Nowadays, spinning requires a much larger knowledge of the raw materials and process, in a sufficient detail to develop mechanisms that will manipulate the "input" without human intervention.

With the spread use of HVI/HVT Systems (in the fibers and yarns quality field) and the arrival of artificial intelligence, patterns of binary data may be analyzed to detect deviations from patterns produced under "statistical control" and making recommendations concerning the root cause of the problem and the appropriate "corrective" action.

Expert systems are suitable for understanding agregate data, comparing what is seen today with patterns from historical databases, can select from several procedures the one or ones that are appropriate with today's data and determines when the parameters of the model describing the data or model being used for predictive purposes should be changed.

In this paper we shall present some work developed in the application of Artifical Neural Networks for the blending of cotton fibers and in the view of real-time quality decision and assurance.