Real-time control of parts input in flexible production systems by the pattern classification and analogy process

Rizauddin Ramli, Hidehiko Yamamoto

Research output: Contribution to journalArticle

Abstract

In this paper, we consider a parts input problem in a flexible production system consisting of two production lines (upstream and downstream) that are linked by buffers (BUFF). The parts input rules for the upstream production line is based on the observation of the parts lineup situation on the BUFF. Using the parts observation, a real-time control of parts input by pattern recognition called Real-Time Parts Input by Pattern Recognition System (RT-PIPS) is proposed. Pattern recognition consists of the pattern classification and pattern identification processes. Two types of pattern identification processes (Sum Square Method (S2M) and Rank Replacement Method (R2M)) are considered. Simulations of virtual production systems are carried out to verify that RT-PIPS is useful in real-time parts input control, although the production orders are different from the predecided production scheduling.

Original languageEnglish
Pages (from-to)3-25
Number of pages23
JournalInternational Journal of Product Development
Volume10
Issue number1-3
DOIs
Publication statusPublished - 2010

Fingerprint

Real time control
Pattern recognition
Pattern recognition systems
Flexible production
Scheduling

Keywords

  • Flexible production system
  • If-then rules
  • JIT
  • Just-in-time
  • Pattern recognition
  • Real-time control

ASJC Scopus subject areas

  • Business and International Management
  • Economics and Econometrics
  • Strategy and Management
  • Management of Technology and Innovation

Cite this

@article{89de8499b3e64d85aff81c979d60112f,
title = "Real-time control of parts input in flexible production systems by the pattern classification and analogy process",
abstract = "In this paper, we consider a parts input problem in a flexible production system consisting of two production lines (upstream and downstream) that are linked by buffers (BUFF). The parts input rules for the upstream production line is based on the observation of the parts lineup situation on the BUFF. Using the parts observation, a real-time control of parts input by pattern recognition called Real-Time Parts Input by Pattern Recognition System (RT-PIPS) is proposed. Pattern recognition consists of the pattern classification and pattern identification processes. Two types of pattern identification processes (Sum Square Method (S2M) and Rank Replacement Method (R2M)) are considered. Simulations of virtual production systems are carried out to verify that RT-PIPS is useful in real-time parts input control, although the production orders are different from the predecided production scheduling.",
keywords = "Flexible production system, If-then rules, JIT, Just-in-time, Pattern recognition, Real-time control",
author = "Rizauddin Ramli and Hidehiko Yamamoto",
year = "2010",
doi = "10.1504/IJPD.2010.029984",
language = "English",
volume = "10",
pages = "3--25",
journal = "International Journal of Product Development",
issn = "1477-9056",
publisher = "Inderscience Enterprises Ltd",
number = "1-3",

}

TY - JOUR

T1 - Real-time control of parts input in flexible production systems by the pattern classification and analogy process

AU - Ramli, Rizauddin

AU - Yamamoto, Hidehiko

PY - 2010

Y1 - 2010

N2 - In this paper, we consider a parts input problem in a flexible production system consisting of two production lines (upstream and downstream) that are linked by buffers (BUFF). The parts input rules for the upstream production line is based on the observation of the parts lineup situation on the BUFF. Using the parts observation, a real-time control of parts input by pattern recognition called Real-Time Parts Input by Pattern Recognition System (RT-PIPS) is proposed. Pattern recognition consists of the pattern classification and pattern identification processes. Two types of pattern identification processes (Sum Square Method (S2M) and Rank Replacement Method (R2M)) are considered. Simulations of virtual production systems are carried out to verify that RT-PIPS is useful in real-time parts input control, although the production orders are different from the predecided production scheduling.

AB - In this paper, we consider a parts input problem in a flexible production system consisting of two production lines (upstream and downstream) that are linked by buffers (BUFF). The parts input rules for the upstream production line is based on the observation of the parts lineup situation on the BUFF. Using the parts observation, a real-time control of parts input by pattern recognition called Real-Time Parts Input by Pattern Recognition System (RT-PIPS) is proposed. Pattern recognition consists of the pattern classification and pattern identification processes. Two types of pattern identification processes (Sum Square Method (S2M) and Rank Replacement Method (R2M)) are considered. Simulations of virtual production systems are carried out to verify that RT-PIPS is useful in real-time parts input control, although the production orders are different from the predecided production scheduling.

KW - Flexible production system

KW - If-then rules

KW - JIT

KW - Just-in-time

KW - Pattern recognition

KW - Real-time control

UR - http://www.scopus.com/inward/record.url?scp=77950662679&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77950662679&partnerID=8YFLogxK

U2 - 10.1504/IJPD.2010.029984

DO - 10.1504/IJPD.2010.029984

M3 - Article

VL - 10

SP - 3

EP - 25

JO - International Journal of Product Development

JF - International Journal of Product Development

SN - 1477-9056

IS - 1-3

ER -