The camera can generate a 3D image of the contents of a volumetric area equivalent to a US or Euro pallet. It helps to pick or items from pallets or trays, and to check that the items have been removed.
The camera incorporates a fixed laser and a rotating mirror. It sweep-scans the volume to produce a 3D image of all of the parts including their x/y/z co-ordinates in millimetres, as well as a 2D overlay that enables the parts to be identified against known profiles.
The combined image is processed by image-handling software to tell a robot arm where to start gripping, starting with the item on top of the pile. The arm grips the item at its optimum position and deposits it, for example, on a tray or conveyor.
With a built-in laser light source, the technology is not affected by changes in ambient lighting and provides reliable measurements of parts with differing surface characteristics and colours. There is no need for external light sources, encoders or extra imaging equipment, allowing integrators and machine-builders to design competitive, flexible installations.
Sick's ScanningRuler 3D camera helps robots to pick large items from bins
“The ScanningRuler offers robotic integrators additional speed, versatility and functionality in solving heavy picking duties, such as automotive parts,” explains Neil Sandhu, Sick’s application specialist for imaging and measurement in the UK. “Up to now this task could only be done by having a fixed camera and moving the heavy bin under it – or by moving the camera, which shortens the camera life. Often the only effective solution has been using human labour, risking problems of repetitive stress and strain.”
The ScanningRuler is integrated with robot control networks via a Gigabit Ethernet link to a PC controller. The IP65-protected device has a 1–2m-wide field of view and can scan a volume of up to 1 x 1.2 x 0.7m.
