The objective of this Faculty Early Career Development (CAREER) award is to develop, implement, and teach a systematic and generic methodology for diagnosing product and process variation in a modern data-rich manufacturing environment. The methodology developed in this CAREER plan will provide a means of effectively utilizing the measurement data for manufacturing variation reduction, leading to higher quality manufactured products that boost the competitiveness of our nations industries. The primary focus of the research is on representing, characterizing, and quantifying the precise nature of the temporal and spatial patterns. The intent is that graphical displays of the variation pattern characteristics will serve as powerful diagnostic aids which facilitate the identification and elimination of root causes of variation by human operators. The methods that will be employed have a strong multidisciplinary emphasis. Temporal and spatial representations of variation, which are based on underlying physical and engineering principles, will be incorporated into the proposed statistical diagnosis algorithms in order to effectively extract and interpret meaningful information from the data. Linear and nonlinear spatial representations and joint time-frequency temporal representations will provide general, yet descriptive, characterizations of process variation. As the structure of in-process measurement data becomes increasingly dense, methods developed for signal and image processing applications gain importance as tools for statistical process control (SPC) of manufacturing variation. Concepts from sensor array processing and time-frequency analysis will be used extensively in the research plan. The proposed education plan emphasizes the need for this multidisciplinary approach to SPC. The curriculum will provide training in the physics and modeling of complex manufacturing processes (from a mechanical engineering perspective) and sensing and signal processing technology (from an electrical engineering perspective), as well as in the statistical fundamentals of SPC. Virtual plant tours and case studies from the results of the research plan will be used to illustrate the complexity of advanced manufacturing processes, the prevalence of in-process measurement technology, and the multidisciplinary methods needed to effectively utilize the data.

The methodology developed in this CAREER plan will provide a concrete basis for integrating systematic diagnosis strategies into SPC, which will advance the body of scientific knowledge on SPC for manufacturing variation reduction. To reach a wide audience, the results will be broadly disseminated in applied statistics, quality control, signal processing, and manufacturing conferences and journals. The multidisciplinary emphasis will strengthen the link between the signal processing and SPC disciplines and help to attract students and researchers from other academic areas to the field of manufacturing quality control, bringing with them an infusion of new ideas. The plan involves close collaboration with major representatives of two industries that are vital to the US economy - automobile and electronics manufacturing. Both have demonstrated a strong commitment to ensuring the success of this CAREER plan, and their manufacturing facilities will serve as testbeds for directly implementing the results. It is expected that the results will be broadly applicable in a variety of manufacturing industries.

National Science Foundation (NSF)
Division of Civil, Mechanical, and Manufacturing Innovation (CMMI)
Application #
Program Officer
Janet M. Twomey
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
Texas Engineering Experiment Station
College Station
United States
Zip Code