Elsevier

Computer-Aided Design

Volume 44, Issue 2, February 2012, Pages 146-153
Computer-Aided Design

Technical note
A knowledge-based framework for creative conceptual design of multi-disciplinary systems

https://doi.org/10.1016/j.cad.2011.02.016Get rights and content

Abstract

Designers often have difficulty in fulfilling creative conceptual designs of multi-disciplinary systems due to the lack of sufficient multi-disciplinary knowledge. Therefore, this paper proposes a knowledge-based framework for the creative conceptual design of multi-disciplinary systems through reusing and synthesizing known principle solutions in various disciplines together. It comprises a formal constraints-based approach for representing the desired functions, a domain-independent approach for modeling functional knowledge of known principle solutions, and a heterogeneous-object-state-search-based approach for synthesizing known principle solutions together for achieving the desired functions. A design case illustrates that the proposed framework can successfully achieve creative conceptual design of multi-disciplinary systems. A prototype system, called the Intelligent Creative Conceptual Design Platform, is implemented based on this framework.

Introduction

Creative design is pivotal for manufacturing enterprises to gain competence advantages since it can deliver novel products to the markets. In engineering design, the novelty of a new artifact can be understood from two perspectives. One is that the new artifact has some novel functions, though the Principle Solution (abbreviated as “PS” later) of each function is not new. Here, PS refers to the physical mechanism of a system for achieving a desired function [1]. The other is that it employs novel PSs to fulfill its existing functions, with better performance achieved. Correspondingly, there are two Creative Conceptual Design (CCD) strategies. One is to encourage designers to think openly to discover new functions and integrate their PSs into an artifact. The other is to encourage them to adopt novel PSs to innovate existing artifacts. For both strategies, designers should explore wide multi-disciplinary solution spaces for generating novel or promising PSs.

Such multi-disciplinary exploration tasks, however, are often very difficult for human designers, who are merely taught with knowledge about certain disciplines. In fact, they often have to rely on the solution knowledge in familiar discipline(s) to generate PSs, which, though, can lead to the loss of novel or promising PSs. Today this issue becomes even worse, since many new solution principles are discovered in various disciplines at a very fast speed, making it impossible for human designers to learn and utilize them for creative design in time. Therefore, a Computer-Aided Conceptual Design (abbreviated as CACD later) tool should be developed to help human designers achieve the CCD of multi-disciplinary systems.

However, few studies have been carried out on CCD of multi-disciplinary systems, though many CACD studies have been done. Multi-disciplinary systems here refer to the engineering systems that comprise components from various engineering disciplines, such as mechanical, hydraulic, electrical, electronics, photoelectric, and energy disciplines. Our research will focus on developing a knowledge-based system for CCD of multi-disciplinary systems. Since the whole PS of a complex system can be decomposed into some basic PSs, a reasonable CCD approach is then to reuse such basic PSs and integrate them in a system for achieving a desired function, which is also the approach adopted here. Our multi-disciplinary CCD approach involves first building a knowledge base of such basic PSs, and then selecting suitable known PSs and creatively synthesizing them together. The work reported here is an extension of our functional representation research in [2]. The current paper also develops a novel algorithm for achieving design synthesis of multi-disciplinary PSs. The rest of the paper presents the related work, the proposed approaches and their implementation.

Section snippets

Literature review

Due to limited space, only some typical CACD systems are reviewed here. Interested readers can find more in some review or research papers, e.g. in [3] or [4]. Existing CACD systems can be classified as reuse-oriented systems and creation-oriented systems. The reuse-oriented CACD systems often aim at modeling the artifact knowledge of a total engineering system for guiding subsequent design of similar systems. For examples, the FBS modeler employs the Function-Behavior-State approach to model

Functional representation

Function refers to the general relationship between the input and the output of a system aiming at performing a task [1]. Since function plays a critical role in conceptual design, its representation is critical for a CCD system. Here, functional representation is classified as the representation of desired functions, and the representation of the functions achieved by known PSs, which is called the representation of functional knowledge for brevity here. Since flows are the primary components

A HOSS-based synthesis approach

Depending on how functions are represented, there are two design synthesis approaches, i.e. the flow name-chaining approach (e.g. in [18]), and the traditional state space search approach (e.g. in [9]). Since our functional representation approach differs significantly from previous ones, a new approach, called the HOSS-based synthesis approach, is developed here to achieve CCD of multi-disciplinary systems. Here, HOSS is the abbreviation of Heterogeneous Object State Search, which means that

Implementation

Using ASP.NET in Microsoft Visual Studio 2005 as the web application development tool and Microsoft SQL Server 2005 as a relational database system, a browser/server architecture-based prototype system, called ICCDP (Intelligent Creative Conceptual Design Platform), has been developed for implementing the proposed multi-disciplinary CCD framework. ICCDP is primarily composed of three sub-systems, i.e. the basic data management sub-system, the PS knowledge management sub-system and the

Conclusions and future work

The proposed approach for CCD of multi-disciplinary systems paves a way for designers to integrate PSs from various disciplines into a multi-disciplinary system for achieving a desired function. It is primarily composed of a formal approach for representing a desired function, a domain-independent approach for representing functional knowledge of known PSs, and a HOSS-based synthesis approach for automated CCD multi-disciplinary systems. A prototype system, called Intelligent Creative

Acknowledgements

We are grateful to the reviewers and the editors for their constructive suggestions. This research is supported by Ministry of Science and Technology of China (Granted No. 2008AA04Z108), Natural Science Foundation of China (Granted No. 50975173 and 50935004) and Science and Technology Commission of Shanghai Municipality (Granted No. 09QA1402800).

References (22)

  • K. Ulrich et al.

    Synthesis of schematic descriptions in mechanical design

    Research in Engineering Design

    (1989)
  • Cited by (96)

    • Scalable multi-process inter-server collaborative design synthesis in the Internet distributed resource environment

      2021, Advanced Engineering Informatics
      Citation Excerpt :

      They complete the searching mission within a time span of 11 design flow visits. As for the intensive and serial working mode adopted by the previous works [1,38,47,57,65–67], there is only one process taking all the design flow visits which is 107 in total. The pure concurrent multi-process working mode also consumes very short time span, however, it occupies more computing resource compared with the scalable multi-process working mode.

    • Knowledge creation in the automotive industry: Analysing obeya-oriented practices using the SECI model

      2020, Journal of Business Research
      Citation Excerpt :

      Finally, the case study is discussed with specific reference to the modes of the SECI model, before the implications of the research and our conclusions are presented. Previous studies on knowledge creation have developed interesting research questions about the conditions that facilitate knowledge creation, how to encourage knowledge creation (Tsai & Huili, 2007), and organizational mechanisms that may allow individuals to develop knowledge collectively within the organization (Chen, Liu, & Xie, 2012). However, there are relatively few contributions discussing opportunities and limitations on context-specific mechanisms of knowledge creation in particular industries (Yang, Fang, & Lin, 2010).

    View all citing articles on Scopus
    View full text