Wireless Sensor Network, 2010, 2, 402-410
doi:10.4236/wsn.2010.24052 Published Online May 2010 (http://www.SciRP.org/journal/wsn)
Copyright © 2010 SciRes. WSN
An Auto-Configuration of 4M Group Management Using
Wireless Sensor Networks
Suk-Keun Cha1, Jeong-Hoon Lee 1, Han Gyu Kim2, Joon Jae Yoo3, Jung Hoon Kang3,
Dong Hoon Kim4, Jun Yeob Song4
1Research and Development Center, ACS Co., Ltd., Seoul, Korea
2Research and Developm ent Center, SM Information and Communication, Seoul, Korea
3Ubiquitous Embedded Fusion Center, Korea Electronics Technology Institute, Seoul, Korea
4U-Production Team, Korea Institute Machinery and Material, Daejeon, Korea
E-mail: {sk_cha, jh_lee}@acs.co.kr, hgk@s mic.co.kr, {yoojj, budge}@keti.re.kr, {kdh680, sjy658}@kimm.re.kr
Received October 9, 2009; revised November 20, 2009; accepted Marc h 19, 2010
Abstract
The first tier of automotive manufacturers has faced to pressures about move, modify, updating tasks for
manufacturing resources in production processes from demand response of production order sequence for
motor company and process innovation purpose for productivity. To meet this requirements, it has to require
absolutely lead time to re-wiring of physical interface for production equipment, needs for change existing
program and test over again. For prepare this constraints, it needs studying an auto-configuration functions
that build for both visibility and flexibility based on the 4M (Man, Machine, Material, Method) group man-
agement which is supports from Wireless Sensor Network (WSN) of the open embedded device called Ma-
chine to Machine (M2M) and major functions of middleware including point manager for real-time device
communication, real-time data management, Standard Application Program Interface (API) and application
template management. To be application system to Reconfigurable Manufacturing System (RMS) for rapidly
response from various orders and model from motor company that is beginning to establishing the mapping
of manufacturing resources of 4M using WSN.
Keywords: Auto-Configuration, Wireless Sensor Network, Reconfigurable, Production Resources of 4M,
Tinyos, Machien to Machine, Middleware, Application Template Manager
1. Introduction
The increasing popularity and applicability of wireless
sensor networks (WSNs) in large-scale applications such
as environmental monitoring and real-time data integra-
tion for manufacturing activities (we called Ubiquitous-
Manufacturing) motivates the development of new
high-level abstractions for programming.
Typically, a sensor network co nsists of a large number
of nodes deployed in an environment being sensed and
controlled. Each node includes one or more sensors, may
contain actuators, has limited memory, power and com-
putational capacities.
Wireless sensor networks can be considered distrib-
uted computing environments with severe constraints of
CPU speed, memory size, power, and bandwidth. Indi-
vidual nodes in the sensor network are typically unreli-
able and the network topology may change dynamically.
Sensor networks also differ because of their tight interac-
tion with the physical environment via sensors and ac-
tuators. Because of this interaction, sensor networks are
very data-centric. Due to all of these differences, many
solutions developed for general distributed computing
platforms and for adhoc networks cannot be applied to
sensor networks. However, sensor networks also exhibit
characteristics of both embedded systems and gen-
eral-purpose systems. They must consume low sensor
power and be robust to environmental conditions, while
also providing common services that make it easy to
write applications [1].
4M Group management service plays a key role in
WSNs as it provides support to high level middleware
services such as real-time device communication, real-
time data management, standard Application Program
Interface (API) and application template management,
security, fault-tolerance, power management, and so on.
S. K. CHA ET AL.403
However, the way a group is managed may vary either
from service to service or according to manufacturing
activities based on 4M resources (e.g., worker activities,
machine running status, error code, tracking of work in
process (WIP) and product, manufacturing operation
procedures). For example, the group management used
by the security service may differ from one used by the
fault-tolerance service. In addition, the group manage-
ment may be reconfigured due to the power level of
some nodes in the network.
This paper is organised as follows. Section 2 intro-
duces basic principles of WSNs, 4M group management
and reconfiguration. The next section presents our pro-
posal on reconfigurable 4M group management. Section
4 illustrates how the proposed service may be used to
support a particular service. Finally, the last section pre-
sents some conclusions and guidelines for further work.
2. Basic Concept
This section presents some basic concepts of production
resources of 4M, WSNs, group management and recon-
figuration in WSNs.
2.1. Production Resources of the 4M
Viewpoint of integration for production resources, 4M
has consisted of following major categories as referred
Figure 4;
- Man
- Machine
- Material
- Method
For data collection from various type of machines,
every machine has standalone device controllers as PLC
(Programmable Logic Controller), CNC (Computerized
Numerical Control), FMS (Flexible Manufacturing Sys-
tem) and robot controller that is automatically interface
through standard communication of RS232C or Ethernet
with associate protocol s b y real-time.
For data collection form workers, it must manually
report for d efects, non-oper ation reasons and lo t numbers
that is decision made by workers using a touch screen,
bar code reader and mobile devices. It requires tracking
of material logistics including Work in Process (WIP)
and real-time inventory level from receipt of raw materi-
als to shipping of finished goods.
For data collection of method, bar code reader, IC card,
RFID/USN will be able to digitalize for production status
information including actual production, quality level,
equipment running, optimal scheduling and material lo-
gistics that provides all associated production activity
tracking for PL (Product Liability) and performance
analysis. [2]
The method of digitalization for 4M defined 3 ways as
well as automatic, semi-automatic and manual using
wired/wireless multi-sensor data fusion as referred Fig-
ure 1.
1) Automatic data collection method: stand alone de-
vice controllers has standard communication port such as
RS-232C or Ethernet, in this case, associated protocols
easy way to automatic data collection by real-time for
current status of actual production operation as well as
machine running status, actual production report and
quality level.
2) Semi-automatic collection method: There are 2 type
of method will be av ailable. First, embedded data collec-
tion system is cabling with process I/O from PLC I/O
point or device controllers with 4M digitalization pro-
gram. Second, installed additional multi-sensor in device
controller or production processes that interface wired/
wireless sensor communication such as RFID/USN.
3) Manual data collection method: workers can di-
rectly data entry by touch screen, bar code, RFID and IC
card for simple way to operating it in harsh environment.
2.2. Wireless Sensor Networks
A wireless sensor network consists of a large number of
small devices with computational power, wireless com-
munication and sensing capability [3]. Sensor nodes are
usually scattered in an observation region. Each sensor
node in the observation region is responsible for extract-
ing data from the environment (such as temperature, hu-
midity, pressure and lumino sity), processing and sending
them through on e or more sink nodes, which are respon-
sible for transmitting these data to the final user.
The design of sensor network applications is highly
influenced by resource scarcity (e.g., battery, memory
and processor), communication models and application
requirements. In wireless transmission, the power of a
signal decreases proportionally with the square of the
Figure 1. Digitalization for production resources of 4M.
Copyright © 2010 SciRes. WSN
S. K. CHA ET AL.
404
distance. Thus, if the communication between sensor
nodes and the sink will be carried through in a single hop,
it will be necessary to adjust signal strength to make pos-
sible message exchange in the network. This procedure
will cause a power consumption increase and, conse-
quently, a reduction of network lifetime.
One solution would be to adopt a multi-hop commu-
nication, where some intermediate nodes act as bridges
between the message source and its destination. However,
in a dense WSN, various nodes often detect common
phenomena leading to unnecessary redundancy of data
and transmissions. This fact makes power consumption
even more critical. One technique that helps to save en-
ergy is data aggregation [4,5]. The id ea is to combine the
data coming from different sources eliminating the re-
dundancy of the transmitted data, minimizing th e nu mber
of transmissions and thus saving energy [1]. Applica-
tion-specific requirements also affect the network’s re-
sources in d ifferent ways.
For example, the consumption of energy for sending
messages to the network tends to be high er in monitoring
applications than in surveillance systems since environ-
ment violation is a rare event (report by exception).
When a violation happens, the application must guarantee
the delivery of the event notification in an adequate time.
In a tracking application, only the sensors closer to the
observed object need to send information to the network.
The application-specific requirements combined with
resource restrictions make application development for
sensor networks a challenging task. The design of mid-
dleware for WSNs must consider these issues in order to
hide the complexity from the application developers.
However, due to the mission of supporting and optimiz-
ing for a broad class of applications, tradeoffs need to be
explored between the degree of application-specific and
generality of the middleware.
2.3. Group Management WSNs
The Group Management service is responsible for the
creation and maintenance of groups of sensors/actuators
nodes. Basic in this service is the group management
model that defines the group structure and how it is ma-
naged. Two kinds of group structure are usually adopted:
hierarchical or flat. The hierarchical structure has been
widely used [6-9] and specifies the existence of a group
leader, which is the group representative to interact with
other groups and to centralize the group information.
The group leader is also responsible for inter-group
communication. In the flat group structure each member
knows the other members.
The design of a 4M group management service for
WSNs should follow some principles similar to those
applied to Mobile Ad Hoc Networks [10], such as:
1) Resource define: the overhead of group manage-
ment should be affordable for resource-constrained de-
vices;
2) Distributed: group management cannot accommo-
date a centralized solution where a single node is always
responsible for managing the group. This is due to node
mobility and the necessity of load balancing;
3) Dynamism: the group management service must
accommodate to the highly dynamic group members of
the network. These group members dynamics are caused
by network topology changes, power level changes and
group management behaviour.
In order to satisfy the aforementioned design princi-
ples, the group management service must implement
three basic functions: member discovery, group forma-
tion and dynamic management.
The member discovery functionality is responsible for
discovering the nodes that are eligible for membership
according to the group management rules. The group
formation task is responsible for identifying the need for
a new group. Finally, the group management functional-
ity refers to an ability to update group membership ac-
cording to the group management rules and network to-
pology changes.
Group Management has been used in sensor networks
to support services such as: Fau lt tolerance (e.g., replica-
tion groups), security (e.g., consensus groups), time
synchronization (e.g., synchronization groups), object
tracking and power management.
The use of a management group is particularly impor-
tant to power management. It is responsible for mini-
mizing power consumption and making nodes power rest
for long durations (from months to years). A promising
approach to reduce power consumption in sensor net-
works is to control the nodes execution. This approach
consists of making a small subset of nodes remain active
to maintain the network services running while the other
nodes sleep or enter in an energy saving mode. This ser-
vice should be responsible for minimizing energy con-
sumption and maximizing network lifetime. This service
can be developed using a Group Management service
that is responsible for managing the nodes that remain in
the group of nodes responsible for maintaining the net-
work services.
2.4. Reconfiguration in WSNs
Reconfiguration and self-adaptation are vital capabilities
of sensor networks that are required to operate in dy-
namic environments that impose varying functional and
performance requirements.
Dynamically adaptive software comprises of tasks that
detect internal and external changes to the system, reflect
on the event occurrences, and adapt to the new condi-
tions. Ad hoc wireless sensor networks, in particular,
Copyright © 2010 SciRes. WSN
S. K. CHA ET AL.
Copyright © 2010 SciRes. WSN
405
must be designed with adaptation cap abilities that enable
them to handle a multitude of operating conditions. Re-
configuration in such systems presents significant chal-
lenges because of the severe constraints in energy, com-
putation, and communication resources. Runtime tech-
nologies that allow software to evolve as system re-
quirements and/or its environment change are critical to
the development and deployment of such systems. This
is in contrast to the current state-of-the-art in that it does
not allow embedded software to evolve at runtime.
3. 4M Group Management Middleware:
Auto-Configuration
A reconfigurable 4M group management service that we
called auto-configuration should perform the following
operations:
1) Layer structure configuration: it provides an identi-
fication of the 4M Group Management services available
for reconfigura ti on;
2) 4M management: It define 4M resources and pro-
vides 4M data entry and mapping functions between 4M
resources for veri ficat ion;
3) Tag Mapping: it returns an identification of the
available configurations of a registered tag from 4M re-
sources with editable function and new registration;
4) 4M Dynamic configuration: it consists of 4M dy-
namic mapping the internal behaviour of the Group
Management service currently in use by drag & drop
operation.
The next sections present the design an implementa-
tion of the reconfigurable management middleware ser-
vice-Auto-configuration.
3.1. Design
This section presents the design of the proposed recon-
figurable group management. The reconfigurable group
management is made up of 4 main elements: point man-
ager, real-time data manager, application template man-
ager and auto-configuration (see Figure 2). The RDB
stores the group management components available to be
used in the reconfiguration process. Meanwhile, the au-
to-configuration is responsible for the reconfiguration
itself. The application asks the auto-configuration to
change the group management service to another one
existing in the RDB.
The UML diagram shown in Figure 3 presents the
elements that make up the configurable group manage-
ment and their relationships. An auto-configuration has
pattern structure of the MVC (Model-View-Controller).
MainForm has composites work center for the user in-
terface and control for the internal operation. Each work
centers provides 4M Resource Definition, Resource
Scheduling, Resource Allocation, Executing & Data
Collection and it can define production type including
Batch, Continuous, Discrete for detailed implementation.
Figure 2. Overview of the auto-configuration.
S. K. CHA ET AL.
406
Figure 3. UML diagram of the middleware service.
General Implementation element consisted of process
parameters, run/down status data from production equip-
ment and operation plan, etc.
Figure 4 shows operation sequence from Auto-Con-
figuration, it can define resources from activity model
based on manufacturing execution management and
planning activity model and it runs monitoring step in
real-time for the tag definitions to data collection in
real-time from the 4M resources.
Function for execution of the Application Template
Manager provides plan, assign and define from 4M re-
sources in the work center. It runs under the mapping by
the point manager for data collection from defined tags
of the 4M resources. After this, the 4M data stores into
Figure 4. UML diagram of the middleware sequence.
Copyright © 2010 SciRes. WSN
S. K. CHA ET AL.
Copyright © 2010 SciRes. WSN
407
the real-time data manager.
The UML diagram shown in Figure 3 presents the
elements that make up the configurable group manage-
ment and their relationships. An auto-configuration has
pattern structure of the MVC (Model-View-Controller).
MainForm has composites work center for the user in-
terface and control for the internal operation. Each work
centers provides 4M Resource Definition, Resource
Scheduling, Resource Allo cation, Executing & Data Col-
lection and it can define production type including Batch,
Continuous, Discrete for detailed implementation.
3.2. Implementation
The u-Manufacturing IT layer consisted of 4 tier layers
as following required functions [11];
1) Layer 0: Interface layer provides wired/wireless
network through interface with manufacturing resources
of 4M using RFID/USN technology, touch screen and
mobile devices. Interface layer must take into considera-
tion high reliable multi-hop function for QoS (Quality of
Service), wireless security for 128 bit AES (Advanced
Encryption Standard) and how to digitalize 4M for pro-
duction information.
2) Layer 1: Execution layer provides gateway function
runs under middleware such as point manager for com-
mon protocol handling of various device controllers,
real-time data manager of 4M data handling by real-time,
standard API manager of integration with other applica-
tion and auto configuration of manufacturing r esource of
4 M. Functionality of Executi on lay er provi des work-order
processing, lot handling, scheduling handling, non-run
handling, material handling based 4 zero (It means Zero
Inventory, Waiting-time, Defect, Down-time) approach.
3) Layer 2: Management layer provides each plant of
ISO Standard MES KPI (Manufacturing Execution Sys-
tem Key Performance Indication) based on managing
factors for quality, cost and deliver, which are supported
functions, such as quality management, plan & schedul-
ing management, production execution management,
productivity management, inventory management, over
equipment efficiency management, production history
management, base data management in standard applica-
tions that are link data for 4M in production resources.
4) Layer 3: Planning layer provides standard applica-
tion function for ERP/SCM/CRM/PDM and so on.
In case of Layer 0 layer of interface, it runs under the
Tiny OS with WSN nodes and open embedded Linux
device called M2M system for data collection of 4M
resources interface by RS232C serial communication,
Digital I/O, Analog I/O and Ether- net with protocol
drivers such as TCP/IP, MIMOSA (Machinery Informa-
tion Management Op en System Alliance) and OPC (OLE
for Process Control) of interoperability stan dard commu-
nication.
Auto-Configuration in the Layer 1 of execution layer
with middleware services runs under the Windows NT
operating system with .Net framework.
The middleware provide user oriented customization
function with WPF (Windows Presentation Foundation)
that brings easy to use and faster response for recon-
figurable work center in production resources from 4M
against with changing various production orders and
models.
Figure 5. The u-manufacturing IT layers.
S. K. CHA ET AL.
Copyright © 2010 SciRes. WSN
408
3.3. Operation of Auto-configuration
Figure 6 shows user benefits of before and after about
operation of the Auto-configuration. Traditional opera-
tion environment, it is inevitable to repeat tasks of the
redefine for device controller’s profile with re-cabling
between device controllers and data collection devices
and testing with modification of existing contro l program
in servers when it becomes reconfiguration of 4M re-
sources.
-Defini t i on of 4M T ag par am ete r s
(COMM_ID=INJECTION01&Equipme
nt=INJECTION01&E quipPointTyp e=I
NTERFACE&ClassName=ProcessV3)
-Defint i on of T a g dat a
-Ne e ded a ddt i o n a l w o r k s fo r
equipment moving and
changing i n production
processes
Definition
of 4M dat a
-Refer to similar applicat ion
-Modofying pr ogram
-Program testing
Modifying
exsiting
program
Moving &
Changing Define for T ag
Change
Moving
Change of
production
model
Ca bli ng and
Test
Once
definition
of 4M dat a
Au to-
Conif ugurat
io n
Change &
move
-Automatic generate tags based on 4M parameters
-Only Drag & Drop
Change
Moving
Change of
productio n
model
-Needed additonal works for equipement
moving and changing in production
processes
-Definition of 4M Tag paramete rs
(COMM_ID=INJECTION01&Equipment
=INJ E CT ION0 1&E qui pP ointT y pe =INTE
RFACE&ClassName=ProcessV3)
-Defintio nof Tag data
Figure 6. Before and after for using an auto-configuration.
S. K. CHA ET AL.409
Figure 7. An example for operation procedures of the auto-configuration.
Such a painful tasks from view point of the production
technology, it requires knowledge based and know how
for fine turning. Major functionality of the Auto-con-
figuration is able to handling above painful repeat work
that provides reconfigurable 4M group management.
Users does not need how to define the profile of de-
vice controller based on 4M resources, it needs just drag
and drop on the grap hic screen of produ ction pro cess if it
once define the profile of 4M resource into the Au-
to-configuration.
Figure 7 shows simple operation procedures of the
Auto-configuration. Base management of the Auto-con-
figuration is work cen ter in production that link to struc-
ture of lines, plants. This screen presents how to define
the 4M resources based on work center. Therefore, user
does not pay attention to definition of 4M resources in-
cluding man, machine, material, method information that
has been on ce defi n ed 4M resou r ce s .
4. Conclusions
This paper has presented a reconfigurable group man
agement middleware service. The proposed service is
designed according to selected guidelines and then im-
plemented in nesC (TinyOS).
The application of the reconfigurable 4M group man-
agement middleware service is illustrated through the
implementation of an Auto-configu ration app lication that
uses the proposed service. In this particular case, the re-
configurable group management middleware service is
used to manage the groups that monitor the 4M resources
being tracked.
Major function of the Auto-configuration for 4M
group management with middleware services including
point manager, real-time data management, standard API
and application template management using wireless
sensor network provides RMS for rapidly response from
various production orders and models from customers to
be establishing manufacturing collaboration.
5. Acknowledgements
This work is supported by the Industry Foundation pro-
ject from the Ministry of Knowledge Economy in the
Korean Governm e nt .
6. References
[1] B. Krishnamachari, D. Estrin and S. B. Wicker, “The
Impactof Data Aggregation in Wireless Sensor Net-
works,” Proceedings of the 22nd International Confer-
ence on Distributed Computing Systems, Vienna, 2-5 July
2002, pp. 575-578.
[2] SMBA, “Report on E-Manufacturing Project,” Korea Insti-
tute of Industrial Technology Evaluation and Planning
Copyright © 2010 SciRes. WSN
S. K. CHA ET AL.
410
(SMBA), Seoul, Vol. 1, June 2004, pp. 34-40.
[3] I. F. Akyildiz, W. Su, Y. Sankarasubramaniam and E.
Cayirci, “A Survey on Sensor Networks,” IEEE Commu-
nications Magazine, Vol. 40, No. 8, August 2002, pp. 102-
114.
[4] S. Lindsey and C. S. Raghavendra, “PEGASIS: Power
Efficient Gathering in Sensor Information Systems,”
Proceedings of the IEEE Aerospace Conference, Mon-
tana, 9-16 March 2002, pp. 1125-1130.
[5] W. Heinzelman, A. Chandrakasan and H. Balakrishnan,
“Energy-Efficient Communication Protocol for Wireless
Microsensor Networks,” Proceedings of the IEEE Hawaii
International Conference on System Sciences, Hawaii, Vol.
8, 4-7 January 2000, pp. 8020-8029.
[6] J. Liu, J. Reich, E. Cheung and F. Zhao, “Distributed
Group Management in Sensor Networks: Algorithms and
Applications to Localization and Tracking,” Telecommu-
nication Systems, Vol. 26, No. 2-4, 2004, pp. 235-251.
[7] J. Liu, J. Reich, E. Cheung and F. Zhao, “Distributed
Group Management for Track Initiation and Maintenance
in Target Localization Applications,” Proceedings of the
2nd International Workshop on Information Processing in
Sensor Networks (IPSN’03), Palo Alto, Vol. 2634, 22-23
April 2003, pp. 113-128.
[8] L. Luo, T. F. Abdelzaher, T. He and J. A. Stankovic,
“EnviroSuite: An Environmentally Immersive Program-
ming Framework for Sensor Networks,” ACM Transac-
tions on Embedded Computing Sy stem (TECS), Vol. 5, No.
3, August 2006, pp. 543-576.
[9] T. Abdelzaher, B. Blum, Q. Cao, Y. Chen, D. Evans, J.
George, S. George, L. Gu, T. He, S. Krishnamurthy, L.
Luo, S. Son, J. Stankovic, R. Stoleru and A. Wood, “En-
viroTrack: Towards an Environmental Computing Para-
digm for Distributed Sensor Networks,” Proceedings of
the 24th International Conference on Distributed Com-
puting Systems, Tokyo, 23-26 March 2004, pp. 582-589.
[10] J. Liu, D. Sacchetti, F. Sailhan and V. Issarny, “Group
Management for Mobile Ad Hoc Networks: Design, Im-
plementation and Experiment,” Proceedings of the 6th
International Conference on Mobile Data Management
(MEM’05), Ayia Napa, 9-13 May 2005, pp. 192-199.
[11] S.-K. Cha, J.-S. Choi, C.-S. Choi and W.-S. Kim,
“Counter Plan and Strategy for Small & Medium Enter-
prises for Ubiquitous Age,” KIMI, December 2004, pp.
63-171.
Copyright © 2010 SciRes. WSN