|
Activity Number:
|
515
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 7, 2008 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Quality and Productivity
|
| Abstract - #301310 |
|
Title:
|
Assessing a Manufacturing Customer's Complaint Using Supplier Process Data
|
|
Author(s):
|
Jon M. Lindenauer*+
|
|
Companies:
|
Weyerhaeuser
|
|
Address:
|
32910 Weyerhaeuser Way S, Federal Way, WA, 98001,
|
|
Keywords:
|
Partial Least Squares-Discriminant Analysis ; Multivariate Analysis ; Score Plot ; Loading Plot ; Variable Importance Plot ; Simca-P
|
|
Abstract:
|
A manufacturing plant has had periodic run rate issues with a supplier's product. From time to time, the supplier product would cause the customer to slow down their machines. The customer identified supplier product deemed as having "good" or "bad" runability. The process data associated with these runs was extracted from the supplier mill database. The good and bad process data was unified into a single dataset. The multivariate method of Partial Least Squares-Discriminant Analysis (PLS-DA) was used to test if the supplier process was operating differently for good and bad runs on the customer's machine. PLS-DA was able to use the process variable data to identify and delineate good and bad supplier product. This paper focuses on communicating a sophisticated statistical analysis method to both supplier and customer mill management, engineering and operations staff.
|
