| Summary: | 碩士 === 義守大學 === 管理科學研究所 === 85 === Customer response for your product is the best indicator of the performance of quality control. Any complaint, which represents defection of quality control, usually means loss of the customer''s rights and damage of company''s reputation. Customer''s complaints may be caused by bad product quality, nonconformity of specifiations, or defects of product functions. In this research, an integrated customer complaint handling system, which takes advantage of artifical neural network (ANN) and expert system (ES, is developed for supporting quality workers in the production workshop of automobile braking system to systematically deal with customer complaints and for the further applications in the industry and researches.
The aim of the ANN module using General Regression Neural Network (GRNN) in the system is to predict the possible reasons of the problem for quality workers while receiving complatint telephone or fax from customer. With the help of the ANN, the system can perform the earlier diagnosis based on the symptoms provided by the customer and prepare some suggested actions before leaving for handling the abnormal lots of product in the customer side. If the problem can not be handled in time or has never happened before, the ES module can be applied to identify the causes of problem by means of a series of measurement and functional tests, and providing conclusion and suggestions. After that, the results obtained with the support of the ES can be used as training samples of the ANN for improving the accuracy of prediction. With the synergy of the artificial intelligent techniques in the system, the effort of measurement and tests in the domestic factory can be significantly, reduced so that much time and other resources are saved. Furthermore, we believe quality worker quickly and effectively deal with customer complaints and increase the satisfaction of customers.
In practive, this prototype system can support quality workers to quickly analyze the rensons of failure from after-sale services and customer complaints in order to find out what should be improved in the future. In addition, this system can be applied as a useful tool to support for re-education and training new quality workers in short period of time.
|
|---|
