At machine-builder Phasor (Sunderland, England), four separate vision systems have been integrated into one fully automated machine to inspect the surface of oil seals for shock absorbers. The rubber-molded seals are checked for scratches, marks and surface blemishes, inclusions, and imperfections.
Peter Hage, Phasor's head of design, developed the system using the Omron (Schaumberg, IL) F150-3 vision sensor because of the device's surface-defect algorithm. In its final form, Phasor's machine feeds individual seals along an input conveyor to a pneumatic arm that transfers them one at a time to a rotating spindle, which forms the first of two inspection stations. Here, three F150-3 cameras focus on the main seal lip, the main body, and the side of the seal.
Each camera views a 30° segment of the seal, and the spindle rotates around in discrete steps, so that the whole of the seal is viewed in a series of 12 images. The F150-3 uses pixel-counting techniques to detect imperfections of the seal in each of the 36 different images. Each image is analyzed in 110 ms.
Next, a second pneumatic transfer mechanism inverts the seal onto another 12-step rotating spindle. A single camera searches the underside of the seal for surface blemishes. The machine is set so that both spindles are loaded and analyzing seals simultaneously, achieving a cycle time of less than 4 s per seal. This requires analyzing 48 different images in 4 s.
So that each seal is associated with the correct 48 images, Omron's Vision Composer marries image and object data. Four F150-3s are controlled by an Omron CPM1 microcontroller. Image-analysis data from the cameras are fed to an Omron CQM1 programmable logic controller that controls safety switches, conveyor drives, and the systems' pneumatics. Omron's CQM1 also controls a pneumatically powered reject gate to divert any defective seals to a recycle bin.
