Machine Vision Waste Sorting
Maciej Bledowski @ Maciej Bledowski
In a waste sorting facility, advanced cameras and computer systems visually identify and target specific objects based on colors, shapes, textures, and logos, to guide robotic arms over conveyor belts carrying waste. Robotic attachments with sensors are attached to the arms to pick out recyclable items and place them in respective bins at speed, using vacuum grippers. Machine vision waste sorting robots can accurately target and process a vast range of objects, regardless of their condition, picking up to 6,000 pieces per hour.
Machine vision and deep learning allow decision-making to be automated, with the technology becoming more accurate over time. Separate robots can be deployed to fulfill different functions, for example, sorting materials and quality control. Some human oversight is still required, which can be done from a control room.