Manufactura

INFORMS Journal on Computing
Vol. 19, No. 2, Spring 2007, pp. 302–312 issn 1091-9856 eissn 1526-5528 07 1902 0302

informs

®

doi 10.1287/ijoc.1050.0171 © 2007 INFORMS

Toward Automated Intelligent Manufacturing Systems (AIMS)
School of Electrical and Computer Engineering, Purdue University, Electrical Engineering Building, 465 Northwestern Avenue, West Lafayette, Indiana 47907-2035,USA {hoiming@alumni.virginia.edu, ersoy@purdue.edu} Krannert School of Management, Purdue University, 403 West State Street, West Lafayette, Indiana 47907-2056, USA {herbm@purdue.edu, kemal@purdue.edu}

Hoi-Ming Chi, Okan K. Ersoy

Herbert Moskowitz, Kemal Altinkemer

I

nformation technology (IT) has been the driver of increased productivity in the manufacturing and service sectors,bringing real-time information to decision makers and process owners to improve process behavior and performance. Thus, organizations have invested heavily in training their employees to use IT in a disciplined, scientific way to make process improvements. This has spawned such popular initiatives as Six Sigma, yielding significant returns, but at considerable investment in training instatistical-analysis and decision-making tools. Can aspects of the decision-making process be automated, letting humans do what they do best (create, define, and measure) and machines (e.g., learning machines) do what they do best (analyze)? We propose an automated intelligent manufacturing system (AIMS) for analysis and decision making that mines real-time or historical data, and uses statistical andcomputational-intelligence algorithms to model and optimize enterprise processes. The algorithms employed involve a regression support vector machine (SVM) for model construction and a genetic algorithm (GA) for model optimization. Performance of AIMS was compared to Six-Sigma-trained teams employing statistical methodologies, such as design of experiments (DOE), to improve a simulated manufacturing operation, athree-stage TV-manufacturing process, where the objectives were to maximize yield, minimize cycle time and its variation, and minimize manufacturing costs, which were affected by conflicting defects and their causes. AIMS generally outperformed the teams on the above criteria, required relatively little data and time to train the SVM, and was easy to use. AIMS could serve as a productivityspringboard for enterprises in existing and emergent technologies, such as nanotechnology and biotechnology/life sciences, where environment and miniaturization may make human monitoring and intervention difficult or infeasible. Key words: support vector machine; genetic algorithm; Six Sigma; design of experiments (DOE); response-surface designs; machine learning; computational intelligence; intelligentsystems History: Accepted by Michel Gendreau, Area Editor for Heuristic Search and Learning; received April 2004; revised February 2005, May 2005, June 2005, October 2005; accepted December 2005.

1.

Manufacturing industries are concerned with producing higher-quality products, while minimizing production cost and time. However, many factors or drivers can affect this goal, and theirinteractions and effects are often difficult to uncover. In a TV production line, factors affecting quality and process performance might include environmental working conditions, design of the printed-circuit board, machine technology, and operator skills. Like total quality management (TQM) in the 1980s, which changed quality thinking, information technology (IT) is transforming quality-improvementpractices from art to science by providing real-time data on the state of a process to owners for improvement. This spawned a business strategy and a disciplined, informationintensive problem prevention/solving process known as Six Sigma.
302

Introduction

First advanced by Motorola around 1986 (Buetow 1996), Six Sigma is a scientific approach to problem prevention/solving that involves rigorous...