Are computers taking away the creativity in design?
Computers are an integral part of today’s design process. They, computers represent a time and cost saving device that aids designs in rendering shapes and looking at a broader parameter of possibilities than would be possible otherwise. In looking at design, one must be aware that it represents a part of a business process that has competitive considerations. The foregoing includes costs, materials, innovation, uniqueness, distinction, functionality, and utility in gathering clients and well as customers. The contribution of computers in this highly charged environment has helped to drive down costs, while developing innovation as well as approaches to creativity.
As such, designers have been able to expand their creativity through the ability of the computer to permit them to look at more possibilities in their search to arrive at solutions that fit within the preceding parameters. As software programs advance on a technological basis, newer and more powerful programs are increasingly taking on more roles in the design process. This represents the foundation of this examination in that generative design has captured a new part of the preceding in that software can generate designs based upon input parameters. Is the preceding innovative? The answer would have to be a yes. Does the foregoing aid in the business aspects in terms of broadening the range of potential possibilities, and helping to keep costs low? Again, yes is the answer.
So, computers do have their place in increasing the efficiency and range of design outputs. The question is, has their role starting to become so large that they, computers are taking over the design function, relegating the designer to a computer operator?
Chapter 1 – Introduction
In equating the question as to whether computers are taking away creativity in design, an exploration into the key words of the examination is seemingly in order. The Houghton Mifflin (2007) dictionary defines generative as “Having the ability to originate, produce, or procreate”. Design, represents the process whereby one creates, fashions, executes and or constructs according to a plan (Merriam Webster (2007). The design process, depending upon the application that is being utilised, represents trial and error in working through the steps to the final design that fits the parameters of the project (Brown, 2001, p. 2). In order to focus in on the context, design as it relates to architecture, construction and new products such as vehicles, and machines shall represent the core of the examination as represented by the question as opposed to the design of clothing, packaging, and related forms whereby the product is designed based upon primarily internal considerations as opposed to client, and or competitive, and market considerations.
As brought forth by McDonagh et al (2004, p. 13):
“As established products have become more similar in technology, functionality, price and quality, companies have turned to design to differentiate their offerings through human-centred innovation and to create stronger emotional connections with their customers. More companies have followed the example of Apple, Braun and Philips, recognising design as a strategic function in their business-not one subservient to marketing, manufacturing or engineering”
The design process has increasingly become more competitive in terms of differentiating products, and appealing to clients and customers, as “…companies seek competitive advantage through more integrated offerings, with differentiation through all points of customer contact that express their brand” (McDonagh et al, 2004, p. 13). Thus, design is a critical function most businesses, taking differing forms, thus the election to restrict this examination to the aforementioned categories.
Rittel and Weber (1973, p. 158) tell us that design problems are more than just complex, they, design problems, represent what they term as “wicked problems”. Moran and Carroll (1996, p. 4) in elaborating on the aforementioned advise that design problems “…be “stated” per se or “solved” in the sense of definitive answers, because the criteria for evaluating goals and outcomes are innumerable, subjective, and conflicting”. They add that (Moran and Carroll, 1996, p. 4):
“Any solution will generate “waves of consequences” that interact among themselves and with other problems, changing the problem situation in irreversible and unknown ways. Thus, each wicked problem is merely a symptom of further wicked problems; their solutions cannot even be finally evaluated.”
In understanding the nuances as well as ramifications of design, it is necessary to note that design is a process that it usually proceeds under conditions that are represented by a high degree of uncertainty, whereby answers to critical facets cannot be had (Moran and Carroll, 1996, p. 4). The design process also is constrained by real world considerations as represented by time, and budgets, thus the boundaries of the process, design, are not unlimited. Thus, in view of constraints, the design process needs both ingenuity as well as creativity. As brought forth by Rittel and Weber (1973, p. 158), the complexity of design problems make design projects too large for on individual to handle as multiple “…technical disciplines are required, as well as management discipline, in addition to creative and integrative skills” (Moran and Carroll, 1996, p. 5).
Evidence supporting the technical difficulties involved in the design process is presented by Brown (1998, pp. 45-46), who states the failure rate for new buildings of all types before construction starts is around 10%. For new products, such as electronics and related categories, the failure rate as a result of design is generally within Brown’s (1998, pp. 45-46) range, however there are exceptions, such as the 33% failure rate that Microsoft’s new Xbox 360 is experiencing (DailyTech, 2007). Pressures to get it right represent a critical business decision, whether there is a client involved, or if the design process represents an internal process for the company’s own products. Thus, time, cost, simplicity, and failure rate minimisation are critical business concerns that the designer must operate within. Thus, the design function, while being creativity, is also subject to the foregoing pragmatic considerations.
Generative design software represents the processes whereby new designs can be automatically produced at the push of a button (Mass Customization & Open Innovation News, 2006). The preceding represents computer software taking various design specifications and formulating them into a final design matrix based upon the input variables within the software program (Mass Customization & Open Innovation News, 2006). The basic forms, patterns and or objects is modified automatically by an algorithm, thus permitting faster trail and error processes (Mass Customization & Open Innovation News, 2006). The process of automatic design generation permits thousands of differing designs to be produced, as well as permitting new ones, as the design process is not restricted to the designer’s imagination (Mass Customization & Open Innovation News, 2006).
It, generative design is “… the power or function of generating, originating, producing, or reproducing” (Merriam Webster, 2007). The designing of architecture, and products represent processes that are three-dimensional (Beilharz, 2004). Parameters as represented by colour, texture, utility design input constraints, space, regulations, gravity, materials, heat (in the case of certain products) and costs are the boundaries to the structure as well (Beilharz, 2004). The application of generative systems to design affects the design process phases and “… integrates the macrocosmic and microcosmic relations of the design system”. For the purpose of clarification, macrocosmic represents “and large or complex system or structure made up of similar smaller systems or structures …” (allwords.com, 2007), whereas microcosmic is defined as “ a little world … a community or other unity that is an epitome of a larger unity” (Merriam Webster, 2007).
This examination shall look at whether or not computers are taking away creativity in design by focusing on generative design, and how it affects the designer. In delving into the preceding, a number of important areas shall be examined in order to understand the design process, what it entails, along with what creativity is and how it is utilised.
Chapter 2 -The Design Process
In designing a product, building, vehicle or machine the principle output as represented by the design process is its specification as represented by either an annotated CAD rendering and or a schematic (Moran and Carroll, 1996, p. 324). The design rationale represents the why the design is done in the manner that it is, with the foregoing including various types of information (Moran and Carroll, 1996, p. 324). Fry (1999, p. 22) takes a wider view of design, stating that “… it is one of the most powerful ways to understand how a world is prefigured, made and acts”. Mitchell (1990, pp. 67-71) explains the design process as one that depending upon the context, takes on differing forms. He explains that the most usual computational variations represent transformations, which he terms as unary, as well as binary operations of shapes as represented in either two dimensional drawings, and or three dimensional geographic models (Mitchell, 1990, pp. 68-69). Mitchell (1993, p. 25) states that there are areas in computer-aided design that fail to support creativity, citing shape emergence as an example. Mitchell (1993, p. 25) argues that:
“that design intentions evolve through the course of a creative design process, that these intentions determine how emergent shapes in drawings will be recognized, interpreted, and reinterpreted, and that interpretation (and reinterpretation) of emergent shapes plays a crucial role in directing design explorations. Traditional computer-aided design systems do not effectively support creative design because they provide only very limited and inflexible ways of interpreting shapes.”
In elaborating on the foregoing Mitchell (1993, p. 25) adds “…Computer-aided design systems can, however, be developed on an alternative foundation that provides the necessary flexibility”. Jun and Kim (2003) have a differing opinion of the preceding. They argue that shape semantics in CAD systems offer the potential for the emergence of shape semantics. In presenting their view, Jun and Kim (2003) offer the following rationale:
“Drawings in the early phase of design support both continuity and change in a process of design through the use of two different types of drawings (1): context drawings which hold the evolving design decisions, and exploration drawings which are abstracted from the context drawing and act as graphic probes to investigate selected issues. The notion of emergence in design is found here. The ambiguity of the exploration drawings enables a designer to read more out of a drawing than he or she puts into it; that is, to generate new meanings within the design task. Through exploring drawings that are ambiguous until concept formation (2) is reached, various interpretations, in particular in visual aspect, are possible. As a consequence new drawings emerge. This process plays a crucial role to develop conceptual form in architectural design.”
The limitations found in CAD systems are based in the fact that they are fixed, in terms of the representation that are embedded (Jun and Kim, 2003). They argue that there are three types of emergence in the architectural field, “… shape emergence, shape semantics, and style emergence” (Jun and Kim, 2003). The rationale for this brief look at emergent shape semantics is that it represents a visual design concept, thus representing a segment of creativity (Jun and Kim, 2003). An alternative view of computers and creativity is offered by Kathleen Gibson, an associate professor of design and environmental analysis, who believes that computers actually aid in increasing creativity (Winter, 2003). Professor Gibson’s unconventional approach is based upon the rationale that the designer can look at a multitude of differing examples, shapes, configuration and styles as part of their own internal creative processes and potential come up with ideas, shapes, and approaches that they may not have worked on without the aide of a computer to run through variations (Winter, 2003).
Mathias (1993) advises that the more time and input designers spend on conceptualising through drawing, visualising as well as re-evaluating designs, the more they, in general, generate new information through multiple reviews and analysis, also resulting in reasoned explanations concerning problems, and proposed solutions. Lawson (1980, p. 6) advises that design represents a mental process that is highly organised, that is capable of manipulating differing as well as many kinds of information. He adds that the preceding blends that data into a set of ideas that is coherent, and finally results on the generation of ideas related to the process (Lawson, 1980, p. 6). Design, as shown throughout this examination, is a process, involving creativity, mental reviews, the manipulation of data and ideas, along with design possibilities, and then synthesising the process toward the end solution. In looking closely at the foregoing, it can easily be deduced that the limitations of the foregoing are represented by the breathe of experience, exposure and mental foundation bank of the designer. This thus further explains why Mathias (1993) makes his statement that the more time designers spend on ‘drawing, visualising as well as re-evaluating designs, the more they, in general, generate new information through multiple reviews and analysis’, the more they see other approaches, solutions and ways in which to accomplish the task, thus heightening their creativity.
The explanation of the design process foregoing represents the use of non-creative facets, such as a systematic approach to the design issues and problems (Browne and Smith, 1993, pp. 1209-1218). Hertz (1992, pp. 396) brings forth what he terms as the mental synthesis-creation cycle, which represents the process of exploring alternatives in the development of the design that solves the end product and or building / structure. This represents the same approach, in general conception, as offered by Lawson (1980, p. 6).
Chapter 3 – Creativity
Creativity, as stated by Turner (1994, p. 21) represents “… the bringing forth of an original product of the human mind …”, which also has what he terms as its “… mundane side as well”. The creative process is present in all of us. It represents our approach to problem solving for issues that we have not experienced before, through utilising past knowledge combined in new ways to result in a solution (Turner, 1994, p. 21). In equating creativity Weisberg (1986, p. 10) tell us that a solution is creative if it has significant novelty, and is useful. In order to qualify as creative, the new approach, and or solution must be new as well as different from those that preceded it, and, those differences need to be significant (Weisberg, 1986, pp. 12-13). The utility, usefulness, is the second facet present in a creative solution, it must solve the problem in a new, and better manner than those that preceded it, that can include at less cost, fewer parts, faster response time, etc. (Weisberg, 1986, pp. 12-13).
In the design process, Cross (1986, p. 15) advises that the special ways in which designers think is embedded in their visual thinking process. Their process of creativity is based in lines, relationships, drawings and formulas representing their approach to problems (Cross 1986, p. 17-18). It must be noted that Dorst et al (1991, pp. 39-40) brought forth that the field of study in terms of design thinking was found lacking in three important areas. Dorst el al (1991, p. 42) identified the preceding as 1. a lack of research clarity, 2. lack of consistency in the tools, methods and theories utilised, and 3. the lack of a common unifying purpose. In defining design thinking, Dorst et al (1991, p. 43) advise that it represents the totality of cognitive activities during the design. The preceding is called ‘design reasoning’ that is distinguished from inituition as a result of the conscious as well as predictable use of rules representing inference for the outcomes of using and manipulating design information. The preceding brings forth the facet of intuition as an important part of the process (Dorst et al, 1991, p. 46). It, intuition, resides in the subconscious, which represents the synthesis of experiences, ideas, new approaches and concepts to result in approaches (Dorst et al , 1991, p. 46).
In subjecting the preceding to experimentation, Dorst et al (1991, p. 46-48) utilised four designers in a reasoning task that was comprised of a series of architectural drawings for buildings of modest size containing 9 errors. The designers had to locate the errors in a think aloud mode to let their processes be known, which was being recorded, and also making sketches, along with marking the drawings (Dorst et al, 1991, p. 46-48). Later analysis of the recordings resulted in the formulation of two groupings. One represented the mental process of new design generation through changes in the drawings and designs (Dorst et al, 1991, p. 46-48). The second category revealed that the verbal statements made led to new information generation that built upon prior statements they uttered aloud (Dorst et al, 1991, p. 46-48). Their prior histories and experiences represented the database to ascertain the errors, and pose solutions. In commenting upon this Dorst et al (1991, p. 46-48) noted that the resulting changes were not novel, nor creative, but did solve the errors. The purpose of the foregoing is that the designers were only given a limited time frame in which to conduct their reviews, which demonstrated visual thinking as well as visual reasoning processes (Dorst et al, 1991, p. 46-48).
The importance of the preceding is that it brought forth the manner in which designers use visual thinking to stimulate their processes. Wiggins and Schon (1992, pp. 45-51) provide illumination to this direction in advising that drawings and representation represent the manner in which designers communicate, not only with themselves, they also use it to communicate with others in furtherance of their ideas. Important in the preceding, is that designer’s employ visual thinking that consists of three types of visual imagery. The preceding is represented by what they see, what they imagine, and what they draw (McKim, 1980, p. 26). The foregoing is a highly important point in this examination, thus the rationale for the exploration these areas. Drawings and other visual representations aid in the stimulation process. It is the combination of these factors that are utilised by designers in the process of creating new designs and creative solutions.
Further elaboration on this process is important in understanding the manner in which visual representations aid the design, and creative processes. Wiggins and Schon (1992, pp. 102-122) tell us that drawings help the designer in seeing, interpreting what can be potentially moved, re-evaluated, and or transformed. Tovey (1989, pp. 26-31) argues that seeing is an important part of the design process as it aids the designer in imaging, which prompts creativity and drawing. The preceding three processes work together to encourage visual thinking. The connection between the use of computers as an aid in this process will be further explored to draw upon examples that either support or refute if generative design impacts the designer, and if it aids or diminishes creativity.
Herbert (1988, pp. 26-40), in his exploration of the design process states that once a designer begins to perceive the task, images, and or pictures representing solutions start to be generated in the designer’s mind. He adds that in the beginning of the process, these images, pictures, ideas and thoughts are not really well defined, thus needing further development (Herbert, 1988, pp. 26-40). The next stage on the process moves towards more refinement as potential solutions, a means to these ends and the routes and ideas as to how to arrive there start to develop (Herbert, 1988, pp. 26-40). The foregoing was brought forth by Mathias (1993) who advised that designer tends to utilise drawing as a means to move to developing their first ideas, and in later stages of the process, drawings are utilised for synthesis. This is what Mathias (1993) describes as stepping back and then forward in the mental processes to be engaged, and then reflect on what has been thought, to re-engage the process again. Mathias (1993, pp. 113) diagrams the preceding as follows:
Diagram 1 – Designer’s Framework for Idea Development
(Mathias, 1993, p. 113)
Analysis of problem statement
Mathias (1993) and McKim (1980) follow the same conceptual foundation in terms of imagery (drawings), representing a foundational facet in the design process. The preceding is part of the creative mental synthesis process whereby ideas begin as a result of engagement with the project. Verstijnen (1997), conducted experiments based upon the research methods of Finke (1990) as well as Helstrup and Anderson (1993). The six experiments conducted by Verstijnen (1997) used undergraduate industrial design engineering and psychology students to investigate emergent figures and drawing / sketching relationships. The experiment called for some students to utilise sketching and drawing, and the others to use mental processes only in resolving the problem of wire frame drawings that had figures embedded (Verstijnen, 1997).
The second experiment represented the investigation of creative mental synthesis tasking asked to generate shapes that were creative based upon a cube, sphere and cone. The results indicated that the restrictions of memory was not a motivating factor in sketching as an aid in creative mental synthesis experiments (Verstijnen, 1997). In the reinterpretation of shapes, the task proved difficult based upon the utilisation of imagery alone, thus suggesting an aid was needed, drawing (Verstijnen, 1997). She added that sketching for use in synthesis may not play a significant, and or important role, however in creative mental synthesis, sketching helped to yield a higher number of creative forms (Verstijnen, 1997). A comparison of the engineering students, and non-students was not included in her study (Verstijnen, 1997).
From the foregoing, the data thus far reviewed seemingly indicates that visual thinking represents an important facet of creative mental synthesis, which is important in the design of new objects. The foregoing also seems to point to the fact that the manipulation of visual images in a mental mode is an important aspect of mental creative synthesis. In addition, the preceding seems to also point to the fact that drawing is important in supporting the process of creative synthesis in new design. The use of the word seemingly and seem are utilised as there is little to no empirical evidence to support the preceding, thus the views are based upon the views as expressed by a number of authors and researchers.
Chapter 4 – Generative Design
In generative design, the description and design of relationships and components is accomplished by the use of powerful algorithms (MacDonald et al, 2005). These algorithms permit users to manipulate geometry and dynamically model through the application of rules that capture relationships in geometric features, along the defining of complex forms (MacDonald et al, 2005). Under a generative design program, the designer, utilising a CAD tool that specifies parameters and the restraints, the program then generates a number of outcomes that the designer then utilises as input, and or for another generation of shapes, forms or approaches, solutions (MacDonald et al, 2005). The generative design approach is applicable in the whole design process, once the parameters have been input and modified (Gatarski and Pontecorvo, 1999).
Generative design has been employed in the development of cars, cell phones, structures and other areas. CAD and Design Automation software has long been utilised to optimise the assembly of differing design elements. Through the use of visualisation software, designers are able to see the results without having to build prototypes. In the generative design methodology, outcomes are rendered after a detailed and high level of input specification, thus saving time over the CAD and Design Automation approach that requires time in that each instance of an idea has to be imagined, manually expressed, and evaluated (Gatarski and Pontecorvo, 1999). On the most basic level, generative design systems consist of four elements (Gatarski and Pontecorvo, 1999):
- design representation,
- generation engine,
- expression engine,
- mechanism for evaluation and selection of the new generated specifications
The following, further explains the workings within these steps (Gatarski and Pontecorvo, 1999):
- design representation,
The design representation are input as a set of parameters, along with the corresponding constraints. In the preceding, the parameter sets represent the genetic design elements, defining the form as well as structural aspects. The constraint set controls the aesthetic as well as the fabrication facets, meaning the limitations as afforded by the material dynamics to be utilised. The foregoing constraints, and rules thus place a limit on the range of the generated design.
- generation engine
This aspect of the generative design process represents the internal process that generates the new design descriptions. It represents a set of prototype design descriptions that are also termed as ‘parents’, and then utilises the algorithm to take the parameter sets and combine them into new descriptions, or children. In general, the algorithms use operations that are based upon the concepts found in the mutation and crossover aspects found in genetics. The preceding thus ensures that the descriptions, children, are drawn from the parameter values as contained in the prototype set.
- expression engine
In the process of generative design, the expression engine interprets the descriptions, rendering them into a structure or model. The process can be set so that it is modular, thus permitting it to be able to support a host of alternative interpretations.
- mechanism for evaluation and selection of the new generated specifications
The generative design approach is based upon having an objective function that evaluates the fitness output of the specifications of the design. In creative design area this facet is usually embedded in the human using the system. The foregoing is termed the human user in the loop as it provides a more intelligent and analytic capability as opposed to computed functions. The preceding is described as being a better match for the ranges of possibilities offered by the generative approach.
Generative design offers the ability for designers to try as many ideas as possible, a critical facet when time is a constraining factor. It permits designers to thus present a variety of solutions that can thus be evaluated internally and or through focus group testing to refine the process and move to the final design variables. The foregoing recognises the fact that the design process has always been a process that is subject to the acceptance of the end user, which is either the client, or the general public. It, generative design, thus represents a means to work through a wide variety of approaches to achieve faster approximations of what will work.
Chapter 5 – The Role of Computers
If computers are potentially limiting, and or taking away design creativity, it would seem appropriate to example the role of computers in the design process and attending questions. In the pragmatic sense, unlimited time to render and use a design is not a function of the design process. Whatever the function, be it architecture or products, there is a limited time frame for the design process to take shape, evolve, become proven through mathematical, structural, component material and cost / build considerations. The preceding represent constraints on the process that are real as well as binding. Computers have advanced the process of design through software such as CAD that enable users to perform a number of functions in time saving fashion. The preceding takes in such functions as (SAP, 2006):
- wire frame geometry creation,
- solid modelling, 3D parametric,
- freeform surfaces,
- automated assembly design,
- engineering drawings from solid models,
- re-utilisation of design components
- ease on design modification and the generation of multiple versions,
- automated generation of design components,
- design simulation without prototype building,
- data exchange,
- as an aid in visualisation process for areas such as rotation, shading, etc.,
- design studies
as a few of the more important, or used functions. In equating this segment of the examination, the question of creativity resurfaces. As previously mentioned by Turner (1994, p. 21), creativity brings ·… forth of an original product of the human mind …”. In the design process, it aids in the production of multiple view, as well as ideas. Gero (1991) indicates that there are, in his view, five creative design processes that result in the introduction of new variables in the design prototype. These are (Gero, 1991):
- First Principles, and
The following sets forth the preceding in more detail
Gero (1991) tells us that ‘combination represents the combining of two or more prototype designs, with the new resulting variable introduced into the original from the former. Within this process, mutation represents the alteration of variables as accomplished by external agents (Gero, 1991).
Within the preceding process, mutation represents the alteration of variables as accomplished by external agents (Gero, 1991). Through mutation, new variables can result due to extrapolation, and or combination of the variables (Gero, 1991).
This is also known as ‘case-based reasoning’, representing structural elements applied by one prototype design into another design problem, as well as the use of past processes for a new design problem (Gero, 1991).
- First Principles
The above is a process whereby new relational knowledge is logically derived from the behaviour of existing structures without the knowledge of the prototype (Gero, 1991).
The process of inferring structural elements that are new through the extension of elements that exist, represents emergence (Gero, 1991).
Indurkhya (2002), in defining creativity from a different perspective, provides insight into Gero’s (1991) concept of new variable introduction. H