CAREER: Collaborative Information Processing of Distributed Sensor Networks for Manufacturing Quality Improvement

该学院的早期职业发展(CALEAR)课程建议为制造质量改进中的分布式传感器网络开发、教授和实施协作信息处理方法，以支持集成的研究和教育计划。最近的研究集中在硬件(传感器设备)和网络技术的发展上。现在的挑战是提供决策支持能力，使分布式传感器网络在制造质量控制方面的全部潜力得以实现。为此，提出的方法包括：(1)网络化传感器系统的协同故障诊断；(2)在役传感器系统的自诊断；(3)面向诊断的传感器分布优化设计。职业项目将采用多学科方法，将工程学知识、多元统计(如时空分析)以及控制理论(如奇偶空间方法)结合起来，以应对技术挑战。该职业项目将开展各种合作努力，为本科生和当地高中的教师/学生扩大研究机会。如果成功，这个职业项目将通过为分布式传感器网络的信息处理能力贡献新的概念、标准和算法，并为下一代制造诊断系统创建一种使能方法，从而推动分布式传感器网络的发展。实现这一目标将带来显著的成本降低，因为质量的提高和停机时间的减少，提高了美国工业和我们国家经济的竞争力。职业教育计划将通过课程和实验室开发、教学创新和其他外展活动为劳动力培训做出积极贡献。与一所为少数族裔服务的大学和当地高中的合作将加强对代表不足的群体的工程教育。通过会议/期刊出版物和国际/行业合作进行广泛传播，将使由此产生的方法广泛应用于对国家经济增长至关重要的许多其他分布式传感器网络。

This Faculty Early Career Development (CAREER) proposes to develop, teach, and implement collaborative information processing methodologies for distributed sensor networks in manufacturing quality improvement to support an integrated research and education program. Recent research has focused on development of hardware (sensor devices) and network techniques. The challenge now is to provide decision support capabilities that will allow the full potential of distributed sensor networks to be realized for manufacturing quality control. Toward that goal, the proposed methodology includes: (i) collaborative fault diagnoses in networked sensor systems; (ii) in-service sensor system self-diagnosis; and (iii) diagnosis-oriented optimal design of sensor distribution. The CAREER project will employ a multidisciplinary approach, where engineering knowledge, multivariate statistics (such as spatio-temporal analysis), as well as control theory (such as parity space approaches) will be integrated to address the technological challenges. The CAREER project will undertake various collaborative efforts to expand research opportunities for undergraduate students and teachers/students from local high schools. If successful, this CAREER project will advance the state of the art on distributed sensor networks by contributing new concepts, criteria, and algorithms to its information-processing capabilities and create an enabling methodology for next-generation manufacturing diagnostic systems. Accomplishing the goal will lead to a remarkable cost reduction as a result of improved quality and reduced downtime, boosting the competitiveness of US industries and our nation's economy. The CAREER education program will make a positive contribution to workforce training through curriculum and lab development, teaching innovations, and other outreach activities. Collaboration with a minority-serving university and local high schools will enhance engineering education for underrepresented groups. Broad dissemination through conference/journal publications and international/industrial collaborations will lead to wide application of the resulting methodology to many other distributed sensor networks that are of vital importance to the nation's economic growth.