An application of neural networks to prediction of SL laser package coupling stability characteristics

Author

Rangwala, Sabbir S. ; Goodwin, Gregory L.

Author_Institution

AT&T Bell Labs., Breinigsville, PA, USA

fYear

1993

fDate

1-4 Jun 1993

Firstpage

169

Lastpage

174

Abstract

Artificial intelligence (AI) is fast becoming a practical and cost effective approach to data and sensor based decision making processes. This report discusses the application of neural networks, a branch of AI, to device quality evaluation. The specific product considered is the SL-560 laser package. We demonstrate the application of neural networks for making decisions on device quality, based on observation of manufacturing process data patterns. The current work aims at using neural networks to model the opinion of human experts. These experts are part of a pedigree review group which examines a large variety of data on each device and then assigns a grade to each device. The specific characteristic considered here is the coupling stability of the laser package through manufacturing process

Keywords

artificial intelligence; expert systems; neural nets; optical workshop techniques; packaging; semiconductor lasers; stability; SL laser package coupling stability characteristics; SL-560 laser package; artificial intelligence; cost effective; coupling stability; data based decision making processes; device quality evaluation; human experts; manufacturing process data patterns; neural networks; pedigree review group; sensor based decision making processes; specific characteristic; Artificial intelligence; Artificial neural networks; Costs; Decision making; Intelligent sensors; Laser modes; Laser stability; Manufacturing processes; Neural networks; Packaging;

fLanguage

English

Publisher

ieee

Conference_Titel

Electronic Components and Technology Conference, 1993. Proceedings., 43rd

Conference_Location

Orlando, FL

Print_ISBN

0-7803-0794-1

Type

conf

DOI

10.1109/ECTC.1993.346836

Filename

346836