INTERNATIONAL CONFERENCE ON ENGINEERING DESIGN ICED 99 MUNICH, AUGUST 24-26, 1999 THE ECOLOGY OF INNOVATION IN ENGINEERING DESIGN Andrew Milne and Larry Leifer Keywords: Innovation, Ecology, Environment, Engineering Design 1 Introduction This paper explores an approach for studying innovation in engineering design. We propose a framework that suggests design activity takes place within an ecosystem composed of populations, environments, resources, and activity cycles. Using our construct, we note aspects of the framework that seem critical to the fostering of innovative activity within engineering design teams. The paper is presented as a philosophical discussion grounded in research results reported in the literature as well as the authors experience. 1.1 The Nature of Creativity and Innovation Innovation is closely associated with creativity. In this paper, we distinguish creativity from innovation, asserting that creativity involves the act of generating a new idea or solution concept, while innovation refers to the act of either applying some creative idea, or creatively applying a familiar idea, in such a way as to create value. In design, innovation refers more specifically to the application of solution concepts to fulfill a set of stated and unstated design requirements. Does innovation come primarily from individuals, or is it the results of team interactions? Tatsuno[1] considered this issue in comparing Western ideals about creativity, which focused on individual-based fission process, to those of Japan, which emphasize team-oriented fusion process. A fission-oriented view places the process is at the mercy of the few creative individuals who are able to generate ideas and facilitate implementation. A fusionoriented perspective, however, implies opportunities for encouraging innovation through proactive management of interactions between and among designers and their environment. It is this latter view that motivates our discussion. 1.2 Foundations of Ecology In the first issue of Ecology, Barrington Moore inaugurates the new journal with a reflection on the embryonic discipline for which it was named: Since the biological field has been reconnoitered and divided into its logical parts, it becomes possible to see the interrelations and to bring these related parts more closely together. Many sciences have been developed to the point where, although the field has not yet been fully covered, contact and cooperation with related sciences are essential to full development. [2] Ecology is a branch of science concerned with the interrelationships of organisms and their environment. Design activity takes place within an ecosystem of its own, one comprised of a
complex, dynamic community of populations and their environments that function as a unit. This may be an alternate form of design science that the research community could embrace the study of relationships and interactions between people, information, environment, and artifacts rather than the formalization of procedures and strategies. 1.3 Relevance to Design Research Engineering design researchers are faced with a similar situation to that Moore and his colleagues faced in 1920. Our discipline has benefited from years of research into specific questions regarding the nature of engineering design, but the nature of the design activity has made it difficult to completely characterize the complexities of design work. Nevertheless, the literature does present findings that support, indirectly, the notion that a synergy exists that is broader than many of the reductionist examinations of design activity. This in turn suggests the need develop research approaches that account for the interactions between elements of the ecosystem, allowing for independent investigations of each element such that the results can later be combined to develop knowledge about the overall system's behavior. 2 Structural Considerations In order to study the nature of innovation in design, it is first necessary to determine the extent of the investigation. Here it is useful to consider a new structural model, one which we base on the notion of a design ecosystem. 2.1 Suggestions in the Literature Some findings that have been presented in the literature lead us to contemplate this broader framework for design. Several authors have observed that design is a complex social activity. Minneman[3] presented a social construct that moved beyond procedure-oriented models of design activity. Minneman s work served to broaden the discussion of design process beyond consideration of merely the artifact and procedure, encompassing as well aspects of human interaction. Brereton[4] added to the discussion through her analysis of the interactions between designers and hardware that is, between designers and prototypes that in some way have a bearing on the design problem and possible design solutions. This reinforced the observations made by Cross[5], who noted that, In most design situations it is not possible or relevant to isolate analysis of the problem from the solution concepts. Brereton analyzed the connection between designer and physical artifacts, rather than abstract solutions, and concluded that designers do indeed make use of such opportunities throughout the design activity. Design, then, involves interactions both between and among human designers and technical artifacts. Other factors also influence the activities of designers and design teams. Culley and Owen[6] identified a series of factors that designers reported constrained their innovation in a longterm project. These included process disruption, lack of think time, lack of awareness of new techniques, and inadequate time for developing prototypes. Together, these factors suggest that design cycle cadence and availability of stimulation affect innovation in design. Frankenberger et al.[7] identified several factors influencing success in design activity, among them group organization, group climate, and availability of information.
If the technical artifacts of design are considered to be a part of the physical environment, these other issues can be considered aspects of the intangible environment. The findings suggests that in addition to human designers influencing their environment, aspects of the environment can influence the designers and their design activity. The study of this interaction, and the characterization of their interrelationships, we call design ecology. 2.2 A Proposed Framework Pioneers in the field of ecology sought to find relationships between scientific disciplines that were connected by the common thread or organisms and their environment. The design research community has a similar opportunity to stitch together elements of practice-inspired and our research-confirmed knowledge about design. The process of engineering design follows an analogous system model to that of a biological ecosystem, where a collection of elements affect the nature of activity, including population, environment, climate, activity cycles, and resources. Each of these has a relationship to one or more elements in the system, creating interdependencies that sustain the activity. Consider one interpretation of these elements in an engineering design context. POPULATION Several sub-populations participate in engineering design. The first is the design team, whether a group or a team of one. The complexity of the design team in terms of background, discipline, geographic distribution may require consideration of the team as a collection of smaller groups. Another sub-population category would include any support organizations marketing, sales, administrative support, and others who in some way interact with the design team in the course of its activities. The final sub-population we will mention is the client or customer who, although perhaps not directly interacting with the design team or its processes, drives the design indirectly by exhibiting particular preferences and behaviors that the design team will attempt to appease. ENVIRONMENT Like population, the design environment can be considered a collection of several subsets, organized according to the categories of physical, organizational, and virtual environments. In the physical sense, design environments consist of meeting spaces, work surfaces, physical design artifacts and prototypes, architectural features, and other settings or objects that physically interact the population. Aspects of the organizational environment would include corporate mission and values, division of responsibility among individuals and work groups, work processes, and organizational structure. Recently, the virtual environment has become increasingly important as electronic CAD and communication tools have become prevalent in design. Efforts to improve knowledge management and the increasing use of electronic collaborative workspaces mediated through Internet-based technologies are moving much of the design work environment into a digital realm that should also be examined as an aspect of the designer s environment. CLIMATE For a given population and environment, design outcomes can be affected by changes in the climate in which design activity takes place. Climatic elements might include market conditions, morale of the design team, and volatility of design requirements. ACTIVITY PATTERNS Another aspect of the design ecosystem is the nature of activities in which members of the population engage. The notion of cyclical patterns is a familiar one in design, whether it be iterations during a particular phase of the design process and the reapplication of the same design process to a new product. Additional cycles of activity in design need to be investigated, including the recycling of engineering design information and the premature re-design of products to maintain competitive advantage.
RESOURCES Much like biological ecosystems, design ecosystems need to be nourished with the natural resources that sustain their populations. Those resources include, but are not limited to, appropriate personnel, timely market information, computing resources and other design support tools, relevant engineering design knowledge, and financial support. 3 Ecological Characteristics of Innovation The process of innovation involves creating connections between ideas and solution concepts that are often applied outside of their normal contexts. Many of the creative solution-finding techniques commonly proposed among them morphological analysis, brainstorming, synectics are designed to force out-of-context thinking by using structured protocols to influence the cognitive activity of individuals. Team-based design, rather than focusing exclusively on structured methods for developing concepts, allows the manipulation of environmental conditions to encourage innovation through interactions between individuals. The following section makes some initial suggestions about what this type of environmental manipulation might involve. 3.1 Promoting Innovative Design Ecosystems The authors have independently built team-based engineering ecosystems within academic settings, taking steps to encourage innovative outcomes from those systems. Leifer established a graduate-level design course, ME210, in which student design teams work with industrial clients to develop design solutions to client-defined problems[8]. Milne founded the Envisioneering program, which simulates an innovative industrial setting in which engineering student teams design and implement projects[9]. Both programs have been recognized for their valuable results. From these experiences, and the literature, we draw the observations given below. While not a set of recommendations, these observations demonstrate use of the framework to examine aspects of successful innovative ecosystems. The interrelationship between these and other elements merits more rigorous study. POPULATION The importance of bringing fresh ideas and perspectives to encourage innovation suggests the need to build diverse design teams. Variety based on technical discipline, work experience, experience level and sub-population membership are commonly encouraged. In the ME210 course, diversity according to personality preference is also promoted[10]. Beyond diverse composition, the Envisioneering experience demonstrated that strong social bonds within the design population can promote risk-taking associated with suggesting new ideas. ENVIRONMENT The establishment of a skunkworks or research & development division in industry allows designers to work on design challenges outside of the normal operations of the production environment, permitting an informal organizational structure that reflects the informal connections that are made in connecting ideas. The ME210 course provides this informality, along with a virtual environment consisting of design knowledge capture and team communication tools to support the design process, allowing designers to innovatively re-use design knowledge. ME210 also provides a dedicated physical environment the design loft that is highly reconfigurable and provides both social and work space within an informal setting.
CLIMATE Hallman identified a number of cultural influences that discouraged creativity, thus suggesting guidelines for providing positive influences, among them a playful atmosphere and acceptance of failure[11]. In the Envisioneer program, organizational values and team culture were arranged to promote a dynamic environment, and strong social bonds. The outcomes suggest that a proper cultural climate can promote innovation without the use of structured cognitive techniques. ACTIVITY PATTERNS Culley and Owen s observation that think time was a constraint of innovation in design resonates with creativity process models that note the importance of an incubation period for new ideas. Brereton s results indicate some form of prototyping process is also valuable in the early phases of design. The pressure of a defined deadline and limited time, however, also seems to inspire greater risk-taking, and often greater innovation. RESOURCES Cannon et al. have studied the use of different electronic media in the design process to communicate design information, in particular noting how different media impact the design ecosystem[12]. The access to past design information has led to successive innovative improvements to the designed artifacts. The authors have also observed one highly-regarded Palo Alto design firm that uses a stuff box of interesting technologies to stimulate design ideas. These cases illustrate how the form and content of one resource, engineering design knowledge, can influence designers and promote innovative design. 3.2 Implications for Design Research The ecosystem model suggests a need to consider how disruptions in one area might cause perterbations elsewhere in the model. Clearly limits on resources impact the design process, restricting the availability of prototyping, personnel, and design information. Less wellunderstood are how the quality of the environment, in its many forms, may impact design activity, in particular design activity aimed at developing innovative products. The ecosystem framework presents a series of research questions that have yet to be addressed. While examination of design teams (population), electronic tools (virtual environment) and processes (activity patterns) are common in the literature, other areas of design ecology are not being extensively investigated, particularly as they relate to innovative design. Some pivotal questions that remain include: How do populations use different forms of stimulation to inspire design innovations? In what way can the physical and virtual environments proactively affect design activity? What resources are most critical at different stages of the design process? What are the critical climate parameters that encourage innovation in design? The most promising approach to understanding the ecological aspects of design activity may involve creation of observation-verifiable simulations. Christiansen reported success with one such tool, The Virtual Design Team (VDT) [13]. VDT is capable of simulating design team communication based on the characteristics of tools with which designers interact and exchange information and the usage patterns engineers exhibit with these tools. VDT successfully models a single dimension of the complex interaction within a design ecosystem population. Additional simulation tools are needed to more completely model the complexity of interactions among all elements in a system.
4 Conclusion The ecosystem framework attempts to organize a number of research thrusts into a more unified model of design. While the challenges involved with developing a working simulation model have not been solved, the design ecosystem construct maintains their visibility. The authors are currently engaged in research that will study the effects of physical and virtual environments on engineering design activity using prototype technology-augmented design environments. The design of the protoype environments will itself be studied in an attempt to develop more information about the interactions within design populations and techniques they use for stimulating design concepts for innovative design artifacts. References [1] Tatsuno, S. Created in Japan, Harper & Row, New York, 1990. [2] Moore, B., The Scope of Ecology, Ecology, Vol. 1 (1), 1920, p. 3-5. [3] Minneman, S. The Social Construction of a Technical Reality: Empirical Studies of Group Engineering Design Practice, Ph.D. Thesis, Stanford University, 1991. [4] Brereton, M. The Role of Hardware in Learning Envingeering Fundamentals: An Empirical Study of Engineering Design and product Analysis Activity., Ph.D. Thesis, Stanford University, 1998. [5] Cross, N., Engineering Design Methods, John Wiley & Sons, New York, 1989. [6] Culley, S. and Owen, G. Drivers and Constraints to Innovation An Overview, Proceedings of the 11 th International Conference on Engineering Design in Tampere,Volume 3, Schriftenreihe WDK 25, Tampere 1997, p. 559-562. [7] Frankenberger, E., Badke-Schaub, P., Birkhofer, H. Factors Influencing Design Work, Empirical Investigations of Teamwork in Engineering Design Practice, Proceedings of the 11 th International Conference on Engineering Design in Tampere,Volume 2, Schriftenreihe WDK 25, Tampere 1997, p. 387-392. [8] Description of course content and structure is maintained at http://me210.stanford.edu [9] Milne, A. Organizational and Educational Innovation: ENVISIONEERING at Penn State, Proceedings of the ASEE Annual Conference, Champaign-Urbana, IL.. 1993, p. 98-101. [10] Wilde, D. Using Studente Preferences to Guide Design Team Composition, Proceedings of ASME Design Engineering Technical Conferences, DETC97/DTM-3890, Sacramento, CA 1997. [11] Hallman, R. Techniques of Creative Teaching, Journal of Creative Behavior, Vol. 1 (3), 1967, p. 325-330. [12] Cannon, D., Mabogunje, A., Leifer, L., Moving the Internet into the Design Media Mix, Proceedings of the 11 th International Conference on Engineering Design in Tampere,Volume 2, Schriftenreihe WDK 25, Tampere 1997, p. 351-356. [13] Christiansen, T. Modeling Efficiency and Effectiveness of Coordination in Engineering Design Teams, Ph.D. Thesis, Stanford University, 1993. Andrew J. Milne Stanford University Center for Design Research Bldg. 560, Stanford, CA 94305 USA Phone: 650.723.3803 Fax: 650.725.5916 E-mail: amilne@cdr.stanford.edu