CAREER: A Methodology to Systematically Characterize and Diagnose Manufacturing Variation with In-Process Measurement Data

职业生涯：一种利用过程中测量数据系统地表征和诊断制造偏差的方法

• 批准号: 0093580
• 负责人: Daniel Apley
• 金额: $ 37.5万
• 依托单位: Texas A&M Engineering Experiment Station
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 2001
• 资助国家: 美国
• 起止时间: 2001-06-01 至 2004-05-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0093580&HistoricalAwards=false
• 关键词: CAREER; Methodology; Systematically; Characterize; Diagnose

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.

该学院早期职业发展（CAREER）奖的目标是开发、实施和教授一种系统的通用方法，用于在现代数据丰富的制造环境中诊断产品和流程变化。 该职业计划中开发的方法将提供一种有效利用测量数据来减少制造变异的方法，从而生产出更高质量的制成品，从而提高我们国家工业的竞争力。 研究的主要重点是表示、表征和量化时间和空间模式的精确性质。其目的是，变化模式特征的图形显示将作为强大的诊断辅助工具，有助于操作员识别和消除变化的根本原因。 将采用的方法具有很强的多学科重点。基于基本物理和工程原理的变化的时间和空间表示将被纳入所提出的统计诊断算法中，以便有效地从数据中提取和解释有意义的信息。线性和非线性空间表示以及联合时频时间表示将提供过程变化的一般但描述性的表征。随着过程中测量数据的结构变得越来越密集，为信号和图像处理应用开发的方法作为制造偏差的统计过程控制 (SPC) 工具变得越来越重要。研究计划将广泛使用传感器阵列处理和时频分析的概念。拟议的教育计划强调了对 SPC 采取这种多学科方法的必要性。该课程将提供复杂制造过程的物理和建模（从机械工程角度）、传感和信号处理技术（从电气工程角度）以及 SPC 统计基础知识方面的培训。研究计划结果中的虚拟工厂参观和案例研究将用于说明先进制造工艺的复杂性、过程中测量技术的普遍性以及有效利用数据所需的多学科方法。该职业计划中开发的方法将为将系统诊断策略集成到 SPC 中提供具体基础，这将推进 SPC 减少制造偏差的科学知识体系。为了吸引更广泛的受众，研究结果将在应用统计、质量控制、信号处理以及制造会议和期刊中广泛传播。多学科的重点将加强信号处理和SPC学科之间的联系，并有助于吸引其他学术领域的学生和研究人员进入制造质量控制领域，为他们带来新的想法。 该计划涉及与对美国经济至关重要的两个行业——汽车和电子制造——的主要代表密切合作。 两家公司都表现出了确保这一职业计划成功的坚定承诺，他们的制造设施将作为直接实施结果的试验台。预计该成果将广泛适用于各种制造行业。