KNOWLEDGE BASED SERVICES FOR DEVICES
IN AUTOMATION
Muhammad Baqar Raza, Robert Harrison and Thomas Kirkham*
Loughborough University, Loughborough, LE11 3TU, Leicestershire, U.K.
*The University of Nottingham, NG7 2RD, Nottingham, U.K.
Keywords: Ontology, SOA, Semantic web service (SWS), Knowledge based system.
Abstract: The integration of manufacturing design and production processes around aggregated shared knowledge
improves production efficiency. In this paper, planning level faults on the assembly lines or conflicts in
product design are identified and picked up in real time via the use of integrated knowledge based services.
Issues with the supply chain can also be fed into the model by linking the services to Enterprise Resource
Planning (ERP) systems. In the production process, errors and/or faults are fed back into the knowledge
base system to aid confident future planning. This approach allows more targeted alerts and reports of
failures, empowering the production operative and allowing more problems to be solved at the source of
origination.
1 INTRODUCTION
Service Oriented Architecture (SOA) has emerged
as a reliable distributed computing method. Web
services are considered the best implementation
method of SOA as they are loosely coupled and
platform independent. Nevertheless to construct
machine processable ‘XML‘ in a complex product
manufacturing enterprise, a higher level of semantics
is required which is provided by ontology. An
ontology is commonly defined as:
a formal, explicit
specification of a shared conceptualization (Gruber,
1993). More specifically, an ontology is an
engineering artifact composed (i) of a vocabulary
specific to a domain of discourse, and (ii) of a set of
explicit assumptions regarding the intended meaning
of the terms in the vocabulary for that domain.
In the manufacturing domain, the relationship
between the design phase of a product and its
creation on a production line is vital for
manufacturing efficiency. Errors in the design or
failure to create an assembly line that suits the
product design can create delays in the process as re-
design or re-configuration occurs. These revisions
have impacts on both the supply chain and overall
manufacturing/assembly output.
The Loughborough University in collaboration
with Ford UK have been investigating how
ontologies and SWSs could be used to improve Ford
production output. These investigations lead to the
development of a software framework, which will
facilitate the integration of product design and
production line configuration. This framework
builds on existing web services developed during
projects such as SOCRADES (Kirkham, 2008).
2 RELATED WORK
The research work on ontologies and the semantic
web has started intruding into real industrial
problems from pure academia work. SOA and its
implementation using Web Services (WS) have
raised significant interest as a technology facilitator
for encapsulating industrial devices as loosely
coupled and interoperable units (Jammes, 2005).
Semantic web is based upon ontology which
provides semantics and reasoning support for
intelligent retrieval and discovery of manufacturing
resources (Ming, 2003).
The use of ontologies in the manufacturing
domain to form intelligent semantic web services to
improve productivity are emerging (Lukibanov,
2005); (Ajit et al, 2004). These ontologies are
applied to a variety of points in the manufacturing
lifecycle ranging from design and production phases.
133
Baqar Raza M., Harrison R. and Kirkham T..
KNOWLEDGE BASED SERVICES FOR DEVICES IN AUTOMATION.
DOI: 10.5220/0003422901330138
In Proceedings of the International Conference on Knowledge Engineering and Ontology Development (KEOD-2011), pages 133-138
ISBN: 978-989-8425-80-5
Copyright
c
 2011 SCITEPRESS (Science and Technology Publications, Lda.)

These innovations are significant for companies like
the Ford Motor Company which is currently facing
challenges in maintaining competitive advantage
when faced with competitors who can produce
products on a large scale and lower cost from less
developed economies. To counter this, western
manufacturing has looked towards innovation in
manufacturing process embracing movements such
as agile manufacturing (Yusuf et al., 1999). The
enablement of high quality and customized
production through agile manufacturing requires
changes in production process. Central to the
production line is the ability for it to support change
and reconfiguration (Harrison et al., 2006). Time
saved during this process has a direct impact on the
efficiency of an organization and is a key focus of
research.
To date re-configuration and assembly line
flexibility has been managed in a variety of ways.
Ford’s Powertrain manufacturing/assembly follows
this pattern with assembly lines consisting of a
variety of vendor specific machinery and control
software. Thus integration of machines and lines is a
challenge and integration with enterprise
management such as ERP software is rarely
achieved.
3 DEVICE LEVEL SERVICES
By encapsulating the functionality of devices as a
service, new systems can be created by composing
simple services into complex applications with a
minimum of programming efforts (Lastra et al.,
2006). Semantic Web is a very strong and flexible
knowledge representation method which can express
entity structure, properties and the causal link
between entities explicitly and concisely (Zhihong,
2002). Semantic web services can foster the
integration of heterogeneous production devices and
of a mix of architectures in systems which would be
chaotic from an ICT perspective (Lastra et al.,
2005).
Embedded web services have developed around
the initial innovation presented by tools such as the
GSOAP toolkit (Engelen and Gallivan, 2002) to a
more detailed and defined toolkit in the form of
DPWS (OASIS, 2009); (Schlimmer, 2010). This
web service toolkit although largely based on
GSOAP, presents support for standards to enable
‘eventing’, subscription and notification of events
enabling a more efficient lower layer of device
based communication. The ability for services to be
present a manufacturing device level allows the use
of semantic web services at real time production
level. This semantic management will impact the
management of the line by injecting the ability to
link the data from the line along with its behaviour
into knowledge based services.
This greater level of management and knowledge
based reasoning on the line will enhance wider
production processes. For example, device level
service behaviour can be factored into ERP and
supply chain management services to help plan
production output.
4 CONFIGURATION
MANAGEMENT
In order to support the execution of services on the
assembly line, supporting services need to be in
place before, during and after device level execution.
In order to make this process dynamic and
transferable, knowledge of service design and
functionality is required. In the BDA project
ontologies were developed to serve this specific
requirement using semantic web services.
Machine, line, component and product data has
been captured into ontologies using Protégé and
OWL (Knublauch et al., 2004); (Martin, 2004).
Ontologies are a key part of a broader range of
semantics based technologies and automated
inference that arose within the Artificial Intelligence
community. Many different representation
formalisms have been explored and reasoning
engines developed. There are different kinds of
ontologies, the ontologies developed for the current
research can be classified as domain ontology as
well as task ontology. Task ontology is developed
for a specific task within a certain domain.
Creating ontology for a domain provides an
opportunity to analyse domain knowledge, make
domain assumptions explicit, separate domain
knowledge from operational knowledge, provide
common understanding of the information structure
and enable reuse of domain knowledge.
The ontologies in the current research represent
how the Powertrain assembly line is constructed and
example ontology of a production task on an
assembly line can be seen in Figure 1.
A visualization of a simple ontology to represent
a typical assembly task is illustrated in Figure 1. An
ontology in the manufacturing domain is used for
controlled vocabulary, consistency in data quality
and navigation through disparate information
sources through mediation. Ontology is an
KEOD 2011 - International Conference on Knowledge Engineering and Ontology Development
134

intensional formal semantic structure which encodes
the implicit rules constraining the structure of a
piece of reality (Bachimont, 2000).
Figure 1: Ontology for a production line task.
An ontology is seen as a meta-level specification
of a logical theory, in the sense that it specifies the
‘architectural components‘ or ‘primitives‘ used
within a particular domain theory (Wielinga, 1994).
Therefore an ontology is a representation of
components and their allowed interactions, with the
purpose of providing an explicit framework in which
to elaborate the rest of the system. Ontologies
overcome data heterogeneity. Data sources can be
heterogeneous in syntax, schema, or semantics, thus
making data interoperation a difficult task. Syntactic
heterogeneity is caused by the use of different
models or languages. Schematic heterogeneity
results from structural differences. Semantic
heterogeneity is caused by different meanings or
interpretations of data in various contexts. To
achieve data interoperability, the issues posed by
data heterogeneity need to be eliminated. Currently,
the basic ontology model is defined along three
axes: relationships, hierarchy, and abstraction, as
shown in Figure 2.
Figure 2: Ontology structure outline for PPR domains.
On the ontology level all the specific domain
concepts, attributes, constraints, and rules are
defined for the related entities i.e. products,
processes and resources (PPR). Ontologies expose
implicit knowledge that has been previously hidden
in domain assumptions or in the implementation of
an application. Figure 3 describes 04 major
functions of an ontology.
Figure 3: Ontology functions.
To support the decision making, an ontology
model is proposed that include three representation
levels: the underlying knowledge representation
level, the ontology concept level and the
instantiation level. The domain ontologies are
divided into generic and user specific concepts. For
example, a library of physical models may be
represented with the help of an ontology. To each
component in the library, information is attached
that describes the function of the component (for
example, an engine has the function “generate
power”). If this information is represented explicitly,
an engineer may be able to search the library for
component models that fulfill a certain function,
rather then designing a new one. In the ontology-
based approach, domain specific ontologies are
modeled to specify semantics of terminology
systems in a well-defined and unambiguous manner.
A model is a simplified representation of a system
KNOWLEDGE BASED SERVICES FOR DEVICES IN AUTOMATION
135

intended to enhance our ability to understand,
predict and control the behavior of a system. Based
upon the presented categorisation, ontology of the
complete assembly line was modeled and an
example of the abstract concepts is shown below in
Figure 4.
Figure 4: Ontology for a production line - snapshot of line
in Protégé 4.0.
Key concepts introduced into the ontology are
work piece, work piece characteristics, workstation,
workstation characteristics, assembly operation,
operator and end product. Based upon these
concepts, relations among products, processes and
resources are established into the ontology which
constitutes manufacturing backbone. The current
assembly line reconfiguration approach is largely
based on the skill and knowledge of engineers rather
than the actual process involved. Whenever there is
any change in the product it is then essentially
engineer’s responsibility to examine the needs of the
reconfigured system to support the new product
(Raza et al., 2011).
Current automation systems fail to meet business
requirements (Raza and Harrison, 2011). The
assembly lines, such as powertrain assembly line for
automotive engine, have a limited capacity to
produce variety of products (engines). The built in
capability to deal with variety has to be limited to
justify investment and is a trade-off between the
inevitable but unpredictable changes and the
increased cost of flexibility (Raza and Harrison,
2011). One of the main reasons for automation
systems and especially the assembly automation
systems in automotive sector not fulfilling the
business requirements is that the relational
knowledge base among products, processes and
resources is either non-existent or not properly
designed / used. There is a genuine industrial
requirement to establish relationships among PPR
domains to readily assimilate the requisite
information for any change in product at any time
(Raza and Harrison, 2011).
Therefore it is imperative to create a knowledge
base among PPR domains to fully utilise the
available knowledge. Ontologies are not only useful
for helping solving the information overload
problem, but can be used for a variety of different
applications, such as sharing explicit knowledge,
increase communication, and help in natural
language understanding. With the help of practical
knowledge in ontology, a quick evaluation of many
potential resource configurations is possible as well
as the best suited one for a changed product (Raza
and Harrison, 2011). There are no platform
independent application tools available for
modelling the PPR information explicitly neither
does any tool exist to link PPR relational
information unequivocally (Raza and Harrison,
2011).
Ontology assists the mapping of data from a
variety of vendors and sources that constitute a
production line. This enables a layer of knowledge
to exist over the distributed data sources that are
needed to help configure, support and run the
production process / assembly line. For example a
breakdown on the line can be followed up by a
detailed report on faulty components based on line
data and distributed component design records from
various vendors. This information could aid the
recovery and repair time for the line as shown
below:
KEOD 2011 - International Conference on Knowledge Engineering and Ontology Development
136

Figure 5: Knowledge based service use in assembly line
management.
Thus a layer of distributed knowledge is wrapped
and exposed as semantic web services as shown in
Figure 5. These services abridge knowledge to data
to be represented in web service processes thus
aiding re-configurations of lines. Another example is
the integration of new product specifications in
existing line designs. Here the new specification can
be accessed using web service calls and compared
using an OWL ontology against the existing line. To
automate (fairly) the task of assembly line
configuration / reconfiguration, product and resource
(line) link points need to be defined at early stages
of design and made available easily to be searched,
analysed and implemented on ‘when and where
required’ basis (Raza et al, 2011). This will enable
automated comparison of new products against
existing line data also expressed as ontologies. This
method will ensure conflicts in the product design
with the line can be found and picked up at earlier
program stages. It provides a broad encompassing
structure of the domains that is handy and practical,
a knowledge based support on every engineer’s desk
for making intrigue calculations and quick relational
constraints. As a result, a PPR domains can be
modelled realistically and constitute a complete
knowledge management cycle of the automotive
domain for the assembly line design activity, as
depicted in Figure 6.
Figure 6: Product, Process and Resource integration
constitute manufacturing knowledge backbone.
5 CONCLUSIONS AND FUTURE
WORK
In this paper we explored a methodology that
incorporates and improves distributed intelligence at
the shop floor level. Currently the ontologies are
used to aid the re-configuration of the assembly lines
for new products. Live use of the ontologies has
been
limited to a few basic error conditions. As the
development and use of the knowledge based
services evolve they will be used as support in the
entire production lifecycle. However, to make it a
reality the data has to be captured or represented into
ontologies from a wide variety of proprietary and
legacy data sources throughout Ford, UK.
Using knowledge based services a new layer of
manufacturing management can be envisaged that
will help the entire production lifecycle. This layer is
enhanced by the use of device level web services
which will enable the live use of knowledge in
automated decision making on assembly lines. To
date, the use of knowledge based services has been
limited to the design phase of machines. By using
the technology in SOCRADES and BDA projects
this approach can be widened out to encompass the
whole process. This will standardise production
management and responses to errors that will reduce
cost and improve manufacturing efficiency.
The recently started AESOP (ArchitEcture for
Service Oriented Process Monitoring and Control)
project EU FP7 (AESOP, 2011) is now extending
the application of the approach in the process control
domain. AESOP deals with several key challenges
that arise such as real-time web services,
interoperability, plug and play, self-adaptation,
reliability, cost-effectiveness, energy-awareness,
KNOWLEDGE BASED SERVICES FOR DEVICES IN AUTOMATION
137

high-level cross-layer integration and cooperation,
event propagation, aggregation and management.
ACKNOWLEDGEMENTS
The authors gratefully acknowledge the support of
UK EPSRC and Loughborough University’s
IMCRC through Business Driven Automation
(BDA) project, the EU SCORADES FP6 and the
ongoing EU AESOP FP7 project. We would like to
thank all the project participants and engineers who
have contributed in this research from
Loughborough University and participant companies
especially Ford Motor Company and Schneider
Electric.
REFERENCES
Ajit S., Sleeman D., Fowler D., and Knott D., ConEditor:
Tool to Input and Maintain Constraints, 14th
International Conference on Engineering Knowledge
in the Age of the Semantic Web, Proceedings of EKAW
2004, pp 466-468.
Bachimont B., Engagement sémantique et engagement
ontologique: conception et réalisation d'ontologies en
ingénierie des connaissances, In Charlet, Zacklad,
Kassel, Bourigault, Ingénierie des connaissances
Evolutions récentes et nouveaux défis. Eyrolles, 2000.
Engelen R. A. van and Gallivan K.., “The gSOAP Toolkit
for Web Services and Peer-To-Peer Computing
Networks” in the proceedings of the 2nd IEEE Intl
Symp.on Cluster Computing and the Grid, pages 128-
135, May 21-24, 2002, Berlin, Germany.
Gruber, T., A translation approach to portable ontology
specifications. Knowledge Acquisition 5:199-220;
1993
Harrison, R., A. W. Colombo, A. A. West and S. M. Lee,
Reconfigurable modular automation systems for
automotive power-train manufacture. International
Journal of Flexible Manufacturing Systems, 2006.
18(3): p. 175-190.
Jammes F., Smit H., Service-Oriented Paradigms in
Industrial Automation, IEEE Transactions on
Industrial Informatics, Vol 1(1), pp. 62-70, Feb 2005.
Kirkham T., Savio D., H. Smit, Harrison R., R. P.
Monfared, and P. Phaithoonbuathong, "SOA
Middleware and Automation: Services, Applications
and Architectures" in 6th International Conference on
Industrial Informatics (IEEE, INDIN 2008): IEEE
Comp. Soc., 2008.
Knublauch H., Fergerson R., Noy, N., and M., Musen. The
protégé OWL plugin: An open development
environment for semantic web applications. In Proc.
of ISWC, 2004, number 3298 in LNCS, pages 229–
243.
Lastra J. L. Martinez, "On Future Self-Orchestrating
Manufacturing Systems", Position paper for the FP7
Workshop on The Agile, Wireless Manufacturing
Plant, Brussels, February 10th, 2005.
Lastra J. L. Martinez, I. M. Delamer, "Semantic Web
Services in Factory Automation: Fundamental Insights
and Research Roadmap", to appear in IEEE
Transactions on Industrial Informatics, Feb 2006.
Lukibanov O., Use of Ontologies to Support Design
Activites at DaimlerChrysler, 8th Intl. Protégé
Conference, 2005.
Martin D., Burstein M., Hobbs J., Lassila O., D.
McDemott, S. McIlraith, S. Narayanana, M. Paolucci,
B. Parsia, T. Payne, E. Sirin, N. Srinivasan and K.
Sycara. Ming C., L. Lin, and J. Dong. “Intelligent
Service Matching in Internet-Based Manufacturing.”
CIMS, vol.9, no. 7, pp. 589-594, March 2003.
OASIS DPWS standard page: http://docs.oasis-
open.org/ws-dd/ns/dpws/2009/01
OWL-S: Semantic markup for web services. Member
Submission 22, W3C, November 2004.
Raza M. B., Harrison R., Kirkham T., –Knowledge Based
Flexible and Integrated PLM System at Ford –Journal
of Information & Systems Management, Volume 1,
Number 1, March 2011, pp 8-16
Raza Muhammad Baqar, Robert Harrison, Information
modeling and knowledge management approach to
reconfiguring manufacturing enterprises, International
Journal of Advanced Information Technology (IJAIT)
Vol. 1, No. 3, June 2011.
Raza Muhammad Baqar, Robert Harrison, Ontological
Knowledge Based System for Product, Process and
Resource Relationships in Automotive Industry,
Proceedings of the 1st International Workshop on
Ontology and Semantic Web for Manufacturing
(OSEMA, 2011), co-located with the 8th Extended
Semantic Web Conference, ESWC2011, Heraklion,
Crete, Greece, May 29, 2011.
Schlimmer J., A Technical Introduction to DPWS
http://msdn2.microsoft.com/enus/library/ms996400.as
px
Wielinga, B. and Schreiber, A. Th. and Jansweijer, W. and
Anjewierden, A. and van Harmelen, F. Framework
and formalism for expressing ontologies. ESPRIT
Project 8145 KACTUS, Free University of
Amsterdam, deliverable DO1b.1, 1994.
Yusuf, Y. Y., Sarhadi, M. and Gunasekaran, A., 1999.
Agile Manufacturing: The drivers, concepts and
attributes. International Journal of Production
Economics, 62, pp. 33–43
Zhihong Deng, Tang Shiwei, Zhang Ming, Overview of
Ontology. Acta Scicentiarum Naturalum Universitis
Pekinesis, 2002,38(5):730-738.
KEOD 2011 - International Conference on Knowledge Engineering and Ontology Development
138