Development of an innovative production management system in spinning mill

Abstract

Cotton selection is critical to the operation cost and yarn quality. However, selecting appropriate raw cotton from the wide range of cotton varieties and types is quite difficult for spinners. Converting highly variable raw cotton into a very consistent linear fibre strand is more difficult and challenging for spinners. The research project aims at developing an innovative production management system to help spinners to optimise cotton selection so that they can produce yarn with the required properties at the lowest cost. In the research project, a new and potential approach was adopted for predicting the quality of yarn, spun at a particular yarn count, from the fibre properties of a particular cotton. The approach is called Case-Based Reasoning (CBR) system, which is a methodology to model human reasoning and thinking. CBR system was also adopted for understanding the relationship between the blend components and the quality of the resultant blended yarn. It is revealed that CBR system is sufficiently transparent for spinners to understand how the yarn quality can be derived, and can give a satisfactory prediction of yarn quality. By incorporating the established models and the database (containing fibre and yarn data) into a computer program, an intelligent system has been built. The intelligent system allows spinners to access the information conveniently and easily with the user-friendly interfaces. Not only the system can help spinners to predict the unknown yarn properties from fibre properties of a particular raw cotton, but it can also enable spinners to select the most appropriate cotton for yarn production, by providing spinners with a number of alternative raw cotton similar to a particular raw cotton. With the use of the intelligent system, spinners will not be compelled to buy more expensive cotton from overseas during bad seasons as they can consider other types of cotton available in their inventory or buy other types of cotton at lower price. In addition, production cost can be minimised by optimising the selection of cotton that allows the production of yarn with the required properties at the lowest cost. The speed of response can also be enhanced by minimising 'time expensive' experimental error. All these result in an improvement in competitiveness and an increase in profitability.