Let’s take the example of a carabiner on which we would like to perform a simple check. Press the latch and make sure it opens.
If there is sometimes the possibility to create a tooling that will wedge the part in the right position, computer vision is an efficient way to quickly identify its position. From there, it is possible to guide a gripper to perform the test.
The same computer vision system can then verify that the part behaves as expected during the functional test.
Finally, during the same pass, another algorithm can perform a visual check of the part’s appearance.
In a single operation, all manual and visual controls can be carried out, at very high speeds. This makes it possible to go as far as a unitary control of the parts at lower costs, with an important benefit on the quality, whether it is during production or on the delivery to the customer.