Prototyping to Leverage Learning in Product Manufacturing Environments

Abstract

Rooted in the automotive industry, this article discusses the topic of leveraging tacit knowledge through prototyping. After first providing an overview on learning and knowledge, the Socialization, Externalization, Combination and Internalization (SECI) model is discussed in detail, with a clear distinction between tacit and explicit knowledge. Based on this model, we propose a framework for using said reflective and affirmative prototyping in an external vs. internal learning/knowledge capturing and transfer setting. Contextual examples from select automotive manufacturing R&D projects are given to demonstrate the importance and potential in applying more effective strategies for knowledge transformation in engineering design.

Full text

Available online at www.sciencedirect.com
ScienceDirect
Procedia CIRP 00 (2016) 000–000
www.elsevier.com/locate/procedia
2212-8271 © 2016 The Authors. Published by Elsevier B.V.
Peer-review under responsibility of the scientific committee of the 6th CIRP Conference on Learning Factories.
6th CLF - 6th CIRP Conference on Learning Factories
Prototyping to Leverage Learning in Product Manufacturing Environments
Jorgen A. B. Erichsena*, Andreas L. Pedersena, Martin Steinerta, Torgeir Weloa
aNTNU - Norwegian University of Science and Technology, Richard Birkelands Veg 2B, Trondheim NO-7491, Norway
* Corresponding author. Tel.: +47-416-46-804; fax: +47-735-94-129. E-mail address: jorgen.erichsen@ntnu.no
Abstract
Rooted in the automotive industry, this article discusses the topic of leveraging tacit knowledge through prototyping. After first providing an
overview on learning and knowledge, the Socialization, Externalization, Combination and Internalization (SECI) model is discussed in detail,
with a clear distinction between tacit and explicit knowledge. Based on this model, we propose a framework for using said reflective and
affirmative prototyping in an external vs. internal learning/knowledge capturing and transfer setting. Contextual examples from select automotive
manufacturing R&D projects are given to demonstrate the importance and potential in applying more effective strategies for knowledge
transformation in engineering design.
© 2016 The Authors. Published by Elsevier B.V.
Peer-review under responsibility of the scientific committee of the 6th CIRP Conference on Learning Factories.
Keywords: knowledge transfer; internal reflective prototyping; prototyping; tacit knowledge; integration events; product development
1. Introduction and Background
In this article, we argue for the use of explorative and
analytical approaches in product development processes by
discussing tacit knowledge accumulation and transfer through
prototypes. With this intention, we attempt to make several
contributions to current literature.
Firstly, we present a mapping of relevant literature on the
topic of knowledge, especially related to product development.
In this section, we are exploring organizational and individual
knowledge, the differentiation of tacit and explicit knowledge,
in addition to some current practices on the transfer of (tacit)
knowledge.
The second contribution is to present a model of prototyping
categories, with special emphasis on the differentiation
between learning and verification as the main intent for
prototyping activities. A model of four prototyping categories
is proposed, and discussed in relation to dealing with known
and unknown problems concerning tacit knowledge in product
development.
The article closes by exemplifying the previous two sections
by providing insights from two industry cases. The use of
analytical and explorative approaches to prototyping are
discussed, and several possible research opportunities are
presented.
The automotive industry—an industry with steadily
increasing demand for faster development cycles and higher
quality products—is subject to increasing competitive pressure.
Making mistakes is costly in an industry where product life
cycles are in the order of five to ten years, and late-stage design
changes have major implications for manufacturing planning
and processes. In addition, automakers need to rely on previous
experience, and cannot start from scratch in each development
project. The use of process and part standardization within the
product technology platforms is a well-established practice to
reduce the burden on the development teams. Hence, much
research is currently targeting knowledge and learning
mechanisms in new product development. Examples include
knowledge-based development (1)—a method for extracting
basic principles of Toyota’s product development processes
(2).
In this paper, we focus on analytical and explorative
approaches, and their relation to both creation and transfer of
tacit knowledge in product development.

2 Author name / Procedia CIRP 00 (2016) 000–000
2. Theory: Knowledge in Product Development
In (3), Ulonska presents numerous definitions of knowledge
found in product development. Rowley differentiates
knowledge and wisdom (4) by defining knowledge as
application of data and information (“know-how”), whereas
wisdom is defined as elevated understanding (“know-why”).
Additionally, it can be argued that knowledge can be further
divided into individual and organizational knowledge (5). The
sum of what is learned, experienced, discovered or perceived
(by individuals) during a project (in the organization) defines
organizational learning. The interactions of individuals are the
main ingredients of organizational knowledge, and the
knowledge of these individuals is called individual knowledge.
This is categorized in three categories; experience-based,
information-based and personal knowledge (6). Nonaka and
Takeuchi argue that the organizational knowledge exists
between (and not within) individuals (7).
2.1. Defining Integration Events and Knowledge Owners
Most product development organizations use stage-gates for
decision making. The stage-gate model is a financially-based
governance method, which leverages the importance of
financial decisions during development. However, this type of
process governance often makes event-based technological
decisions harder. Hence, there is a call for a more event-based
governance model in product development (8). An example on
such events can be the emerging trend of hosting ‘integration-
events’. These events are so-called learning cycle gates, and
aim at ensuring better insights and information while
preserving previous project know-how and learnings. This
way, large product development organizations aim at
transferring project (individual) knowledge into organizational
learning. Here, informal knowledge is formalized (made
explicit), and formal knowledge is interpreted (by the
individuals). The key to successful organizational learning is a
mutual exchange of these two kinds of knowledge.
Some companies employ key experts or learning facilitators
as catalysts for the exchange of knowledge within their
organization. These so-called knowledge owners are usually
technical or functional managers, who help preserve and
facilitate the learnings and insights. Examples of key experts
are Toyota’s functional managers who owns the technology.
The functional managers employ existing knowledge within
projects, while so-called chief engineers challenge the existing
standard by being the customer representative. By spending
time with and on the development team, these key experts gain
experience and insights, which in turn will contribute to
organizational learning inside the company.
By taking a closer look at learning mechanisms in product
development in Fig. 1—first introduced by Eris and Leifer (9),
and then further iterated by Leifer and Steinert (10)—the
distinction between formal and informal knowledge is
clarified. Key experts are usually working in the informal area
(i.e. learning loops two and three), whereas the organization as
a whole operates in the formal area (i.e. learning loop one).
2.2. Tacit and Explicit Knowledge in PD
The terms tacit and explicit knowledge are closely linked to
formal and informal knowledge. Explicit knowledge consists
of information, facts and numbers that have been formalized
(learning loop one from Fig. 1) (11), and they can be
summarized into so-called ‘knowledge artifacts’ (12).
Examples on these knowledge artifacts include the widespread
use of A3 sheets in the Toyota product development system
(2,13), which usually contain condensed explicit information
about a project or system. Tacit (or informal) knowledge
includes everything non-explicit, hereunder learnings, know-
how, craft and skill of the product engineering individuals,
Figure 1 - Learning mechanisms in product development, adopted from (9) and (10).

Author name / Procedia CIRP 00 (2016) 000–000 3
accumulated in learning loops two and three (14). We argue
that one key dimension of tacit knowledge is the interaction
with (and use of) objects and experiences in the product
engineering processes, often referred to as prototypes in one
form or another.
2.3. The SECI-model and Transfer of Knowledge in PD
First proposed by Nonaka, Toyama and Konno (15) as a
prevalent model for enhancement of knowledge creation
through conversion of tacit and explicit knowledge, the SECI
process (Fig. 2) can be used for describing the different stages
of knowledge transfer. The SECI model consists of four stages,
including socialization, externalization, combination and
internalization, and is used to describe how various knowledge
is transferred (in an organization) by spiraling through the four
stages. Four knowledge assets are presented as facilitators of
knowledge creation, and are categorized as experimental,
conceptual, systemic and routine. The latter has gotten
increasing support since its first appearance, and a study by
Chou and He (16) concludes conceptual knowledge assets (i.e.
PD insights) to have the most effect on knowledge creation.
By further studying the model, we can categorize the three
stages socialization (tacit-to-tacit), internalization (explicit-to-
tacit) and externalization (tacit-to-explicit) as forms of either
creation or transfer of tacit knowledge in development teams.
The last stage, combination (explicit-to-explicit), can be
described as an implemented knowledge repository, where the
formalized knowledge within the organization might be
distributed to sub-groups that require this knowledge. In the
context of transferring tacit knowledge, socialization includes
creating a work environment that encourages understanding of
expertise and skills through practice and demonstrators.
Externalization, or the act of formalizing the tacit knowledge,
aims at feeding this into the organization. Similarly,
internalization aims at interpretation of formal knowledge, and
includes conducting experiments, sharing results, and
facilitating prototyping as a means of knowledge acquisition
(15). Chou and He (16) also conclude that conceptual
knowledge assets—i.e. “knowledge articulated through
images, symbols and language” (15)—are the most efficient
tool for facilitating externalization and internalization.
2.4. A Proposed Model of Prototyping Categories
In (17), prototypes are defined as “An approximation of the
product along one or more dimensions of interest”, thus
including both physical and non-physical models. Examples
include (but are not limited to) sketches, mathematical models,
simulations, test components and fully functional pre-
production versions of the concept (18).
We argue that prototyping can be divided into four different
categories (Fig. 3) (19). The horizontal axis—the intent of the
prototype—is split into two sub-categories; “reflective” and
“affirmative”. The vertical axis, displaying the target audience
of the prototype, is spit into “internal” and “external”. This two-
by-two matrix gives four different prototyping categories
which will be briefly explained below.
2.4.1. External, affirmative prototyping
Typically used for making pre-production models, this kind
of prototyping approximate a nearly finished model, and are
often termed alpha and/or beta prototypes (20) intended for
validation or showcase purposes. These prototypes are high
fidelity (i.e. highly detailed) models, used for external
validation (e.g. certification test etc.), marketing, or in-depth
customer interaction. In an automotive setting, these may be
the cars subject to road testing, being pre-production cars tested
on closed test circuits by external users.
2.4.2. Internal, affirmative prototyping
Focused in terms of function, this type of prototyping is
intended for function, reliability and feasibility testing.
Examples include combinations of subsystems, fatigue testing
of conceptual prototypes or project milestones to validate team
progression. Although high in fidelity (regarding function and
complexity), these prototypes are still rarely shown to public
audiences. Automotive examples on this kind of prototyping
includes running lifecycle testing of components, like shock
absorbers, axles and other moving parts.
2.4.3. External, reflective prototyping
Companies often seek feedback from external sources by
showing off concepts. User interaction is carefully observed
and recorded for further study, and responses and reactions are
used for further improving other concepts. This kind of
prototyping is used for observing interaction with external
sources, enabling the design team to take a step back and learn
from the observations. In the automotive industry, automakers
often show off one-of-a-kind concept car projects at large
automotive venues to gather external feedback and reactions.
2.4.4. Internal, reflective prototyping
Internal, reflective prototyping is a learning activity, used by
the product development team to learn and conceptualize ideas.
These prototypes are rough, made for exploring, understanding
Figure 2 - The SECI-model, with blue areas highlighted as areas of interest,
adopted from (15).

4 Author name / Procedia CIRP 00 (2016) 000–000
and experimenting with functionalities that are essential for
product success, with the aim of creating new insights within
the product development team (21). Typically, internal,
reflective prototypes have low fidelity (22), and therefore
regarded as waste after a project is finished. These prototypes
may prove especially useful when facing high complex
problems, like the component layout of an automotive engine
bay.
By using terminology from the Tacit Knowledge
Framework (23,24), we use the terms ‘knowns’ and
‘unknowns’; Both affirmative prototyping categories are linked
to analysis, as they are dealing with known problems and
requirements—the ‘known knowns’ (i.e. known articulated
problems with known possible solutions). Adversely, reflective
prototyping categories aim at exploration, and thus at dealing
with unknown problems—the ‘unknown unknowns’ (i.e. non-
articulated problems with unknown solutions). Coming from
this perspective, we argue that known problems are best solved
analytically, while unknown problems are best solved
exploratively.
3. Examples: Learning from Prototyping
In the following subsections, the theory presented in the
previous section will be accentuated to show the influence of
internal, reflective prototyping in product development. The
first case considers applying a physical prototype to an analysis
for evaluating the numerical method and consequentially
learning about the method and saving time in the process. The
second case presents a failed crash box, once designed for a
new car model that was well analyzed—but still failed due to
an overlooked design-manufacturing detail. A discussion of the
mistakes is made in light of the theory presented.
3.1. Case I: Applying Physical Computation for a Rotational
Spiral Spring
In (25), a case illustrates the effects of combining numerical
computations with testing a physical representation of the
design. The time required to design a concept by using
analytical tools in complex cases can be greatly reduced by
applying a physical prototype for testing and comparison, as
proposed in the article.
The case studies a rotational spiral spring that is analyzed by
setting up a numerical model (using mechanical spring theory),
predicting stiffness and maximum stress of the rotational spiral
spring. Meanwhile, a physical model is made with MDF
(Medium Density Fibreboard) and tested (Fig. 4). The output
data reveals a striking similarity, though the stiffness is
somewhat overestimated in the analysis. Although the results
are not identical, the combination of the physical and numerical
computations shows the numerical analysis to be transferable
to the physical dimension and may be scaled further.
Combined, these methods yield satisfactory results in a very
short time.
This case shows very well how time can be saved by
applying internal, reflective prototyping early in the product
development process to facilitate faster learning. This approach
may prove especially applicable for complex cases, reducing
complexity by understanding which analytical tools might be
appropriate—and saving time by doing so. As for all internal,
reflective prototyping, the prototype used for the physical part
of the computation is not applicable in the finished product.
However, it facilitates the designers’ learning of how their
analytical problem transfers into the physical domain. Internal,
reflective prototyping is used to learn from internally, either
individually or as a collaborative group, as they typically are
low fidelity in nature, but educational and time saving.
3.2. Case II: Crash Box Failure Due to Lack of Variability
Testing
In this case, we use an example from a large European
automaker, which had designed a crash box for topological
optimization, to be fit into a new car model. Crash boxes,
separate deformation elements between the front bumper and
the front longitudinal rail, are designed to deform on low-speed
impact to prevent damage to the rest of the car to reduce the
repair cost. The production method of the crash box was
extrusion of one open cross-section that was bent, cut, pierced,
and welded into a closed box configuration with an integrated
foot plate mounted to the rails.
The Danner crash test (26) rates cars at the impact of
collision in their ability to minimize costs of repair at 0-
15km/h, for the purpose of evaluating the car’s properties to set
an insurance premium base. In the Danner test, the crash box
Figure 4 - MDF prototype with markings used to estimate the flex of the
rotational spring (25).
Figure 3 - A proposed model of four prototyping categories.

Author name / Procedia CIRP 00 (2016) 000–000 5
of the said model was expected to crush in a controlled manner
upon collision test impact without damaging expensive
components or activate the air bags, which are the costliest to
replace. In the numerous FEA simulations done to optimize the
system, the welding configuration was assumed to be
geometrically perfect, starting at the very end of the box.
However, in production (MIG) welding, start and stop of the
weld seam tend to create minor groove of varying magnitude at
the very end, depending on dimensional accuracy of the
individual part, and other control parameters. Hence, the
accuracy of the FEA model was not capable of capturing the
local stress state in the vicinity of the grove (as illustrated in
Fig. 5). Instead of failing by controlled crushing as predicted in
the FEA model, occasionally, the weld seam failed like a zipper
starting from the very end of the box once the bumper folded
and contacted the very end of the crash box. The fluctuations
(in the force deformation curve) triggered the air bag sensors,
resulting in the airbags deploying in low speed tests at 15 km/h.
This type of failure is considered catastrophic as a consequence
of the repair costs associated with replacing the airbags.
The influence of small variations imposed by manufacturing
(welding) is a very complex matter. Sensitivity testing of the
crash box with the same production-intent premises as the
serial produced product would have prevented encountering a
failure such a long time after launch. This clearly demonstrates
the risk of failing to integrate the product development process
and the manufacturing process. The design engineers did not
know this would be an issue, and the unspecified ‘parameter’
related to end configuration (of the weld) remained an
unknown until several vehicles were retested after launch.
If the team had engaged in internal reflective prototyping
activities, the influence of such critical design features could
have been uncovered. The learning outcome in this case could
have led the team members to acquire the necessary knowledge
to see the disconnection between the manufacturing process
and the intended design, possibly identifying a low-cost
solution (process or design change) to such a fairly fixable
problem.
In this case, properly done internal, affirmative prototyping
could have uncovered the problem. However, we would argue
that doing internal, reflective prototyping in the early stages of
the development process would have facilitated important
learning. As a result, the early development process would be
less complex, and problems not otherwise perceived as
problems would be uncovered. Hence the value of prototyping
and testing to learn—not only to verify—could have
significantly saved time, money and averted the ultimate failure
of the design.
4. Research Potential of Using Explorative and Analytical
Methods for Learning in Product Development
Furthermore, the insights, experience and learnings present
a unique research opportunity, since improved understanding
of the creation and transfer of tacit knowledge will alter how
we facilitate the product development process. Hence, there is
a call for more research concerning how tacit knowledge
influences the development of products with high levels of
complexity, especially when dealing with many unknown
unknowns.
As identified in (27), there is a gap between professional
knowledge and real-world practice. In his works, Simon
applies methods of optimization from statistical decision
theory, thus laying a foundation for a scientific approach to
treating knowledge. Adversely, Schön (28) argues that the real
challenge lies not within the treatment of well-formed
requirements, but rather the extraction of such requirements—
practically unknown unknowns—from real world situations. In
(29), Schön presents reflective iteration rounds as a learning
tool of great potential. Taking this perspective, we argue that
reflective prototyping may be used as a learning tool in
handling unknown unknowns in product development.
Ultimately, we argue that, in reality, product development
requires balancing of the tacit and the explicit, the explorative
and the analytical. We have seen that disconnection between
product development and manufacturing processes cause major
implications for entire value chains. In hindsight, exploration
and experience of manufacturing techniques and challenges
could have led to the discovery of potential risks and problems
in the product development process (unknown unknowns),
and—if so—how to best balance analysis and exploration for
uncovering these unknowns in a cost and resource efficient
manner?
5. Conclusion
The purpose of this paper has been to accentuate the
possibilities of using prototyping in product development for
manufacturing settings. An attempt has been made to map
future opportunities, both for industry and academia, and a call
for the recognition of prototyping as a time saving learning
tool. The potential of applying exploration by interaction with
prototypes related to knowledge capture, transfer and learning
is demonstrated in the context of the automotive industry. Thus,
a call for increased focus on mixing analytical (e.g.
simulations) and explorative (e.g. prototyping) approaches is
presented as a viable direction for further efforts in both
industry and academic communities.
Altogether, the importance of understanding the interplay
between (tacit) knowledge, explorative and analytical
Figure 5 - Exemplification of a crash box, with highlighted area of interest.

6 Author name / Procedia CIRP 00 (2016) 000–000
approaches to problems in product development and
manufacturing, and the role of prototyping for learning are
topics that require further pursuit.
Acknowledgements
This research is supported by the Research Council of
Norway through its user-driven research (BIA) funding
scheme, project number 236739/O30.
References
[1] Ringen G, Welo T. Knowledge based development practices in systems
engineering companies: A comparative study. In: Systems Conference
(SysCon), 2015 9th Annual IEEE International. 2015. p. 353–8.
[2] Morgan JM, Liker JK. The Toyota product development system. N Y.
2006;
[3] Ulonska S. A Knowledge-Based Approach for Integration of System
Design Methodology and Documentation in Advanced Multi-
Disciplinary NPD Projects [Internet]. NTNU; 2014 [cited 2015 Oct 29].
Available from: http://brage.bibsys.no/xmlui/handle/11250/276639
[4] Rowley JE. The wisdom hierarchy: representations of the DIKW
hierarchy. J Inf Sci [Internet]. 2007 Feb 15 [cited 2015 Oct 29];
Available from:
http://jis.sagepub.com/content/early/2007/02/15/0165551506070706
[5] Schubert P, Lincke D-M, Schmid B. A global knowledge medium as a
virtual community: the NetAcademy concept. AMCIS 1998 Proc.
1998;207.
[6] Løwendahl BR, Revang Ø, Fosstenløkken SM. Knowledge and Value
Creation in Professional Service Firms: A Framework for Analysis.
Hum Relat. 2001 Jan 7;54(7):911–31.
[7] Nonaka I, Takeuchi H. The Knowledge-Creating Company: How
Japanese Companies Create the Dynamics of Innovation. Oxford
University Press; 1995. 320 p.
[8] Ringen G (1), Welo T (2). Towards a more event-driven NPD process:
First experiences with attempts of implementation in the front-end
phase. 75-1 Proc 19th Int Conf Eng Des ICED13 Des Harmon Vol1
Des Process Seoul Korea 19-22082013. 2013;
[9] Eris O, Leifer L. Faciliating product development knowledge
acquisition: interaction between the expert and the team. Int J Eng
Educ. 2003;19(1):142–52.
[10] Leifer LJ, Steinert M. Dancing with Ambiguity: Causality Behavior,
Design Thinking, and Triple-Loop-Learning. In: Gassmann O,
Schweitzer F, editors. Management of the Fuzzy Front End of
Innovation [Internet]. Cham: Springer International Publishing; 2014
[cited 2015 Dec 2]. p. 141–58. Available from:
http://link.springer.com/10.1007/978-3-319-01056-4_11
[11] Ryle G. The Concept of Mind: 60th Anniversary Edition. Routledge;
2009. 377 p.
[12] Holsapple CW, Joshi KD. Organizational knowledge resources. Decis
Support Syst. 2001 May;31(1):39–54.
[13] II DKS, Smalley A. Understanding A3 Thinking: A Critical Component
of Toyota’s PDCA Management System. CRC Press; 2011. 186 p.
[14] Polanyi M, Sen A. The Tacit Dimension. University of Chicago Press;
2009. 129 p.
[15] Nonaka I, Toyama R, Konno N. SECI, Ba and Leadership: a Unified
Model of Dynamic Knowledge Creation. Long Range Plann. 2000 Feb
1;33(1):5–34.
[16] Chou S-W, He M-Y. Knowledge Management: The Distinctive Roles
of Knowledge Assets in Facilitating Knowledge Creation. J Inf Sci.
2004 Jan 4;30(2):146–64.
[17] Eppinger SD, Ulrich KT. Product design and development. 1995. 1995;
[18] Elverum CW, Welo T. On the use of directional and incremental
prototyping in the development of high novelty products: Two case
studies in the automotive industry. J Eng Technol Manag [Internet].
2015 [cited 2015 Oct 29]; Available from:
http://www.sciencedirect.com/science/article/pii/S0923474815000405
[19] Erichsen JAB, Pedersen AL, Steinert M, Welo T. Using Prototypes to
Leverage Knowledge in Product Development: Examples from the
Automotive Industry. 2016 IEEE International Symposium on Systems
Engineering (ISSE). In press. p. 6.
[20] Otto KN, Wood KL. Product Evolution: A Reverse Engineering and
Redesign Methodology. Res Eng Des. 1998 Dec;10(4):226–43.
[21] Lim Y-K, Stolterman E, Tenenberg J. The Anatomy of Prototypes:
Prototypes As Filters, Prototypes As Manifestations of Design Ideas.
ACM Trans Comput-Hum Interact. 2008 Jul;15(2):7:1–7:27.
[22] Bryan-Kinns N, Hamilton F. One for All and All for One?: Case
Studies of Using Prototypes in Commercial Projects. In: Proceedings of
the Second Nordic Conference on Human-computer Interaction
[Internet]. New York, NY, USA: ACM; 2002 [cited 2015 Sep 30]. p.
91–100. Available from: http://doi.acm.org/10.1145/572020.572032
[23] Gervasi V, Gacitua R, Rouncefield M, Sawyer P, Kof L, Ma L, et al.
Unpacking Tacit Knowledge for Requirements Engineering. In: Maalej
W, Thurimella AK, editors. Managing Requirements Knowledge
[Internet]. Springer Berlin Heidelberg; 2013 [cited 2016 Jan 28]. p. 23–
47. Available from: http://link.springer.com/chapter/10.1007/978-3-
642-34419-0_2
[24] Sutcliffe A, Sawyer P. Requirements elicitation: Towards the unknown
unknowns. In: Requirements Engineering Conference (RE), 2013 21st
IEEE International. 2013. p. 92–104.
[25] Foehr AGC, Stuecheli M, Meboldt M. EFFICIENT DESIGN
EVALUATION THROUGH THE COMBINATION OF
NUMERICAL AND PHYSICAL COMPUTATIONS. 80-6 Proc 20th
Int Conf Eng Des ICED 15 Vol 6 Des Methods Tools - Part 2 Milan
Italy 27-300715. 2015;
[26] AZT Automotive :: en [Internet]. [cited 2016 Jan 30]. Available from:
http://azt-automotive.com/en/
[27] Simon HA. Administrative behavior. Cambridge Univ Press; 1965.
[28] Schön DA. Educating the reflective practitioner. 1987;
[29] Schön DA. The Reflective Practitioner: How Professionals Think in
Action. Basic Books; 1983. 388 p.