Mapping Knowledge about Product Lifecycle Engineering for Ontology Construction via Object-Process Methodology

Mapping Knowledge about Product Lifecycle Engineering for Ontology Construction
via Object-Process Methodology

D. Dori1, M. Shpitalni2 (1)
1Faculty of Industrial Engineering and Management
2Laboratory for CAD and LCE, Faculty of Mechanical Engineering
Technion - Israel Institute of Technology
32000 Haifa, Israel

Abstract
Knowledge mapping is a first and mandatory step in ontology definition. This paper considers the lifecycle of
products and systems, and discusses the creation of a knowledge-based ontology. With respect to the life cycle
of products and systems, knowledge refers to the processes involved in their creation (design manufacturing
and assembly), use and maintenance, and end of life (EOL). Hence, this knowledge should consider what a
product is comprised of (its structure), how it operates (its dynamics), and how it interacts with the environment.
A clearly defined and consistent mapping of knowledge regarding structure, operation and interaction is
necessary to construct an effective and useful ontology. Yet, in order to obtain the required knowledge and to
organize it in a consistent and useful form, an appropriate ontology must be used. An interactive, iterative and
consistent method is needed to cope with this complex and circular problem. In this paper, the Object-Process
Methodology (OPM) approach is considered, along with OPCAT [1], a tool for OPM-based knowledge modeling.
OPM is a systems-modeling approach that represents knowledge about systems concurrently and bi-modally
through graphics (a set of Object-Process Diagrams, OPDs) and text (Object-Process Language, OPL, a subset
of English), yielding a single, unified and consistent view. In this paper we propose a foundational modeling and
ontology construction approach for a generic product that incorporates hardware and software components.
The ontology can serve as a basis for a knowledge model to cover the entire product lifecycle, from inception to
EOL, and can be applied in the VRL-KCiP Network of Excellence.
Keywords
Knowledge management, Lifecycle ontology, Object-process methodology



1 INTRODUCTION
complex, spanning a growing number of disciplines,
Advances in information technologies along with
development of an ontology that can serve as a basis for
innovations in production systems have dramatically
a rigorous methodology for lifecycle engineering is still in
changed processes and engineering practices related to
its evolutionary stages. In particular, the early lifecycle
stages of most systems are still the least structured and
product development [2]. These changes, examined in
least coherent. They suffer from ambiguity and lack a
terms of research challenges and opportunities by Kimura
clear and direct translation of customer needs/wants into
et al. [3], have enhanced global manufacturing capabilities
precise product specifications.
to a point far exceeding the demand. As a consequence,
competition among manufacturers has dramatically
All phases comprising this lifecycle, including the end of
increased. At the same time, new issues, most notably
life (EOL) phase, must be taken into account when
environmental consciousness, are becoming increasingly
mapping the knowledge and defining the ontology.
critical, imposing additional constraints on new products.
Interaction between the phases must also be considered,
bearing in mind that products are becoming more
The need for a shift in paradigm is described by Takata et
complex, and more software-based components are being
al. [4], who claim that the scale of industrial activities has
integrated into products.
already exceeded its limit. Hence, the manufacturing
Following the early work by Tipnis [10] on product lifecycle
paradigm must be changed, as suggested by Kimura [5].
economic models, several methods have been proposed
To a large extent, this change in paradigm is based on the
for modeling and/or evaluating all or part of a lifecycle
transition from selling of products to selling of services
system. Takata et al. [4] proposed flexible means to
and on accommodating a holistic lifecycle approach [6], as
represent technical information relevant to manufacturing
suggested by Alting and Jorgensen [7] for sustainable
facility management, especially with respect to
industrial production and later reexamined by Hauschild et
maintenance. Krause & Kind [11] proposed a reference
al. [8]. Recent studies [9] indicate that future products and
model to assure supply of appropriate information
systems will be increasingly knowledge-driven rather than
according to specific requirements pertaining to all
energy-driven, underlining the urgent need to develop the
lifecycle phases. In 2001, Brissaud & Tichkiewitch [12]
proper tools and environment for extracting existing
proposed a comprehensive product model that can be
knowledge and generating new knowledge. This goal can
used to globally optimize the phases comprising the
be achieved only if a comprehensive ontology is defined
product lifecycle. Concurrently, Westkämper [13]
to consider the entire lifecycle of products and systems.
proposed a platform for integrating the management of
Knowledge mapping is a first and mandatory step in the
assembly/disassembly and product lifecycle. Tools to
process of ontology definition. This paper considers the
support conceptual design, development, and assessment
lifecycle of products and systems, and discusses the
in terms of environmental cost and impact have been
creation of a knowledge-based ontology.
proposed by Park et al. [14], by Kaebernick et al. [15], who
While systems and products are becoming ever more
proposed a simplified lifecycle assessment for the early

---

design stages of products, and more recently, by Duflou et
Links can be structural or procedural. Structural links
al. [16]. Proactive support tools for lifecycle engineering,
express static, time-independent relations between pairs
as well as a simulation system for lifecycle process
of entities. The four fundamental structural relations are
planning, have been proposed by Takata and Kimura [17].
aggregation-participation, generalization-specialization,
Zhang et al. have proposed web-based applications, such
exhibition-characterization, and classification-instantiation.
as a web-based system for reverse manufacturing and
Procedural links connect entities (objects, processes, and
product environmental impact assessment, which takes
states) to describe the behavior of a system.
end-of-life dispositions into consideration [18].
System behavior can be manifested in three major ways:
(1) processes can transform (generate, consume, or
In defining ontology, the knowledge must first be mapped.
change the state) of objects; (2) objects can enable
Knowledge, in turn, is about systems: their function, their
processes without being transformed by them; and
structure, their dynamics, and their interaction with their
(3) objects can trigger events invoking processes if some
environment, a collection of systems in itself.
conditions are met. Accordingly, a procedural link can be
System architecture is the underlying combination of
a transformation link, an enabling link, or an event link.
structure (expressed in terms of objects and their
A transformation link expresses object transformation, i.e.,
attributes) and behavior (expressed in terms of processes
object consumption, generation, or state change. An
and how they transform objects) that enables a system to
enabling link expresses the need for an object to be
attain its intended function. System architecting is
present, possibly at a certain state, in order for the
therefore of paramount importance to successful design,
enabled process to occur, but the enabled process does
implementation, and evolution of artificial complex
not transform the enabling object.
systems in general and products in particular.
An event link connects a triggering entity (object, process,
Having recognized the commonality among a large variety
or state) with a process that it invokes. The event types
of systems, product lifecycle engineering combines a host
supported by OPM include state entrance, state change,
of scientific, technological, ecological, and human aspects
state timeout, process termination, process timeout,
in a highly multidisciplinary fashion. Recently, the notion of
reaction timeout, and external events. External events can
Product Lifecycle Management (PLM) has started to make
be clock events or events triggered by environmental
headway, not just in academic echelons but also in
entities, for example a user or an external device.
leading enterprises which have realized the necessity of a
comprehensive approach to ensure that systems and
2.1 The Bimodal Graphic-Text OPM Representation
products be managed by best industry practices backed
Two semantically equivalent modalities, one graphic and
by sound methodology. A clear, consistent and
the other textual, jointly express the same OPM model.
comprehensive mapping of knowledge about systems
The graphical, visual OPM formalism consists of a set of
must account for system structure, dynamics, and
interrelated Object-Process Diagrams (OPDs), showing
interaction with the environment.
portions of the system at various levels of detail. Each
From cognitive science [19] we know that humans access
OPM element is denoted in an OPD by a symbol. The
and process knowledge via two main channels: the visual
OPD syntax specifies correct and consistent ways by
channel and the non-visual channel. The former relies on
which entities can be connected via structural and
graphics and diagramming, while the latter is based on
procedural links, each having its specific, unambiguous
language, both spoken and written. Hence, an ideal
semantics. OPM assigns special graphical symbols to a
knowledge mapping and representation system should be
selected set of relations, as in Unified Modeling Language
capable of presenting knowledge using graphics as well
(UML) class diagrams, but for a larger set of relations [21].
as at least a subset of natural language text.
OPCAT [1] is a Java-based software environment that
In this paper we apply an OPM-based approach to
supports OPM system modeling and evolution. As shown
modeling product lifecycle knowledge. In particular, we
in the toolbox at the bottom of OPCAT's GUI in Figure 1, a
focus on knowledge organization and ontology creation for
triangular graphical symbol along the line connecting two
the Virtual Research Lab for a Knowledge Community in
items (objects or processes) is assigned to each of the
Production (VRL-KCiP) Network of Excellence (European
four fundamental structural relations mentioned above,
Community 6th Framework Programme).
symbolically represented as follows: black triangle for

aggregation, white triangle, as in UML, for generalization,
black on white triangle for characterization, and black
2 OBJECT-PROCESS
METHODOLOGY
circle in white triangle for instantiation. Like associations in
In complex products, structure and behavior, as well as
UML class diagrams, other structural relations become
hardware and software, are typically highly intertwined and
textual tags (labels) recorded along the arrow connecting
interdependent. Motivated by this observation, Object-
the two entities, such that concatenating the source entity
Process Methodology, OPM [20] has been developed as a
with the tag and then with the destination entity yields a
holistic approach to the study and development of
meaningful sentence.
systems, integrating object-oriented and process-oriented
Object-Process Language (OPL) is the textual counterpart
paradigms into a single frame of reference. Structure and
modality of the graphical OPD set. OPL is a dual-purpose
behavior, the two major aspects exhibited by any system,
language, oriented towards humans as well as machines.
co-exist in the same OPM model without highlighting one
Catering to human needs, OPL is designed as a subset of
at the expense of the other. Due to its structure-behavior
English that serves domain experts and system architects
integration, OPM provides a solid basis for modeling
jointly engaged in analyzing and designing a system or
complex products. The OPM ontology consists of entities
product. Every OPD construct is expressed by a
and links. Entities include things and states. Things are
semantically equivalent OPL sentence or phrase. This
stateful objects and processes, both of which can be
dual representation of OPM increases the processing
physical or informatical. Objects exist, possibly at some
capability of humans, according to the cognitive theory of
state, while processes transform objects by generating or
multimodal learning proposed by Mayer [19]. By catering
consuming them, or changing their state.
to the modality principle of this cognitive theory, OPM
enables modeling systems to operate both graphically and
textually.

---

2.2 Top-Level View of Product Lifecycle Engineering
stored, sold, or retired." is a state enumeration sentence
Figure 1 presents a snapshot of OPCAT's graphic user
listing all possible states of the object Product throughout
interface, showing a top-level view (System Diagram, or
its lifecycle. States can be initial or final, as expressed in
SD) of the Product Lifecycle Engineering system
the sentences "pre-tested is initial." and "retired is final."
considered here. OPCAT translates each OPD construct
The corresponding graphic symbol is a thick frame for the
into its equivalent OPL sentence, yielding an OPL
initial state and a double frame for the final state.
paragraph that is a collection of OPL sentences specifying
A major problem with most graphic modeling approaches
the knowledge represented graphically in the OPD.
is their scalability. As system complexity increases, the
Conversely, typing an OPL sentence complements the
graphic model becomes loaded with shapes and cluttered
OPD, such that at any point the graphic and textual
with links crossing each other in all directions. According
representations are completely equivalent and can be
to the cognitive principle of limited channel capacity [19],
reconstructed one from the other. With OPCAT, users can
there is an upper limit on the amount of detail humans can
model complex systems and products at all levels of
process before becoming overwhelmed. OPM addresses
granularity, and express and query knowledge related to
this principle and implements it in OPCAT using three
these models. In the OPCAT GUI in Figure 1, the
abstraction/refinement mechanisms: (1) unfolding/folding,
hierarchy of diagrams and things (i.e., objects and
for refining/abstracting the structural hierarchy of a thing
processes) appears on the left, the graphic (OPD) window
and applied by default to objects; (2) in-zooming/out-
at the top right, the text (OPL) window at the bottom right,
zooming, to expose/hide an item's inner details within its
and the palette with the various OPM entities and links at
frame, applied primarily to processes; and (3) state
the bottom. Some OPL sentences, all generated
expressing/suppressing, to expose/hide an object's states.
automatically by OPCAT, are shown in the OPL window.
These mechanisms enable complexity management by
Objects are denoted as rectangles, processes by ellipses,
enabling the creation of interrelated OPDs (along with
and object states by rounded rectangles inside the object.
their corresponding OPL paragraphs) that are limited in
Most of the objects depicted are physical, as denoted by
size, thereby avoiding information overload and facilitating
the shading of their representative rectangles. These
convenient human processing. Flexible combinations of
include Product and Manufacturer, also expressed in the
these three mechanisms enable OPM to specify a system
sentences "Product is physical." and "Manufacturer is
to any desired level of detail without loss of legibility or
physical." The entire OPL paragraph is as follows:
clarity of the resulting specification. The complete OPM
Product is physical.
system specification is expressed graphically by the
Product can be approved, distributed, used, pre-
resulting set of consistent, interrelated OPDs, and
textually by the corresponding OPL script, the union of
tested, rejected, stored, sold, or retired.
information expressed in the OPL paragraphs. Like OPD,

pre-tested is initial.
each OPL paragraph is a collection of sentences spanning

retired is final.
no more than a single page, so that humans can
Product is made of Raw Material.
comfortably read and digest the knowledge it expresses.
Product benefits User.

Manufacturer is physical.
Manufacturer makes & supports Product.
3 OPM AND PRODUCT LIFECYCLE ENGINEERING
Environment is environmental and physical.
The Product Lifecycle Engineering process is in-zoomed
Environment consists of User, Market Demand, Raw
in Figure 2 to show subprocesses, from Design to End of
Material, Technology, and Competition.
Life, executed in a top-to-bottom order.

User is environmental and physical.
3.1 The Design Process

Market Demand is environmental.

Raw Material is environmental and physical.
The Design process is further in-zoomed in Figure 3 to

Technology is environmental.
show subprocesses ranging from Requirement

Competition is environmental.
Engineering through Conceptual and Detailed Design to
Transfer to Production.
Product Lifecycle Engineering is physical.
Product Lifecycle Engineering requires Environment.
3.2 Simulation through Animation
Product Lifecycle Engineering affects Product, User,
OPCAT is capable of simulating a system through vivid
and Manufacturer.
animation. Figure 4 shows the animated Design process,
These sentences show how knowledge that combines
with Detailed Design in action, generating the Hardware,
structure and behavior can be represented both by
Software, and EOL models. After Design comes
intuitive (yet formal) graphics and by human intelligible
Manufacturing. One of the outcomes of Detailed Design is
text, OPL, a subset of English. Users not familiar with
the Software Model, which, together with Hardware
OPM graphic notation can verify their specifications by
Models and the EOL Model, constitutes the Product
inspecting the OPL sentences generated or edited with
Model. The Software Model is the instrument for Software
each graphic input. For example, the OPL sentence
Module Developing, embedded within the Making process.
"Product Lifecycle Engineering affects Product, User,
In turn, the Making process is embedded within
and Manufacturer." is equivalent to the relevant part of
Manufacturing. Figure 5 shows the details of the Software
the OPD showing the three effect links (denoted by
Module Developing process, revealing Analysis, Design,
bidirectional arrows) connecting the objects Product,
and Implementation as its three major subprocesses.
User, and Manufacturer to the process Product
Software, as part of the entire product, undergoes a "mini
Lifecycle Engineering, giving rise to the OPL reserved
lifecycle," similar to the grand product lifecycle, albeit on a
word "affects". The effect link connects a process
smaller scale. When software constitutes a significant and
(Product Lifecycle Engineering) to an object, or in our
especially a major part of a product, the situation might be
case, each one of the three affected objects, Product,
reversed. Hardware design and manufacturing, if present,
would then be embedded as part of the encompassing
User, and Manufacturer. States are situations in which
software product lifecycle model. Here, however, we
an object can exist. The sentence "Product can be
consider a case in which software still plays a less
approved, distributed, used, pre-tested, rejected,
significant role in the overall product architecture.

---

Figure 1: Top-level view (System Diagram, SD) of the Product Lifecycle Engineering system, showing the OPD window (top),
part of the corresponding OPL window (bottom), and the OPD tree (left).

Figure 2: The Product Lifecycle Engineering process is in-zoomed to show subprocesses from Design to End Of Life,
executed in a top-to-bottom order.

Figure 3: Design process is in-zoomed to show subprocesses from Requirement Engineering to Transfer to Production.

---

Figure 4: Animated Design process, showing Detailed Design in action, as it generates Hardware, Software, and EOL models.

Figure 5: Software Module Developing process, showing Analysis, Design, and Implementation as three major subprocesses.

Figure 6: End of Life process, with subprocesses: Reclaiming, Salvage Decision Making, Disposing, Recycling, Refurbishing.

---

3.3 Commerce, Using, and End of Life
and Selling of Services on Product Design, Archives
Commerce, the next process in the product lifecycle,
of Mechanical Technology and Automation (Special
comprises Storing, Marketing, Distributing, and Selling.
Issue), 24/2:251-261.
The following process, Use & Service, zooms into Using,
[7] Alting, L., Jorgensen, J., 1993, The life cycle concept
Customer Relationship Management, Maintenance,
as a basis or sustainable industrial production,
and Retiring. Using is the ultimate objective of the entire
Annals of the CIRP, 42/1:163.
product lifecycle, for all processes activated before and
[8] Hauschild, M., Wenzel, H., Alting, L., 1999, Life cycle
after it are aimed at making the Using process longer,
design - a route to the sustainable industrial culture?,
Annals of the CIRP, 48/1:393-396.
more productive, and more appealing to the User. Using
[9] Manufuture: A Vision for 2020, Manufuture 2004,
changes the Benefit which is a User attribute, from
http://europa.eu.int/comm/research/industrial_techno
potential to materialized.
logies/pdf/manufuture_vision_en.pdf.
Retiring is the process that occurs when Product is no
[10] Tipnis, V.A., 1991, Product life cycle economic
longer usable, when using it causes more trouble than the
models, Annals of the CIRP, 40/1:463-466.
benefit from the service it is supposed to provide, or when
[11] Krause, F.-L., Kind, Chr., 1995, Potentials of
newer, more attractive and economic alternatives are
information technology for life cycle oriented product
available. Retiring changes the state of Product from
and process development, in Proceedings of the IFIP
used to retired, marking its entry to the End of Life
WG5.3 international conference on life cycle
process. As described in Figure 6, the End of Life
modeling for innovative products and processes, H.
process includes the subprocesses of Reclaiming,
Jansen and F.-L. Krause, Eds., Chapman & Hall:
Salvage Decision Making, Disposing, Recycling, and
Berlin, 14-27.
Refurbishing.
[12] Brissaud, D., Tichkiewitch, S., 2001, Product Models
for Life-Cycle, Annals of the CIRP, 50/1:105-108.

[13] Westkämper, E., 2002, Platform for the Integration of
4 SUMMARY AND FUTURE WORK
Assembly, Disassembly and Life Cycle Management,
Using Object-Process Methodology and OPCAT, the
Annals of the CIRP, 51/1:33-36.
OPM-supporting software tool, we have laid the
[14] Park, J.-H., Seo, K.-K., Wallace, D., Lee, K.-I., 2002,
foundations for ontology of a generic lifecycle product that
Approximate Product Life Cycle Costing Method for
spans the entire spectrum of the product lifecycle, from
the Conceptual Product Design, Annals of the CIRP,
Design to End Of Life, and includes hardware and
51/1:421-424.
software components. The ontology specifies the
[15] Kaebernick, H., Sun, M., Kara, S., 2003, Simplified
hierarchy of processes and the objects they transform.
Lifecycle Assessment for the Early Design Stages of
This ontology can serve as a basis for specialized product
Industrial Products, Annals of the CIRP, 52/1:25-28.
lines and their characteristic lifecycle processes. It can
[16] Duflou, J., Dewulf, W., Sas, P., Vanherck, P., 2003,
also serve as a basis for knowledge mining via
Pro-active Life Cycle Engineering Support Tools,
technologies such as inexact graph matching used as
Annals of the CIRP, 52/1:29-32.
Web services. As future work, we will examine a number
[17] Takata, S., Kimura, T., 2003, Life Cycle Simulation
of representative cases typical of manufacturing
System for Life Cycle Process Planning, Annals of
processes matching the expertise of the research groups
the CIRP, 52/1:37-30.
in the VRL-KCiP NoE. We will integrate these cases as
[18] Zhang, H.C., Li, J., Shrivastava, P., Whitley, A.,
specializations of the generic product lifecycle model.
Merchant, M.E., 2004, Web-Based System for
Reverse Manufacturing and Product Environmental

Impact Assessment Considering End-of-Life
5 ACKNOWLEDGMENTS
Dispositions, Annals of the CIRP, 53/1:5-8.
This research has been supported in part by the Minerva
[19] Mayer, R. E., 2001, Multimedia Learning, Cambridge
Schlesinger Laboratory for Automated Assembly and the
University Press.
EU VRL-KCiP NoE.
[20]
Dori, D., 2002, Object-Process Methodology-A

Holistic Systems Paradigm, Springer Verlag, Berlin,
6 REFERENCES
Heidelberg, New York, www.ObjectProcess.org.
[21] Dori, D., 2002, Why Significant Change in UML is
[1] Dori, D., Reinhartz-Berger, I., Sturm, A., 2003,
Unlikely, Communications of the ACM, 82-85.
OPCAT – A Bimodal CASE Tool for Object-Process
Based System Development, Proc. IEEE/ACM 5th
International Conference on Enterprise Information
Systems (ICEIS 2003), Angers, France, 286-291.
(www.ObjectProcess.org)
[2] Bar Cohen, A., 1995, Mechanical Engineering in the
Information Age, Mechanical Engineering,
117/12:66-70.
[3] Kimura, F., Lipson, H., Shpitalni, M., 1998,
Engineering Environments in the Information Age:
Research Challenges and Opportunities, Annals of
the CIRP, 47/2:87-90.
[4] Takata, S., Kimura, F., van Houten, F.J.A.M.,
Westkämper, E., Shpitalni, M., Ceglarek, D., Lee, J.,
2004, Maintenance: Changing Role in Life Cycle
Management, Annals of the CIRP, 53/2.
[5] Kimura, F., Suzuki, H., 1995, Product Life Cycle
Modelling for Inverse Manufacturing, Life Cycle
Modelling for Innovative Products and Processes,
Chapman & Hall, 80-89.
[6] Shpitalni, M., 2004, Impact of Life Cycle Approach

---