The premise of our work is that designers, in the process of doing their work, create models of various kinds, for various purposes, and that it is the negotiation of the structure and content of these models that comprises the bulk of the task of doing design. We give here an overview of a framework for enabling designers to capture and structure as much of the information they use and generate as is possible. We have designed and implemented such a system for creating models in a computer that can be shared with other designers in the course of an ongoing design, made persistent for future recall, classified and categorized so as to facilitate both the study of how design is done in a given organization and the study of design in general.

Arguing that design is a social process, we expand the meaning of modeling and analysis to include all activities facilitating continual refinement and criticism of the design requirements, process, and solutions. We do not assume any a priori methods for modeling or analysis; rather, we provide a framework and an approach to study designers and give them whatever modeling and analysis capabilities they choose. Our approach is the basis for a support tool, n-dim, currently under development. AI EDAM, 1992, 7(4):257-274 Subrahmanian et al. (1992) 1 The Objective of Modeling and Analysis Design as a social process involving designers, customers, and other participants consists of creating and refining a shared meaning of requirements and potential solutions through continual negotiations, discussions, clarifications, and evaluations. This shared meaning, crystalized as the design artifact and made persistent as shared memory forms the basis of accumulated experience upon which subse...

To be effective in practice, concurrent engineering requires access to and organization of knowledge accumulated over time, product versions, and customers. More important, separate knowledge resources have to be shared and coordinated over space and time if successful design is to be accomplished. In this paper, we address the nature of communication in design, especially across disciplines, and the support systems that facilitate better communication. While a lot of research effort is being expended on sametime communications within a group, we consider, as well, the need for, and the problems associated with, different-time, different-place communication. We present these views in connection with an on-going development effort, n-dim. 1 Introduction A large scale design project must coordinate expertise from many different disciplines. Empirical studies conducted by us and others indicate that design is a continual negotiation of constraints, terminology and trade-offs for the crea...

: Research is a design activity whose decisions involve the ways in which research is carried out and its results interpret. These activities comprise what is referred to as research methodology. This paper brings the concepts of capturing design rationale and machine learning to bear on the design of research itself. Therefore, design decisions concerning research must be recorded to allow for understanding feedback and updating of research strategies. In addition, successes as well as failures of research decisions must be reported to facilitate learning about research. This requires a shift in the way research is carried out and reported. This paper illustrates this shift in the context of a specific project on machine learning in design. Reich (1993) 1 Introduction There are two senses of science according to McMullin (McMullin, 1987). In the first sense, denoted by S 1 , science is a collection of hypotheses, theories, observations, data, etc. that are the end products of a long...

: Design activities are fundamental to technological progress. Current design research holds tight to positivism, abandoned and critically opposed to by philosophers, mostly those outside the U.S. Maintaining the positivist view when conducting research leads to significant deficiencies in the quality of research, and to problems in transferring research results to practice. In spite of significant research efforts, the improvement of practice is slow. This improvement, in turn, does not necessarily reflects the diffusion of research results into practice, but rather, the development of ideas by practitioners. This paper analyzes this theory-practice problem of technology from practical, cultural, and philosophical perspectives. It proposes a research methodology of design and briefly shows how this methodology can shed light on some problems related to technology. The paper also discusses the fundamental role of design in technology, thereby viewing the research methodology proposed a...

Studies on design research methodology are infrequent, although there is a consensus that more effort is needed for improving design research quality. Previous calls for exercising better research methodology have been unsuccessful. As numerous studies reveal, there is no single scientific methodology that is exercised in science or in any other research practice. Rather, research methodologies are socially constructed. Since some constructions are better than others for different purposes, it becomes valuable to study different methodologies and their influence on research practice and results. Proposals for such studies are offered. 1 The state of design research methodology In many disciplines, research methodology is seldom discussed by researchers. Such neglect may result from several attitudes towards research methodology including indifference or ignorance. Researchers may be indifferent because their research is well received by the community therefore they need not change or w...

This paper describes the development of computational support tools for practically successful engineering techniques. The paper reviews the requirements for manual Quality Function Deployment techniques, presents them, and discusses their limitations. It argues that computational support tools can alleviate most of these limitations and that a graph-based information representation for such techniques is an excellent choice for supporting both QFD techniques and their integration with other external CAD-related computational services. The paper presents an architecture for a computational QFD (CQFD) tool based on the graph-based modeling environment n-dim. It shows how this architecture supports most of the requirements for QFD techniques, in addition to providing many additional functionalities, and briefly illustrates how the CQFD tool will be used. Keywords Quality Function Deployment, TQM, design practice, 7 management tools, graph representation, n-dim, collaborative design, d...

Bridge databases contain significant information that can assist in future bridge management decisions. The effective utilization of this information requires facilities for the flexible extraction of usable knowledge in the form practitioners can use. We propose machine learning technology as a means to perform this knowledge extraction and to support the maintenance of bridge database information. An integrated system is proposed for supporting practitioners in the practical use of this technology. INTRODUCTION The premise of this research is that bridge databases, containing historical data on bridge performance and maintenance decisions, contain knowledge that can be used to enhance future decision making. The major impediment on using this vast amount of data in practice is the problem of making the knowledge embedded in the data explicit for decision makers. While the development of new database technologies, comprising the majority of present efforts in bridge management resea...

this paper is immaterial. What few researchers may do to patch the above flaw is to replace practical testing with testings on benchmark problems. To be relevant to practice, benchmark problems must be representatives of the problems encountered in practice in all their complexity. Unfortunately but clearly, benchmark problems are never complex as practical problems and their relevance to practice is doubtful (e.g., in optimization research (Haftka and Sobieski, 1992)). Replacing practical testing with benchmark testing removes any hope for obtaining even a semi-independent yardstick for measuring research quality or practicality. Gaining control over the practical relevance of research is the theme of this paper. Therefore, I do not discuss nor advocate for a particular CAE research topic; such topics rise and fall as fashions change. In this position paper, I advocate for an approach for doing CAE research that has better chances of impacting practice. In what follows I review three observations that influence the quality and practical relevance of research. The observations are not new, their seeds can be located in very early CAE research projects; however, their impact on research is often ignored. In contrast, I incorporate these observations into a hypothesis about improving CAE research. In the spirit of this workshop, I use examples from Fenves' research: the development of STRESS and the research on decision tables. 3 Observations about research: A CAE perspective

Abstract not found