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
Prof. ZHAO Jianghong,
School of Design, Hunan University, Changsha, China
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
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 . 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 . 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 . 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 .
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 . 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.
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.
 Zhao Jianghong, Wu Chao. Internet-based Computer Aided Industrial Design, China Mechanical Engineering, 10:53-54, 1999.
 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.
 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.
 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.
 Van Someren, The Think Aloud Method: A Practical Guide to Modeling Cognitive Processes. London, The Academic press, 1994.