|
Abstract:
|
If the product needs to be inspected from many angles, one camera is usually not sufficient. The traditional solution is to increase the number of cameras and light sources attached to fixed positions on the production line. Inspecting one to three features in this way could be considered an easy task to solve. However, more features than that usually makes this approach far too complex and expensive, or even impossible, as all inspected features still cannot be covered. The camera should be free, with the ability to recalibrate its position and its application for every product.The developed management software controls all the components: robot, camera, light ring and interface with a CAD program. Using an interface with a CAD program such as DELMIA enables creating an offline programming application. The model of the product and the production line are modeled in the CAD environment and then simulated, finally creating the instructions that the real robot will receive for executing the inspection task. Running simulated robot programs in a model environment first prevents any unnecessary risks of collisions and increases the running time of the production line.By using these devices together, it is possible to inspect different features of the product and also different products on the same production line without re-programming the camera or the robot or the need of having different camera fixtures for every product. The common steps that are used in the programming of the system start from simulating robot movements with a CAD program. After the creation of the application in CAD, the task file is sent to the management software that converts the offline instructions into real commands for the robot. The testing of the movements of the robot follows and then the light ring is programmed. As a final step, the intelligent camera is trained to identify the defects in the product, and finally, the application is run. |