Variation Reduction in a Continuous Web Process

Variation has long been an enemy to manufacturing quality and productivity. As technology increases our ability to record and process data we explore what kinds of improvements can be made with the new tools now available. The Tenneco Packaging facilities in Tomahawk, Wisconsin offer a particularly rich source of data with which to test these tools. Three good data sets have been assembled and explored as a result of this internship with varied success. One of the results of this work is the development of a design of experiments for machine speed on the mill's largest paper machine. Junehee Lee and Geoffrey Lauprete will perform further analysis on the data in conjunction with their doctoral studies. Exploration for other potential sources of important variation was also performed including variation in first stage processing, raw material variation, and high frequency variation in final processing. Several incremental improvements have been suggested as per these studies and will hopefully lead to improved production, raw material cost savings, and reduced wear on the machinery. Thesis Advisors: David H. Staelin, Department of Electrical Engineering and Computer Science Arnold Barnett, Sloan School of Management