A Framework
of e-based Quality Management for Distributed Manufacturing System
Iraj Mahdavia,
Namjae Chob,
Babak Shirazia
a
Department of Industrial Engineering, College of Technology,
Mazandaran
University of Science & Technology, PO Box734,
Babol, Iran
irajarash@rediffmail.com
b
School of Business, Hanyang University
17, Haegdang-dong,
Seongdong-gu, Seoul, Korea
ABSTRACT
Advanced
manufacturing systems need to be developed for an enterprise to survive
in the increasingly competitive global market. Statistical e-based quality
control approach combines statistical quality analyses and reporting
capabilities with web technology to deliver process optimization solutions.
In this paper we develop a framework for e-based quality management
to provide the capacity to access required data anywhere. It helps enterprises
to develop customized quality information systems, create and distribute
reports via the internet, and to provide real-time display of quality
profiles for processes monitoring. Quality engineers and managers have
been dependent on IS departments to secure access to such data. e-based
quality profile is designed to bridge the gap between raw dataand genuine
quality improvement efforts by providing a powerful web-based solution
for real-time quality process.
Keyword: Distributed
manufacturing system; e-based quality profile; Real-time process; Distributed
quality management
INTRODUCTION
The global
diffusion of advanced manufacturing systems naturally resulted in the
needs for distant product development, which in turn requires the provision
of production information and knowledge on the place, time and configuration
across product life cycle. In response to this need the research community
has come with a solution called distributed manufacturing system, which
is defined as ‘‘an Internet based computational architecture that
supports the sharing and transfer of knowledge and information about
the product and manufacturing process amongst geographically distributed
companies to support right engineering decisions in e-based environment
(Rodriguez and Al-Ashaab, 2002). A number of research initiatives related
to e-based manufacturing systems have been undertaken by several authors.
Information models and engineering applications are integrated to form
a framework in a structured and transparent manner using communication
protocols shared among elements of the system (Molina and AL-Ashaab,
1995). The traditional manufacturing control systems have little capacity
to adapt and react to the dynamic changes in environment. For this reason,
new approaches are called for to provide the capability to adapt to
changes without external interventions (Leitao and Restivo, 2000). The
emergent manufacturing control architectures should use the multi-agent
technology to support the development of autonomous and adaptive control
architectures. Nowadays, we also observe a trend for a high level of
product customization to fulfill market demands, shifting the manufacturing
paradigm from the era of mass production to the era of mass customization.
The major requirements that emergent manufacturing systems should comprise
include: enterprise integration, distributed organizational architecture,
heterogeneous environments, and the fault tolerant integration of human
elements with the structure of modern organization. The network-type
operation and processing further facilitates the adoption of concurrent
systems in distributed computing environments. Significant changes
have been made to enterprise strategies and manufacturing paradigms,
particularly for companies working together in the global marketplace
(Jagdev and Thoben .2001) recognize three types of enterprise collaboration:
supply chain1, extended enterprise2
and virtual enterprise3. Information and Communication Technologies4
have stimulated the emergence and evolvement of various enterprise collaborations.
A virtual enterprise is a network of independent organizations that
jointly form an entity committed to provide a products or services.
From the customer’s perspective, as far as that product/service is
concerned, these independent organizations, for all practical and operational
purposes, virtually act as a single entity/enterprise (Hao et al., 2005).In
this paper we review the concept of advanced manufacturing and its effect
on the structure of quality management for distributed manufacturing
system. The goal of this paper is to improve the quality of production
and reduce loss through systematic tracking and the use of information
as continuous feed-back to production lines.
ADVANCED
MANUFACTURING ENVIRONMENT
The advanced
manufacturing environment integrates a wide span of components of a
factory as shown in Figure1. Integrated manufacturing systems span from
a single equipment automation system to more sophisticated systems such
as Manufacturing Execution System5, Yield Management System6,
Equipment Engineering Capability7
and Enterprise Resource Planning8 (Su et al., 2002) .One of the major
concerns for a manufacturing company at present is to obtain product
consistency. Naturally the importance of the quality process is highlighted.
Quality management practices focus on reducing process variance, which
has a direct impact on supply chain performance including inventory
and lead time (Flynn et al., 1995).
If process
variance is reduced, there is a less need for safety stock and cycle
stock (Flynn and Flynn, 2005). The quality management problem has many
different formulations while a set of core characteristics can be identified
(Dean and Snell, 1991; Anderson et al. 1994; Mehra et al., 2001).
Figure1- Advanced
Manufacturing System Components
A manufacturing
execution system shares its wealth of information with other parts inside
a company as well as with key customers and suppliers. Such an extensive
sharing of information leads to an increased level of quality assurance
and visibility into operations.
DISTRIBUTED MANUFACTURING SYSTEM MODEL
For an integrated system be effective throughout a network of distributed
manufacturing system, it must be connected to Internet (or any world-wide
information area network) .The e-manufacturing system would include
a mixture of both traditional functional systems and newly transformed
network-based components as shown in Figure 2.
Internet
Production
Equipment
Connection
to Network
Factory
Operation
Advance
Process
Control
Recipe
Management
AMHS System
Yield Management
System
Spare
Parts
E-Diagnostic
System
ERP
Scheduler
Dispatcher
WIP tracking
System
Fault
Detection
Equipment
Management
Tracking
Process
Tool data
Connection
Figure 2- A
Model of e-Manufacturing System
Any internet
based system architecture is structured around a three-layered framework:
information, application and end user layer (Rodriguez and Al-Ashaab,
2005). The end user layer is connected to the application web server
(application layer), which in turn is connected to the information databases
.information layer. as shown in Figure 3. The application layer consists
of two elements: decision support applications and information management
tools. The web-based interface of the end-user layer helps users to
view and use different decision support applications and tools. The
product data is produced and used by different engineering applications
throughout the entire production processes. The data is usually stored
in what is called a product model. The structure of a product model
is related to the type of engineering application that it supports.
Distributed
Manufacturing
Knowledge
Model
Firewall
End-User
Information
Application
Users
Decision
Support Applications
- Project timing
-QFD
- Design
-Product
-Fabricating
-Customer key -FMEA
-Equipment
Requirement
- Cost - Process
Parameter
Project
Management
Information
Management
Knowledge
Management
Product
File Access
Team
Management
Specification
Definition
Product
Engineering
Process
Engineering
Tool
Making
Product
Model
Figure 3- Three
layer of e-based distributed manufacturing systems
But in this
paper we also focus on applications that concentrate on quality design
for product via internet based system. Designing a quality strategy
to handle data analysis and reporting processes is a vital activity
for manufacturing systems. Despite the enormous amount of data that
companies have the data is not readily available for users to spot trends
quickly and identify underlying causes.
STATISTICAL
e-BASED QUALITY CONTROL MODEL
Quality engineers and managers have
been dependent on IS department to gain access to information such as
the number of units produced vs. planned. Quality managers have used
quality control systems such as statistical process control, production
part approval process, failure mode effects analysis, gage calibration
and document control. Being mostly stand-alone applications these individual
applications have could not meet quality objectives required for today's
complex manufacturing processes. Their inability to provide effective
links to enterprise integrated management systems has become a major
drawback to be used effectively in a distributed organization. For example,
many PC-based automation software
products in the past were incapable of interfacing with other industrial
automation systems. Failing to use standard protocols for communication
and lacking the technology for seamless connections to the corporate
database, these software systems turned into islands of information
and lost much of their designed benefits and savings.
Recent factory-floor
automation systems provide native links to the administration so that
mission-critical data collected from the factory-floor are captured
and stored. Manufacturers currently measure process and performance
in order to improve quality in production. Today, many companies are
committed to extending the application of quality measures to product
design, customer service, finance, and procurement. Sometimes such effort
further extended to facilities in different geographic locations as
well as to suppliers, strategic partners, and sales channels. Statistical
e-based quality control enables continuous quality improvement through
easy and cost-effective access to quality data. It requires an Internet-enabled
software solution to allow individuals to access the quality data for
various analyses.
A statistical
internet-based quality control also needs a dynamic web-based application
that serves as a quality portal, so that users can easily and cost-effectively
integrate a wide array of information for quality improvement. From
within a web browser, authorized users should be able to analyze their
operations real-time. It must be configurable to meet a wide range of
needs for information and in-depth quality analysis.
Statistical
Internet-based quality control must have a continuous quality improvement
strategy to ensure quality products and services. Empowering individuals
with quality analysis and reporting capabilities shifts the responsibility
for quality beyond the hands of a few individuals or a specific division
to the whole organization. Everyone in the organization takes on the
responsibility to continuously assess, improve, and monitor quality.
Statistical e-based QC has combined statistical quality analysis and
reporting capabilities with web technology to deliver a process optimization
system that will give everyone in an enterprise .from the manufacturing
floor to the executive suite. the information they need to monitor,
analyze, and improve important processes. With this system, an enterprise
could access data anywhere and provide analyses everywhere. Statistical
e-based QC must be designed to help enterprise to bridge the gap between
raw dataand genuine quality improvement. It must provide a powerful
web-based solution for real-time data to be shared anywhere in the world
any time in asecure environment.
Requirements
for this system include:
A. Client/server
architecture with native drivers that allow the enterprise SQL database
to be connected to the quality control database without administrative
intervention.
B. Connectivity
to appropriate PC and mainframe versions of an SQL platform.
C. Close to
zero administrative functionality as far as possible. Installation,
configuration and upgrades should occur at a single station, and all
client stations should be able to self-synchronize.
D. The ability
of clients to run off-line (store data locally) in the event of database
or network failure, as well as automatic upload and resynchronization
upon reconnection.
E. The ability
to recover individual clients from the server in the event of client
failure or loss.
F. Adherence
to inter application communication standards, including OLE automation
and OPC.
According to
these requirements we could design a system as shown in Figure 4:
Internet
Internet
Application
(Web Server)
Information
End User
Client
Active
Server
Page
E-QC
Info.
-User
-Form
-Reports
Action
Browser
e-based
QC
HTML
R
E
Q
H
T
M
L
Analysis
Data
-Specs
-Defects
-Measures
Reporting
Tools
Statistics
& outputs
ODBC
e-Server
Business
Logic Layer
Data Layer
Presentation
Layer
System Services
Figure4- Components
of e-based Quality Contorl
Statistical
e-based QC utilizes statistical e-server to perform all data access,
transformation, analyses and output creation. Users request an action
from within the GUI in a browser .the presentation layer. By hitting
a button, selecting a menu or completing a form. Once the request is
received by the presentation layer, the presentation layer (Active Server
Pages) Handles the request and directs it to the application .QC Business
Logic. for processing. The application processes the request and determines
appropriate actions needed to complete the request. If computation,
analysis or graphics are needed, the application sends the request to
platform. Requests directed to platform include:
Read data from
a database, perform computations and return results in the form of HTML
table or data
file.
Accept data
from the application, process the request, for example generating an
SPC chart, and return the resulting graph in the form of image file
or HTML. Sampling the data, check SPC rules, and make alarm by sending
rule violation information to the application. Results are returned
to the application on request in the form of graph image, transformed
data, and formative expression or, a complete HTML page with multiple
graphs and analysis. The presentation layer returns results back to
the browser.E-Server is a module used for statistical analysis and seamless
integration of new or existing applications. In addition to highly customizable
SPC charting capabilities, e-Server should have the capacity to provide
an extensive suite of graphics and statistical analysis methods as well
as the dynamic reporting capability via internet, intranet or extranet
so as to support continuous process improvement. It should also be able
to perform real-time display of quality control charts to monitor processes.
It should be a flexible, full-featured analysis required for continuous
quality improvement.
DISTRIBUTED
QUALITY MANAGEMENT
Quality management
for an enterprisewide e-based QC system focuses on analyzing data to
make decisions that affect future production and revenue. The analysis
includes real-time enterprise wide view that levels of operations, and
batch report according to predetermined time period. E-based quality
control system over a distributed environment should satisfy the following
activities as shown in Figure 5:
• Remote monitoring and
intelligence • Remote diagnostics and maintenance
• Remote
configuration and debugging • Remote
sensing and controlling
• Remote modeling
and tooling
Office#1
Remote monitoring
Remote diagnostics
Remote debugging
Remote sensing
Model tool
behavior
Office#2
Remote monitoring
Remote diagnostics
Remote debugging
Remote sensing
Model tool
behavior
internet
Factory
Floor
Factory
Floor
e-QC enabled
controller
Firewall
Figure 5- Distributed
model of e-based quality control
CONCLUSION
Word-wide competition
leads to the need for new systems of control for distributed manufacturing
systems. The integration of information systems and self-organization
factors provide us with opportunities to fulfill such needs to adapt
quickly to the changes in environment. The new control systems should
incorporate the integration of new technologies, tools and paradigms.
The emergent control architecture relies mainly on the multi-agent technology
to support the autonomous and adaptive control. A multi-agent system
is suitable especially to the distributed manufacturing environment,
which presents modular, decentralized, changeable, ill-structured and
complex issues. To deal with disturbances that deviates the process
from the original plans the system should respond dynamically and quickly
using a mechanism to find out the best plan to handle the disturbance
based on pre-defined rules and knowledge acquired through past experience.
In this paper, we introduced a framework of e-based quality management
that would provide a web-based solution for real time process. The suggested
logic is useful when we want to use a wide variety of quality characteristics
as key terms. As a result of using statistical e-based quality profile,
anyone in an enterprise can contribute to quality improvement efforts.
Web-enabled quality control system will present an extensive connectivity
outside a plant. A customer’s engineer could tour the plant site and
check the profile online. It is used mostly within a factory today,
but after some successes, and with the process in place, access to quality
data will be extended across the supply chain and to customers. One
should be able to obtain information about the batch he ordered and
see how it conforms to specifications. Based on the research and framework
done in the distributed manufacturing systems we developed a framework
on e-based quality control. A mathematical model for an e-based statistical
control on the basis of e-based quality profile can be elaborated in
future research.
REFERENCES
Anderson, J.
& Rungtusanatham, M. & Schroeder, R.G. 1994. A theory of total
quality
management underlying the Deming management method, Acad. of Manage.
Rev., 3, pp. 472-509.
Dean, J.W. &
and Snell, S.A. 1991. Integrated manufacturing and job design: moderating
effects of organizational inertia, Acad. of Manage. J., 34, pp.776-804.
Flynn, B.B.
& and Flynn, E.J. 2005. Synergies between supply chain management
and quality management: emerging implications, International Journal
of Production research, Vol 43.16., pp.3421-3436.
Flynn, B.B.
& Sakakibara, S. & Schroeder, R.G. 1995. Relationship between
JIT and TQM: Practices and Performance, Academy of Management
Journal, Vol 38.5., pp.1325-1360.
Hao, Q. &
Shen, W. & Wang, L. 2005. Towards a cooperative distributed manufacturing
management framework, Computers in
Industry, Vol 56, pp.71-84.
Jagdev, H.S.
& Thoben, K.D. 2001. Anatomy of enterprise collaboration, Production
Planning and Control, Vol 12.5., pp.437-451.
Leitao, P. &
Restivo, F. 2000. A framework for Distributed Manufacturing Applications,
Proceedings of ASI’2000 International Conference, Bordeaux, France.
Mehra, S. &
Hoffman, J.M. & Sirias, P. 2001. TQM as a management strategy for
the next millennia, Int. J. Op. & Prod. Manage, Vol 21.5/6.,
pp.855-876.
Molina, A. &
Al-Ashaab, A. & Ellis, T., and Young, R. 1995. A review of computer-aided
simultaneous engineering systems, Research
in Engineering Design,
Vol. 7, pp.38–63.
Rodriguez, K.,
and Al-Ashaab, A., 2002, A review of internet based collaborative
product development systems, Proceedings of the International Conference
on Concurrent Engineering: Research and Applications, Cranfield, UK.
Rodriguez, K.
& Al-Ashaab, A. 2005. Knowledge web-based system architecture for
collaborative product development, Computers in Industry, Vol
56, 125-140.
Su, D. &
Ji, S. & Li, J. & Chen, X. 2002. Web-enabled collaborative
environment for integrated design and manufacture, Proceedings of
the International Conference on Concurrent Engineering: Research and
Applications, Cranfield, UK.
1
SC
2
EE
3
VE
4
ICT
5
MES
6
YMS
7
EEC
8
ERP