Meta-knowledge: A Method to Encode and Decode Products’ DNA in Knowledge
Lifecycle
WANG Wei, Ph.D
Candidate
School
of Design, Hunan University, Changsha, China
e-mail:yien_killer@hotmail.com
Prof. ZHAO
Jianghong,
School of Design, Hunan
University, Changsha, China
Abstract
Meta-knowledge
is the knowledge to structure and manage design knowledge and user knowledge in
knowledge lifecycle, and is also called “the knowledge about knowledge”. This
paper describes an effective method used in computer aided concept design and
generative design based on meta-knowledge, and discusses the acquisition of
meta-knowledge from designers’ protocol experiment, and presents ICAID system
which is an application of meta-knowledge.
Keywords: Meta-knowledge, Products’ DNA, Knowledge
lifecycle, CAID
1.
Introduction
With the
improvement of information technology, many theories and approaches are
introduced into concept process of product design based on computer and Web.
Usually we can compartmentalize them as two kinds, systems supported by database
and systems without database. Identifying products’ DNA, or called a genetic
code, is a key task in the implementary mechanisms of both systems. The theory
of knowledge lifecycle has been presented as an effective method to study the
whole processes of products design and identify products’ DNA since 1990s [1]. In
recent years, this theory has been applied in ICAID system, which is an
internet-based design system combined with CAID (Computer Aided Industrial
Design) and KBS (Knowledge-based System) technology, funded by the Chinese
“Tenth Five-year” National Science and Technology Project.
2. The
Acquisition of meta-knowledge
In the
conventional model of products design, knowledge of design is generated in the whole
products design process and encoded in the product. The knowledge lifecycle
model in Figure 1 introduces the concept of knowledge of design presented by Keiichi
Sato [2]. There are two kinds of knowledge in the model. One is design
knowledge, which is mainly a kind of domain knowledge including design experience,
design value, background knowledge and expert insight. The other is user
knowledge, which is a kind of idiographic knowledge including the knowledge of
users’ decoding, operation and experience. And the whole knowledge lifecycle can
be regarded as an encoding/decoding process of products’ DNA.
Figure
1 Knowledge Lifecycle in Product Lifecycles
Meta-knowledge,
also called “the knowledge about knowledge”, is the knowledge to direct and
manage knowledge of design in knowledge lifecycle. It is an effective bridge
between the knowledge lifecycle in products lifecycle and the knowledge-based
CAID system [3]. The function of meta-knowledge is decoding product’ DNA in
certain product lifecycle and encoding it into a CAID system in Figure 2.
Figure
2 The relationship between meta-knowledge and Products’ DNA
In ICAID
project, we experiment on the actual NCMT (Numeric Control Machine Tools)
design process to acquire meta-knowledge. The experiment (performed in Hunan
University from Dec. 23 to 27, 2002) includes: Subject (five machine tools designers),
Experimenter (three ICAID system designers), Task (a shape project of vertical
machining center), Environment (design software based on Windows and pattern
data supports), and Assistant (five software operators). The experiment process
is: Subject, who have little experience in NCMT industrial design, are required
to independently design a shape project of certain vertical machine center in
five hours; Experimenter records the whole process by digital video,
observation and protocol (a method of Experimental Psychology); and Assistant
only can give Subject the assist of software operation but not of ideas and
estimates about design.
Analyzing
the experiment’s result, the meta-knowledge in NCMT design includes two parts
as followed:
The meta-knowledge of management is the knowledge focused on structuring
and managing design knowledge and user knowledge in knowledge lifecycle. In
ICAID system, it includes the planning of ICAID system platform, the structure
of each function parts and the users’ flow.
The meta-knowledge of problem solving is the knowledge of object reasoning
and problem solving in industrial design. It includes the projection from
structure to shape in Figure 3, and the relationship between shape and human’s aesthetic,
psychology and cognitive in Figure 4.
Figure
3 Different methods of the projection from structure to shape
Figure
4 The relationship between shape and psychology
Additionally,
we must notice that different species of products have different structures of
meta-knowledge. For example, the most influencing factor in NCMT shape design
is the projection from structure to shape. But it becomes more complicated when
it is a brand design and vehicle styling [4].
3. ICAID
system: the application of meta-knowledge
Aimed to
industrial designs of NCMT, ICAID system is a viable mode evolved from meta-knowledge
and products’ DNA. The prime purpose of ICAID system is to provide a design
tool to aid designers working on NCMT design. Because NCMT are high tech and
precise machines, the process of conceptive design requires both an intelligent
aid from ICAID system and a close cooperation in multidisciplinary workgroup.
Therefore
we present that the whole ICAID system should includes three parts under an
environment: User (designers), Computer Aid (provided by ICAID system
software), Expert Aid (provided by industrial experts), and System Platform (a
software environment). Based on ICAID system platform, it makes possible that the
users’ design process is favoring, and the cooperation between users and expert
is easy. This kind of cooperative product design process is real time,
long-range, and computer aided. (Presented by Figure 5)
Figure
5 The structure of ICAID system
ICAID
system is an interactive and intelligent environment based on Internet. Its
basic software is based on Web & Server technologies such as HTML, Java,
ASP, VRML, Flash ActionScript, and product design software such as Solidworks (a
CAD/CAM software). ICAID system includes three main parts as followed:
Client-Server Mode Software is a product design tool, an
entrance of Knowledge Base and Database, a tool to acquire, abstract and renew
knowledge, and a communication environment between users and experts.
Knowledge Base is a structure to deposit Design Knowledge and
User Knowledge, including descriptive facts and exercisable rules. It can give
users computer aid and estimate through Client-Server Mode Software.
Knowledge’s acquirement, abstraction and renewal also depend on the software.
Database is a structure to store meta-model data,
process data and user information data.
In the
concept design stage, ICAID system provides two CAID methods to meet the
different cognitive process of designer [5]. One is Draft Processor (Fig. 6)
and the other is Image Scale (Fig. 7). Draft Processor is a bottom-up
processing tool of knowledge reasoning, which is a design process from
structure to form and from parts to body. Image Scale is a top-down processing
tool of knowledge reasoning, which is a design process from concept to shape
and from semantic stimulation to visual information.
In the
latter design stages, contrasting to other generative design system, ICAID is
supported by database and cases base. Users develop their designs based on the
meta-models which encode meta-knowledge and store in database. This kind of
knowledge’s expression is a rapid and effective method to achieve users design requirement.
4.
Conclusion
Meta-knowledge
is an effective method to manage design knowledge and user knowledge in design
process and decode/encode Products’ DNA into a CAID system. As a new concept,
ICAID is an application of meta-knowledge founded on knowledge base and model base
which are supported by internet. This kind of implementary mechanism provides a
beneficial solution towards other CAID problems. However, owing to different
structures of meta-knowledge based on different species of products, the
structure of ICAID should be adapt to design object when use in new domains.
References
[1] Zhao Jianghong, Wu Chao. Internet-based Computer Aided Industrial
Design, China Mechanical Engineering, 10:53-54, 1999.
[2] Keiichi Sato. Creating a New Product Paradigm between Media Space
and Physical Space. In：Proceedings of International Council of Societies
of Industrial Design. Seoul，Korea，362-368，2001.
[3] PLANT, ROBERT T.; GAMBLE, ROSE. Using meta-knowledge within a
multilevel framework for KBS development. International Journal of
Human-Computer Studies 46 (4): 523-547, April, 1997.
[4] Jay P. McCormack, Jonathan Cagan, Craig M. Vogel. Speaking the Buick
language: capturing, understanding, and exploring brand identity with shape
grammars. Design Studies 25 (1): 1-29, January, 2004.
[5] Van Someren, The Think Aloud Method: A Practical Guide to Modeling
Cognitive Processes. London, The Academic press, 1994.