Applying Ontology-based Knowledge Methodology in Product
Innovative Collaborative Conceptual Design Framework
Janus S. Liang
National Taiwan Normal University, 162 Heping East Road Section 1, 10610, Taipei, Taiwan, Taiwan
Keywords: Product Innovative Design Ontology, Function Representation, Collaborative Design, Semantic Web.
Abstract: In this research, an ontology-based knowledge methodology is utilized in distributed design environment. It
is composed of a regular restrictions-based method for expressing the aspired functions, a field-independent
method for constructing functional knowledge of given criterion solutions, and a heterogeneous-substance-
search method for combining given criterion solutions to reach the aspired functions. This research presents
that the capability of function design ontology (FDO) to be inferred can acquire the design intentions by an
exposition with a real toy product. Finally, this research proposes a novel framework of information sharing
in product innovative design and a design viewer for collaboration in product generation.
1 INTRODUCTION
To develop a sturdy function-oriented model needs a
comprehension of component geometry and its
substantial outcomes. The morphological
characteristics are also importance except the
geometric properties. They are results of the
principles substantial processes and of the design
intents. However, present solid modeling software
offer incomplete product descriptions and are not
able to perform in the light of the semantic content
of the models (Zhang, 2010). Furthermore, the
semantic web sustains combined and similar entry to
services and message sources as well as to
intelligent utilizations by the definite expression of
the semantics involved in an ontology. While
designer turns into increasingly knowledge-intensive
and collaborative, a concurrent crucial for
computational platforms to enable conceptual
product generation, by effectively supporting the
regular expression, recovery, and reuse of product
information will be achieved. (Kissel et al., 2012)
This research addresses a collaborative
knowledge-based product innovative conceptual
design mode that function design ontology (FDO)
performs in a regular, definite description of a
shared conceptualization of product design
modeling. The FDO, in which is derived from AsD
ontology (Kim et al., 2006), makes design
knowledge accurate and machine-interpretable. In
this research, implicit design constraints (including
criterion solutions) are definitely expressed using
Semantic Web Rule Language (SWRL) and Web
Ontology Language (OWL). A meta-pattern,
function interconnection pattern (FIP), which was
initially constructed to acquire the relations in
function engineering, is promoted utilizing the
ontology techniques to meet the requirements in
collaborative engineering. The implicit conditions of
function engineering relations are expressed by
utilizing SWRL norms and OWL triads. The FDO
also acquires a design rationale, including the
function intention, assembly, and spacial relations.
This work generates an information sharing
circumstance in innovative conceptual design that
can inquire FDO messages selectively.
2 RELATED WORKS
Existing computer-aided conceptual design systems
can be classified as reuse-oriented systems and
creation-oriented systems. The former often aim at
modeling the product knowledge of a total
engineering system for guiding subsequent design of
similar systems. For instances, the FBS modeler
employs the Function-Behavior-State approach to
model the function-related knowledge of existing
criterion solutions for reuse (Cabrera et al., 2009).
Based on an existing functional taxonomy, several
studies have developed component taxonomy for
archiving, searching and reusing component
324
Liang J..
Applying Ontology-based Knowledge Methodology in Product Innovative Collaborative Conceptual Design Framework.
DOI: 10.5220/0004616603240330
In Proceedings of the International Conference on Knowledge Engineering and Ontology Development (KEOD-2013), pages 324-330
ISBN: 978-989-8565-81-5
Copyright
c
2013 SCITEPRESS (Science and Technology Publications, Lda.)
knowledge (Joshi and Pathak, 2011). The latter
generates a criterion solution for a desired function
through recalling basic components from its
component library and synthesizing them together
for a desired function. Based on the fact that systems
in different disciplines can be represented with
similar graphs on the mathematical meta-level, the
infused design approach employs multiple graph
representations to identify suitable problem-solving
methods from related disciplines for human
designers. However, it cannot integrate criterion
solutions in multiple disciplines into one system.
The above literature analysis discloses there are still
no suitable computer-aided conceptual design
systems for fulfilling innovative conceptual design
in a collaborative environment.
Observing the product from the constructional
point of view gives a product break down structure
(product, assembly, subassembly, component, and
feature) each of which requires be designing and
therefore calling as product design units (PDEs). A
PDE at component building level is a reusable
design information unit representing a potential
solution means for a function requirement. Rehman
and Yan (2003) proposed a generic function to PDE
mapping process model for supporting decision
making in the conceptual design stage. Baxter et al.
(2007) categorized existing work in which the
design process has a relationship to design
information management or design reuse mainly
into: (i) Design process with the information
management at its core; (ii) Integrating design
rationale process; (iii) Design methodology as
design process description or management method;
(iv) Design information capture and representation
through design processes. Panchal et al. (2006)
developed an approach for the integrated design of
materials, products, and design processes which was
based on the use of reusable interaction patterns to
model design processes, and the consideration of
design process decisions using the value of
information metrics. It was shown that the integrated
design of materials and products can be carried out
more efficiently by considering the design of design
processes. However, their style still has implicit
restraints and relations that cannot be analyzed by a
calculator in a semantic web circumstance. This
research will develop the FDO style using ontology
technology. The FDO will definitely express
conceptual design restraints and the calculator will
can reason the surplus implicit ones. Meanwhile, the
function interconnection pattern (FIP) will involve
the motion characteristics in product innovative
conceptual design except the static assemble
conditions.
3 SYSTEMATIC
REPRESENATION AND
SYNTHESIS APPROACH
A HeOS-based (Heterogeneous Object Search)
synthesis method is generated in order to accomplish
ICD in collaborative environment. The agent
concept (Russel and Norvig, 2009) is utilized to
illustrate the HeOS-based synthesis method. It is
supposed that there is a SCD (Smart Conceptual
Design) agent that involved all criterion solutions in
its knowledge base. For SCD agent, the given
criterion solutions can be considered as its behaviour
devices for converting import flows, the import
flow(s) derived from an aspired function can be its
original condition, and its export flow can be the
target situation. This agent can sense its condition
automatically, and can choose proper criterion
solutions from its knowledge base to perform it,
until the target conditions is achieved. The condition
of SCD agent consists of the flows that SCD agent
can detect. Besides, each flow is also expressed with
a name of flow, along with a group of attribute-
values. Following the attribute restraints on the
import flow(s), SCD agent can create original
conditional flow(s) by integrating whole probable
values of each attribute completely, and place them
into its condition. After a given criterion solution is
specified to perform a conditional flow, some export
flows can then be created that are also place into the
condition for farther design synthesis.
After a qualified criterion solution has been
recognized, SCD agent will then decide how it
performs the present conditional flow, where the
main attention is what flow(s) the criterion solution
will export. The functional knowledge of the
criterion solution is utilized to predict the export
flow(s). In a state space of heterogeneous substance,
what SCD agent should predict originally is the
name of the export flow. Based on the import-export
flow name pair(s) of a given criterion solution, SCD
agent knows the name of the export flow. Then SCD
agent has to decide the value scopes of the attributes
for the export flow. This procedure includes two
portions. One is to decide the value scopes of the
attributes in accordance with the attribute-mapping
norms of the criterion solution. For instance, when
the criterion solution Crank-slider mechanism is
utilized to perform an import “Angular_velocity”
ApplyingOntology-basedKnowledgeMethodologyinProductInnovativeCollaborativeConceptualDesignFramework
325
flow with its attribute Axial_orientation set as Z, the
attribute orientation of the export “Linear_velocity”
flow can then be predicted as either X or Y. Another
is to decide the value ranges of the attributes that are
not restricted by the attribute-mapping rules, using
the attribute constraints of the selected criterion
solution on its export flow. Eventually, SCD agent
creates all possible export flows, using the value
scope determined for each attribute. This can be
done by extracting each value decided for each
attribute and combining them together in an
thorough way. This action process can be shown
with an example shown in Figure 1, where the
allowable value scopes that SCD agent has
determined for the attribute of the export flow are
shown in the right side. Based on these allowable
values, SCD agent generates two probable export
flows by means of a thorough conjucntion process,
i.e. “Linear_velocity {Stability: variable;
Orientation: X; Direction: reciprocating;
Motionstate: continuous}” and “Linear_velocity
{Stability: variable; Orientation: Y; Direction:
reciprocating; Motionstate: continuous}”.
Figure 1: An example of SCD agent’s action process.
In order that SCD agent can investigate in
collaborative solution environment for discovering
the conjunctions of novel and potential criterion
solution, the thorough search approach is utilized to
generate the automated ICD synthesis algorithm.
The process of design synthesis can then be
implemented after a designer imports an aspired
function with the restraints on the import and export
flows. Besides, it is important that a designer takes
care of the most potential criterion solution selected
for refining farther design after various
combinatorial criterion solutions are exported. If
SCD agent cannot create a potential criterion
solution with the assigned the maximal depth for
searching, the designer may add the maximal depth
for searching and re-begin the process of design
synthesis.
Suppose a designer intends to develop a toy dog
that can shake its head and tail after turning on the
switch. A solar energy is chosen for driving the toy.
Utilizing the proposed method of functional
expression, the import flow, i.e. the solar light, can
be expressed as “Lighting_energy {Stability:
constant || variable; Motionstate: continuous; Type:
Hot_light}”, while the export flow, i.e. the head and
tail-shaking action of the toy, can be expressed as
“Angular_velocity {Stability: variable;
Axial_orienation: X || Y; Direction: reciprocating;
Motionstate: continuous}”. To illustrate this
example, several criterion solutions are chosen to
create the criterion solution knowledge base, which
are Crank_slider, Spur_gear_pair, Crank_rocker,
Rack_pinion, Transformer, Solar_array, AC_motor,
DC_motor, Electrical_inverter, and
Light_emitting_diode. Meanwhile, the process of
design synthesis is depicted below: (1) SCD agent
converts the restraints on the import flow into some
detailed circumstance. Consequently, two original
environmental flows are generated. (2) SCD agent
starts to perceive its circumstance. The first
conditional flow perceived is the flow
“Lighting_energy {Stability: constant; Motionstate:
continuous; Type: Hot_light}”. SCD agent then
explores the restraints of the given criterion
solutions on the import flows to discover qualified
criterion solutions. Hence, the criterion solution
Solar_array is recognized as a qualified one. (3)
SCD agent applies the functional knowledge of the
recognized criterion solution to perform the present
flow. Based on its name pair of import-export flow,
SCD agent finds that the criterion solution will
export an “Electrical_current” flow; in the light of
the relevant attribute-corresponding norms and the
attribute restraints on the export flow, then SCD
agent can decide the value scopes of the attributes
for this export flow, and then create an export flow,
“Electrical_current {Stability: constant; Motionstate:
continuous; Direction: positive; Type:
Direct_current}”, which is also place into its
condition. Subsequently, SCD agent keeps on
detecting its condition until whole conditional flows
have been searched. When a flow is detected as
unsearched, it will then discover suitable criterion
solutions to perform it, and the recently-generated
export flows will be increased to its environment. (4)
when the search process terminates, SCD agent then
searches for the environmental flows that can meet
the restraints on the export flow (goal), and get
whole relevant flows by a backtracking process,
with a effect of some conjunctions of given criterion
solutions. For instance, SCD agent will then get the
relevant flows and bind the corresponding criterion
solutions as a combinatorial criterion solution when
the flow “Angular_velocity {Stability: variable;
KEOD2013-InternationalConferenceonKnowledgeEngineeringandOntologyDevelopment
326
Motionstate: continuous; Axial_orientation: X;
Direction: Reciproating}”. The combinatorial
criterion solution is listed “Solar_array
DC_motor Crank_rocker”.
The above design project shows that SCD agent
can create combinatorial criterion solutions for a
aspired function by integrating given criterion
solutions from diverse disciplines into a
collaborative environment. For instance, when the
maximal depth for searching is set as four, a
combinatorial criterion solution, “Solar_array
Electrical_inverter AC_motor Crank_rocker”,
can be created, which is included criterion solutions
from diverse fields, e.g. mechanical engineering,
electrical engineering, electronic engineering, etc.
Furthermore, the SCD agent cannot only determine
whether various criterion solutions meet each other
in flow names, but also can verify if they are
congenial regarding the elaborated properties.
4 FRAMEWORK OF
INNOVATIVE CONCEPTUAL
DESIGN
4.1 Function Relation Model
To fulfill the desired functions, the mechanism
design portion also has to be involved in the
innovative conceptual design system except the
above mentioned approaches. Thus, a FIP is
generated in this research. It is a meta-pattern with
XML format, developed by utilization of the FDO
style. All geometric elements in the FIP are
connected to a relevant solid model, namely, the
output of the style is an XML file deriving from
geometric elements in a CAD model. By utilization
of FDO style, a designer can assign linking modes,
linking conditions, and spacial relations between
components in a CAD circumstance. Figure 2
illustrates a Unified Modeling Language for
showing a FIP static structure. Function has
“part_of” relations with the two features (assembly
and component), and component has “part_of”
relations with the feature form. The feature assembly
is crucial factor for realizing the function intent. It
involves the characteristics below: coupler,
matching, degree of freedom, and material. The
FDO style that acquires motion, assembly, and
linking relations of a product, is included several
aspects: extraction of feature matching, description
of spacial relation, extraction and generation of
feature coupler, generation of feature assembly as
well as relation extraction of function engineering.
Figure 2: The diagram of function interconnection
pattern – UML format.
4.2 Ontology-based Function Design
In this framework of ontology-based function,
product properties, assembly, function, and linking
procedures are generated in an ontology. Different
collaborative designers can entry the function
messages by utilizing a semantic inquiry while
applying collaborative design tools (called eCol-
tools). The FDO has been generated based on a FIP,
which was depicted in Section 4.1; the ontology-
based FIP acts as a meta-pattern that provides a
connection between CAD models and function
relations. Besides, the FDO ontology also offers
facts by utilization of a reasoning module.
In innovative conceptual design, OWL and
SWRL are utilized for establishing relationships.
This study indicates that SWRL principles can be
inferred to supply semantic inquiries and
information requirements in a collaborative ICD
circumstance. To develop the FDO and principles,
Protégé SWRL editor is applied for this research.
There are three types of relations, spacial relations,
and linking relations expressed in OWL and SWRL.
First of all, the relations between a feature form and
a feature component are expressed the “Inhere”. The
Inhere relations infer two constraints. The inferred
constraints can be transformed to two alleged
conditions.
Meanwhile, the relations among form features
are expressed the “Inter-Conjunction”. The
constraint (Rc
uv
) means the relation between two
form features in the hierarchy of form feature. The
generated SWRL rules are used to add new facts and
utilized in a fact-adding mode only in this research.
Finally, the relations between form features that
inhere to different components are defined the
“linking”. The implied constraints are represented in
SWRL rules.
The designer’s initial intention forced on
function design can be investigated through a
ApplyingOntology-basedKnowledgeMethodologyinProductInnovativeCollaborativeConceptualDesignFramework
327
definite linking manner and the planned degree of
freedom. For instance, if designers intend to
perpetually bind two components to move
simultaneously, they specify spacial relations to
fasten them. A designer concerns a fastened coupler
and assigns a screwing utilization for a linking
manner, then the degree of freedom consistent with
the screwing utilization can be reasoned by applying
the ontology inferring ability, and utilized to check if
this screwing utilization will meet the designer’s
intention on the function. The SWRL principles to
express spacial relations, degree of freedom and
inference instances are depicted (Kim et al., 2006):
(i) aligned spacial relation: Two components that
aligned between collinear lines along the direction y,
have degree of freedom of {lin_y :: rot_y}. (ii)
against spacial relation: Two components that have
against spacial relation between planar surface along
the direction x, have degree of freedom of {plan_x ::
rot_x} - in other words, plane degree of freedom
along x-direction of the coordinate of the instance
component and rotational degree of freedom within
its own coordinate system (x-direction). (iii)
decrease in degree of freedom: When multiple
spacial relations are assigned, the result of
interaction of degree of freedom should be decided
to realize the resultant degree of freedom. (iv) bolt
with nut or rivet: If a component is joined by more
than on bolt with nut or rivet, the component has
restricted degree of freedom.
5 THE FRAMEWORK OF
DESIGN INFORMATION
SHARING
A novel framework of design information sharing
and design viewer generated by utilizing the
standard widget toolkit (SWT), Java, and the Jena
ontology application program interface. The design
viewer delivers just an essential group of design
information in a collaborative environment and
hence it conquers the obstructions of present
conceptual design systems that designers have to
find either with finite sustains of system or manually
to find appropriate design information. As an
instance, let us consider the scenario that tail
mechanism of toy dog is being concurrently
generated by multiple collaborators in distributed
environment.
In the distributed framework, the function
ontology is corresponded to the FIP, and the FIP is
definitely connected to a consistent solid model.
Figure 3 shows a structure for the information
sharing in conceptual design. First of all, product
relationships, i.e. function, are acquired from the
function design models that are created by different
CAD tools, and the consistent FIPs are created. This
FDO module executes the selection and FIP creation
as the FDO generator. The FIP is corresponded to
the consistent CAD model. The FDO is created by
utilizing an ontology tool, e.g. Protégé, and the OFM
generator creates an ontology-based function model
(OFM) that an ontological expression of FIP, and
executes the corresponding with the FIP messages.
The OFM generator converts the normalized
contents in FIP to OFM, when reasoning the pre-
constructed ontology. OFM strictly follows the
ontological rules in spite of the FIP created from
different CAD systems may be heterogeneous. The
inferring unit performs the rules in FDO, and puts
recently detected facts to the OFM base. The FDO
viewer treats embedded inquiries and customized
inquiries by utilizing a query unit based on the
OFM. The unit spread the outcomes through the
ontology and then transmits the outcomes to the
conceptual design browsers.
Figure 3: The framework of conceptual design information
sharing.
6 IMPLEMENTATION OF
SYSTEM
In this framework, the elements in the ontology are
persistently connected to a corresponding solid
model by FIP, which is format and application
independent. In this way, the designers do not need
to know a specific CAD format, and loss of
KEOD2013-InternationalConferenceonKnowledgeEngineeringandOntologyDevelopment
328
information is relatively low since design
information is represented explicitly in the FRM.
Figure 4 shows the interface of conceptual design
browser, designers can search for product models
that contain relevant design information by using the
interface. There are several views of the product
information provided on the browser: a hierarchical
view of the ontology classes, a product geometry
view (assemble/exploding condition), and view of a
model data structure and XML structure. Through
browsing the ranking system, a designer can decide
that classes are definitely expressed in the model and
their sub-class relations. The XML view offers the
designer the chance to examine the structure in that
the model will be delivered from one system to
another. Meanwhile, the viewer of conceptual design
explores for relative elements by utilizing keyword
that investigates class structure and file name of the
relevant ontology model. The relevant model comes
back files that involve the particular design
messages if the search rules are met. Besides, other
search rules are also shown to make supplementary
searches on similar items for designer. For instance,
if a designer searches for linkage assembly that
involves screwing information displayed in Figure 4,
other search rules such as riveting, welding, or
adhesive bonding may be returned by utilizing the
ontology model because they are also joining
methods, which is a result of a semantic query.
Figure 4: Screen snapshot of SICD system – FDO viewing
browser.
7 CONCLUSIONS
In this research, the proposed approach for ICD
system paves a path for designers to integrate
criterion solutions from various disciplines into a
collaborative framework for fulfilling a desired
function. It is mainly composed of a formal
approach for representing a desired function, a
domain-independent approach for representing
functional knowledge of known criterion solutions,
and a HeOS-based synthesis approach for automated
ICD system in collaborative environment.
Furthermore, the author has also proposed a
collaborative ICD structure sustained by an
ontology-based knowledge methodology. The
generation of a FDO mentions the requirement that
product design has for high-level semantic
modeling. High-level semantics and a normailization
of function notions as well as relationships among
those concepts give the calculator the capability to
infer messages for the instances of particular
function model. In this research, function knowledge
has been categorized into a ranking system of
function concepts. They inserted in conceptual
design have been defined utilizing particular names
and hence have generated a normal vocabulary to
depict functions. This is essential for concerted
function design that in various disciplines
symbolically include. This research obtains most
benefits of OWL and SWRL techniques to express
complex relationships that are spread among form
properties and components. Because the restrictions
have been normalized, they can be explained and
kept in a normal way in the collaboration period.
The FDO and viewer can be applied in different
activities regarding the collaboration in product
design modeling.
ACKNOWLEDGEMENTS
This research is supported in part by the National
Science Council in Taiwan under contract number
NSC 101-2511-S-003-059-MY2 and NSC 100-
2511-S-003-068-MY2.
REFERENCES
Baxter, D., Gao, J., Case, K., Harding, J., Young, B.,
Cochrane, S., & Dani, S. (2007). An Engineering
Design Knowledge Reuse Methodology Using Process
Modelling, Research in Engineering Design, 18, 1, 37-
48.
Cabrera, A. A. A., Erden, M. S., & Tomiyama, T. (2009).
On the potential of function-behavior-state
methodology for the integration of modeling tools,
Proceeding of the 19th CIRP Design Conference
Competitive Design, Cranfield University, 30-31
March, 412-419.
Joshi, R. M., & Pathak, M. V. (2011). A functional
taxonomy of SMSBIR systems. 3 rd International
Conference on Electronics Computer Technology,
Kanyakumari, India, 8-10 April, 166-171.
ApplyingOntology-basedKnowledgeMethodologyinProductInnovativeCollaborativeConceptualDesignFramework
329
Kim K. Y., Manley D. G., Yang. H. (2006). Ontology-
based Assembly Design for Collaborative Product
Development. Computer-Aided Design, 38, 12, 1233-
1250.
Kissel, M., Bradford, N., Kreimeyer, M., & Lindemann,
U. (2012). Product structure management as the
backbone of engineering design: exploration of a
reference model. International Design Conference,
Dubrovnik, Croatia, May 21-24, 1709-1718.
Panchal, J. H., Choi, H. J., Allen, J. K., McDowell, D. L.,
& Mistree, F. (2006). Designing design processes for
integrated materials and products realization: A
multifunctional energetic structural materials
example, The 32nd ASME Design Automation
Conference (DAC), Philadelphia, Pennsylvania, USA,
10-13 Sept..
Rehman, F., & Yan, X. T. (2003). Product design elements
as means to realize functions in mechanical conceptual
design. Proceedings of 14
th
International Conference
on Engineering Design, Stockholm, Sweden, 19-21
Aug., 213-214.
Russell, S. & Norvig, P. (2009). Artificial intelligence: A
modern approach (3
rd
ed.). NJ: Prentice Hall.
Zhang, D. (2010). A framework for ontology-based
product design knowledge management. Seventh
International Conference on Fuzzy Systems and
Knowledge Discovery, Shandong, China, 10-12 Aug.,
1751-1755.
KEOD2013-InternationalConferenceonKnowledgeEngineeringandOntologyDevelopment
330