Ceramic Tile Process Modeling for Quality Improvement Using ANN

This paper proposes a method of improving the quality and timeline of ceramic products through development of neural network model relating multiple quantitative and qualitative design parameters to product quality. Various processes in a ceramic tile manufacturing company were studied in this research. An intelligent software module to demonstrate the usefulness of such an approach is developed. The input variables describe the raw materials, ambient conditions and line settings. The neural network predictive model assigns one quality value to the final product based on the input data. This prediction can be used to make a priori adjustments to materials and line settings so that a product of required quality is produced without trial and error.