Human Behaviour in Design

“Human Behaviour in Design” is supporting the discussion concerning the role of human beings and their strategies in engineering design.

It was of high value, to have just a few speakers and to spend some time with the posters of all the participants and to have a lot of time for discussion. By the excellent posters all the authors have concentrated their key messages for the presentation and the following discussion. All the posters are published in the web [www.pe.mw.tum.de/hbid/].

To identify successful design strategies for early stages of the design process, individual design procedures have been investigated with the aim to determine the applicability of Design Methodology in these phases of the product development process. First results have been published at ICED’01 (Bender et al. 2001b). This paper provides results of a detailed analysis of the observed design procedures of N=71 participants (83 evaluable cases) and their effects on design performance. Based on the results of this empirical study a re-interpretation of conceptualisation is proposed as a contribution to the advancement of design theory.

Initially a quotation of Prof. G. Roth (Institute for Brain-Research, University of Bremen): “Will, thinking and behaviour of man are governed to a great extent by the limbic centres, which work fundamentally unconscious. Our rational thinking has a very limited access to these processes”. (Roth 2000).

In this chapter, we explore the effect of strategic knowledge in conceptual design process by examing the differences in cognitive processes and groupings between a novice and an expert. In our previous studies, we found that the expert’s cognitive activity and productivity (in terms of image generation) were three times as high as the novice’s in the overall design process (Kavakli et al. 1999). We investigated the structure of cognitive actions in the design protocols, and found that there is evidence for the coexistence of the cognitive actions (Kavakli and Gero 2001). Certain groups of cognitive actions increase and decrease in parallel with each other in the protocols of the novice and expert designers. We suggested that the differences in the performance of designers could be attributed to the differences in the structure of those concurrent cognitive actions. Investigating the concurrent cognitive actions, we found that the expert’s cognitive actions are well organized and clearly structured, while the novice’s cognitive performance has been divided into many groups of concurrent actions. (Kavakli and Gero 2002). In this chapter, we focus on explaining the difference in performance between the expert and novice in terms of their respective strategic knowledge.

Whether we define designing as systematic problem solving or, conversely, as free-wheeling artistic creation, we postulate that there is agreement about the following:

During the process of designing, designers continually reason about prospective features of the designed entity and the rationale for accepting or rejecting them. This is taken to be true whether the designer is an individual working by him or herself or a team working on a design assignment jointly.

This paper deals with the elementary analysis of cognitive processes in mathematical problem solving for better diagnostics. The present experiment was designed to analyse the internal process and to localize the neural substrates involved in solving mathematical tasks, using EEG-coherence. The internal process is revealed by a sequence of microstates. Microstates are defined here as coherence maps which remain stable over time. The difference in performance between gifted and normal subjects in solving mathematical problems is reflected by the difference between the sequential and topographical properties of microstate sequences. Sequential properties were measured by means of entropy reduction. A higher entropy reduction was found in mathematically gifted subjects in contrast to the normal subjects. Topographical properties were measured by means of difference maps of microstates between mathematically gifted and normal subjects. Gifted subjects exhibit a higher coherence left and right parietal and left frontal. This result corroborates the double representation hypothesis.

A core area of scientific thinking is explaining. This means answering to the “why-questions and how questions” (Hempel 1965). Why does Sam have a fewer? Why did an organization fail abroad? Why a structure is able to support the weight of snow? How more effective valves for an engine can be designed? How to make computer games more attractive for female users? These are typical examples of design problems, all of which should be based on scientific explanation, i.e., what should be answered based on the laws of nature or as is becoming increasingly more evident, based on the laws of the human mind.

It is said, though not without controversy, that what distinguishes design from art is function. Design is for a purpose, usually a human one. As such, design entails both generating ideas and adapting those ideas to intended uses. This occurs iteratively. Form and function. Studying how people go about both these tasks gives insights that can facilitate the design process. Two relevant projects will be described. The first investigates how designers and novices get ideas from sketches and applies those insights to suggestions for promoting generation of ideas. The second seeks to develop computer algorithms for designing individualized visualizations, algorithms that are informed by cognitive design principles.

In a collection of papers presenting the Developments in design methodology, Cross (1984) presented “prescription of an ideal design process” as the first of the four stages that he distinguished. This stage, qualified as the period of “systematic design”, was reflected by texts dating from 1962 to 1967. Stages two, three and four were “description of the intrinsic nature of design activity” (1966–1973), “observation of the reality of design activity” (late 1970s), and “reflection on the fundamental concepts of design” (1972–1982). Nearly twenty years after 1984, normative, prescriptive models of design (exemplified by Pahl and Beitz 1977, 1984) are still very powerful —even if the Human Behaviour in Design Symposium (2003, this book), organised in Pahl and Beitz’ Germany with its particularly strong methodological tradition, testifies to a movement that might correspond to the second and third stages identified by Cross (1984): Are they stages of a new cycle?

The participants in stream I consisted of a very good mixture of people from different disciplines, several of which had not met before. As a result, the short poster presentations, gave rise to numerous discussions, showing not only overlap in results and ideas, but also a strong discipline-dependent use of terminology. The poster presentations and the initial set of questions resulted in a large number of interesting issues, which were grouped and discussed, resulting in three main questions that formed the basis for the second part of the session.

Due to its depictive nature mental imaging can handle spatial issues as good as sketching. Besides, imagery is a mental experience and is perhaps an ideal tool to synergistically work with the process of thinking — that too without the load of overt sensory-motor operations that the action of sketching demands. It is no wonder therefore that most designers talk about rich imagery experiences that they go through. Yet, mental imagery has not been used as a pedagogic tool in the design studio environment.

Technical development and globalisation impose problems on engineering design which require highly efficient processes due to time, quality and cost. Products, processes and knowledge change so quickly that there are hardly any standard procedures or routine solutions adequate for various situations. Overnight the best solutions may become inappropriate if parts of the problem context are changing. This situation is even more difficult because the designer is integrated in a social context, and thus, processes of division of labour, organisation of tasks, planning of interfaces, as well as continual linking individual work with group work are concurrent requirements in product development.

The purpose of design research is to support development of better products by developing a better design process. A design process is initiated with the recognition of a need, leading to the establishment of requirements for the intended product. Therefore, establishing requirements is essential, and should be a central issue in design research (Chakrabarti 1994). A requirement is defined here as characteristics which a designer is expected to fulfil through the eventual design. Several categories of requirements have been proposed and their importance proclaimed (Birkhofer 1992). Several methods are suggested in design literature for aiding requirement establishment, from check-lists and QFD methods to combination of both as a software support ((Pahl and Beitz 1988; Pugh 1991; Kruger and Knoll 1994). Design process has been investigated in many studies and requirements remain part of these studies (e.g., Blessing 1994; McGinnis et al. 1989). However, detailed investigation as to how requirements get identified, analysed and used in the design process and how these influence the quality of its outcome — the emergent design - has not been undertaken before. The findings outlined in this article are based on four separate but related empirical studies of the design process. The first two studies (Morgenstern and Knaab 1996; Chakrabarti et al. 2002; Nidamarthi et al. 1997; Nidamarthi 1999; Nidamarthi et al. 2001) were aimed specifically at understanding how requirements are identified and satisfied during design, while the other two were focused on how shape evolves during design (Baumgartner 1995; Chakrabarti and Wolf 1995). In this article, results from these four studies have been brought together to identify the activities and mechanisms for requirements identification and application, and evaluate their effectiveness in developing products with a high degree of requirement satisfaction.

Designing is question intensive. However, our knowledge of the role of question asking in design is limited. The research presented in this article is a summary of the significant findings of a doctoral dissertation that addresses this limitation (Eris 2002). When the findings are considered together, they constitute the conceptual framework of a question-decision centric design thinking model.

The product development process has undergone significant changes in the last two decades. Consumers have become more sophisticated in their choices, products have become more complex, and the barriers to entry to competitors have been significantly lowered. All this has resulted in an increased emphasis on creative teamwork and shorter product development times. In turn this has led to an increased number of descriptive studies of the engineering design process in the research community as well as the development of various tools and methods aimed at improving the process. Success in this endeavor has been a difficult battle for the research community and several authors have made attempts to analyze the obstacles responsible for the difficulty. Blessing for example notes:

“... identifying whether this [a method or tool] indeed contributes to success is far more difficult and the results are not easy to generalize. Success is difficult to measure other than in a real, industrial situation and action research in an industrial situation is notoriously difficult, let alone comparative action research. Furthermore, the success of a method or tool depends on the context in which it is being used. This context is different for every design process, because every design project is unique” (Blessing et al. 1998).

Product and process development is time dependent and complex e.g. as there are many “agents” taking part in the processes who occasionally can cause new situations to suddenly appear. Such agents are managers, internal and external specialists, team members, test persons, customers, users, inventors, controllers, etc. Product and process development is also complicated as there are many knowledge areas involved, such as engineering, business administration, computer technology, sociology, psychology, management, etc. Also to take in account when forming good conditions for successful development is that chaotic situations can develop rapidly as e.g. new players can occur by chance, can new technologies suddenly appear, can the financial support completely change the grounds for development, etc.

Globalisation has resulted in increased multinational cooperation. To compete in international markets, small and medium-sized manufacturers need to exploit their particular specialisms (e.g., in terms of personnel, knowledge and resources) by making strategic alliances with companies having complementary areas of expertise. By making alliances with designers located in primary international markets, indigenous manufacturers hope to achieve an in-depth understanding of the target market and their requirements. Cross-national alliances offer the potential for mutual benefits, such as sharing information and resources, reducing costs, enlarging markets, etc. In addition, the geographically dispersed team members may be supported by a number of communication infrastructures (e.g., networked, multimedia environments), which can not only reduce lead-time and the cost of product development, but also facilitate collaboration with partners (Siemieniuch and Sinclair 1999).

The industrial world is changing rapidly and engineering organisations need to adapt fast to remain in business. Some of the pressures faced by engineering managers include: (1) the trend towards globalisation and the reliance on IT; (2) the increasing commercial pressures which demand continual improvements in quality, shorter lead times and reduced costs; (3) the increasing complexity of both products and processes; (4) the problems of retaining and retrieving knowledge and experience; and (5) the need to move towards sustainable development.

Considering the multitude and variety of factors that determine interaction between individuals, it is sensible to establish a common understanding of the issue as a starting point for a detailed discussion. This common understanding can be achieved by a shared mental model of interaction between individuals, i.e. a system with related elements (Fig. 6). This model, which is based on the holistic view in Fig. 5, can also be used for emphasizing independent objects of research and facilitating the demarcation of research areas and their objectives.

Investigations of how the design process actually proceeds have shown (B&B Reports 1994–2001) that engineering design can be made both more efficient and better at fulfilling the original purpose if the design process flow is based on order and clearness. This applies to the description of the objects being designed at the various stages of their development; it applies, likewise, to the formulating of rules, regulations and critical paths with which it will be necessary to support a methodical approach, and it also applies to their use. Objects being designed are to be represented in a variety of external forms, but, frequently, the full possibilities of such representations are not exploit. For the concrete stages of the design process, technical drawings with their laws and norms are one well-established and useful type of representation. Another is geometrical modelling, whether with two-dimensional or three-dimensional CAD systems. However, there is no set of representational and other support tools to assist in the earliest stages of engineering design. In hopes of closing this gap to some extent, the paper concentrates on cognitive outsourcing (see introduction of the chapter) and helps for conceptual design, with the aim of offering suitable methods and computer-assisted tools which may support the designers conceptual work.

At the institute of product development of the Technical University of Munich, a workshop was held to bring together designers and CAD software developers. At the workshop, a designer, who worked for an automotive supplier for clutches, complained that CAD restricted him in being creative. On this complaint, the representative of a company that distributed a major CAD-system responded that “for a supplier, restrictions from automotive manufacturers are so severe, that there is no room for creativity, anyway.” This statement may reveal a potential misunderstanding according to creativity. The designer may not be able to invent a new principle for clutches, but does he therefore not need to be creative?

When people are asked about their definition of VR, they might say something like “VR is being within artificial worlds or dreams”, “VR is the technical creation of real experiences” or “VR is interacting naturally with non-real entities”. Those quotes are not very precise, but they already contain the “three I”, introduced by (Burdea and Coiffet 1994) for a definition: immersion, which is the attempt to provide the VR user with feedback that helps to create the feeling of “being there” (presence); interaction, which integrates the user even more into the virtual environment offering him the power to change and use it; imagination, that is the ability of human beings to believe something is real even if its representation is somewhat unreal. Another, more precise and technical definition would be “VR is a human-machine interface, which allows for perceiving an artificially generated environment as a reality with multiple senses involved in that process.”, which is based on (Hennig 1997).

The research has two backgrounds. One is the recognition that design can be regarded a largely knowledge-based activity. Historically, research in design knowledge began with knowledge representation and reasoning. Design knowledge representation looked at knowledge about design objects and design processes. While the former is obvious, design process knowledge is all about how to design and is related to design reasoning that deals with how to use design object knowledge. In contrast, design reasoning was mostly concerned about how to computationally arrive at conclusions, but not how to use which knowledge, which is the central question of knowledge deployment. As theories and technologies to deal with knowledge made progresses, research focuses gradually extended to such topics as modeling, acquisition and capturing, learning, discovery, data mining, maintenance, management, reuse and sharing. These addressed, however, little about how to make knowledge well ready for use.

When designers need to generate ideas, they tend to take paper and pencil, and start to produce idea sketches. Or, they may call an idea generation meeting. The available techniques for generating ideas in such idea generation meetings are mainly based on writing as a working medium. This research project started with the notion that idea generation techniques for design could be enhanced if sketching could be included in such idea generation meetings. For, many researchers regard sketching to be instrumental not only to the designer’s creative process (e.g. Fish and Scrivener 1990; Goldschmidt 1991; Schon and Wiggins 1992; Goel 1995) but also to the design team’s creative process (Bly 1988; Tang and Leifer 1988; Scrivener and Clark 1994).

It is a common idea that designers do not take into account the people who are confronted with the their ideas and products. Many studies about designers activities (e.g. in engineering and software design) show that especially in the early stages of the design process (clarification of the task, conceptual design) usability aspects are neglected (e.g. Dylla 1991). Especially novices concentrate mainly on technical aspects (Ahmed 2001). Even if designers work for other designers, this is a major problem. This is shown by the research of Andreasen and others about integrating the perspective of other designers while developing design tools (Araujo et al. 1996; Andreasen and Mc Alonee 2001; Araujo 2001). Often instructions for customers are based on a deficiant model of the users knowledge, motives and habits (Hacker 1991). This situation is criticized by many engineers. Concepts like “usability engineering” (e.g. Nielsen 1993) show the importance and the necessity for improvements in this field. This can happen in several ways. (a) Methods for designers should be developed and evaluated which integrate the user’s perspective in designers thinking and work process. (b) Users should participate in the design process not only by the formulation of requirements (if customer and user are identical) and by the testing prototypes and products. He should also be involved in other stages of the design processes e.g. conceptualisation of first ideas. By that way designers should get solutions which fit in a better way to users needs.

Session III worked primarily in the area of the following three main topics:

While making notes during the conference and trying to summarise what had come out during two days of intensive discussion, it became more and more clear that the three issues had started to merge. The rather distinct questions that were formulated before the conference, were reformulated and refined during the first discussions into a set of questions that guided the subsequent discussions. Although this was done seperately in each of the thematically different streams, some overlap started to emerge. The subsequent discussions, formal as well as informal, showed that from which point of view this research area was considered - individual, group or computer support - common issues emerged that turned out to be concerned with the effectiveness and efficiency of the process of doing design research, rather than its research questions.