A neural network and CBR-based model for Sewing Minute Value

Sewing minute value (SMV) is a benchmark to measure the production efficiency. It is an important business data for costing, production planning as well as key reference for product development. The traditional method to calculate SMV is complicated and not robust. There are so many rules to calculate SMV such as the speed of the machine or the motion of the machinist. It lacks randomness and flexibility. We proposed a new model which is based on neural network forecasting theory and case based reasoning technique. We will demonstrate the advantage of using the new method in this paper with better result representing an increase of 11% in accuracy compared to that of the old method.