Research Papers

Throughput Bottleneck Prediction of Manufacturing Systems Using Time Series Analysis

[+] Author and Article Information
Lin Li1

Department of Mechanical and Industrial Engineering, University of Illinois at Chicago, 3057 ERF, 842 W. Taylor Street, Chicago, IL 60607linli@uic.edu

Qing Chang

Department of Mechanical Engineering, New York Institute of Technology, Harry Schure Hall, Old Westbury, NY 11568qchang@nyit.edu

Guoxian Xiao

Manufacturing Systems Research Laboratory, General Motors R&D Center, 30500 Mound Road, Warren, MI 48090-9055guoxian.xiao@gm.com

Saumil Ambani

Department of Mechanical Engineering, University of Michigan–Ann Arbor, 1210 H. H. Dow, 2300 Hayward Street, Ann Arbor, MI 48109-2136sambani@umich.edu


Corresponding author.

J. Manuf. Sci. Eng 133(2), 021015 (Apr 04, 2011) (8 pages) doi:10.1115/1.4003786 History: Received April 23, 2009; Revised February 22, 2011; Published April 04, 2011; Online April 04, 2011

Throughput bottlenecks define and constrain the productivity of a production line. The most cost-effective way to improve system throughput is to mitigate bottlenecks toward a balanced system. Most of the currently used bottleneck detection schemes found in literature utilize long-term analysis to identify the bottlenecks for a known period and ignore the operation dynamics leading to bottleneck shifts. This paper proposes a method for predicting the throughput bottlenecks of a production line using autoregressive moving average (ARMA) model. We consider the production blockage and starvation times of each station to be a time series used to predict throughput bottlenecks. It is realized that the blockage and starvation times of a production line are critical indicators reflecting the production system dynamics and its internal material flow. As the first attempt in literature for throughput bottleneck prediction, the results demonstrate that the ARMA model can accurately predict blockage and starvation information of each station and hence can accurately predict the system throughput bottleneck, which will lead to the most significant production improvement.

Copyright © 2011 by American Society of Mechanical Engineers
Your Session has timed out. Please sign back in to continue.



Grahic Jump Location
Figure 5

Layout of a production line

Grahic Jump Location
Figure 6

Comparison between raw data and predicted data

Grahic Jump Location
Figure 1

A serial line with five machines and four buffers

Grahic Jump Location
Figure 7

ARMA orders change when M1 changes its downtime

Grahic Jump Location
Figure 8

ARMA orders change when M12 changes its downtime

Grahic Jump Location
Figure 9

Comparison between baseline PM policy and prediction based PM policy

Grahic Jump Location
Figure 2

Blockage and starvation status of a 5M4B line

Grahic Jump Location
Figure 3

Bottleneck detection results after eliminating the downtime of M3

Grahic Jump Location
Figure 4

Data processing for machine i



Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In