Starting out with a treadmill, a camera and a pile of waste, two entrepreneurs - engineers Victor Dewulf and Peter Hedley – are commercialising their intelligent waste sorting system via Recycleye, a business that has raised millions of euros in funding.
The system is made up of waste recognition and 6-axis robotic sorting solutions that can be used separately or in tandem. Recycleye CTO Hedley explained that Recycleye Vision uses a low-cost camera with RGB sensors to identify each waste item on a conveyor belt using the same visual characteristics as the human eye.
“It achieves this through the machine learning algorithms which power the system, involving models being trained on thousands of images of waste on belts, meaning the system learns, like a child, to recognise items by the correct category,” he said via email. “The vision system operates at 60 frames per second and classifies every item on the belt 120 [times] to increase certainty.”
He added that the optimum pick sequence and coordinates of detected objects are sent from Recycleye Vision to Recycleye Robotics via Microsoft Azure IoT Edge. The robot’s pneumatically controlled gripper then picks up an item, rotates towards an appropriate bin and shoots the item into it. It can do this up to 55 times per minute.
Hedley said the robot contains a pressure sensor to avoid wasting time on empty and unsuccessful picks and that the pneumatics system is designed so that the gripper is the narrowest diameter that incoming items will travel through inside the robot.
“There is minimal chance of blockage because any object that is sucked up through the gripper will be able to travel through the whole system to the filter,” said Hedley. “There is also a blockage detector system to indicate when assistance is required.”
Founded in 2019, the company’s robotic solution currently generates the most revenue. The robotic picker is 75 per cent lighter than any comparable solution on the market, so it can be installed over a weekend during plant downtime at no additional cost.
“Before developing our own robotic picker with FANUC, specifically for dry mixed recyclables sorting, we heard from clients that existing DMR sorting robots were large and heavy, meaning they faced installation costs nearly as expensive as purchasing costs,” said Hedley.
The potential for vision capabilities in the recycling industry could, however, see Recycleye Vision pull ahead of its robotics solution in the next five-to-ten years,
“Our AI’s ability to detect to object level, for example between an aluminium can and aluminium aerosol, means it could open a new market for pure valuable food-grade packaging bales, which would be highly profitable for MRFs [Materials Recycling Facilities],” said Hedley. “Meanwhile, Recycleye Vision also has the potential to be integrated with other sorting hardware beyond robotics.”
EPO President António Campinos believes the waste recognition and sorting solutions are making a vital contribution to reducing the world’s waste and moving towards a circular economy.
“The speed at which they have not just developed these innovations, but also turned them into reality, is remarkable and we look forward to seeing their story unfold,” said Campinos.
The Young Inventors prize recognises innovators aged 30 or under who have developed solutions to tackle global problems and help reach the United Nations Sustainable Development Goals. The winners will be announced at the European Inventor Award virtual ceremony on 21 June.
Five ways to prepare for your first day
If I may add my own personal Tip No. 6 it goes something like this: From time to time a more senior member of staff will start explaining something...