENDT CORPORATION’s Close.
Regardless of where testing occurs, all recommend in-person visits to the facility when their material is running. “It is beneficial for customers to see their material being tested and how the different technologies are employed to help reach their goals,” says Lehmann. Close adds, “We know it works. It’s best if the customer can see, learn and understand how the testing works.”
TOMRA process engineers examine the material and consider the customer’s objectives to determine the necessary technology or blend of technologies required. They measure production capacity, determine air demands, and evaluate the material recovery and purity. Generated reports consist of flow charts with the recommended sorting technologies and the required steps to reach goals.
“Depending on material complexity and objectives, it can take as little as a day or up to a week or longer to conduct testing,” explains Lehmann. On average, it takes between 8 to 10 weeks from first meetings to test completion and reporting.
Push to automate
As the industry advances automation, testing is key to success. “We are always pushing the boundaries, and we now have plants that sort to furnace-ready aluminum,” says Close. “There is a vision of full plant automation with workers in a control room that receive and react to data, where quality control hand picking is handled by robots.”
Deep learning and artificial intelligence are assisting the industry to advance this sorting automation. “With TOMRA GAIN, we use deep learning to sort more complex material fractions, and we can tie the entire system – or multiple systems from different locations – together with TOMRA Insight data reporting,” mentions Rhoad.
Wolf agrees, adding that the industry has always set its sight on reducing labor through more automation. However, he sees limitations with today’s technology in reaching the fully autonomous plant. “Full automation,” he mentions, “drives up the investment, which may negatively impact the ROI of the overall project. We have seen this with the advent of robots. They are not the silver-bullet answer to everything as originally thought, and it’s all application dependent.”