A new process control system to be incorporated into a data-intensive manufacturing process has been developed by combining the stochastic models and the digital coding techniques. The system developed could also extract, retain, and synthesize only the essential information to characterize the salient qualities of yarns and fabrics, without having to process and store the redundant vast amounts of data acquired on-line. Several stochastic models such as Poisson, Gamma, and extreme distribution were used and their parameters were estimated and stored every 50 m sub-block of yarns. The parameters were coded sequentially using a digital technology. Most of the essential process informations are stored in the files without the unnecessary information in real-time.