On an indexing rotary table with a diameter of 5.6 m and 12 nests, up to 8 motor vehicle window panes are presented to the vision system in a measuring cabin.
Through the PLC the vision system receives information about the type of window pane currently lying at the measuring position and identifies with appropriate measuring methods the sub-type of the pane.
6 cameras and 6 switched LED light sources help to measure the X-Y position as well as the dislocation of the window pane on the work piece holder. Through a serial interface four robots request and receive the measuring results.
Using these coordinates they clean the window panes and put clips on certain pane positions.
The results derive from a calculation of the measuring test results, the nest-offsets and the single robot-offsets.
In case of bad results none of the robots will work on the pane.
For calibration of the whole system, one work piece holder is equipped with marked pieces. The screen positions of the single calibration marks need to be taught manually.
Finally the position of the pane on the work piece holder is determined by the rubber edges, which can be identified by the vision system.
Since each single nest and each robot is calibrated, an absolute relation to the robot coordinates can be determined.
Each of the eight pane types need a separate program and therefore also a separate camera/ lighting combination.
