DATA INTEGRATION SOLUTION FOR PAPER INDUSTRY
A Semantic Storing Browsing and Annotation Mechanism for Online Fault Data
Sergiy Nikitin, Vagan Terziyan
Agora Center, University of Jyväskylä, Mattilanniemi 1, Jyväskylä, Finland
Jouni Pyötsiä
Metso Automation
Keywords: Semantic Web, Semantic Storage, Information Integration, Paper Industry.
Abstract: A lot of IT solutions exist for simplification and time saving of industrial experts’ activities. However, due
to large diversity of tools and case-by-case software development strategy, big industrial companies are
looking for an efficient and viable information integration solution. The companies have realized the need
for an integrated environment, where information is ready for extraction and sophisticated querying. We
present here a semantic web-based solution for logging and annotating online fault data, which is designed,
and implemented for a particular business case of a leading paper machinery maintenance and automation
company.
1 INTRODUCTION
Rapid changes and discontinuities in the 21st
century business environment will challenge
companies with the growing demand for competitive
advantages within their business solutions. To
ensure high flexibility, sustainable growth and
profitability, companies have to search for new
innovative approaches to products and services
development. New innovative business solutions call
for strong integration of automation technology,
information and communication technology (ICT),
and business processes. At the same time, embedded
intelligence in different machines and systems gives
new possibilities for automated business process
operation over the network throughout the machines
and systems life cycles.
The new emerging remote service solutions
imply that products transform into life cycle services
and these services, in turn, transform into customers'
service processes. Business messages coming from
intelligent machines and systems drive these
processes, utilizing embedded intelligence and ICT
solutions.
In the future, a variety of collaborative resources,
like intelligent machines, systems and experts, will
create a huge amount of new information during the
life cycles of machines and systems. Message flow
management and compression to on-line knowledge
are already a demanding issue for the logging of
product-related activities. On the other hand,
optimization requirements demand more effective
knowledge utilization and the speeding up of
network-based learning in the process of
collaboration between different resources.
Industry challenges the IT-sector with the new
requirements that are dictated by the need to offer
essentially new services to customers in order to be
competitive in the market. These requirements may
become hard to meet using conventional tools and
approaches. The growth in the information volumes
we want to store and process by integrating data
from different sources leads to an unprecedented
level of complexity. Modern Enterprise Resource
Planning (ERP) systems are trying to provide
integrated solutions for large companies. However
the installation and adjustment of such systems may
take a half a year, involving hundreds of consultants
and subcontractors.
The current trend towards more open service-
based computing environments is a new approach to
componentization and components distribution.
Service-oriented architecture aims to achieve a new
191
Nikitin S., Terziyan V. and Pyötsiä J. (2007).
DATA INTEGRATION SOLUTION FOR PAPER INDUSTRY - A Semantic Storing Browsing and Annotation Mechanism for Online Fault Data.
In Proceedings of the Fourth International Conference on Informatics in Control, Automation and Robotics, pages 191-194
DOI: 10.5220/0001621801910194
Copyright
c
 SciTePress

level of reusability and business process flexibility;
however, to ensure the interoperability between the
components we need a common “glue” that would
adjust the semantics of the data being exchanged.
The Semantic Web technology (Berners-Lee, T., et
al., 2001) introduces a set of standards and
languages for representation of a domain model with
the explicit semantics. The main instrument of
domain model construction is ontology, which
allows for domain data representation in a
formalized and unified way.
In this paper we present a solution that utilizes
Semantic Web technology to provide a tool for
online maintenance data browsing, analysis and
annotation. We utilize experience obtained in the
SmartResource project (SmartResource, 2006) and
apply the General Adaptation Framework (Kaykova
et al., 2005) to align data with the domain ontology.
The paper is organized as follows: In the next
section we describe the paper machinery ICT
infrastructure, and Section 3 presents the solution we
have developed using Semantic Web tools and
standards. We end with conclusions and future work.
2 IT INFRASTRUCTURE IN
PAPER INDUSTRY
Metso Corporation is a global supplier of process
industry machinery and systems as well as know-
how and aftermarket services. The corporation's core
businesses are fiber and paper technology, rock and
minerals processing, and automation and control
technology. Metso's strategy is based on an in-depth
knowledge of its customers' core processes, close
integration of automation and ICT, and a large
installed base of machines and equipment. Metso's
goal is to transform into a long-term partner for its
customers. Based on the remote service
infrastructure, it develops solutions and services to
improve efficiency, usability and quality of
customers' production processes throughout their
entire life cycles.
2.1 Remote Service Infrastructure
Metso's remote service infrastructure consists of a
service provider's Central Hub and several
customers' Site Hubs, which are integrated over the
network (see Figure 1).
The key issues in a Site Hub solution are: open
standards, information security, reliability,
connectivity and manageability.
GPRS/
GSM
Single Devices
Site
Hub
Legacy
Systems
Local Administrator
Interface
Customer Site 1
Firewall
Legacy
Systems
Site
Hub
Local Administrator
Interface
Customer Site n
Firewall
Site
Hub
Local Administrator
Interface
Customer Site 2
F
i
rewal
l
Central
Hub
Global
Administrator
Interface
Metso Site
Firewall
Service Network
Figure 1: Site Hub network architecture.
Message Center is the main component of the
Site Hub. It checks the validity of messages and
routes them to the correct receivers. Messaging in
Site Hub is based on Web Services technology.
Hub-based integrated infrastructure combined
with secure connectivity allows easy incorporation
of new business logic on both customer and Metso
sites. A messaging mechanism between customers
and Metso provides a very flexible medium for
information exchange and new service provisioning.
3 LOGGING AND ANNOTATION
OF MAINTENANCE DATA
The main purpose of the system we present here is
to store alarm data, generated by paper machine
monitoring systems. When an alarm happens, a
SOAP/XML message (SOAP, 2003) is generated
and sent to the Site Hub, which then forwards it to
the Central Hub. We have established a message
flow from the Central Hub to the computer at the
university network, where messages are processed
by our system.
3.1 Architecture of the System
The system can be divided into two main
subcomponents – Message Handler and Message
Browser (see Figure 2).
Message Handler receives and processes
SOAP/XML messages from customers. It invokes
the Adapter to transform the XML content into an
ICINCO 2007 - International Conference on Informatics in Control, Automation and Robotics
192

RDF-graph (RDF, 2004) object and store it in the
Sesame RDF storage (Sesame).
Web
Service
Adapter
MESSAGE
BROWSER
HTML
Query
Results
XML
SeRQL
XML
Graph
MESSAGE
HANDLER
Browsing and Annotation tool
Central
Hub
Metso Site
RDF graph
Application server
SOAP
message
Figure 2: Architecture of the system.
The RDF storage contains an Ontology that plays
the role of a schema for all data within the storage.
Based on the analysis of SOAP/XML messages, we
have defined main concepts (classes) with the
corresponding properties (see Figure 3).
Message
-messageUID
-securityLevel
-time
-hash
-hasAlarm
-receiverGroup
-hasMessageType
-messageSender
-messageReceiver
ExpertAnnotation
-annotationName
-numOfMessagesReferred
-annotationTime
-annotationDescription
-madeByExpert
-messageReference
Expert
-expertName
-hasAnnotation
Alarm
-value
-failureDescription
-lowLimit
-situationDescription
-tag
-alarmTime
-highLimit
-status
-productionLine
-alarmSource
-customer
-measurementUnit
Figure 3: Ontology classes.
The Message class describes such message
properties as message sender and receiver, message
reception time, etc. The Message class also refers to
an Alarm class, which contains information about
the reason for message generation, such as
measurements of sensors, status data and exact
module of the production line where the alarm
happened. The ExpertAnnotation class defines the
structure for labelling groups of messages with an
expert’s decision and has references to instances of
the Message class.
The Message Browser component provides a
web-based interface for browsing and filtering
messages stored in the RDF-storage, according to
user-defined filtering criteria.
The purpose of message filtering is to distinguish
the groups of messages leading to exceptional
situations. The expert provides annotations for
message groups which are stored to the RDF-storage
and that can be used as samples for machine learning
algorithms. The client-server interaction is
implemented using AJAX technology (Garrett,
2005), which provides a more dynamic script-based
interaction with the server (see Figure 4). When a
user performs any action that requires invocation of
server functionality, the script on a client side wraps
the required parameters into XML format and sends
it to the server. For example, in order to filter the
messages, a user selects the needed parameters and
specifies parameter values within the corresponding
textboxes (see Figure 4).
MESSAGE BROWSER
Application server
Web
Browser
XML
HTML
Control
Servlet
AJAX
Script
HTML
Query
Results
XML
SeRQL
Browsing
XML
Graph
Annotation
Figure 4: Client-server interaction.
On the server side, the Control Servlet handles
the XML document. For filtering, it generates a
SeRQL query (Broekstra, 2004) and executes it. On
the client side, a dedicated callback script function
processes the response and shows the result in a web
browser.
3.2 Integration with the Agent
Platform
We realize that the extension of the system will
challenge the complexity of development and
maintenance. That is why, following the autonomic
computing paradigm (Kephart, 2003), we have
tested agent-based scenario (see Figure 5)
implemented on a JADE agent platform
(Bellifemine; 2001). We have assigned an agent to
manage RDF-storage activities (Metso Storage
Agent) and provided a Metso Expert Agent to
interact with a maintenance expert.
DATA INTEGRATION SOLUTION FOR PAPER INDUSTRY - A Semantic Storing Browsing and Annotation
Mechanism for Online Fault Data
193

Web
Service
MESSAGE
BROWSER
HTML
Query
Results
XML
SeRQL
XML
Graph
MESSAGE
HANDLER
Browsing and Annotation tool
Central
Hub
Metso Site
Application server
SOAP/XML
METSO
STORAGE
AGENT
METSO EXPERT AGENT
KML
SOAP
Real-Time
Monitoring
tool
Adapter
SOAP/XML
RDF graph
Figure 5: Agent-enabled system.
The messages coming from customers are
handled by the Metso Storage Agent, which
incorporates Adapter to perform transformation and
storage. Then, the Metso Storage Agent sends the
message to the Metso Expert Agent, which updates
the situation on a Real-time Monitoring Tool and
provides an expert with the message content and a
link to the browsing and annotation tool.
4 CONCLUSIONS
Although we have succeeded with the
implementation of the solution presented here, there
are still many issues to cope with in order to meet
key industrial requirements, such as scalability,
maintainability and robustness. RDF-storages can
handle billions of triples, but there are no mature
semantic storage-oriented development patterns or
guidelines. Nevertheless, the simplicity and
efficiency of querying, as well as model extending,
provide incontestable arguments in favour of
semantic data storages. The ontological domain
model brings more benefits to customers when there
are more sources integrated. However, the
complexity of such a system, if developed using
conventional approaches, will be too burdensome to
maintain and extend. In order to distribute the
complexity, we introduce self-manageable entities in
the agent-based communication scenario.
ACKNOWLEDGEMENTS
This research has been supported by the
SmartResource project, funded by TEKES, and the
industrial consortium of Metso Automation,
TeliaSonera and ABB. The preparation of this paper
was partially funded by the COMAS graduate school.
REFERENCES
Bellifemine, F., Poggi, A., and Rimassa, G. 2001. JADE: a
FIPA2000 compliant agent development environment.
In Proceedings of the Fifth international Conference
on Autonomous Agents (Montreal, Quebec, Canada).
AGENTS '01. ACM Press, New York, NY, 216-217.
DOI= http://doi.acm.org/10.1145/375735.376120
Berners-Lee, T., Hendler, J., and Lassila, O. (2001) The
Semantic Web, Scientific American, Vol. 284, No. 5,
pp. 34-43.
Broekstra, J., Kampman A., and F. van Harmelen.
Sesame: An Architecture for Storing and Querying
RDF Data and Schema Information. In D. Fensel, J.
Hendler, H. Lieberman, and W.Wahlster, editors,
Semantics for the WWW. MIT Press, 2001.
Broekstra, J., Kampman, A. SeRQL: An RDF query and
transformation language. In Proceedings of the
International Semantic Web Conference, ISWC 2004,
Hiroshima, Japan.
Garrett, J., Ajax: A New Approach to Web Applications
(white paper).,http://www.adaptivepath.com/publica-
tions/essays/archives/000385.php, February 2005.
Kaykova O., Khriyenko O., Kovtun D., Naumenko A.,
Terziyan V., Zharko A., General Adaption
Framework: Enabling Interoperability for Industrial
Web Resources, In: International Journal on Semantic
Web and Information Systems, Idea Group, ISSN:
1552-6283, Vol. 1, No. 3, July-September 2005,
pp.31-63.
Kephart J.O., Chess D.M., 2003. The vision of autonomic
computing, IEEE Computer, Vol. 36, No. 1, pp. 41-50
RDF – Resource Description Framework, A W3C
Recommendation, Feb 2004, http://www.w3.org/RDF/
SmartResource – a TEKES funded project, http://www.
cs.jyu.fi/ai/OntoGroup/SmartResource_details.htm
SOAP – Simple Object Access Protocol, A W3C
Recommendation, 2003, http://www.w3.org/TR/soap/
ICINCO 2007 - International Conference on Informatics in Control, Automation and Robotics
194