The acquisition of certain process data is done in a visual way. For example, a change in color, surface condition, shape, positioning on a conveyor, a heat map…
While some of this data can be processed in a simple way by a color sensor and an algorithm, some requires more complex acquisition and processing. Is the paint applied to the entire piece? Is the baking done? Are the consumables too worn?
Under these conditions, it is possible to mobilize a neural network, analyze the images and other data, and then make the right decision for the next step.
Another advantage of vision is that it reduces the need for expensive industrial equipment for data acquisition. While the use of a color sensor, a probe, requires guiding the part precisely underneath, vision is able to acquire this information in-process with no more installation than a camera.