A PCA-FBPN Approach for Job Cycle Time Estimation in a Wafer Fabrication Factory

A PCA-FBPN Approach for Job Cycle Time Estimation in a Wafer Fabrication Factory

Toly Chen (Feng Chia University, Taiwan)
Copyright: © 2012 |Pages: 18
DOI: 10.4018/ijfsa.2012040103
OnDemand PDF Download:
$30.00
List Price: $37.50

Abstract

Variable replacement is a well-known technique to improve the forecasting performance, but has not been applied to the job cycle time forecasting, which is a critical task to a semiconductor manufacturer. To this end, in this study, principal component analysis (PCA) is applied to enhance the forecasting performance of the fuzzy back propagation network (FBPN) approach. First, to replace the original variables, PCA is applied to form variables that are independent of each other, and become new inputs to the FBPN. Subsequently, a FBPN is constructed to estimate the cycle times of jobs. According to the results of a case study, the hybrid PCA-FBPN approach was more efficient, while achieving a satisfactory estimation performance.
Article Preview

Introduction

Semiconductor manufacturing is a highly competitive industry. To become an agile supplier, shorting the cycle time of every operation is the most critical step. For this purpose, a number of strategies are applicable, such as downsizing, lean production, better scheduling, and others. In order to allocate limited resources in these strategies, precisely estimating their benefits is a pre-requisite, which relies on the accurate cycle time estimates.

A number of studies (Chen, 2003, 2008a; Pai et al., 2004) have stressed the importance of cycle time estimation to the management of a semiconductor manufacturing factory. However, in a reentrant production system such as a semiconductor manufacturing factory, the cycle time of a job depends not only on the progresses of jobs that have been released, but also on jobs that will be released in the future, which constitutes a complex issue involving much uncertainty.

Semiconductor manufacturing is generally split into four main phases: wafer fabrication, wafer probe, packaging, and final test. This study is focused on the wafer fabrication phase, which usually takes several months and is the top priority for improvement.

The existing approaches to estimate the cycle time of a job in a wafer fabrication factory can be classified into the following categories (Chen, 2006): statistical analysis (Raddon & Grigsby, 1997; Hung & Chang, 2002), simulation, artificial neural networks (ANN) (Chang & Hsieh, 2003; Haller et al., 2003; Chang et al., 2005; Chen, 2009; Kuo et al., 2010), case-based reasoning (CBR) (Chang et al., 2001), fuzzy theory (Haller et al., 2003; Chen, 2006; Hsiao et al., 2005; Chen et al., 2010), and hybrid approaches (Chen, 2006). A comprehensive comparison of these approaches can be found in reference (Chen, 2007a). In addition, Chen et al. (2009) considered a special case in which the cycle time of a job in a ramping-up wafer fabrication factory is to be estimated. Chung and Huang (2002) considered the special condition in a wafer fabrication factory with engineering lots. Moreover, an internal due date is usually based on the estimated cycle time. Therefore, research on internal due date assignment should also be investigated (e.g., Wilamowsky et al., 1996; Behnamian et al., 2009).

Complete Article List

Search this Journal:
Reset
Open Access Articles: Forthcoming
Volume 6: 4 Issues (2017)
Volume 5: 4 Issues (2016)
Volume 4: 4 Issues (2015)
Volume 3: 4 Issues (2013)
Volume 2: 4 Issues (2012)
Volume 1: 4 Issues (2011)
View Complete Journal Contents Listing