University of Oulu

Quality forecasting tool for electronics manufacturing

Saved in:
Author: Juuso, Esko, Jokinen, Timo, Ylikunnari, Jukka, Leiviskä, Kauko
Organizations: University of Oulu, Faculty of Technology, Control Engineering Laboratory
University of Oulu, Faculty of Technology, Department of Process and Environmental Engineering
University of Oulu, Faculty of Technology, Control Engineering Laboratory
University of Oulu, Faculty of Technology, Department of Process and Environmental Engineering
University of Oulu, Faculty of Technology, Control Engineering Laboratory
University of Oulu, Faculty of Technology, Department of Process and Environmental Engineering
University of Oulu, Faculty of Technology, Control Engineering Laboratory
University of Oulu, Faculty of Technology, Department of Process and Environmental Engineering
Format: eBook
Online Access: PDF Full Text (PDF, 1.2 MB)
Persistent link: http://urn.fi/urn:isbn:9514275071
Language: English
Published: 2000
Publish Date: 2000-09-21
Description:

Abstract

In electronics manufacturing, the package information and their inspection and repair reports are collected to databases. This data can be useful in improving quality and reliablity of new products. At design phase manufacturability can be influenced efficiently; later in the production delays and costs for changes will gradually increase. Quality Forecasting Tool (QFT) has been built to forecast quality of products by joining package lists of the product and known quality levels of these packages. The quality forecasts obtained with the QFT system for three products have been tested by comparing the results to the real quality of these products on a period of seven months. This statistical analysis shows the components with biggest problems.

Variation of the quality differences between products and periods of time supports the importance of intelligent systems in these applications. Obtaining more accurate quality forecast requires knowledge on the characteristics of the product, for example component and joint densities as well as their differences. The quality levels obtained from the databases describe average quality in the production line, and this data must be modified on the basis product and process knowledge to improve the quality estimate. Intelligent methods can be applied to model product quality on the basis of product characteristics. To use these methods more queries are needed to extract data about real quality in different situations. By this way it is possible to find factors influencing to the quality. The intelligent methods combine these risk factors in modifying the quality levels of the packages before calculating the quality forecast of the product. The intelligent system had very promising results which support real time application in the future.


Series: Control Engineering Laboratory. Report A
ISBN: 951-42-7507-1
ISBN Print: 951-42-5599-2
Issue: 12
Subjects:
Copyright information: This publication is copyrighted. You may download, display and print it for your own personal use. Commercial use is prohibited.