A Model of Technological Imagination and Creativity: Cognitive Task Analysis

Abstract

An integrated model of cognitive tasks involved in the process of a technological innovation was proposed based on these theories: 1. CDIO theory of technological innovation, 2. Wallas’s creative thinking processes, 3. Khalr & Simon’s theory of scientific discovery, and 4. the conceptual combination theory of imagination. The central theme of this model is the proposition that three cognitive conditions are necessary for technological imagination and innovation: 1. cross-domain knowledge, 2. simple heuristics, and 3. pattern recognition ability. Although the required domain knowledge and implementation methods are different across domains, heuristics that lead to a breakthrough at each phase of CDIO in a technological innovation are similar, with conceptual combination as the cognitive engine for generating original and imaginative ideas.

1. Introduction

From the problem conception in terms of a functional demand to finding a solution, the designing and making of the product, its manufacturing and marketing, to taking social responsibility, technological innovation is a complex problem-solving process, requiring the cooperation of a chain of different industries and experts. For example, the semiconductor industry was formed around 1960 and consists of an aggregate of companies engaged in the design and fabrication of semiconductors and semiconductor devices, such as transistors and integrated circuits (IC). It is in turn the driving force behind the wider electronics industry for such products as power electronics, consumer electronics, and e-commerce. The semiconductor market segments include networks and communications devices. A need for technological innovation in the IC manufacturing industry may be driven by demands or orders from the users, the IC design, the IC fabrication and manufacturing, IC packaging and testing, logistical supports to deliver products to customers, and user services. The size of silicon wafers determines how many IC can be produced per wafer in IC manufacturing. The semiconductor industry has been continuously striving to increase the wafer size. Although foundries used to produce 3-inch wafers, today’s common wafer size is 12 inches. Increasing wafer size is not a trivial process. In fact, silicon wafer manufacturing technologies have to be re-designed and re-engineered in order to increase the wafer size. Innovation is a common daily demand of all companies involved [1].

The CDIO theory proposed by Crawley and colleagues [2, 3] for improving engineering education pinpoints 4 critical phases in technological innovation, whether it is a new product, procedure, or system [4]:

1.1 Conceptualization

The need for a certain functional requirement of a product/system that is meaningful and valuable to the society is sensed, such as how to reduce CO₂ emissions from the production process. This raises a design problem, the next phase of innovation.

1.2 Design

The initial functional requirement is abstract in nature. It must be developed into the tangible possible designs of a physical device or system. Based on the design of the current product, the industrial designer may sketch several different designs of how the product/system will look like when new functional requirements are added. The most promising design is chosen, given current technological constraints, and a prototype is constructed.

1.3 Implementation

A prototype is merely a conceptual design presented in 3D. How the prototype of a conceived new product or system can be physically realized must be studied and implemented by a team of RD workers and engineers. That means various components of the product will be made, assembled, and contextualized into a real product that can work in a specified context.

1.4 Operation

When the first new product is created, it must be reliably reproducible with an equal quality and in a certain quantity. The manufacturing facilities must be set up so that parts can be manufactured, assembled, tested, packaged, and delivered to customers in a timely manner with a reasonable price and with supports for maintenance, repair, and even recycling.

Technological innovation is a form of complex problem-solving with humans as the primary problem solvers. Its chances of success heavily rely on deep knowledge about the product, the industry of the domain [5], and imagination and creativity. Creative ideas always spring from an individual’s mind. Although CDIO may represent activities from different industries or individuals, the underlying cognitive tasks necessary for a person to come up with an innovative idea are similar. Understanding how the human mind stores, retrieves, transfers, combines, and transforms ideas in the human cognitive system is important to facilitate the generation of original and satisfactory solutions at each CDIO phase.

The purpose of this study is to propose a model of creative processes in each CDIO phase of technological innovation that incorporates Wallas’s four stages of creative processes [6]. Conceptual combination theory of imagination was used to account for what happened during the incubation stage that later brought about the breakthrough ideas at the illumination stage [7]. This model of imagination and creativity in technological innovation will later be shown to be analogous to Klahr and Simon’s [8] theory on the cognitive process in scientific discovery in which domain knowledge, simple heuristics, and pattern recognition are three cognitive tasks indispensable for making scientific breakthroughs.

Specifically, we propose that in order to achieve something creative, Wallas’s four stages of creative problem-solving processes (i.e., preparation, incubation, illumination, and verification) will occur during each CDIO phase. The conceptual combination theory of imagination is a simple heuristic to generate imaginative and innovative ideas during the incubation stage, which then dawn on the individuals at the illumination stage. As shown in Table 1, this model of technological imagination and creativity can provide a handy framework for analyzing the cognitive tasks required in every stage at a given CDIO phase and also for designing educational programs and procedures to facilitate engineering imagination and creativity.

Phase1	C	D	I	O
Stage2/goal	Problem finding	Solution finding	Making the first product: contextualization and specification	Building a production system
Prp.	Memory search and knowledge acquisition about problems related to: The current products and user experiences.	Memory search and knowledge acquisition about: solutions of the problem, and user experiences.	Memory search and knowledge acquisition about: methods, tools and materials for making the product, contexts of use, and user characteristics.	Memory search and knowledge acquisition about: Production methods, materials, machinery, human resources, users, markets, law, management, and etc.
Inc.	Automatic spreading activation of memory and conceptual combination.	Automatic spreading activation of memory and conceptual combination.	Automatic spreading activation of memory and conceptual combination.	Automatic spreading activation of memory and conceptual combination.
Ill.	Pattern recognition	Pattern recognition	Pattern recognition	Pattern recognition
Vrf.	Testing by logical analysis, experiment and data analysis.	Testing by building a prototype and discerning user satisfaction.	Testing by pilot runs of the product and learning levels of user satisfaction.	Testing by users of the production system and market.

Table 1.

A model of creative processes in technological innovation.

¹

C, conceptualization, D, design, I, implementation, O, operation.

²

Prp., preparation, In., incubation, Il., illumination, Vrf., verification.

2. Creative problem-solving processes in technological innovation

Creativity is a criterion for evaluating all technological innovation. Imagination is often treated as a synonym for creativity. In the present study, the distinction between imagination and creativity will be stressed so that conditions that enhance engineering imagination and creativity can be better understood. Creativity is defined as the process by which a socially valuable new product is produced [9], whereas imagination is defined as the process by which a new idea emerges in one’s mind. This distinction is important because the creation of an appropriate but novel product valued by society is often a very long and complex process (e.g., the designing and manufacturing of an electrical vehicle).

Factors affecting technological creativity include not only an individual’s personal characteristics such as ability, personality, and motivation but also the social, intellectual, and technological resources necessary to support such endeavors [10]. Thus, it is argued that highly creative achievement is determined by a multiplicative rule composed of many personal, social, and environmental factors. The absence of one required factor would render the creative effort futile [11]. Imagination, on the other hand, occurs in one’s head and is relatively easier to manage and foster through education and training.

Klahr and Simon [8] reviewed studies on historical accounts of scientific discoveries, psychological experiments with nonscientists working on tasks related to scientific discoveries, direct observation of ongoing scientific laboratories, and computational modeling of scientific discovery processes. They proposed that cognitive processes in scientific discovery consist of three components: search through domain knowledge, use of some simple heuristics to consider and generate ideas, and pattern recognition when a solution emerges in the mind. In the present study, it is proposed that although very different kinds of knowledge are required for creative problem-solving activities during each CDIO phase, the creative-thinking processes are nevertheless similar. The conceptual combinations are the simple heuristics or cognitive mechanisms that humans use to imagine and go beyond what is given in our memory.

In brief, every episode of creative accomplishment will go through Wallas’s four stages of the creative problem-solving process [6]: first is the preparation stage, during which an individual senses the problem and searches for available solutions but fails; second, in the incubation stage, because of the failure in previous problem-solving attempts, the individual puts aside the problem and shifts attention to some unrelated activities and appears to be disengaged from the problem; third, in the illumination stage, after a short or long period of incubation during which the person does nothing deliberately, the solution to the problem may suddenly pop into the individual’s mind; and the last stage is verification, during which the individual performs required tests to see if the idea will indeed work.

These four stages require different cognitive engagement and social participation. For example, extensive learning and searching for information and examples of past and currently available products or practices are the major cognitive activities in the preparation stage. Implementation of the ideas and testing how they accomplish the goals usually require collaborative work and investment of material and financial support from society. Of these four stages, the incubation stage and the illumination stage occur only in an individual’s head.

How the human mind worked used to be regarded as a black box. However, with recent advances in neuroscience and neuroimaging technology, the mind is no longer as black as it used to be. The present study proposes that the way the brain works during incubation to bring about illumination is through the act of imagination. Imagination is a kind of mental activity through which existing knowledge in one’s associative brain networks can be activated, combined, and interpreted so that new ideas may be generated. Conceptual combination is thus proposed to be the cognitive mechanism, a simple heuristic, by which the human brain creates new ideas during the incubation stage [12].

3. Imagination: conceptual combination and interpretation

Conceptual combination refers to joining two or more concepts to generate a new idea [13]. It is believed to be the most wonderful merit of the human mind [14]. At the neurological level, conceptual combination is made possible because any neuron (concept) in human brains can be linked to many other neurons unconsciously or consciously. The meaning of an object or event is thus determined by its activation pattern, a created network, of neurons [15, 16].

In engineering, many innovative ideas were indeed conceived and created through conceptual combination. For example, the electrical bicycle is made possible by joining an electricity-powered mechanism and the manual bicycle. Personal computers originated from the idea that the traditional mainframe computer could be carried with you wherever you go, like a briefcase. The Blackberry phone was first conceived by joining of two concepts: cell phone with email [17]; mirrors were added to elevators as a means to distract people’s attention from the slow speed of elevators [18]. In other areas such as language, new words and phrases are constantly created via a combination of already existing words [19, 20]. In reading comprehension, the meaning of a sentence is achieved by activation and the integration of different word meanings in a sentence so that a gist in the form of a proposition for the sentence, a macroproposition for the paragraph, and a situation model for the entire text could be derived [21, 22].

Chance or luck has always been considered an important element in creative achievement [23]. For example, Campbell’s theory of blind variation, selection, and retention [24, 25] argues that random variation is the basic mechanism that organisms use to cope with an uncertain environment. For example, mutation is the major mechanism by which a virus adapts to harmful environmental factors. Humans are blessed with a brain that can perceive the environment, store learning and experiences, and retrieve the information from memory to help understand and solve new problems that arise in the present and future. By manipulating information in the brain, simulating possible environmental changes, or just letting the brain wander randomly, humans, unlike lower organisms, can think and plan their coping responses in advance. The way human brains work to create new knowledge is by a random combination of ideas in the brain. Ideas may automatically collide with each other by chance, and some patterns may emerge unexpectedly to shed light on the puzzled mind [24]. However, mere activation and association of previously stored information in the memory is not enough to generate original ideas from conceptual combination. Some novel ideas that are remote associates from other knowledge categories must be found to help one get out of a rut [26], and interpretations must be made to meaningfully link the concepts [27], either automatically or deliberately.

After reviewing extensive empirical evidence from psychological literature, Simonton [28] used the theory of constrained stochastic process to account for scientific creativity. He pointed out that the road to scientific discovery was filled with uncertainty and luck. Logics and systematic thinking do not provide much help. They are primarily useful after the discovery to provide proofs and explanations for the validity of the discovery. Based on his analysis, Simonton pointed out two factors that contributed to scientific discovery. The first is the domain knowledge of which each scientist can only sample a small portion as the target of the study. The second factor is the scientists who evaluate the existing knowledge in the field of their study and who then try to find new and useful knowledge via the combination of knowledge shared in their scientific community.

In the theory of constrained stochastic process, the random combination of ideas that leads to scientific discovery proceeds in two steps. One is ideation, which refers to the generation of possible ideas; the other is elaboration, which refers to further explanation and refinement of the ideas. Because scientists usually work on more than one project at a time, the cross-talk from these diverse projects may provide chances for novel reformulations and discoveries. Blind variations and combinations of existing knowledge are thus the brain’s mechanisms to generate original ideas for different types of creativity [25].

Technological innovation is founded on scientific knowledge. The difference between technological innovation and scientific innovation is that the former is less constrained [28] because there usually exists numerous possible solutions to a technological problem. Empirical evidence shows that random variation and conceptual combination are crucial for the emergence of original technological solutions. For example, Zeng et al. [18] found that the remote association of concepts from different domains spawned more creative mash-up web service design ideas in information technology than were generated from intradomain mash-up web service design. It is thus proposed that conceptual combination is the simple heuristics to promote imagination and creativity in all four CDIO phases in technological innovation.

4. Interpretation: making sense of novel experiences

Emergence of original ideas from conceptual combination is not merely a process of memory retrieval from free association [29], neither is it the remote association [26] of concepts in one’s knowledge network. When two seemingly unrelated ideas are retrieved from memory, an interpretation process is automatically activated in the human cognitive system, aiming to make sense and find meaning for the co-occurrence of these objects or events [19]. Finding a way to meaningfully connect two terms is the simplest form of relational thinking, a kind of abstract thinking [30]. It is this interpretation process that generates and adds new nodes to the memory network and makes the knowledge network grow by itself. On the one hand, because there are many possible ways to interpret the co-occurrence of concepts, objects, or events, the interpretative processes involve divergent thinking, thus making evaluation of the fluency, flexibility, originality, and elaboration of emerged ideas possible. On the other hand, the act of interpretation also requires convergent thinking because a new idea must be generated and chosen to account for the reason why these concepts co-occurred [7]. Three criteria were raised to evaluate whether the emerged new concepts from conceptual combination indeed made sense: diagnostic feasibility, plausibility, and informativeness. That is, the new concepts must be derived from distinctive feature or meaning of the given concepts, be a plausible explanation of the co-occurrence of these given concepts, and add something new when combined [31]. For example, “that lawyer is a shark” would automatically generate an image of a (greedy and reckless) lawyer with sharp teeth and biting, rather than a lawyer with a fish tail.

People use three kinds of interpretation to associate two unrelated words: 1. conjunctive interpretation, 2. property transfer interpretation, and 3. relational interpretation [27]. In conjunctive interpretation, new ideas emerge from finding property overlap of concepts to be combined. For example, “vitamin C” is the concept that emerges from the combination of the concepts “banana–apple.” In property transfer interpretation, new concepts emerge by giving the unique property of one concept to another concept. For example, the combination of the two concepts “lily–light” may produce the response “lily-shaped desk lamp.” In the relational interpretation, a mediating concept is introduced to link two concepts such that the initially unrelated concepts become related. For example, “lavender–beer” may produce a response such as “beer party in a lavender garden.” Two subtypes of conjunctive interpretation were also observed [32]. Other than finding a common property for the two concepts in conceptual combination, the mapping/conjunction between two concepts can also occur at a more global level or at a structural level. This type of mapping is usually accomplished by analogy. For example, the concept of an airplane is structurally analogous to a bird generated from “fly like a bird.” Another unique type of conjunctive interpretation is “negation;” the new concept is obtained by a negative interpretation of the common property found between two concepts, for example, “darkness” as a response counter to the property “brightness” derived from light-rationality pair. Originality of new concepts emerging from different types of interpretation may be quite different in terms of whether the original concepts are modified or whether the new concept entails using the original concepts to create a broader meaning.

5. Conditions that foster originality in conceptual combination

However, not all ideas that emerge from a combination of novel concepts are original [25]. Several factors might affect the outcome of a conceptual combination, including novelty or semantic distance of concepts [33, 34], the abstractness of the concepts to be combined [35, 36], types of interpretation required [37, 38, 39], age [40], number of iterations [41], or even the ontological category of the concepts. For example, Bock and Clinton [39] found that noun-noun pairs of the natural kind elicited significantly more property-related interpretations (e.g., moon-orange → round) than noun-noun pairs of the artifact kind. Likewise, noun-noun pairs of the artifact kind elicited significantly more relational interpretations (e.g., knife-bike → cutting the edge) than noun-noun pairs of the natural kind. Last but not least, recognition and decisions regarding which ideas to select and retain from a pool of interpretations are also important [8, 24].

Imaginativeness of an emerged idea is thus a function of availability and distance among the ideas to be combined in one’s brain, the type of interpretation chosen in conceptual combination, and pattern recognition to identify the desired final idea. This model of technological imagination and innovation can therefore be described by integrating Klahr and Simon’s [8] model of scientific discovery and Simonton’s [25, 28, 42] blind variation, selection, and retention (BVSR) theory of creativity. Namely, cognitive processes involved in technological innovation include three components: 1. availability of cross-domain knowledge, where concepts from two or more than two domains need to have a chance to meet in one’s head; 2. a simple heuristic, namely conceptual combination, is applied to make one or more than one interpretation to combine seemly unrelated concepts; and 3. some utility functions, or selection and retention criteria are applied so the individual can recognize and choose certain types of interpretation yielded by conceptual combination (Figure 1). Imaginativeness of an idea that emerges from conceptual combination is a product of these three factors within an individual. According to this model, cognitive tasks for an individual engaged in the CDIO processes of a technological innovation are analyzed as follows.

Figure 1.

Cognitive process in technological innovation.

6. Cognitive task analysis of the creative processes in CDIO

Completion of CDIO tasks in a technological innovation involves the collaboration of many experts from different industries. Experts with different domain knowledge and experiences must cooperate to solve the problem at each CDIO phase and across several CDIO phases. However, for execution and completion of goals at CDIO, even though the required domain knowledge and experiences for innovation are different, the creative processes are similar. That is, at each CDIO phase, the problem is solved by going through Wallas’s 4 stages of creative problem-solving processes [6]. Cognitive tasks at each problem-solving stage are summarized in Table 1. In the following task analysis, the conceptual combination processes are illustrated primarily with examples from the creation of the Book of Chanel N^o 5 [43, 44].

The Book of Chanel N^o 5 was commissioned by the owner of the perfume company Gabrielle Chanel to Irma Boom, an Amsterdam-based designer, who has made more than 250 volumes of books. About 20 percent of them are in a permanent collection at MoMA. The 300-page book has no ink. Each of the crisp white pages is embossed with a drawing or quotation that helps the story of Gabrielle Chanel unfold. The book structure is housed in a black box. The book won the Dutch Design Awards in 2013 and is part of the permanent collection at MoMA [45]. The creation of the Book of Chanel N^o 5 is chosen as an example because it provides an interesting, creative example of a small but complex technological innovation that is comprehensible for most readers.

6.1 Cognitive tasks for conceptualization phase

The goal at the conceptualization phase of a technological innovation is to find a problem that is worth of future efforts. Creative ideas for a product first emerge in one’s mind as a problem or a functional requirement. For example, how can I fly like a bird? Raising a question like this may drive further efforts to solve the problem. The question must be meaningful and have potential utility to the individual. For an innovative technological problem, its meaning and utility must be recognized not only by someone, usually the users, but also by society and technological professionals, to attract resource investment. Following Wallas’s theory, the cognitive tasks for emergence of creative ideas from an individual’s mind at the conceptualization phase are explained as follows:

1. Preparation stage. For any creative and innovative task, activation of prerequisite knowledge about the product and the search for new knowledge are the primary cognitive tasks in the preparation stage. Technological innovations always build on previous technology. Features and functions are added to or removed from the previous products. For any innovative problem, there is no ready answer in the current knowledge base. One must go beyond the boundary and search in other related or unrelated domains. Cross-domain knowledge must be pooled via collaboration of the RD team. The individual’s memory search can be performed either deliberately or subconsciously by the automatic spreading activation nature of the human brain [16]. A technological innovation often begins from the conceptualization of a need from users. For example, how to manufacture a wafer so that more IC could be produced per wafer? To sense the need of the problem and to solve the problem, deep knowledge and experience in wafer manufacturing is necessary. In the case of the Book of Chanel N^o 5, the need for the creation of the Book of Chanel N^o 5 was from the founder of the fashion company, G. Chanel, who has the knowledge of and experience with the perfume, perfume making, and the birth of Chanel N^o. 5 but who is without book-making expertise. She turned to Irma Boom, who is an expert in bookmaking but is without expertise in the history and making of Chanel N^o. 5. When Boom began working on the book, she totally immersed herself in the subject. The fashion brand provided her with as much information as possible and let it percolate until the breakthrough idea struck her. Boom spent time in Chanel’s Paris apartment and studied Gabrielle Chanel’s life. She made field trips to the factory and gardens and witnessed the bottling process and even joined the Chanel team as they picked roses in Grasse, a village in the Provence region of France. What she was searching was: “what is unique about Chanel N^o 5?” The answer would be the key to the design idea: “what is unique about the Book of Chanel N^o 5?”

2. Incubation stage. To come up with a new and original idea, one will certainly encounter a period of bottleneck, that is, a time when there is no idea for the solution of the problem. This period can either be relatively short or be very long, depending on the difficulty level of the problem or the knowledge state of one’s memory. When one experiences failure, it is usually followed by a short or long period of breaking away. Attention is shifted to other activities. However, the brain does not stop working during this period. Remote ideas from different domains may be automatically activated unconsciously and collide with each other to form unexpected patterns or configurations of ideas [24].

3. Illumination stage. Once the activated memory network associated with the book’s concepts (book—Chanel—perfume—fragrance—smell—invisible) was formed, as illustrated in Figure 2, a new idea (an invisible book) for the solution popped up suddenly after a period of incubation. Boom’s conception of the unique feature of Chanel No. 5 was inspired by the nature of fragrance when Boom was in the rose garden one day and found what she smelled there was so intense, and exciting, but not visible. The idea for the book suddenly struck her: to create a book of Chanel N^o. 5 is like to creating a book of fragrance: a book without ink, but readable, an analogical mapping between the book and the perfume.

4. Verification stage. Although the conception is abstract in nature, the idea must be subject to logical analysis and reality testing. Is it really possible to make a book without ink but for it to still be comprehensible to readers? Previous examples of books without ink such as Braille must be retrieved from memory and other external sources to verify the plausibility of the current idea.

Figure 2.

Illustrated possible memory activation given book of Chanel N^o 5 as cue.

6.2 Cognitive tasks for design phase

The abstract idea needs to be transformed into tangible, visual representations so that rough sketches of the objects in their finished form can be visualized by people, its building or construction feasibility can be examined, and alternative designs can be compared so that the better one can be chosen. The cognitive tasks involve moving from the pure abstract idea to sketches of the possible appearances of the object and then to 3-D prototypes, to envision how the new functional requirements can be materially realized. For example, how can a current 10-inch wafer manufacturing line be extended into a 12-inch wafer manufacturing line? General design knowledge and engineering knowledge about how to realize an abstract functional requirement into a concrete physical product must work together. In the case of designing a book of Chanel N^o 5, the design problem is how to make a book that is invisible but readable. Boom’s solution to creating a book that was without ink but readable was by embossing. The cognitive tasks to generate the answer “embossing” include:

1. Preparation stage. Knowledge of and experience with the current technology in product making is not sufficient to solve the problem; new knowledge and experience must be imported from other domains. The design is guided by the new functional requirements of the end product and existing engineering knowledge of the to-be-made product. In the example of designing the Book of Chanel N^o 5, knowledge and experience in bookmaking with printing or without printing must be retrieved and compared. With a goal to create something original, previous cases of books without print must be analyzed, retained, or ruled out in the process of finding something unique. Irma Boom is an expert in bookmaking. She has created over 300 books and holds an excellent reputation for her artistic autonomy within her field. Creating a sensory tactile experience when designing and making books is very important to Boom, and she aims to inspire discovery and interaction. She analyzes every little detail in bookmaking to maximize a book’s engagement potential in contrast to its digital counterparts [45].

2. Incubation stage. After extensive review of previous examples of books without ink, some original ideas must be generated. Many ideas and experiences regarding bookmaking will be searched and then abandoned. A period of mental block and sense of failure are inevitable. A period of breakaway from the problem is called for. However, the combination of concepts: book—no print—readable may subconsciously keep activating experiences and knowledge about books or reading under special circumstances such as for the blind, the hearing impaired, and so on. Consequently, Boom’s experiences of making book covers by embossing were activated in the memory network in the incubation stage. Figure 2 is a simplified illustration of possible concepts activated during the process of the information searching in the brain.

3. Illumination stage. The idea of using embossing to make the book pages emerged! It could be used to make the content of the book semi-invisible but still readable.

4. Verification stage. Designing a technological product is half-conceptual half-physical in nature. A prototype must be made to communicate the idea in a physical form. Boom has long used embossing on the cover of her books, but using it as the only source of printing was a new concept with unique challenges. A prototype must be produced to show how the book would appear to G. Chanel. It is a miniature of the product to be created. It shows the structure and configuration of the components of the product in its finished form. The plausibility of making the real object can then be evaluated and discussed among experts, engineers, and users. Revisions and redesign are relatively easy at this stage.

6.3 Cognitive tasks in implementation stage

A prototype of the product is still conceptual in nature. Can the prototype design work in reality? A real, physical object must be built by RD engineers. Technological constraints are the major challenge at this phase. The nature of engineering work is constraint satisfaction or contextualization [46]. For example, to create an electrical bike, adding a power system with a power control system and user interface mechanism to a bike body is the major challenge. Moreover, the final product must be contextualized to the persons and conditions under which the bike is to function. Every detail must be specified before the product can be made. Engineering knowledge, industrial knowledge, and knowledge of the potential markets and users must be searched and available for consultation. The goal of the implementation stage is to make a real object work as designed and that can operate properly in a prescribed context. The making and manufacturing of the first technological product is not only a very tedious and time-consuming process, but it is also very expensive. Other than materials and machinery for making the object, a variety of engineering knowledge, equipment, and procedures must be recruited. In the case of building the first electrical bike, technological problems involve making of the bike body, battery, power control device, man-machine interface device, and other auxiliary parts. Most important of all is putting these elements together and making sure that the bike can ride for a reasonable distance with reasonable labor and speed. RD workers and experts in the related domains must be recruited and work together to solve the new problems involved in making the first instance of the new electrical bike.

1. Preparation stage. In engineering, the product is first broken down into components and parts, each of which is manufactured separately. Different lines of engineers work at about the same time, and their outputs are then integrated to form the newly innovated product. New designs invariably create new technical problems for making the product. For example, embossing is commonly done with materials that are hard enough to endure the press. How can embossing be done with a sheet of thin and fragile paper? How to bind the book without damaging the content? New materials and/or new embossing techniques must be found to solve the problems. The size, weight, and content of the book may also affect how many pages of embossed content can be bound. New solutions must be found. In Boom’s case, making an embossed book involved combining both printing and embossing techniques. Using embossing as the only source of printing was a unique challenge for the printing industry.

   Typically, books are bound and cut, but the pressing process would render the embossing flat, so it was necessary to figure out a different way to ensure the subtlety of Boom’s designs would keep their form. The publisher ended up using an old letterpress machine, with the ink removed. Each page was first designed on an aluminum plate and turned into a mold that the pages would then press against. Numerous technical details like this one needed to be surmounted in making the first book. The solutions needed to be found before a first copy of the Book of Chanel N^o 5 could be produced in its final completed form. These cognitive tasks cannot be performed by any individual alone. It requires the combined efforts of many experts from different domains. Some problems may need to be fed back to the conceptualization and design phases for reworking.

   Knowledge about the making of a product is thus distributed among different experts. Therefore, collaborations among individuals are essential. For example, more than one thousand steps are involved in wafer fabrication. It is not possible for one person to have the complete knowledge needed to create a new wafer manufacturing line. In the case of bookmaking, other than how to print the content of the book onto pages, the content of the book must be written, and the size, weight, thickness, and anticipated demands of the targeted readers and so on must be specified before the book can materially exist.

2. Incubation stage. A period of breaking away from the problem and the opportunity to exchange ideas among different experts are critical for breakthroughs in solution finding. Problems and a failure to find the proper solution may occur for each subproblem. The individuals who work on the problem may have to wait for some period before another’s problems are solved. Communication among RD teams and their scientific and technological communities becomes vital. Various solutions are searched and activated and shared in the brains for random variations. Conceptual combination can thus occur within one’s brain or between multiple brains via communication.

3. Illumination stage. Solutions for a given problem always emerge first from a team member’s brain. The ideas must be voiced and heard and have a chance to be discussed. In the case of Chanel N^o 5, the final book is 5 cm thick, a nod to the perfume’s name, and each design was hand drawn. The 300-page book was printed devoid of ink. It was instead embossed with text and images, creating a semi-invisible narrative for Gabrielle Chanel. Although there was a temptation to infuse the pages with the smell itself, Boom rejected the idea because it was too literal, too obvious. For the readers, the concentration is on the images, text, and tactile sensation when one leafs through the book, creating an experience through which one can almost smell the perfume. To Boom, that is much more interesting and thought-provoking than the real smell of the perfume. The final product is a book filled with solid white, textural pages. The content of the book was taken from the world of Mlle, a book on Coco Chanel: “Mademoiselle: Coco Chanel and the Pulse of History” [44, 47].

4. Verification stage. The reliability and quality of each technological innovation must be tested via experimentation, running simulation, or field testing to make sure each part functions as designed and meets the end user’s demands. In the Book of Chanel N^o 5’s case, the final book had to be semi-visible but completely readable by the readers and, at least, be satisfactory to G. Chanel.

6.4 Cognitive tasks in operation phase

The ultimate goal of a technological innovation is to compile and assemble a production procedure so that the product can be reproduced with reliability and flexibility and in a certain quantity upon order at any time. The possibility of manufacturing automation is also a concern. It is not just the act of creating and making a new technological product; it is the act of creating a new industry. For example, more than one thousand steps are involved in the wafer manufacturing procedure. A huge quantity of material and human resources must be invested to create a production line. Problems such as inventory control, testing, packaging, delivery, recycling, and customer service are also essential components of production line management. Cognitive tasks in the operation phase include working out every detail about the routine manufacturing, packaging, testing, and delivery of the finished product to end users while making sure the product can function properly for them. In a manner similar to the implementation stage, many sectors and people are involved in the execution of the tasks. For example, in the semiconductor industry, a wafer manufacturing line consists of an engineering chain, a supply chain, a manufacturing execution system, an equipment engineering system, and an internet system to support the interactions of these separately operated systems. Problems, such as a bottleneck in scheduling, machine failure, defects in products, or consumer complaints, need to be dealt with constantly. Daily routine operations are accomplished by many people and facilities. Some mechanical failures or quality problems may call for innovative solutions, redesigning, or reengineering. The whole cycle of CDIO may be repeated again and again. In addition, with use, users can detect problems that were missed by the designers or manufacturers. Their feedback provides invaluable opinions for product improvement and innovation. When setting up the production system for the first time, the creative processes are described as follows:

1. Preparation stage. The goal is to search for possible methods to manufacture the product for the market. Building a reliable, flexible, efficient, and effective manufacturing system is the goal at this phase of product innovation. User friendly, customer satisfaction, safety, ease of maintenance, and environmentally friendly are the ultimate criteria for the evaluation of the success of a product system. For example, a production system for a 5-inch wafer is not the same as the production system for a 3-inch wafer. Many new problems arise for each wafer size. They may be identified not only by experts in manufacturing and industrial management but also by operators at the work site, users, or lawmakers. Besides, because of the constantly evolving nature of science and technology, there is always a continual demand for improvements in production methods. These problems need to be solved by RD engineers or domain experts. In the case of making the Book of Chanel N^o 5, numerous problems could arise regarding the fragility of an embossed book, such as how to bind the book? How to repair or replace a damaged page? How to handle the book so that the embossed content will not be accidentally erased? How long would the embossed content stay legible? How to store or ship the book so that its contents will not be damaged? These problems may come as feedback to the experts in the conceptualization, design, or implementation phases for clarification and solutions. Or they might be solved by personnel such as production engineers, quality control engineers, or industrial engineers at the work site. The search for knowledge regarding production methods, equipment, materials, and procedures is again the first step to take here.

2. Incubation stage. A period of rest and separation from the problems at hand is necessary for remote associations in the memory to be activated and reach the activation level. Engineers and technical support staff need to come up with plans for a manufacturing plant. When actually building the production line, new technical problems may be detected and sent to RD engineers for solutions. Some problems may need to be sent to implementation, design, or even conceptualization phases for further clarification and ideation. In essence, the knowledge for problem-solving is again distributed among many different experts. Time and opportunity for the exchange of ideas and experiences are critical for remote conceptual combinations to occur.

3. Illumination stage. Once the remote idea reaches its activation level, it may pop out of an individual’s head after some period of incubation. The idea must have a chance to be heard, discussed, and accepted by the team. For example, to protect the embossed content of the Book of Chanel N^o 5, the book structure is housed in a black box so that the content of the book will not be damaged during handling and delivery.

4. Verification stage. The efficiency, effectiveness, safety, and environmental sustainability of the manufacturing procedure and the quality of its outputs must be constantly evaluated. User satisfaction, the success of sales, and employee satisfaction and safety are the final tests of the success of technological innovation. Other than the test of the market, the Book of Chanel N^o 5 is now in a permanent collection at the Museum of Modern Art in New York City.

As summarized in Table 1, the above cognitive task analyses show that CDIO represents different domains of knowledge involved in the innovation process of a new technological product. From the conceptualization of a need for a new product, to the design of the physical product, to making the first instance of the product, and, finally, to creating a manufacturing line for reproducing the product, the diversity and complexity of the tasks involved increase tremendously with the progression of the innovation. Cross-domain knowledge and collaboration among experts and professionals are crucial for the success of technological innovation. The criteria and methods to validate the viability of the innovation outcome for each phase of CDIO are different. Heavy reliance on cross-domain knowledge, user experiences, and cross-domain collaboration are unique features of technological innovation. Nevertheless, the creative cognitive processes, namely, the incubation stage and illumination, are similar across the 4 phases and are also similar to other types of creative works such as scientific discovery or art. Specifically, creative ideas are always conceived within and emerge from an individual’s mind. However, it is very likely that the remote associates for conceptual combination may come from another person’s mind. The opportunity to communicate and share thoughts with one another is important in technological innovation. A period of a brief or long pause from the current domain area is necessary for the mind to shift its focus of attention and wander away from the current problem so that remote associates from another domain may gain access to the conscious state. Thus, room to endure the ambiguity and uncertainty of the situation must be provided. An illumination of a solution popping out of the mind occurs when one of the team members recognizes the potential of relating the current problem to a remote associate, and an interpretation is formed and acknowledged by the team. This creative cognitive process can be succinctly delineated by a slight modification of Klahr and Simon’s [8] model of scientific discovery, shown in Figure 2.

7. Conclusions and discussion

In sum, the model shows three conditions for the generation of imaginative and creative technological innovation: 1. the importance of cross-domain knowledge is stressed, not just because an innovative solution, by its definition, cannot be found in the original domain of the product but also because the nature of technological innovation, more often than not, involves combinations or incorporation of new functions from other domains. For example, a smartphone is a combination of an original mobile phone and internet functions. 2. conceptual combination is viewed as a primitive heuristic for information searching and interpretation. The human ability to imagine and create is rooted in the brain’s capability to randomly combine information stored in memory and make sense of these novel conceptual combinations. It highlights the importance of not only memory search and retrieval but also cognitive mechanisms to make sense of the meanings implied by the remote associations of the concepts. Many heuristics and methods for innovative ideas such as Triz [48], SCAMPER [49], and morphological analysis [50] are just more elaborate and structured methods of conceptual combination. 3. Pattern recognition is a cognitive task not very well explained before in terms of how an original idea can be recognized and selected as a plausible solution to the problem. Here, it is proposed that the ability to recognize and select the outcome from conceptual combinations is a joint function of the goal, the interpretation efforts, and the value (bias) held by the individual and the team members. Recognition of a solution in its primitive form is shaped by the goal one is searching for in the creative process. Without it, one can be totally blind to the opportunity [51]. However, the acceptance and adoption of technological innovation are also determined by the group’s value about what the most desirable outcome is and how much risk (the investment) the group would take for its success in the making and selling of the product. Cultural and social factors may also come into play at the verification stage of every phase of CDIO [25, 42, 52].

The plausibility of the model awaits empirical verification. It cannot be tested directly. However, its theoretical and practical implications can be tested by logical analyses, experiments, field studies, or case studies. Because technological innovation is an application of scientific knowledge to the development of artifacts for human use, the importance of scientific knowledge and user experience for innovation in every phase of CDIO is beyond doubt. The nature of technological innovation itself is a combination of different technological products employed to produce a function that has higher utility than the previous one. Conceptual combination is the most primitive form of creative thinking. Empirical studies that examined the effects of conceptual combination in technological innovation were reported in the earlier sections of this paper. Studies can also be done to examine the effects of diversity of a team’s knowledge and its team members’ imaginative abilities on product innovation. For example, in Wang, Lu and Li’s study [53], data were drawn from 49 dyads who were the finalists out of 120 teams of a collegiate saw-design competition. Their task was to conceive and design an unusual use of the saw. The ideas behind the sketches of their design were presented and scored by 3 professors and 2 design professionals according to three criteria: inventiveness (60%), clarity of conceptualization and presentation (20%), and creative strategy for competition and marketing (20%). Results showed that participants’ imagination score measured by a conceptual combination test, efficiency, effectiveness of communication between the dyad, and heterogeneity of the team composition all contributed positively to design performance. The interaction between the imagination score and the heterogeneity of the team suggested that the dyads with higher imagination scores produced more creative designs when their collaborators were from a more different domain. In addition, a behavioral measure of imagination was constructed based on conceptual combination theory with acceptable reliability and validity [7]. The test scores were found to be able to predict design students’ design performance more than the originality measure of divergent thinking ability. Methods of training to enhance engineering imagination based on conceptual combination theory have also been designed and can be incorporated into engineering education [32].

One practical implication that deserves special attention for technological education and the industry is that chance plays a role in creative processes, but it does not come without cost. The magic of incubation and illumination suggests that, given sufficient motivation and prerequisite conditions, human minds may continue to freely search and combine ideas even subconsciously. In an industry, the absorptive ability of a firm, most likely an effect spilled over from its leaders, affects the innovativeness of the firm [54]. In engineering education, efforts should not be limited to only acquiring the CDIO knowledge necessary for technological innovation but also include recognition and cultivation of the important environmental conditions for innovative ideas to be brooded upon, pop out, and be recognized and selected.

Conceptual combination is a simple heuristic for generating imaginative ideas. It is made possible by the way human neurons may randomly combine with other neurons and generate new links between existing nodes. This automatic bottom-up process is accompanied by a top-down interpretation process that makes sense of the possible relation(s) between or among originally unrelated nodes (concepts) and generates new concepts. Life experiences and knowledge are indispensable components in the process, with the goal as the engine that ignites all the processes, decisions, and steps taken in CDIO phases of a technological innovation. A technological innovation project involves problems to be creatively solved for the first time in history. The final product is the joint effort of many experts and professionals. For a firm, heavy workload and time pressures are killers of a firm’s opportunity to think freely and the tendency to interact with one’s coworkers in good humor. In engineering education, providing training regarding the complete CDIO process in technological innovation and execution enables future engineers to gain a broader perspective of their tasks at hand and prepares them with a mind that can solve problems more creatively. The bottleneck in the use of conceptual combination lies firstly in the time-consuming process of knowledge acquisition, not only within one’s own application domain but also across other domains, and secondly in the ability to come up with original interpretations, a kind of abstract thinking and hypothesis generation, for a novel conceptual combination. Broad interests, curiosity, and a quest for knowledge are few of the important traits to be cultivated in engineering education in order to achieve minds that value novel, creative ideas and that are willing to play with ideas imaginatively, take reasonable risks, and use labor wisely to implement new ideas.

Acknowledgments

This work was in part supported by Taiwan’s Ministry of Science and Technology under Grant Number: MOST 107-2511-S-009-00-MY3.