An Expert System Approach for the Diagnosis and Rectification of Yarn Faults

Previous research indicate that expert system has been in use for automating operations in the textile and garment industry. However, its application on yarn faults diagnosis and rectification is inadequately explored which left the troubleshooting of yarn faults operated manually by human experts. This paper aims to explore the implementation of expert system for diagnosis and rectification of yarn faults. Accordingly, experts were interviewed at textile and garment factories, and a model of the domain knowledge is developed using decision tree. Moreover, a prototype system of the knowledge base of yarn faults is developed and evaluated. Results showed that the overall performance of the prototype system is generally higher than the results in previous expert system research, and introducing multi-lingual facility and performing small scale prototyping in expert system is a promising approach for yarn fault diagnosis and rectification.