First and foremost, the quality of the product is increased.
Sample testing can often be replaced by 100 percent quality checks. In the example of paper production this means that every single square inch of paper produced has been reliably checked for flaws "on the fly". The result is a superior product. The same applies to the printing of patterns on textiles or the production of sheet metals: The manufacturer guarantees a 100 percent perfect delivery, which is especially important if products are safety-critical.
Secondly, machine vision can lead to significant cost reductions.
Often, vision systems are employed in the early production stages. Defective parts are immediately removed from the manufacturing process and not finished. In many cases the removed part can be re-introduced into the production process. This saves materials. Defective parts never continue on to subsequent manufacturing stages and therefore incur no further costs. At the same time the system may become "self-learning" in that it recognizes recurrent defects. This statistical information can be fed back into the process to systematically rectify the problem at the point where it originates, resulting in increased system productivity and availability.
Machine vision technology is unique in its ability to resolve the trade-off between raising quality and cutting costs. Examples abound in which machine vision does both jobs at the same time.