Optical sorters

Optical sorters sit wherever a line needs to take out what people used to pick by hand: rotten, green, bruised or discoloured pieces, foreign material such as stones, glass, plastic, wood and insects, and product with the wrong colour, size or shape. On a potato line they work on raw tubers before peeling, on peeled potatoes and on cut strips before the fryer. In nut, grain and seed plants they take out discoloured kernels, shells and stones. In frozen vegetable and fruit plants they remove stalks, leaves, foreign bodies and defects before freezing or packing.

There are two main designs. Belt sorters carry product flat on a fast conveyor and view it as it flies off the end, which suits wet, sticky, large or fragile pieces such as potatoes, fries, vegetables and fruit. Chute or freefall sorters let product slide down a channel or fall through the air in a thin curtain past the cameras, which suits dry, free-flowing product such as rice, grains, seeds, beans, nuts and dried fruit. Lane sorters and camera graders hold each piece in its own lane or cup, turn it and photograph it from several sides, then grade it by size, shape and defects into several outlets, which suits whole potatoes, onions, fruit and other produce sold by grade.

The sensors decide what a sorter can see. Colour cameras see colour and visible spots. Lasers see structure and can tell foreign material from product of the same colour. Near-infrared and hyperspectral cameras see differences in chemistry, such as sugar or moisture, that the eye cannot. X-ray finds dense foreign bodies and some internal defects. Many sorters now add shape analysis and machine learning trained on images of the customer's own product.

Cost depends on the width of the machine and its capacity, the number and kind of sensors, the number of ejectors and output streams, the hygienic design, and the software. Running cost depends mostly on compressed air, cleaning time, and the good product lost with the bad.