主持人：张玺 副教授 & 吴建国 特聘研究员
In-situ Quality Assurance via Multi-Sensor Data Fusion
报告人 Judy Jin
Department of Industrial and Operations Engineering
Director of Manufacturing Program, University of Michigan
Connected manufacturing systems with distributed in-situ process sensing provides the feasibility to acquire and communicate massive process operational data during production. This provides unprecedented opportunities for in-situ quality control for defects prevention. Meanwhile, it also calls for data analytics challenges for modeling the intricate dependency of products quality on massive in-situ sensing signals. This research aims to develop a general functional regression model for predicting a scalar quality response based on mixed types of process sensing data including images and functional waveform signals as well as scalar process setup parameters. To represent a set of time-dependent images, a third-order tensor is taken with the advantage of preserving not only the spatial correlation within one image but also the temporal dependency among multiple images. Simulation studies and a case study will be presented to discuss the proposed iterative estimation method and model inference.
Prof. Jionghua (Judy) Jin is currently a professor in the Department of Industrial and Operations Engineering and the Director of Manufacturing Program at the University of Michigan. Dr. Jin’s research focuses on data fusion and analytics in quality engineering with broad applications in both manufacturing and service industries. She has received numerous awards including the prestigious NSF CAREER Award in 2002 and the presidential PECASE Award in 2004, respectively and 14 Best Paper Awards. She is currently the Editor of Quality and Reliability Engineering for IISE Transactions. She was also the former Vice President of INFORMS in 2010~2013 and the President of Quality Control and Reliability Engineering Division in IIE in 2007~2008. She is an elected Fellow of IISE and ASME, ISI, and a senior member of ASQ. She received her BS and MS in Mechanical Engineering at Southeast University, Nanjing, China in 1984 and 1987, and her PhD in Industrial and Operations Engineering at the University of Michigan in 1999, respectively.
Advances in Data Analytics for IoT Enabled Smart and
报告人 Shiyu Zhou
Department of Industrial and Systems Engineering
Director of IoT Systems Research Center, University of Wisconsin-Madison
Internet of Things (IoT) represents the convergence of three major and irreversible technology trends, namely (i) embedded sensing/smart devices, (ii) pervasive connectivity, and (iii) real-time analytics and contextual intelligence. The ability to collect and share relevant data across a wide range of devices, coupled with the ability to make real-time decisions, results in an unprecedented opportunity for system modeling, monitoring, and prognosis. In this talk, several new data analytics techniques tailored for IoT-enabled smart and connected systems will be introduced, including modeling and prognosis of condition monitoring signals using joint model and multivariate Gaussian convolution processes, and learning of system interdependencies through graphical modeling and transfer learning. The advantageous features of the proposed methods are demonstrated through numerical studies and real world case studies.
Prof. Shiyu Zhou is the Vilas Distinguished Achievement Professor in the Department of Industrial and Systems Engineering and the Director of IoT Systems Research Center at the University of Wisconsin-Madison. His research focuses on data-driven modeling, monitoring, diagnosis, and prognosis for engineering systems with particular emphasis on manufacturing and after-sales service systems. He has established methods for modeling, analysis, and control of Internet-of-Things (IoT) enabled smart and connected systems, variation modeling, analysis, and reduction for complex manufacturing processes, and process control methodologies for emerging nano-manufacturing processes. He has won a large number of highly competitive federal research grants. His research also attracted significant interests from industry and received significant direct funding support from various companies. He is a recipient of a CAREER Award from the National Science Foundation and the Best Application Paper Award from IIE Transactions. Now he’s the director of IoT Systems Research Center at UW-Madison and a fellow of IISE, ASME, and SME.
Data Science in Autonomous Experimentation Platform
报告人 Yu Ding
Mike and Sugar Barnes Professor of Industrial & Systems Engineering
Texas A&M University
In 2016, US Air Force Office of Scientific Research issued a MURI call for developing autonomous material experimentation platform, a smart manufacturing system that can look for new materials on its own, with no or minimum human intervention. A group of faculty members at Texas A&M University responded, made to the final round of five full proposals, but in the end failed to win the competition. This failed MURI effort, on the other hand, catalyzed a strong interest and subsequent collaboration among data scientists and material scientists, which is still ongoing. In the ensuing years, the team had better lucks with other opportunities and is in fact actively researching for a solution related to the original AFOSR’s call. In this talk, the speaker would like to share his thoughts concerning the challenges and possibilities for such an autonomous experimentation platform.
Prof. Yu Ding is the Mike and Sugar Barnes Professor of Industrial & Systems Engineering, Professor of Electrical & Computer Engineering, and a member of Texas A&M Institute of Data Science, Texas A&M Energy Institute, and TEES Institute of Manufacturing Systems. Dr. Ding received his Ph.D. degree from the University of Michigan in 2001. Dr. Ding’s research interest is in the area of data and quality science, and system informatics. Dr. Ding is a recipient of the 2018 Texas A&M Engineering Research Impact Award, the recipient of the 2019 IISE Technical Innovation Award, and a Fellow of IISE and ASME.
A Convolution Framework for Learning and Predicting
3D Printing Shape Accuracy
报告人 Qiang Huang
Department of Industrial and Systems Engineering
University of Southern California
Geometric shape accuracy is an important quality measure for products built by additive manufacturing (AM) processes. With increased availability of AM product data and advances in computing, Machine Learning for AM (ML4AM) has become a viable strategy for enhancing printing performance. We propose a Shape Deviation Generator (SDG) under a novel convolution formulation to facilitate the learning and prediction of 3D printing accuracy. Shape deviation representation, individual layer input function and transfer function for the convolution formulation are proposed and derived. A deconvolution problem for identifying the convolution kernel is formulated to captures the inter-layer interaction effects in the layer-by-layer fabrication processes. The printed 2D and 3D shapes via a stereolithography (SLA) process are used to demonstrate the proposed modeling framework and derive new process insights for AM processes.
Prof. Dr. Qiang Huang is currently an Associate Professor at the Daniel J. Epstein Department of Industrial and Systems Engineering, University of Southern California (USC), Los Angeles. He was the holder of Gordon S. Marshall Early Career Chair in Engineering at USC from 2012 to 2016. He received National Science Foundation CAREER award in 2011 and IEEE Transactions on Automation Science and Engineering Best Paper Award from IEEE Robotics and Automation Society in 2014. He is Department Editor for IISE Transactions, Associate Editor for ASME Transactions, Journal of Manufacturing Science and Engineering, and a member of Editorial Board for Journal of Quality Technology. He also served an Associate Editor for IEEE Transactions on Automation Science and Engineering and for IEEE Robotics and Automation Letters.
Data Science in Precision Medicine of Brain Diseases: Bridging to Monitoring, Diagnosis, and Treatment
报告人 Jing Li
School of Computing, Informatics, and Decision Systems Engineering
Arizona State University
The brain is the most complex organ of the human body. The booming of artificial intelligence was originally motivated by attempting to mimic how the human brain works in problem solving. As much as one can be amazed by how the brain works in healthy individuals, the brain is susceptible to various types of diseases that are extremely difficult to treat – largely due to our lack of understanding of this complex organ. Examples of these diseases include, but not limited to, glioblastoma – an aggressive type of cancer with median survival of only 14 months, Alzheimer’s Disease – currently no cure, migraine – affecting 1 out of 4 U.S. households, traumatic brain injury – contributing to a third of all injury-related deaths, etc.
Just like the rest of the world that is being dramatically changed at the availability of immense amounts of data, medicine is facing enormous challenges of effectively transforming various forms of data into information and ultimately into decisions to revolutionize patient care. In this talk, I will present some unique data science challenges and our attempts in the solution methodologies for providing enabling tools in monitoring, diagnosis, and treatment of brain cancer and diseases.
Prof. Jing Li is an Associate Professor in Industrial Engineering at the School of Computing, Informatics, and Decision Systems Engineering at Arizona State University. She is a co-founder of an ASU-Mayo Clinic Center for Innovative Imaging. Her training and earlier research focused on data analytics for quality and reliability improvement of industrial systems and processes. Her present research focuses more on health and medicine, developing methodologies in data fusion, model integration, and transfer learning to provide decision support in diagnosis and treatment of cancer and neurological diseases.
Detecting Bursts in Water Distribution Systems with
Functional Data Analysis
报告人 Jian Liu
Department of Systems & Industrial Engineering
University of Arizona
Bursts in water distribution systems (WDSs) are a special type of short-term, high-flow water loss. Since WDSs are usually deployed underground, it is difficult to detect bursts before their catastrophic results being observed on the ground surface. Continuous hydraulic data streams collected from automatic meter reading and advanced metering infrastructure systems make it possible to detect bursts in WDS based on data analytics. Conventional statistical process control methods may not be effective. This is because that the temporal correlations and non-stationary shifts imbedded in the data streams are not explicitly considered, leading to high rate of false alarm and/or miss detection. In this research, a burst detection method based on functional data analysis is proposed. The temporal correlations are modeled with functional bases, and the non-stationary shifts induced by bursts are decomposed as anomalies from the data streams of customers’ daily use without bursts. The proposed method significantly reduces the rate of false alarm or miss detection. Its effectiveness is demonstrated with a case study based on numerical simulation of a real-world WDS.
Prof. Jian Liu is an Associate Professor in the Department of Systems & Industrial Engineering at The University of Arizona. Dr. Liu’s research specialty is in the fusion of multi-source, multi-scale and multilevel information in hierarchical and distributed systems for better system design, operation and maintenance. He is a member of INFORMS and a member of IISE. His research has been supported by NSF, AFOSR, among others.
Multisensor Degradation Modeling and Prognostics
报告人 Kaibo Liu
Department of Industrial and Systems Engineering, UW-Madison
Associate Director, UW-Madison IoT Systems Research Center
In condition monitoring, multiple sensors have been widely used to simultaneously collect measurements from the same unit to estimate the degradation status and predict the remaining useful life. In this talk, we propose a generic framework for multisensor degradation modeling, which can be viewed as an extension of the degradation models from one-dimensional space to multi-dimensional space. Specifically, we model each sensor signal based on random-effect models and characterize failure events by a multi-dimensional failure surface which is an extension of the conventional definition of the failure threshold for a single sensor signal. To overcome the challenges in estimating the failure surface, we transform the degradation modeling problem into a supervised classification problem, where a variety of classifiers can be incorporated to estimate the degradation status of the unit based on the underlying signal paths, i.e., the collected sensor signals after removing the noises. As a result, the proposed method gains great flexibility. It is also capable for sensor selection, able to handle asynchronous sensor signals, and easy to implement in practice. Simulation studies and a case study on the degradation of aircraft engines are conducted to evaluate the performance of the proposed framework in parameter estimation and prognosis.
Prof. Kaibo Liu is currently an Associate professor at the department of Industrial and Systems Engineering, University of Wisconsin-Madison. He received the B.S. degree in industrial engineering and engineering management from the Hong Kong University of Science and Technology, Hong Kong, China, the M.S. degree in statistics and the Ph.D. degree in industrial engineering from the Georgia Institute of Technology, Atlanta, respectively. Dr. Kaibo Liu’s research is in the area of system informatics and data analytics, with an emphasis on the data fusion approach for system modeling, monitoring, diagnosis, prognostics and decision making. The significance of his research has been evidenced by the wide recognition in a broad of research communities in Quality, Statistics, Reliability and Data Mining, including several best paper awards from INFORMS and ISERC and several featured articles from IIE and INFORMS magazines. His research has been successfully funded by NSF, ONR, AFOSR, DOE, and Industry. He is the receipt of three prestigious early career awards, including the 2019 Outstanding Young Manufacturing Engineer Award by SME, the 2019 Feigenbaum Medal Award by ASQ, and the 2019 Dr. Hamed K. Eldin Outstanding Early Career IE in Academia Award by IISE. He is currently serving as an associate editor of IEEE Transactions on automation science and engineering and IEEE CASE. More information can be found in his website: http://kaibo.ie.wisc.edu/index.html
Data Decomposition for Analytics of Engineering Systems
报告人 Xiaowei Yue
Department of Industrial and Systems Engineering
Virginia Polytechnic Institute and State University
Data decomposition is an important step for high-dimensional data analytics of complex engineering systems, but it is less emphasized in our current data analytics domain. This paper summarizes the key techniques for data decomposition, and separates them into two categories. One is deterministic decomposition, and the other is stochastic decomposition. The deterministic decomposition captures geometric or algebraic shape from the high-dimensional datasets directly, which is efficient for feature extraction and dimensionality reduction; while the stochastic decomposition provides probabilistic descriptions, and corresponding statistical distributions are estimated from the datasets. A novel methodology framework of data decomposition is proposed to formulate the existing approaches. Based on this methodology framework, some future research opportunities for new methodology development are discussed for data analytics of engineering systems.
Prof. Xiaowei Yue is an assistant professor at the Grado Department of Industrial and Systems Engineering, Virginia Tech. He got his Ph.D. in industrial engineering, M.S. in Statistics from Georgia Tech, M.S. in Engineering Thermo-physics from Tsinghua, B.S. in Mechanical Engineering from Beijing Institute of Technology. His research interests focus on engineering-driven data analytics for advanced manufacturing. The objective is to develop new methodologies for predictive modeling, uncertainty quantification, system optimization, and model based engineering (MBE). He won several best paper awards, e.g. IEEE Transactions on Automation Science and Engineering Best Paper Award, and Mary G. and Joseph Natrella Scholarship from the American Statistical Association, and IISE Pritsker Doctoral Dissertation Award, etc.