Holistic modeling and analysis of multistage manufacturing processes with sparse effective inputs and mixed profile outputs

Abstract

In a Multistage Manufacturing Process (MMP), multiple types of sensors are deployed to collect intermediate product quality measurements after each stage of manufacturing. This study aims at modeling the relationship between these quality outputs of mixed profiles and sparse effective process inputs. We propose an analytical framework based on four process characteristics: (i) every input only affects the outputs of the same and the later stages; (ii) the outputs from all stages are smooth functional curves or images; (iii) only a small number of inputs influence the outputs; and (iv) the inputs cause a few variation patterns on the outputs. We formulate an optimization problem that simultaneously estimates the effects of process inputs on the outputs across the entire MMP. An ADMM consensus algorithm is developed to solve this problem. This algorithm is highly parallelizable and can handle a large amount of data of mixed types obtained from multiple stages. The ability of this algorithm in estimations, selecting effective inputs, and identifying the variation patterns of each stage is validated with simulation experiments.

Keywords:

• Multistage manufacturing process
• profile data
• sparse data

Additional information

Notes on contributors

Andi Wang

Andi Wang is a PhD student in the Stewart School of Industrial and Systems Engineering at Georgia Institute of Technology. He received his BS in statistics from Peking University in 2012 and a Ph.D. from Hong Kong University of Science and Technology in 2016. His research interests include advanced statistical modeling, large-scale optimization, and machine learning for manufacturing and healthcare system performance improvements via process monitoring, diagnostics, prognostics, and control. He is a member of IISE and INFORMS.

Jianjun Shi

Jianjun Shi received BS and MS degrees in automation from the Beijing Institute of Technology in 1984 and 1987, respectively, and a PhD degree in mechanical engineering from the University of Michigan in 1992. Currently, Dr. Shi is the Carolyn J. Stewart Chair and Professor at the Stewart School of Industrial and Systems Engineering, Georgia Institute of Technology. His research interests include the fusion of advanced statistical and domain knowledge to develop methodologies for modeling, monitoring, diagnosis, and control for complex manufacturing systems. Dr. Shi is a Fellow of the Institute of Industrial and Systems Engineers (IISE), a Fellow of American Society of Mechanical Engineers (ASME), a Fellow of the Institute for Operations Research and the Management Sciences (INFORMS), an elected member of the International Statistics Institute, a life member of the Amreican Astatistics Association (ASA), an Academician of the International Academy for Quality (IAQ), and a member of National Academy of Engineers (NAE).