Informs 2015

92 / 552

INFORMS Philadelphia – 2015
90
SB73
73-Room 203B, CC
Data Analytics for Quality Control and Improvement II
Sponsor: Quality, Statistics and Reliability
Sponsored Session
Chair: Xi Zhang, Assistant Professor, Peking University, 5 Yiheyuan Rd.,
Beijing, 100871, China,
Co-Chair: Kaibo Liu, Assitant Professor, UW-Madison, 1513 University
Avenue, Madison, WI, 53706, United States of America,
1 - Multistage Process Monitoring through Mewma Control Chart
with Generalized Smoothing Parameters
Sangahn Kim, Rutgers University, 96 Frelinghuysen Road CoRE
Building, Room 201, Piscataway, NJ, 08854, United States of
America, , Myong K (MK) Jeong,
Elsayed Elsayed
The multivariate exponentially weighted moving average (MEWMA) control
chart is effective in monitoring a multistage process when the residual is applied
to remove variance propagation effect. In this paper, we propose a generalized
model of multivariate EWMA, which uses appropriate non-diagonal elements in
the smoothing matrix based on the correlation between stages and within a stage,
and suggest an optimal design for the proposed chart.
2 - Modeling In-Process Data of Machining Operations:
Time Series vs. Spatial Point Cloud
Mohammed Shafae, PhD Candidate, Virginia Tech, 112 Durham
Hall, Blacksburg, VA, 24061, United States of America,
Lee Wells, Jaime Camelio, Marco Ferreira
Traditional approaches for analyzing machining data revolve around representing
them as time-series. The world of time-series analysis has provided several
techniques for data analysis. However, they are implemented without asking, “Is
time-series the most appropriate way to represent this data, and if not, how
should the data be represented and what techniques need to be developed to
analyze this alternate representation?” Exploring the answer to this question is
the focus of this presentation.
3 - Separation and Prognostics of HRV Based on a
Kernel-distance-based Multivariate Control Chart
Lili Chen, Peking University, 5 Yiheyuan Rd., Beijing 100871,
China, Xi Zhang
Isolating component signals directly from the observed heart rate variability
signals can be a challenge. In this article, a signal separation based on integration
of EMD and ICA is developed to separate the frequency components in HRV. A
prognostics framework with Kernel-distance based multivariate control chart for
disease detection has been developed to monitor the component signals. A real
case study demonstrates the effectiveness of the proposed method.
4 - Data Fusion Approach for Degradation Modeling and Prognostics
with Multiple Failure Modes
Abdallah Chehade, UW-Madison, 1513 University Avenue,
Madison, 53706, United States of America, ,
Xi Zhang, Kaibo Liu
Operating units often suffer from multiple modes of failure in practice. This study
proposes a data fusion approach to online classify the failure mode of the
operating unit based on multiple degradation-based sensor data and then predict
the remaining lifetime of the unit. A case study that involves a degradation
dataset of aircraft gas turbine engines is used to numerically evaluate and
compare the performance of the developed method with existing literature.
SB74
74-Room 204A, CC
QSR Refereed Research Session
Sponsor: Quality, Statistics and Reliability
Sponsored Session
Chair: Haitao Liao, Associate Professor, University of Arizona,
University of Arizona, Tucson, AZ, 85716, United States of America,
Co- Chair: Associate Professor, Pennsylvania State University, 310
Leonhard Building, Industrial and Manufacturing Eng., State College
PA 16801, United States of America,
Co-Chair: Tirthankar Dasgupta, Associate Professor, Harvard University,
Department of Statistics, 1 Oxford Street, 7th Floor, Cambridge MA
02138, United States of America,
1 - QSR Refereed Research Session
Hui Yang, Associate Professor, Pennsylvania State University, 310
Leonhard Building, Industrial and Manufacturing Eng., State
College, PA, 16801, United States of America,
The papers in this refereed session are selected after a rigorous peer-review
process. Four finalists will make presentations in the QSR refereed session. The
winner of best paper award will be announced at the QSR business meeting
during the conference.
SB75
75-Room 204B, CC
Advanced Manufacturing I
Cluster: Advanced Manufacturing
Invited Session
Chair: Yuan-Shin Lee, Professor, North Carolina State University,
400 Daniels Hall, Department of Industri, Raleigh, NC, 27695,
United States of America,
1 - Supply Chain Design Implications of Modular
Production Systems
Satya Malladi, Graduate Student, Georgia Institute of Technology,
755 Ferst Drive NW, ISyE,, Atlanta, GA, 30332, United States of
America, Alan Erera, Chelsea White
Modular/mobile production systems allow production capacity to be transported
closer to demand. We address the following questions: When should some or all
production capacity be modular? When should spatial changes in customer
demand result in logistics adjustments only and when should it also be met by
relocating modular production capacity?
2 - Regenerative Medicine Manufacturing – Challenges and Tools for
Scale-up and Scale-out
Rohan Shirwaiker, North Carolina State University,
406 Daniels Hall, Raleigh NC 27607, United States of
America Binil Starly
Regenerative medicine technologies continue to be developed and successfully
tested in labs, but their translation to commercial-scale viable production
continues to be a significant challenge. This talk will discuss the current state of
the art, hurdles in bench to bedside translation, and manufacturing and systems
engineering tools and approaches that can be used to enable efficient scale-up and
scale-out.
3 - A Simulation Model of the EBM Additive Manufacturing Process
Richard Wysk, Dopaco Distinguished Professor, North Carolina
State University, 400 Daniels Hall, 111 Lampe Dr, Raleigh, NC,
27606, United States of America,
Michael Blum, Ismail Lekorchi, Ola Harryson, Kali Drake,
Christopher Kelly, Timothy Horn
Electron beam melting is a 3D printing process that has deficiencies that need
improvement. This paper defines, and analyzes each step in the EBM process. A
simulation model of the process is created to determine time required to process a
part. This model focuses on the processing steps required to bring about the
production of mechanical metal parts. An investigation to improve the efficiency
of each step has been created to model this additive process. This model and
process is described.
4 - Smart Machining and Manufacturing Systems
Yuan-Shin Lee, Professor, North Carolina State University,
400 Daniels Hall, Department of Industri, Raleigh, NC, 27695,
United States of America,
With recent advancements in manufacturing process, information technology,
sensing and automation technologies, it is envisioned a new kind of smart
manufacturing that can locally deliver customized products with high quality bu
with the cost structure of a mass manufactured product. One of the major
building block is the new generation smart CNC machines. In this talk, we will
discuss the new generation smart machines and manufacturing systems.
SB73