Content uploaded by Frank T Piller
Author content
All content in this area was uploaded by Frank T Piller on May 19, 2018
Content may be subject to copyright.
Economic implications of 3D printing: Market structure models in light
of additive manufacturing revisited
Christian Weller, Robin Kleer
n
, Frank T. Piller
TIME Research Area, Technology and Innovation Management Group, School of Business and Economics, RWTH Aachen University, Germany
article info
Article history:
Received 11 February 2014
Accepted 24 February 2015
Available online 6 March 2015
Keywords:
3D printing
Additive manufacturing
Market structure
Flexible manufacturing
Economic modeling
abstract
Additive manufacturing (AM), colloquially known as 3D printing, is currently being promoted as the
spark of a new industrial revolution. The technology allows one to make customized products without
incurring any cost penalties in manufacturing as neither tools nor molds are required. Moreover, AM
enables the production of complex and integrated functional designs in a one-step process, thereby also
potentially reducing the need for assembly work. In this article, we discuss the impact of AM technology
at both firm and industry level. Our intention is to discern how market structures will be affected from
an operations management perspective. Based on an analysis of established economic models, we first
identify the economic and technological characteristics of AM and distill four key principles relevant to
manufacturers at firm level. We then critically assess the effects of AM at industry level by analyzing the
validity of earlier assumptions in the models when these four principles apply. In so doing, we derive a
set of seven propositions which provide impetus for future research. In particular, we propose that in a
monopoly, the adoption of AM allows a firm to increase profits by capturing consumer surplus when
flexibly producing customized products. Meanwhile in competitive markets, competition is spurred as
AM may lower barriers to market entry and offers the ability to serve multiple markets at once. This
should ultimately result in lower prices for consumers.
&2015 Published by Elsevier B.V.
1. Introduction
Research has shown that technological innovation affects firms
and market structure (Mills and Schumann, 1985; Vickers, 1986;
Geroski and Pomroy, 1990; Khanna, 1995). In particular, the
adoption of flexible manufacturing systems (FMS) has had sig-
nificant implications for manufacturers and market structure (e.g.,
Sethi and Sethi, 1990; Womack et al., 1991). FMS can flexibly
produce a variety of different outcomes using the same (manu-
facturing) resources (Gerwin, 1993). More recently, research has
highlighted the potential of additive manufacturing (AM) technol-
ogy to spark a new industrial revolution by extending the features
of conventional FMS technology (The Economist, 2011, 2012;
Berman, 2012; Mellor et al., 2014). AM refers to “the process of
joining materials to make objects from 3D model data, usually
layer upon layer”(ASTM International, 2013). Colloquially, AM is
often referred to as “3D printing”(Lipson and Kurman, 2013). The
main benefit of AM technology is that it enables the flexible
production of customized products without cost penalties in
manufacturing. It does so by using direct digital manufacturing
processes that directly transform 3D data into physical parts,
without any need for tools or molds. Additionally, the layer
manufacturing principle can also produce functionally integrated
parts in a single production step, hence reducing the need for
assembly activities. Thus, AM technology significantly affects the
costs of flexibility, individualization, capital costs, and marginal
production costs (Koren, 2006; Dolgui and Proth, 2010; Berman,
2012). While AM could simply be interpreted as a new generation
of conventional FMS, earlier research has argued that its economic
characteristics are so different that investment decisions into AM
are highly strategic (Mellor et al. 2014). Nonetheless, the oppor-
tunities of AM still come with a number of limitations: available
materials do not always match the characteristics of conventional
manufacturing processes, the production throughput speed is
rather low, most manufactures still demand an additional surface
finish, and common standards for quality control are not estab-
lished yet (Berman, 2012). Given this trade-off, one motivation of
our paper is to provide support in evaluating a firm's specific
manufacturing context before potential investments in AM
technology.
Additionally, AM technology affects market structure beyond
direct effects on a single firm's production processes. There is a
growing community of “makers”who develop and share 3D
models, sell 3D printed products on marketplaces, and even
develop and provide their own 3D printers for home usage
Contents lists available at ScienceDirect
journal homepage: www.elsevier.com/locate/ijpe
Int. J. Production Economics
http://dx.doi.org/10.1016/j.ijpe.2015.02.020
0925-5273/&2015 Published by Elsevier B.V.
n
Corresponding author. Tel.: þ49 241 80 99176; fax: þ49 241 8092367.
E-mail addresses: weller@time.rwth-aachen.de (C. Weller),
kleer@time.rwth-aachen.de (R. Kleer), piller@time.rwth-aachen.de (F.T. Piller).
Int. J. Production Economics 164 (2015) 43–56
(Gershenfeld, 2005; Lipson and Kurman, 2010; de Jong and de
Bruijn, 2013). Furthermore, a steadily growing number of 3D
printers for home and industrial use extends the scale and scope
of manufacturing options. AM technology is even on the political
agenda, with U.S. President Barack Obama promoting AM as
having “the potential to revolutionize the way we make almost
everything”(Obama, 2013). Only two years ago, industry analyst
Gartner (2012) argued that AM is at its “peak of inflated expecta-
tions,”noting that the technology is still too immature to satisfy
such high expectations. More recently, however, Gartner (2014)
predicted that industrial use of AM is likely to reach a level of
mainstream adaptation between 2016 and 2020.
1
Various early adopters demonstrate AM's potential benefits for
firms in different markets. In the shoe industry, for example,
manufacturers have been using AM technology for many years
to rapidly test new designs and accelerate the innovation process
(Jopson, 2013). But today, customized end products are also
manufactured with AM. For example, Nike offered a customizable
football cleat produced with AM in 2013. At the same time, AM
technology also facilitates market entry into a relatively mature
industry. For example, the Belgian shoe retailer Runners Service Lab
offers affordable, customized running shoes produced with AM
technology. As a demanding industrial application consider the
usage of AM in the aerospace industry, where the elimination of
many conventional design-for-manufacture constraints promises
opportunities for optimized designs to increase performance and
reduce weight of components. Despite the high regulatory require-
ments, AM has already been used for the low volume production
of aerospace components. Boeing uses some thermoplastic com-
ponents produced with Selective Laser Sintering technology on
commercial 737, 747 and 777 programs and has several hundred
components on the 787 aircraft prototype (Mellor, 2014). Further
industrial applications of AM are frequent in medical markets,
defense, automotive and machinery components (Wohlers, 2013).
Despite the current “AM hype”, research on the economic and
business effects of AM technology is still scarce. Most academic
literature on AM is focused on the technological aspects in the
fields of engineering, material science, and computer science.
Therefore, our paper aims to outline some central potential
economic implications of AM on manufacturing firms and mar-
kets. We acknowledge that our assessment is futuristic to some
degree. AM is currently mostly applied on a small scale in niche
markets or in a lab environment. No large-scale empirical data is
yet available to test our propositions. However, current research
continues to enhance the capabilities of AM technology. Industry
experts claim that its maturity will lead to a broad industrial
penetration within the next few years (Lux Research, 2013;
Wohlers, 2013; Gartner, 2014).
This is why an economic assessment and an evaluation of
business implications are crucial. Our work aims to initiate this
academic discussion. By relying on established production eco-
nomic models as the basis of our analysis, we strive to ground our
arguments beyond any hype or speculation. Doing so, we con-
tribute to the existing literature in four ways. Firstly, we distill four
key principles of AM relevant to manufacturing firms, and system-
atically evaluate their potential effects on a firm’s payoff function
(in a monopolistic setting), applying Milgrom and Roberts' (1990)
model of modern manufacturing on AM. Secondly, we analyze
extant literature containing market structure models that assess
advancements in the flexibility of manufacturing systems along
the dimensions of AM technology’s key principles, investigating
the effect of AM on competition at industry level. Based on our
analysis we then develop a set of seven propositions as an impetus
for future research on the economic implications of AM technol-
ogy. Fourthly, we outline relevant implications for industrial
practitioners by summarizing AM's technological opportunities,
applications, and constraints. Our article proceeds as follows:
Section 2 presents an overview of AM technology and its economic
characteristics. Section 3 investigates the effects of adopting AM at
firm level, while Section 4 derives implications for market struc-
ture at industry level. Finally, we present conclusions and implica-
tions for future research.
2. AM characteristics and key principles for manufacturing
firms
2.1. Technological background
AM technology has been in use since the 1980s. Its early
application was limited to the production of prototypes. The
technology’s primary goal was to offer a quick and affordable
way to receive tangible feedback during the product development
process. Primarily, the stereolithography method was used to
harden liquid photo-sensitive polymers with laser light in a
layer-by-layer process (Gibson et al., 2010). This production of
prototypes has become common practice in many firms and
industries and is not the object of our investigation. The far greater
opportunities of AM are in replacing conventional production
technologies for series manufacturing (“rapid manufacturing”,
Gibson et al., 2010). This application is the object of the current
hype and debate. It includes the production of parts and compo-
nents, but also the manufacturing of end products. Market expert
Terry Wohlers forecasts that end products will account for 80
percent of the total AM production output by 2019 (Davidson,
2012). The availability of materials for AM has increased steadily
over the last years, ranging today from various plastics to ceramics,
metals and concrete –basically, any material that can be liquefied/
melted and re-solidified (Gibson et al., 2010).
2
AM offers several technological advantages that enhance con-
ventional FMS. First of all, AM enables direct manufacturing of
digital 3D models stored in a computer-aided design (CAD) file,
without the need for tools or molds (Gibson et al., 2010). AM is
therefore a prototypical example of a flexible manufacturing
technology, as it conceptually enables a vast variety of outcomes
in manufacturing in any given sequence on one stable manufac-
turing system (Berman, 2012). Other than consumables (materials)
and the actual AM machine, only the product's digital 3D model is
necessary for manufacturing (Gebhardt, 2003). Setup and change-
over costs are negligible as only a different CAD file needs to be
uploaded into the machine when changing the product to be
manufactured –neither tools nor molds are necessary (Petrovic et
al., 2010). As a result, AM enables product individualization with-
out cost penalties (Gibson et al., 2010). By adding material layer by
layer until the product is finalized, AM has fewer process and
design restrictions; it therefore also allows for functionally opti-
mized product designs (e.g., lightweight designs, integrated cool-
ing chambers) (Petrovic et al., 2010; Lott et al., 2011). Furthermore,
an increase in design complexity does not mean higher production
1
Indications of market confidence in the sustainability of business models
relying on 3D printing offerings include the recently announced acquisition of the
3D home printer manufacturer and operator of the Thingiverse platform, Makerbot,
by Stratasys for US$403 million (Stratasys, 2013). Furthermore, dedicated invest-
ment funds have been launched that track the performance of the AM sector which
also indicates the growing importance of this industry.
2
Ongoing research activities also enhance specific AM technologies for print-
ing human tissue, as provided for example by the California-based company
Organovo, or for printing edible substances (3D Systems, an AM machinery
provider, recently launched a “food 3D printer”, enabling chefs and bakers to
produce customized edible arrangements).
C. Weller et al. / Int. J. Production Economics 164 (2015) 43–5644
costs, contrary to conventional technologies where production
unit costs usually increase with a higher complexity of product
design (Kota et al., 2000). Moreover, AM enables the production of
functionally integrated product designs in a one-step manufactur-
ing process. Certain functions, such as moving parts or cooling
systems, can be directly integrated into the produced parts with-
out involving additional manufacturing or assembly steps (Gibson
et al., 2010), further reducing manufacturing costs due to a
decrease of production stages.
3
Finally, case studies show that
AM technology can substantially lower required raw materials and
reduce scrap compared with conventional manufacturing technol-
ogy, particularly for metal parts (Petrovic et al., 2010; Lipson and
Kurman, 2013).
However, AM technology still has several restrictions limiting
its application. For example, available materials and the choice of
colors and surface finishes are still constrained (Berman, 2012).
Furthermore, the build space of AM machines sets a physical limit
to product dimensions (Gebhardt, 2003). With current AM tech-
nology, quality issues are also a concern. Parts may lack resistance
to environmental influences and fail with exposure to high
stresses (Petrovic et al., 2010; Berman, 2012). In addition, the
precision of the produced parts still needs improvement. There-
fore, reproducibility of parts cannot be assured, and global quality
as well as testing standards are still to be defined (Petrovic et al.,
2010; Lott et al., 2011). Moreover, design tools have yet to fully
exploit the possibilities AM technology offers. Missing guidelines
currently make it difficult for non-experts to optimize product
designs and attain the necessary know-how (Lipson and Kurman,
2013). Current research aims to overcome existing limitations and
to enhance the capabilities of AM technology. For example, the
Oak Ridge National Laboratory works closely with industry part-
ners to advance AM technology (i.e., reduce costs, energy needs
and emissions) and facilitate adoption by firms. In Europe, the
Fraunhofer Institute for Laser Technology leads in enhancing the
accuracy and efficiency of the Selective Laser Melting technology,
providing a much larger scope of metal based industrial applica-
tions at lower cost. Recently, the Self-Assembly Technologies Lab
at MIT has drawn significant attention by incorporating a fourth
dimension into AM by producing parts that are reactive to their
environment and change their shape over time in the usage stage
(Pei, 2014). Table 1 summarizes the technological opportunities
and limitations of AM. However, further technological progress
may change this picture on both sides.
2.2. Economic characteristics of AM
From an aggregated perspective, AM is most advantageous in
market environments characterized by demand for customization,
flexibility, design complexity, and high transportation costs for the
delivery of end products (Gibson et al., 2010; Berman, 2012; Lipson
and Kurman, 2013). AM facilitates product innovation because
design iterations are relatively inexpensive and parts can be
rapidly produced. AM technology frees up the solution space for
product designs, in essence solely limiting it to designers' creativ-
ity and physical laws. Theoretically, AM technology is capable of
producing any physically feasible product design compiled in a 3D
model, because products are manufactured layer by layer. Thus,
product designs can be optimized according to their desired
function rather than restricted from production technology or
supply chain constraints (Berman, 2012; Lipson and Kurman,
2013). Furthermore, firms are able to offer highly customized
products that match customer preferences. Product customization
potentially yields an increase in customers' perceived product
value and, thus, higher willingness to pay (as demonstrated in the
research on mass customization; see Franke and Piller, 2004;
Franke et al., 2009; Fogliatto et al., 2012). This is why firms can
charge a price premium.
4
AM enables customers to co-design products that perfectly fit
their demand. A number of websites
5
already allow consumers to
create individual products by altering distinct design parameters
within a 3D product configurator –the individualized 3D models
are then directly produced with AM technology. As a result,
product variety can potentially become infinite without incurring
additional costs in manufacturing. In contrast, conventional pro-
duction technologies require modular product architectures to
provide customized goods in an efficient way by combining pre-
assembled, modular parts, resulting in distinct product variants.
Thus, postponement strategies are pursued to gain mass produc-
tion efficiency, usually involving multiple production steps (Tseng
et al., 1996; Brabazon et al., 2010). Nonetheless, each variant
creates additional complexity and costs in a supply chain relying
on conventional production technology, which is not the case with
AM (Berman, 2012). Therefore, AM technology has the potential to
resolve the “scale-scope dilemma”on the cost side: there are no
penalties associated with a higher degree of product variety (Lott
et al., 2011). Moreover, lead times for the production of single
batches of product variants can largely be reduced,
6
while product
variants can be produced in any sequence without additional
changeover time or switching costs. Thus, AM potentially enables
an efficient lot size of one (Gibson et al., 2010) without costly
inventories of semi-finished and finished goods (Berman, 2012).
Thus, the degree to which designers can create products that meet
individual customers' demand is likely to become a key factor for
firms that want to profit from AM (Reeves, 2009; Lipson and
Kurman, 2013).
AM will also have an impact on decisions taken regarding
manufacturing location. Relatively low fixed costs for basic AM
machines and setup, combined with the feasibility of economically
producing small batch sizes, will potentially enable local produc-
tion near or even at the point of use (Berman, 2012). High
transportation costs for the delivery of end products that exceed
transportation costs for raw materials as well as penalties for late
delivery can also shift the manufacturing location toward the point
of use (Kleer and Piller, 2013). As a result, cost advantages of
producing in low-wage countries might diminish in the long run
(Petrovic et al., 2010; Schuh et al., 2011). In addition, new services
facilitate access to local AM manufacturing capacities, which
lowers barriers to market entry. For example, startups like U.S.-
based TechShop and established players like the logistics compa-
nies UPS or the French La Poste provide local manufacturing
capacities, such that small businesses and consumers alike can
produce 3D designs in a local shop equipped with AM technology.
In some industries, penalties for late delivery may become a major
3
However, support material is often required with AM, particularly when
producing moving parts. Today, support material often still must be removed in a
separate production step (Gibson et al., 2010). Many support materials can be
recycled (Lipson and Kurman, 2013).
4
In general, price premium refers to a price increment above a certain
reference price (e.g. marginal costs). In markets with imperfect competition (e.g.
monopolistic competition) producers have the market power to charge a price
premium to their customers. Chamberlin (1962) explains that a producer can
charge prices that are higher than perfect competition would allow, when meeting
specific needs of customers.
5
Several online applications base their offerings on AM technology and co-
design by consumers, such as individualized medals and trophies (www.twikit.
com), personalized 3D-printed figurines (www.cubify.com/store/3dme.aspx) and
individualized dolls configured with a smartphone application (www.makie.me).
6
Although build rates are not currently reaching levels of mass production, the
elimination of additional production processes and tools or molds makes AM a
viable option for small to mid-scale production. Further technical advancements
that enhance production throughput speed and quality will drive its application
toward larger-scale production levels (Atzeni and Salmi, 2012; Mellor et al. 2014).
C. Weller et al. / Int. J. Production Economics 164 (2015) 43–56 45
driver for locally installing AM manufacturing capacities. For
example, local on-demand production of replacement parts at
mining sites (Nicholls, 2013) or industrial production lines where
site locations are remote can reduce costly production downtimes
or excessive inventories for replacement parts (Lipson and
Kurman, 2013).
However, these opportunities are counterbalanced by a number
of limitations. Marginal production costs of AM remain higher
than with conventional technology, owing largely to high material
costs and energy intensity (Gibson et al., 2010). Nonetheless,
material costs are likely to decrease when additional suppliers
enter the market (Lux Research, 2013). Another constraint is
relatively low production throughput speed compared to conven-
tional production methods (e.g., injection molding) (Gibson et al.,
2010). While product variety can be increased without cost
penalties in manufacturing, AM cannot exploit economies of scale
when increasing the production volume of a product variant
(Berman, 2012). Therefore, mass production of standardized parts
will likely remain the domain of conventional manufacturing
technologies. Furthermore, quality issues may discourage some
potential customers from buying products produced with AM
technology. In combination with improved 3D-scanning and
reverse-engineering capabilities, AM also poses severe risks to
the intellectual property rights of product designs (Kurfess and
Cass, 2014). Copying a physical product and converting it into
shareable 3D design data might become as easy as copying a
printed document or sharing ordinary computer files –similar
developments led to disruptive change in the music industry
(Wilbanks, 2013). The issue of property rights in an age of digital
product designs is one of the most severe economic consequences
of AM, and demands further consideration.
In conclusion, Table 2 summarizes the characteristics of AM
from an economic perspective. AM offers a wide range of eco-
nomic opportunities. Manufacturers, however, also face various
limitations. Some of them may diminish as the result of further
research and technological progress (Wohlers, 2013; Gartner,
2014). Others, however, are inherent to the characteristics of AM
technology, and the ability to manage the trade-off between AM's
advantages and challenges will become a major source of compe-
titive advantage.
2.3. AM's key principles for manufacturing firms
Summarizing the preceding discussion, we can derive a num-
ber of key principles that differentiate AM technology from the
capabilities of conventional FMS (summarized in Table 3). We
strive to formulate those principles independently from the rapid
progress of AM technology (Wohlers, 2013). All arguments pertain
to capabilities that are possible today and are likely to improve in
the future.
First, AM technology is a versatile manufacturing machine
(Garrett 2014) that is capable of directly transforming any suitable
digital 3D model into a physical product using standardized data
interfaces for various applications. Secondly, AM offers customiza-
tion and flexibility for free –that is, it does not require tools or
molds, thus eliminating sunk costs before production start. In
addition, flexibility is high because the sequence and volume of
products can be altered without cost penalties in manufacturing.
Thirdly, AM offers complexity for free –that is, it allows for prod-
uct design complexity and a larger number of product variants
with no increase in manufacturing costs. Fourthly, assembly work
Table 1
AM technology’s opportunities and limitations from a technological perspective.
Technological characteristics of AM
Opportunities Limitations
þDirect digital manufacturing of 3D product designs without the need for tools or
molds
–Solution space limited to ‘printable’materials (e.g., no combined materials)
and by size of build space
–Quality issues of produced parts: limited reproducibility of parts, missing
resistance to environmental influences
–Significant efforts are still needed for surface finishing
–Lacking design tools and guidelines to fully exploit possibilities of AM
–Low production throughput speed
–Skilled labor and strong experience needed
þChange of product designs without cost penalty in manufacturing
þIncrease of design complexity (e.g., lightweight designs or integrated cooling
chambers) without cost penalty in manufacturing
þHigh manufacturing flexibility: objects can be produced in any random order
without cost penalty
þProduction of functionally integrated designs in one-step
þLess scrap and fewer raw materials required
Table 2
AM technology’s opportunities and limitations from an economic perspective.
Economic characteristics of AM
Opportunities Limitations
þAcceleration and simplification of product innovation: iterations are not costly and end
products are rapidly available
–High marginal cost of production (raw material costs and energy
intensity)
–No economies of scale
–Missing quality standards
–Product offering limited to technological feasibility (solution space,
reproducibility, quality, speed)
–Intellectual property rights and warranty related limitations
–Training efforts required
–Skilled labor and strong experience needed
þPrice premiums can be achieved through customization or functional improvement
(e.g., lightweight) of products
þCustomer co-design of products without incurring cost penalty in manufacturing
þResolving “scale-scope dilemma”: no cost penalties in manufacturing for higher
product variety
þInventories can become obsolete when supported by make-to-order processes
þReduction of assembly work with one-step production of functional products
þLowering barriers to market entry
þLocal production enabled
þCost advantages of low-wage countries might diminish in the long run
C. Weller et al. / Int. J. Production Economics 164 (2015) 43–5646
requirements are reduced when producing integrated functional
designs in one step.
From the perspective of strategic production management,
choosing an appropriate manufacturing system still remains an
issue in an age of additive manufacturing (Kakati, 1997; Mellor
et al. 2014). Both AM and conventional manufacturing technolo-
gies retain distinct advantages and limitations. This demands a
detailed assessment of both options. While there is a rich body of
literature for conventional manufacturing technologies, economic
analysis of AM is scarce. In the following, we will apply the four
key principles to uncover the economic implications of AM at both
the level of the individual firm (monopoly) and in a competitive
market setting at industry level (market structure effects of AM).
Currently, only a limited number of manufacturing firms rely on
AM technology, making it difficult to conduct empirical research.
To further elaborate the economic implications of AM technology,
we therefore chose to examine a number of economic models
identified in a literature review. These models originally discuss
differences between conventional FMS and dedicated manufactur-
ing systems. In the following sections, we apply AM technology’s
key principles on these established models to identify potential
changes in the models’outcomes.
3. AM’s impact on a manufacturing firm’s payoff function
To systematically analyze the effects of AM as addressed by the
four key principles previously identified, we now examine how the
adoption of AM will affect a firm's payoff function. As a theoretical
foundation, we use the payoff function from Milgrom and Roberts'
(1990) model of “modern manufacturing”. The model allows us to
perform a firm level analysis in the situation of a monopoly, i.e. we
only look at one firm and its relationship with its customers.
However, as argued by Milgrom and Roberts (1990), these models
may also serve as a “building block for oligopolistic markets.”We
will extend our discussion to a competitive industry perspective in
the next chapter.
3.1. Milgrom and Roberts'(1990) model of “modern manufacturing”
One of the most cited, seminal articles on FMS and its impact
on the firm is Milgrom and Roberts' (1990) assessment of the
technologies and capabilities required by a firm applying FMS.
They investigate a firm adopting FMS by examining technology,
strategy, and organization. “Modern manufacturing”encompasses
the use of FMS technology and corresponding design techniques
(CAD). The authors examine a single firm in a monopolistic market
setting with a profit function that features key elements crucial for
assessing the value of flexible manufacturing. The company is able
to control a set of decision variables in marketing, design,
manufacturing, engineering and organization, while facing para-
meters related to product demand, operational expenditures,
capital costs, and time. In general, the company’sprofit comes
from its total operating profit less fixed costs, which consist of
expenses for machine setups, product design, and capital expen-
ditures. The analysis of this model finds that a firm can improve
market responsiveness and quality through technological
advancements in manufacturing, but that the successful exploita-
tion of these opportunities can only be realized with a coherent
business strategy. A firm needs to engage in coordinated decisions
along all functions of its value chain to react to a continuously
changing business environment. Moreover, investments in differ-
ent elements of modern manufacturing are mutually reinforcing;
for example, the utility of flexible machinery increases with the
use of digital order processing. Complementarities in marketing,
design, manufacturing, engineering and organization make it
profitable for a firm to increase the level of modern manufacturing
measures after adoption. However, non-convexities in the profit
function would lead to an erratic rather than smooth adaptation
process.
3.2. AM's key principles applied to the payoff function
The payoff function introduced by Milgrom and Roberts (1990)
extensively captures key elements in order to assess the impact of
manufacturing flexibility on a firm. Table 4 summarizes the
elements of the payoff function and illustrates our proposed
impact of AM technology on these parameters. We conducted this
indicative assessment on the basis of the characteristics of AM
identified in Section 2. Regarding the parameter functions, the
signs indicate whether an increase of one parameter has a positive
or negative effect on the function's value, as defined by Milgrom
and Roberts (1990). For example, the expected wait time for a
processed order to be filled (
ω
¼
ω
(m,þr,n)) decreases with an
increasing number of setups per period (m) but increases with a
higher probability of a defective batch (r) given a certain number
of products (n). Table 4 also highlights the potential impact on
decision variables and parameters if AM technology is adopted.
The “þ”or “–” symbols indicate the qualitative directions (posi-
tive/negative effect) in which the payoff function's elements are
influenced by the adaptation of AM technology. The assessment is
based on the AM technology's key principles, which we use to
systematically evaluate each element of the payoff function in
what follows.
Table 3
Key principles of production with AM technology.
Versatile manufacturing machine
On-demand direct digital manufacturing of 3D product models enabled
End products are rapidly available at constant marginal cost (no economies of scale)
Local availability of versatile manufacturing resources with standardized interface
Customization and flexibility for free
Product designs can be customized without cost or time penalties in manufacturing
Volume and product flexibility without cost or time penalties for machine setup or changeover
No tools or molds needed
Complexity for free
Higher design complexity without cost penalty in manufacturing
Little design constraints for products
No cost penalties for higher product variety
Reduction of assembly work
Direct production of functionally integrated parts (e.g., moving parts, cooling system) possible
Fewer production steps involved
Lower manual intervention throughout production processes
C. Weller et al. / Int. J. Production Economics 164 (2015) 43–56 47
Firstly, prices of products (p) can potentially be increased if AM’s
benefit of costless individualization is used to offer highly customized
or functionally optimized products tailored to the customers' needs.
By skimming customers' higher willingness-to-pay for these custo-
mized products (Franke and Piller, 2004; Franke et al., 2009; Fogliatto
et al., 2012), producers will be able to demand a price premium
compared to non-customized products, resulting in an increase of
their rent. From a manufacturing standpoint, AM enables more
frequent product improvements (q), design changes (d) and setups
(m) because no additional setup costs (e,s) are incurred. In addition,
direct processing of 3D data combined with design-to-order or make-
to-order processes helps to reduce total order fulfillment times (a,b,
ω
,t). As long as the relatively low production throughput speed is not
predominant, processing times can be reduced (a). Shorter waiting
times for customers also lead to less demand shrinkage (
δ
). Integrated
order processes might also result in the elimination of inventories (
ι
)
and lower capital costs (
κ
). A fully automated and digitalized produc-
tion process entails less manual work and reduces the probability
of defective batches (r). Furthermore, scrap is negligible (w),
and necessary rework costs (
ρ
), though dependent on quality
requirements, can largely be eliminated when using AM technology.
Lastly, base demand per product (m)canbeincreasedbythe
technology's ability to flexibly produce highly customized products
that meet specific customer needs.
However, AM also bears high material costs and energy
intensity, which will negatively affect marginal costs of production
Table 4
Effects of AM’s key principles on a manufacturer’s payoff function as defined by Milgrom and Roberts (1990).
Elements of payoff function AM’s key principles and effect on elements AM’s limitations
and effect on
elements
Details on AM’s effects
Versatile
mfg.
machine
Customization
and flexibility
for free
Complexity
for free
Reduction
of
assembly
work
pPrice of each product þþ Buyer might be willing to pay more if products are
customized or functionally optimized.
q(Expected) number of
improvements per product per
period
þþ þ Design iterations are not costly, product designs could be
continuously improved without cost penalties.
aOrder receipt and processing
time
þþ þ þ –(throughput
speed)
Direct digital manufacturing with digitized information
flow possible, reducing total order fulfillment time as
long as low throughput speed does not erode processing
time.
bDelivery time þMight be reduced if local production is implemented.
cDirect marginal costs of
production
–(marginal
costs)
Increased due to higher cost of raw material and higher
energy intensity.
dDesign cost per product
improvement
þþþ Design changes easily adopted in digital 3D model
without cost penalty in manufacturing.
eExtra setup costs on newly
changed products
þþ No extra setup costs incurred for changed products
because no tools or molds are required and higher
product variety without cost penalty.
mNumber of setups per period þþ Can be increased without cost penalty because no extra
changeover efforts occur, increase might be necessary if
product variety is larger because of, for example,
customization.
rProbability of a defective batch þ–(quality issues) Can be reduced due to lower manual work requirements,
but currently still quality issues.
sDirect cost of a setup þþ Setup costs are very low (zero) because no tools or molds
are required.
wWastage costs per setup þþ Negligible scrap and rework requirements for
changeovers.
ρMarginal cost of reworking a
defective unit
–(quality issues) Current quality issues might cause efforts for reworking.
ιCost of holding inventory per
unit
No impact (but ideally, less inventory is necessary with
AM).
ωExpected wait for a processed
order to be filled ω¼ω(m,
þr, n)
þþ –(quality issues) Can be reduced because number of setups can be
increased without additional wait time as long as quality
issues are not predominant.
tTotal expected wait for an
order to be received,
processed, filled and shipped
t¼aþωþb
þþ þ þ –(throughput
speed)
May be reduced because setups can be increased without
extra wait time and assembly steps can be eliminated as
long as lower throughput speed does not erode
processing time.
δDemand shrinkage with delay
time δ¼δ(t,τ)
þþ þ þ –(throughput
speed)
Shrinkage is reduced when total expected wait time is
shortened.
mBase demand per productm¼m
(p,þq,n, τ)
þþ þ þ –(quality issues) Might be higher if qcan be increased and individual
customer need is met without extensive increase of
price.
κCapital cost κ¼κ(a,b,c,
d,e,r,s,w,τ)
þþ þ þ –(throughput
speed, marginal
costs, quality
issues)
Lowering of a, b,c,d,e,s, and wdo not necessarily
increase capital costs with AM technology; capital costs
for AM investment are likely to increase with higher-
quality requirements (lowering of r).
nNumber of products þþ –(material
availability, build
space)
Increased if customization strategy is pursued without
cost penalties in manufacturing but constrained by
available materials and build space.
þ¼Positive effect/improvement of parameter, ¼negative effect/deterioration of parameter.
C. Weller et al. / Int. J. Production Economics 164 (2015) 43–5648
(c). Moreover, quality issues still lead to waste in the production
process. Furthermore, constraints regarding available materials,
build space, and limitations in getting access to know-how and
skilled labor control the scope of applications in which AM can
compete against conventional manufacturing methods (Petrovic
et al., 2010; Berman, 2012).
In conclusion, AM technology seems to decouple several dec-
ision variables from capital cost requirements. In contrast to the
assumptions by Milgrom and Roberts (1990), reductions of order
receipt and processing time, delivery time, marginal cost of
production, design costs per product improvement, extra setup
costs on newly changed products, direct setup costs, and wastage
costs per setup do not necessarily lead to a significant increase of
capital costs after the initial investments in AM systems have been
taken. With AM, direct digital manufacturing allows for digital
information flow along the value chain from product design to
production. In combination with a design or make-to-order
process, total fulfillment time can be reduced and inventories
and setup costs eliminated, while marginal costs of production
remain constant after investments in AM technology. Nonetheless,
higher throughput speeds, better quality, and less defective
batches involve higher capital investments for more advanced
AM machines. Fig. 1 summarizes the effects of AM adoption on the
elements of the payoff function, influencing the profitofa
manufacturer in a monopolistic setting when we assume a
moderate increase in marginal production costs, non-excessive
processing times, and a sufficient quality of produced parts.
Concluding, we propose:
Proposition 1. If AM technology unleashes its potential along all
parameters in a manufacturing firm's payoff function, a monopolist
can increase profits by offering customized products with a price
premium at no cost penalties in manufacturing as long as marginal
costs of production, processing time, and the probability of a defective
batch are not excessively high.
This analytically derived result confirms our intuition about the
positive impact of AM on a monopolist's profit. Adjusting some
elements within the model of Milgrom and Roberts (1990)
indicates that AM has profound economic implications for produ-
cers at the firm level in a monopoly situation. However, we do not
believe that AM will lead to monopolistic markets, as represented
by our base model (Milgrom and Roberts, 1990). Nonetheless, this
analysis allowed us to better understand and demonstrate the
impact of AM on different cost parameters at firm level. We
propose that these impacts are also valid under competition which
we investigate in the following section, where we look on the
effects of AM on market structure (industry level perspective). We
find that the firm-level implications outlined above, in particular
with regard to production costs, still hold under competition when
adopting AM.
4. AM's impact on market structure models
In this section, we first introduce dedicated models to discuss
the impact of FMS on market structure. These models are then
further assessed by applying the key principles of AM technology
to identify market structure implications of AM.
4.1. Market structure models assessing FMS
Various scholars have discussed the economic implications of
manufacturing flexibility driven by new technological advance-
ments (e.g., CAD software, automation, numerical control) and
how those technologies have enabled manufactures to utilize
flexibility as a strategic pillar for profit generation (Womack et
al., 1991; Gerwin, 1993). Research has also explored the effects of
flexibility on manufacturing firms and its investment trade-offs in
different dimensions (Sethi and Sethi, 1990; Gerwin, 1993). Flex-
ibility concepts crucial to the competitive position of a firm usually
include product, volume, new product design, and delivery time
flexibility (Suarez et al., 1996). Product flexibility refers to the
ability of a firm to efficiently shift capacity from the production of
one product to another (Sethi and Sethi, 1990). Volume flexibility
describes a firm's ability to easily change production output at
minimum cost penalties (Chen and Adam, 1991). Flexibility in new
product design enables rapid and efficient change of machinery
setups when product specifications are altered (Sethi and Sethi,
1990). Moreover, shortening of delivery times can be the result of
enhanced manufacturing flexibility (Tseng, 2004). For further
analysis, we identified scholarly articles that incorporate a formal,
analytical model with the features of product differentiation and
manufacturing flexibility.
7
In this literature, two main approaches
for assessing the market structure implications of FMS have been
used: (1) product attribute address models and (2) game-theoretic
models of technology choice. We will briefly present the structure
of four core models in these two domains in the following, as
summarized in Table 5, before then turning our attention to the
assessment of AM impact on those models.
4.1.1. Product attribute address models
Product attribute address models, based on Hotelling (1929)
and Salop (1979), conceptualize the market as a circumference
(single sections can be simplified as a line) on which each point
represents one product variant (Eaton and Schmitt, 1994). Each
consumer demands one most preferred variant in this interval;
any deviation causes disutility, which is proportional to the
distance to the most preferred good. Firms can “anchor”their
base product at one point on the interval and serve a certain
market segment by altering the base product at modification costs,
which increase proportionally to the distance to the base product
(Norman and Thisse, 1999).
Eaton and Schmitt (1994) examine the impact of FMS on
market structure and conclude that production flexibility pro-
motes market concentration through preemption and mergers or
cartels. They treat the key characteristics of flexible manufacturing
as “economies of scope in the production of differentiated goods”
(Eaton and Schmitt,1994, p. 875), defining two levels of economies
of scope: strong economies of scope, which exist if a cost-optimal
production plan involves only one base product, and weak econo-
mies of scope, which in a cost-optimal production plan involve
more than one base product but less than the total number of
goods produced. A company can produce a base product when
investing in product development (sunk costs), and the base
+
+ + +
+ / -
--
Fig. 1. AM’s potential effects on a manufacturing firm’s payoff function ( þ/:
increase/decrease).
7
We selected scholarly articles that contain a formal, analytical model using a
query in the EBSCO database. The search query limited the results to articles
containing the terms “flexible manufacturing”and “market structure”or “market
entry”in the title, abstract, or keywords. A few articles (Farrow et al., 1985; Bischof,
1987; Auer and Speckesser, 1997; Eckel, 20 09; Rogalski, 2012) were not further
analyzed because they do not incorporate a formal, analytical model or do not
pertain to manufacturing technology-driven market structure assessments.
C. Weller et al. / Int. J. Production Economics 164 (2015) 43–56 49
product can be produced at constant marginal cost. By modifying
the base product, variants are produced within an attribute space
involving additional costs. First, switching costs arise from chan-
ging the production process from one variant to another. Second,
the production of one unit of a variant requires costs of modifying
the base product. Eaton and Schmitt (1994) model this effect by
setting costs of modification proportional to the distance in the
attribute space between the base product and the variant. Key
assumptions related to FMS in their model are that the product
development of base products incurs sunk costs, incremental costs
of modifying a base product are proportional to the distance to the
base product, and switching costs between different varieties
are zero.
Norman and Thisse (1999) also examine the impact of FMS on
market structure, focusing on strategic aspects of technology
choice. If firms adopt FMS, the market will be characterized by a
tougher price regime in which manufacturers discriminate on
prices of differentiated products. FMS enable the production of
differentiated products that meet individual customer demand,
and the price of one product variant can be adjusted without
affecting other product variants, enabling firms to discriminate on
prices. This means that flexible manufacturers can precisely deter
entry by locally lowering prices when a threat of entry exists. If
firms use FMS to deter entry, it is probable that the number of base
products offered in the market is lowered.
4.1.2. Game-theoretic models for technology choice
A second group of scholars use game theory approaches to exa-
mine multi-stage models that assess the choice of dedicated versus
flexible manufacturing technology (Chang, 1993; Röller and Tombak,
1993). The focus of this approach is on the equilibria results arising
from comparative static analyses. Assumptions about market demand
and a firm's payoff function are made to reflect the necessary trade-
offs in adopting either flexible or dedicated manufacturing technology.
In this school, Chang (1993) treats FMS as a strategic choice for firms
to deter market entry and concludes that under uncertain consumer
preferences, an incumbent threatened by new market entrants is more
likely to install an FMS than an unchallenged monopolist. Thus,
product design flexibility can serve as a preemptive measure to deter
market entrance. The more consumer preferences fluctuate, the higher
the producer’sbenefit from product design flexibility. In contrast, in an
environment characterized by stable consumer tastes, producers strive
to realize economies of scale in the production of identical products –
here, product design flexibility is of no value. Chang assesses the
monopolist’s rationale in a sequential game between the monopolistic
incumbent and a potential entrant. Both have the choice of adopting
either flexible or dedicated manufacturing technology. The model is
structured as follows. In period one, the incumbent decides to invest in
flexible or dedicated manufacturing technology given a known con-
sumer taste. The incumbent earns monopoly profits.Inperiodtwo,a
new consumer taste is revealed. Then the potential entrant, aware of
the incumbent’s technology, decides on its market entrance. Chang
identifies a “zone of strategic flexibility,”determined by the likelihood
of change in consumer demand and the switching costs for changing
production to another product, in which the monopolist holds excess
flexibility when threatened by potential market entrants.
Röller and Tombak (1993) examine the effects of firms investing in
FMS technology on market structure. They analyze a two-stage game.
First, firms decide on their investment in either flexible or dedicated
manufacturing technology. Second, firms choose their output volume
for two differentiated product markets. Manufacturing flexibility of
FMS captures the ability to serve both markets while firms with
dedicated manufacturing technology participate in one market only.
Röller and Tombak assume that fixed costs for investments in FMS
vary across firms but are always higher than or equal to those of
dedicated manufacturing technology. They conclude that FMS tech-
nology tends to be sustained when markets are large and products are
more differentiated. Furthermore, more firms rely on FMS technology
when markets are concentrated or, in other words, when fewer firms
participate in the market. This is a result of inter-segment competition,
enabled by flexible manufacturing technology that allows firms to
serve multiple markets at once. Thus, when firms adopt FMS in highly
concentrated markets, competition is still spurred. As a result, profits
might not be large, even though fewer firms participate in the market,
as highlighted by Röller and Tombak (1993).
Table 5
Literature related to FMS and its impact on market structure.
Reference Type of model Flexibility
concept
Model setup Key conclusions
Eaton and Schmitt (1994) Product attribute address
model
Product
flexibility
Address model for differentiation of product
attributes; consumer preferences for desired
products; sequential market entry with Bertrand
price game
Production flexibility promotes market
concentration through preemption and
mergers or cartels.
Norman and Thisse
(1999)
Product attribute address
model
Product
flexibility
Firms compete in two-stage location–price game
based on spatial model of product differentiation
with mixed technology configuration of firms
(dedicated and flexible mfg. technology)
FMS lead to a tougher price regime in the
market; markets are characterized by a trade-
off: consumers benefit from lower prices at the
expense of the manufacturers, while FMS can
also be used to deter entry, leading to a
reduction of base products in the market.
Chang (1993) Game-theoretic model for
technology choice
New
product
design
flexibility
Sequential model: in period one, the incumbent
decides on its investment in flexible vs. dedicated
mfg. technology given a known consumer taste.
In period two, a new consumer taste is revealed.
Then, the potential entrant decides on its market
entrance
Incumbents threatened by potential entrants
install FMS rather than a monopolist; new
product design flexibility is installed as a
preemptive measure to deter market entrance.
Röller and Tombak (1993) Game-theoretic model for
technology choice
Product
flexibility
Two-stage game: (1) firms decide on mfg.
technology (FMS vs. dedicated) and (2) firms
choose their production output
Higher market concentration, larger markets
and higher product differentiation lead to a
higher proportion of firms adopting FMS.
Higher fixed costs of FMS lead more firms to
invest in dedicated mfg. technology.
Note: Mfg,¼manufacturing.
C. Weller et al. / Int. J. Production Economics 164 (2015) 43–5650
4.2. AM's key principles applied to market structure models
In this section, the key principles of AM, as derived in Section 2,
are applied to the four market structure models outlined before.
Having shown that AM creates significant impacts at firm level in
Section 3, we now focus on the implications for market structure,
when incumbents and entrants can choose to adopt AM (or not) in
order to extend the capabilities of conventional FMS.
4.2.1. Product attribute address models and AM
In Section 2 we argued that AM is a versatile manufacturing
machine that enables direct manufacturing of digital 3D models.
Moreover, AM leads to an increase in manufacturing flexibility,
enabling firms to fulfill individual needs with customized pro-
ducts. The ease of product changeovers without the need for tools
or molds, the potential reduction of required manual intervention,
and the elimination of assembly steps are all factors that lead to
decreased product modification costs. The cost of modifying a base
product is a central element in the two product attribute address
models outlined before (Eaton and Schmitt, 1994; Norman and
Thisse, 1999). As a result, usage of AM technology will likely enable
firms to realize strong economies of scope in product differentia-
tion as defined by Eaton and Schmitt (1994). Thus, the producer
could serve the entire product attribute space with one base
product. Furthermore, sunk costs associated with the development
of new products might be lower with AM. As a versatile manu-
facturing machine capable of transforming a digital 3D model into
a physical product, AM also helps to lower iteration costs during
the development process of a new product variant.
Fig. 2 provides an illustrative example of the effects of AM on
the product attribute address model as described by Eaton and
Schmitt (1994).Itfirst outlines the original model with either
three independent firms or two merged firms jand hon a section
of the product attribute space with three base products X
i
,X
j
and
X
h
. The incentive for mergers arises from the additional profit that
can be earned. Moreover, the figure shows the impact of AM on
market structure. We simulate an entrant with AM technology by
setting the costs of modification to zero while the marginal
production costs of AM are greater than those of FMS and
dedicated manufacturing technology. This entrant is able to cover
the entire product attribute space on entry into the market and
faces one-time sunk costs for the necessary resources and devel-
opment efforts. Production costs for the AM entrant are chosen in
a way that this firm can serve certain market segments. Therefore,
marginal production costs with AM are lower than the costs for
modifying a base product with FMS for some segments within the
product attribute space (see the dark gray areas in Fig. 2). Thus, the
entrant can sell a product in these segments at a price between its
own marginal cost of production and the second-smallest mar-
ginal cost of producing that product.
8
Incentives for entry with AM technology exist if expected
profits exceed the sunk costs for market entry. Upon entrance
with AM, the effects on the market would be as follows. Market
prices decrease because the marginal cost of production with AM
technology is the upper-price barrier in market segments in which
the AM entrant is not the cost leader. Therefore, the incentives for
mergers or cartelization between the incumbents are also reduced
(as illustrated in Fig. 2, the increased profit for merged firms jand
hcannot be achieved with AM entrant). Firms without AM
technology could only increase market shares by producing addi-
tional base products, while abandoning base products can only
lower profits. Incentives for further entrants with AM could exist if
technology advancements lead to lower marginal costs of produc-
tion and, thus, reduced market prices. Any additional AM entrant
with higher production efficiency (lower marginal production
costs) would set the upper-price barrier to the second most
efficient AM producer across the entire market segments. With
decreasing marginal cost of production, other costs become pre-
dominant, particularly supply and delivery costs. For example,
local production close to the buyer’s location could determine cost
advantages if transportation costs for the delivery of the end
product are extensively high and exceed the sourcing costs of
necessary raw materials (Brody and Pureswaran, 2013; Kleer and
Piller, 2013).
Concluding this discussion based on an assessment of Eaton
and Schmitt (1994), we propose:
Proposition 2. Entry of manufacturers using AM technologies will
lead to lower market prices as the entrant lowers the upper price
barrier.
Proposition 3. In a market where many firms have adopted AM, the
competitive position is determined by other costs than manufacturing
costs; in particular, procurement and delivery costs. If delivery costs
for finished goods exceed procurement costs for raw materials, local
production close to the customer becomes beneficial.
When we apply AM technology's characteristics to the model
proposed by Norman and Thisse (1999), the main impact pertains
to the costs of re-anchoring products. AM’s key principles include
“customization and flexibility for free”and “complexity for free”;
thus, AM theoretically allows infinite product variety and product
design changes at no cost penalty in manufacturing. As a result, re-
anchoring costs would become smaller when translated into
Norman’s and Thisse’s terms. The initially discussed trade-off
between price reduction and entry deterrence becomes moot if
AM technology is adopted by market players. The price mechan-
ism follows the results from the example illustrated in Fig. 2, while
entry deterrence with AM technology is only possible until a new
AM entrant begins serving the market at lower marginal produc-
tion costs. Norman and Thisse (1999) also assume that marginal
production costs are equal for dedicated technology and FMS.
However, AM would most likely lead to higher marginal costs of
production than dedicated manufacturing technology or conven-
tional FMS –at least for the base products and its direct
surrounding in the attribute space (until customization/modifica-
tion costs with FMS offset the initial cost advantage compared to
AM). This is also taken into account in the illustrative example of
Fig. 2. Nonetheless, even if some market players adopt AM
technology, there may still be room for single entrants that
produce niche products with dedicated manufacturing technology,
as Norman and Thisse argue.
Concluding, we propose the following effects for the product
attribute address model by Norman and Thisse (1999):
Proposition 4. In a market with high AM penetration, incumbents
lose their strategic advantage of locally cutting prices to deter market
entrance.
As the previous arguments have shown, product attribute
address models value the advantage of AM in flexibly producing
customized products meeting customers' demand within the
attribute space. Nonetheless, certain model restrictions limit the
validity of AM technology's impact. First, these models do not
consider differences in market size and therefore do not include
8
An implicit assumption is that consumers’reservation price is sufficiently
large. Moreover, we assume that the producer’s marginal cost of modifying the
base product is lower than the consumers' marginal disutility of distance to its
most preferred good rotðÞ. As a result, all goods within the attribute space are
offered. We define the consumers' utility as Ux;pxðÞðÞ¼VpxðÞtxx
n
, where V
is the reservation price for the most preferred good x
n
,pxðÞis the market price of
good xand tis the marginal disutility of distance to its most preferred good (Eaton
and Schmitt, 1994).
C. Weller et al. / Int. J. Production Economics 164 (2015) 43–56 51
the effects of decreasing marginal costs with higher production
output. While only negligible economies of scale occur with AM-
based production, per-unit costs with non-AM technology usually
decrease with higher outputs. Fig. 3 illustrates these effects: the
larger the production output, the less competitive AM becomes
when per-unit costs of conventional FMS decrease (the cost curve
for conventional FMS is adjusted downward as illustrated in
Fig. 3a). Thus, at a certain level of economies of scale, profits of
the AM entrant would completely diminish (in the case of ‘high
economies of scale’in Fig. 3a), eliminating incentives for market
entrance with AM. Secondly, product attribute address models do
not consider differences in the output of production technologies
related to the quality of produced parts, order fulfillment times, or
material constraints that may hinder participation in certain
markets. These limitations of AM technology (compared to con-
ventional FMS) could be incorporated into the product attribute
model as a sort of disutility perceived by customers which is
equivalent to the higher production costs of AM. We have
illustrated this effect in Fig. 3b: if requirements for production
quality and time are too high, there are no incentives to adopt AM.
Further, the models do not consider capital costs related to the
acquisition and setup of machinery or working capital require-
ments, which limits comparability between conventional FMS and
AM-based production. This trade-off in technology choice would
need to be resolved at firm level. Here, our discussion of a firm's
payoff function in Section 2 could help to navigate through
potential impacts of an investment in AM technology, supporting
decision-making. Finally, all analysis within product attribute
XiXjXhXiXjXhXiXjXh
: Incumbents’ profit
: Marginal production cost with conventional FMS
: Market price
: Increased profit for merged firms jand h
: AM entrant’s profit
: Profit loss of incumbents
: Marginal production cost with AM
Three independent firms Merged firms j and h
No AM With AM entrant
Fig. 2. Production costs, market price and profits without/with AM entrant in the product attribute address model (for the two figures on the left hand side, “No AM”, see
Eaton and Schmitt, 1994).
XiX
j
Xh
Low economies of scale
X
i
X
j
X
h
Med. economies of scale
XiX
j
Xh
High economies of scale
XiXjXh
Lower quality/time related
downsides of AM
XiXjXh
Med. quality/time related
downsides of AM
XiXjXh
High quality/time related
downsides of AM
: Marginal production cost with AM
: Marginal production cost with conventional FMS
: AM entrant’s profit
: Market price
Fig. 3. (a) Economies of scale and (b) production quality/time related requirements and its impacts on the product attribute address model with AM entrant (own illustration
based on Eaton and Schmitt, 1994).
C. Weller et al. / Int. J. Production Economics 164 (2015) 43–5652
address models is limited to manufacturing costs, while product
modifications obviously still demand efforts in upstream processes
(e.g., product design, testing, customer interaction, order fulfill-
ment; MacCarthy et al. 2003). The additional cost for these
activities may become smaller when using AM, but they still exist.
Again, we refer to our assessment in Section 2 to understand these
interdependencies.
4.2.2. Game-theoretic models for technology choice and AM
We now turn to the effects of AM on the outcome of the multi-
stage game-theoretical models for technology choice by Röller and
Tombak (1993) and Chang (1993). Several common assumptions in
these models do not necessarily hold any longer when firms adopt
AM technology. First, marginal costs of FMS and dedicated
manufacturing technology are assumed to be equal. As discussed
previously, marginal costs of AM technology are frequently higher
than those for conventional FMS and dedicated manufacturing
systems because of relatively expensive raw materials and high
energy needs. Second, these models assume that fixed costs
associated with the implementation of conventional FMS are
larger than those for dedicated manufacturing technology. With
AM, this assumption needs to be assessed in light of market size
and firms’production output. AM's capability of direct digital
manufacturing without a need for up-front investments in
machine setups, tools, or molds should make production of low
volumes less capital intensive. Moreover, multiple markets could
be served at once, allowing firms to distribute their fixed costs
over these markets. Building on Röller and Tombak (1993) we
hence argue that the effective difference in fixed costs between
flexible and dedicated manufacturing technology becomes smaller
when adopting AM. As a result, the proportion of firms with AM
would be larger than the proportion of firms with conventional
FMS when effective fixed costs for AM are lower due to its
distribution over multiple markets. This also means that another
of Röller and Tombak's central arguments finds support if AM
allows participation in multiple markets: in spite of a larger
proportion of firms adopting AM in more concentrated markets,
profits can still decline as a result of increased inter-segment
competition. Concluding this discussion, we propose:
Proposition 5. With higher penetration of AM technology, competi-
tion will increase as AM technology enables manufacturers to offer a
broad product range, resulting in inter-segment competition.
However, there still is a trade-off: economies of scale can lead to
substantially lower marginal costs of production with FMS or dedi-
cated manufacturing technology compared to AM. Therefore, the
larger the production volume for a certain product variant, the less
beneficial AM becomes (as already argued in Section 4.2.1). Thus, in
large markets with little heterogeneity of demand (product differ-
entiation) and high concentration, conventional manufacturing tech-
nologies would remain predominant. Hence, we propose:
Proposition 6. In large markets where heterogeneity of demand is
relatively low, dedicated manufacturing technology that involves
large economies of scale remains superior to employing AM
technology.
As Chang (1993) argues, the adaptation of more flexible
manufacturing technology can also bear a strategic rationale:
incumbents may increase their product design flexibility to deter
market entry. In the “zone of strategic flexibility”, an incumbent
holds excess flexibility when threatened by a potential entrant
(compared to the absence of a potential entrant). In this regard,
AM could help incumbents preserve their market dominance as it
facilitates the adjustment of production output to meet fluctuating
customer demand. This is because it is a versatile manufacturing
machine that does not involve switching costs. Thus, when
adopting AM technology, incumbents face fewer mobility barriers
in serving different market segments (Caves and Porter, 1977). This
in turn signals potential entrants that newly available market
segments can quickly and easily be served by the incumbent,
making entrance less attractive. Moreover, economies of scope in
the incumbents’value chain (i.e., marketing, procurement) may
further facilitate entry deterrence. Hence, we propose:
Proposition 7. In markets characterized by high uncertainty and
fluctuation of product demand, incumbents adopting AM technology
will dominate the competition with new entrants due to economies of
scope in marketing and procurement.
In summary, we show that existing market structure models
can be used to discuss the effects of AM technology, even though
such models were originally developed to discuss investment
decisions in either dedicated manufacturing systems or conven-
tional FMS. Slight adjustments of the models reflecting AM
technology's characteristic as an enhanced flexibility option,
however, lead to significant changes in the outcomes, as described
in Propositions 2–7. However, similar to the product attribute
models, extant game-theoretic models for technology choice do
not cover all aspects relevant for a full assessment of AM. While
they focus their scope on fixed and marginal costs during
manufacturing, economies of scale and costs related to quality
assurance and order fulfillment times are not reflected in the
models' design. Moreover, product differentiation is not thor-
oughly considered. Röller and Tombak (1993) conceptualize man-
ufacturing flexibility as the ability to participate in two markets of
different products, but they do not consider any further product
differentiation with price premiums in meeting specific consumer
tastes. Chang (1993) incorporates fluctuating consumer tastes but
limits scope on inter-temporal aspects. Further research could
build on these shortcomings and could provide an extended model
design. Still, we believe that our assessment of market structure
implications in a manufacturing environment characterized by the
adoption of AM's key principles has demonstrated important
changes and strategic implications for many markets.
5. Conclusion and implications for future research
As Stigler (1939) proposes, flexibility is not a free good. When
demand is certain and uniform products are sufficient, there is no
value in increasing the flexibility of a manufacturing process.
However, if the market environment is characterized by uncer-
tainty, high product variety, or fluctuating customer tastes, those
firms equipped with flexible manufacturing technology may
obtain a competitive advantage. Therefore, identifying the appli-
cations for which the benefits of AM can best be converted into
additional value for producers and customers is crucial. Those
benefits include the application of AM in product development
(rapid prototyping) in order to lower development costs, shorten
time to market, and reduce capital intensity. In manufacturing, AM
offers new dimensions of flexibility to deliver highly customized
products at no cost penalties and even with little manual assembly
work (rapid manufacturing). In more detail, we identified four key
principles of AM that essentially capture its opportunities for
manufacturing, extending the features of conventional FMS. First,
AM is a universal manufacturing machine that can directly trans-
form a digital 3D model into a physical product. Second, customi-
zation and flexibility are free –no tools or molds are required –
while product designs and volumes can be altered without
incurring cost penalties in manufacturing. Third, complexity is
free because additional design complexity and product variety do
not incur additional costs in manufacturing. Fourth, assembly
C. Weller et al. / Int. J. Production Economics 164 (2015) 43–56 53
efforts are potentially reduced when producing functionally inte-
grated products with AM. By matching these principles to the
payoff function by Milgrom and Roberts (1990) we showed that a
monopolist can increase its profits by producing individualized
products at lower costs (while potentially achieving a price
premium). We also analyzed how AM's key principles affect the
underlying assumptions of existing market structure models.
Table 6 summarizes the main effects of AM's key principles on a
manufacturing firm’s payoff function and the market structure of
its industry.
As a result of these opportunities, AM technology has the potential
to disrupt even relatively mature markets. Recall our initial example
of the footwear industry: established firms speed up their product
innovation process or use AM to directly produce shoes with new
design features, while new entrants offer innovative products based
on customization strategies. These entrants face little barriers when
using AM technology, focusing their value proposition on the provi-
sion of customized shoes at little additional manufacturing cost
(Jopson, 2013). More generally, our analysis allows us to point out
markets where AM technology may be adopted first: markets with
overall lower economies of scale. Markets for AM could hence be
characterized by four patterns:
i. Small production output, as typical for prototyping applications
but also many industrial components or especially spare parts
for older product families still in use.
ii. High product complexity, as typical for lightweight construc-
tions in the aerospace or performance car industries (AM
allows the manufacture of mash structures that provide the
same performance effect by largely reduced material usage),
but also for product designs where current production tech-
nologies like molding or milling cannot provide complicated
internal structures such as e.g., cooling chambers.
iii. High demand for product customization tailored to individual
customers'needs, as typical for many medical or dental applica-
tions (implants, prostheses) but also consumer markets like
jewelry or sport performance products.
iv. Spatially remote demand for products, for example the decen-
tralized production of replacement parts in the mining indus-
try or on exploitation platforms of the oil industry.
Our analysis furthermore proposes that markets for AM will be
characterized by a higher degree of flexibility, leading to more
product variants available on this market. Because capital cost
requirements before the start of production are reduced, market
entry barriers become smaller, leading to more contestable mar-
kets overall. Product attribute address models indicate that AM
technology leads to strong economies of scope in product differ-
entiation. A producer equipped with AM is capable of serving the
entire market, once entered. As a result, markets provide greater
product variety while consumers receive their most preferred
goods. Moreover, market prices and incentives for mergers or
cartels are potentially lowered. Game-theoretical models for
technology choice also demonstrate that AM enables firms to
serve multiple market segments; mobility barriers related to the
technology investment become smaller. Manufacturing firms that
increase their flexibility with AM are capable of serving fluctuating
customer preferences while also strengthening their market dom-
inance over time. Both types of market structure models offer
Table 6
Key principles of AM and their potential effects on a manufacturing firm’s payoff function and market structure models.
AM key principles Effects on payoff function Effects on market structure
Payoff function as defined by Milgrom and
Roberts (1990)
Product attribute address
models
Game-theoretic models for technology choice
Versatile manufacturing
machine
Versatility enables firms to produce
customized products on demand
Total capital costs could be lower because
product changes do not involve one-off costs
AM enables strong
economies of scope in
product differentiation
Market can be served along
entire product range once
entered
Multiple, highly differentiated markets can be
served at once; allowing split of fixed costs
Fewer mobility barriers in serving more market
segments may lead to increased competition
Customization and
flexibility for free
Customized products can be offered without
penalties in manufacturing potentially
resulting in price premium
No time penalties for customization resulting
in higher demand per product
High product flexibility
allowing firms to serve
entire market without cost
penalty
Customers can all be served
with most preferred good,
while market prices can be
lowered
Participation in different market segments
enabled, high incentives to increase flexibility
Production is adjustable to fluctuating customer
preferences
Complexity for free
Product improvements and design changes
can be carried out without cost penalties in
manufacturing resulting in higher demand
per product
Higher design complexity without capital
cost impact
Negligible modification and
re-anchoring costs of base
product: higher variety can
be offered
Higher product variety
without additional costs
Incentives to raise a firm’s degree of flexibility
are increased as long as economies of scale do
not lead to massive cost advantages of non-AM
technology
Reduction of assembly
work
Total order processing time can be reduced
Fewer defective batches due to less manual
work requirements
Negligible modification and
re-anchoring costs of base
product
May reduce marginal costs
for production of assembly-
intensive products
May reduce marginal costs for production of
assembly-intensive products
C. Weller et al. / Int. J. Production Economics 164 (2015) 43–5654
insights into the potential effects when firms adopt AM technology
to enhance their manufacturing flexibility.
Nonetheless, the models investigated in our analysis have
certain limitations. First, several cost dimensions are not reflected.
In particular, capital costs for the acquisition of the technology and
up-front investments for setting up production (e.g., tools, molds)
can significantly vary among conventional FMS, dedicated manu-
facturing technology and AM systems. However, the discussion of
different elements in a firm’s payoff function in Section 2 helps to
navigate through those investment decisions. Furthermore, per-
unit production costs depend on a firm’s production output. As a
result of economies of scale, large-scale production tends to be less
costly with conventional FMS and dedicated manufacturing tech-
nology as we identified before. Other costs related to quality and
production time requirements are not incorporated in the original
models either. Nonetheless, our discussion based on slight adjust-
ments of the models’assumptions illustrates AM technology's
impact in this regard. Additionally, the material and other tech-
nological constraints of AM technology may prevent firms from
entering certain product markets. Moreover, firms need to deter-
mine whether they can increase their rent by skimming higher
willingness to pay when meeting consumers’tastes with custo-
mized products before investing in enhanced manufacturing
technology such as AM. Thus, the choice of the right production
technology remains a trade-off that needs to take into account all
cost dimensions, technological constraints and the manufacturing
firm's strategy.
These limitations bear ample opportunities for future research.
Existing market structure models discuss investment decisions as
a choice between dedicated manufacturing technology and FMS,
failing to reflect all characteristics of AM. However, our assessment
in Section 4 indicated that even slight modifications reflecting
AM’s characteristic as a more advanced flexibility option can lead
to significant changes in the outcome of market structure models.
Therefore, it is vital for future research to further refine or develop
market structure models that sufficiently incorporate AM’s char-
acteristics. In particular, these models should better incorporate
the cost dimensions related to the introduction of AM technology,
such as assembly efforts, quality and production throughput
speed. However, technological constraints, such as the available
materials for AM, may also limit relevant industrial markets and
should be taken into account. Nonetheless, current research
activities give reason to presume that AM technology will continue
to gain maturity and larger industry penetration.
As already proposed by Milgrom and Roberts (1990), coordi-
nated, cross-functional decisions along a company's value chain
are necessary to efficiently leverage AM as a new source of modern
manufacturing. Beyond production management, AM has implica-
tions on adjacent research streams like innovation management,
marketing, but also business policy and strategy. This demands an
agenda for multidisciplinary research, including, in particular,
future research which can gather empirical evidence on AM's
impact on producers and markets. It is imperative to better
quantify AM’s implications on production-related costs as well as
on up- and downward transactions along the supply chain.
Empirical work would likely indicate that the effects of AM on
production costs highly depend on distinct applications and use
cases. Second, research should collect empirical evidence on the
technology’s applications and the value added. Important issues
include the investigation of market demands to be fulfilled by AM,
and the measurement of potential price premiums achievable by
the additional design freedom gained by AM. But the question
remains whether consumers are willing at all to engage in a co-
design of 3D product models, given that the greater degree of
design freedom may also come at the cost of higher cognitive
complexity in the design process, leading to confusion and
uncertainty (Merle et al., 2010). We see plenty of opportunities
for research here at the interface between marketing and opera-
tions management. Our research indicates that results are specific
to certain industries and respective demands of customers. Further
research should therefore analyze the impact of additive manu-
facturing in these industries in more detail. Healthcare (i.e., bio-
printing), automotive and aerospace are the first candidates, but
other such as consumer niche markets (e.g., customized jewelry)
as well. The detailed assessment of such ecosystems could help to
direct further research –both concerning technological develop-
ments and business models –in the right direction. Considering
the rapid development in AM technologies and its applications in
recent years, economic implications of such research and devel-
opment activities could be potentially even more significant than
mere AM itself.
Another interesting field demanding research is the ability of
consumers to become their own producers, competing perhaps
even directly with established manufacturers. Consumers get
access to digital product designs and share these. At the same
time, local AM technology allows them to inexpensively transform
these files into tangible products. To what degree does such a
“democratized manufacturing”disrupt existing market structures,
and what are the implications for manufacturing firms? In this
context, research on the implications of AM for intellectual
property rights is also of particular interest for both firms and
policy makers (Wilbanks, 2013; Kurfess and Cass, 2014; Li et al.,
2014). Will the digitalization of production and products be as
disruptive as it was in the content industries (music, film, news-
papers), or will the uniquely complex nature of manufacturing,
and the complementarities of its elements, prevent such a drastic
shift? Could the implementation of “digital rights management”
for digital production secure the current economic interests of
innovating organizations or professionals? And would this not
hinder or even prevent innovation arising from open user com-
munities (“makers”) or lead users (de Jong and de Bruijn, 2013;
Von Hippel, 2005)? Modeling these interactions between tradi-
tional producers and new user manufacturers (Kleer and Piller,
2013) could become an insightful and fascinating field for opera-
tions management and production economics. We hope that the
propositions derived from our theoretical and conceptual work in
this paper provide a first foundation for these endeavors.
Acknowledgements
The authors thank the German Research Foundation (DFG) for
support within the Cluster of Excellence "Integrative Production
Technology for High-Wage Countries" at RWTH Aachen University.
In particular, we thank Christian Hinke and his Selective Laser
Melting team from Fraunhofer Institute for Laser Technology (ILT)
for the fruitful exchange throughout our research. We would also
like to thank the Editor, Bart L. MacCarthy, as well as two
anonymous referees for constructive feedback on earlier versions
of this manuscript.
References
ASTM International, 2013. F2792: Standard Terminology for Additive Manufactur-
ing Technologies.
Atzeni, Eleonora, Salmi, Alessandro, 2012. Economics of additive manufacturing for
end-usable metal parts. Int. J. Adv. Manuf. Technol. 62 (9–12), 1147–1155 .
Auer, Peter, Speckesser, Stefan, 1997. Labour markets and organisational change:
future working structures for an ageing work force. J. Manag. Gov. 1 (2),
177 –206.
Berman, Barry, 2012. 3-D printing: the new industrial revolution. Bus. Horiz. 55 (2),
155 –162 .
Bischof, Bob, 1987. How cummins adopted the flexible approach with AGVs with
AGVs. Manag. Serv. 31 (6), 8–13.
C. Weller et al. / Int. J. Production Economics 164 (2015) 43–56 55
Brabazon, Philip G., MacCarthy, Bart, Woodcock, Andrew, Hawkins, Robert W., 2010.
Mass customization in the automotive industry: comparing interdealer trading
and reconfiguration flexibilities in open fulfilment. Prod. Oper. Manag. 19 (5),
489–502.
Brody, Paul, Pureswaran, Veena, 2013. The new software-defined supply chain. IBM
Institute for Business Value. available online at 〈http: //ibm.com/business/
value/software-defined-supply-chain〉(accessed 15.07.13).
Caves, Richard E., Porter, Michael E., 1977. From entry barriers to mobility barriers:
conjectural decisions and contrived deterrence to new competition. Q. J. Econ.
91 (2), 241–261.
Chamberlin, E.H., 1962. The Theory of Monopolistic Competition: A Re-Orientation
of the Theory of Value. Harvard University Press, Cambridge, MA.
Chang, Myong-Hun, 1993. Flexible manufacturing, uncertain consumer tastes, and
strategic entry deterrence. J. Ind. Econ. 41 (1), 77–90.
Chen, F. Frank, Adam Jr., Everett E., 1991. The impact of flexible manufacturing
systems on productivity and quality. IEEE Trans. Eng. Manag. 38 (1), 33–45.
Davidson, Paul, 2012. 3-D printing could remake U.S. manufacturing. USA Today
October 7, 2012. Available online at 〈http://usatoday30.usatoday.com/money/
industries/manufacturing/story/2012-07-10/digital-manufacturing/56135298/
1〉(accessed 30.04.13.
de Jong, Jerom P., de Bruijn, Eric, 2013. Innovation lessons from 3-D printing. MIT
Sloan Manag. Rev. 54 (2), 43–52.
Dolgui, Alexandre, Proth, Jean-Marie, 2010. Supply Chain Engineering: Useful
Methods and Techniques. Springer, London.
Eaton, B. Curtis, Schmitt, Nicolas, 1994. Flexible manufacturing and market
structure. Am. Econ. Rev. 84 (4), 875–888.
Eckel, Carsten, 2009. Endogenous sunk costs, flexible manufacturing and the
productivity effects of international trade. Scand. J. Econ. 111 (2), 369–386.
Farrow, Scott, Bovbjerg, Randall R., Vaupel, James W., 1985. Flexible manufacturing:
market structure and market failure. J. Policy Anal. Manag. 4 (4), 583–587.
Fogliatto, Flavio S., da Silveira, Giovani J.C., Borenstein, Denis, 2012. The mass
customization decade: an updated review of the literature. Int. J. Prod. Econ.
138 (1), 14–25.
Franke, Nikolaus, Piller, Frank T., 2004. Value creation by toolkits for user
innovation and design: the case of the watch market. J. Prod. Innov. Manag.
21 (6), 401–415.
Franke, Nikolaus, Keinz, Peter, Steger, Christoph J., 2009. Testing the value of
customization: when do customers really prefer products tailored to their
preferences? J. Mark. 73 (5), 103–121.
Garrett, Banning, 2014. 3D Printing: new economic paradigms and strategic shifts.
Glob. Policy 5 (1), 70–75.
Gartner, 2012. Hype Cycle for Emerging Technologies August 2012. Available online
at 〈http://www.gartner.com/newsroom/id/2124315〉(accessed on 30.04.13).
Gartner, 2014. Hype Cycle for Emerging Technologies August 2014. Available online
at 〈http://www.gartner.com/newsroom/id/2819918〉(accessed 23.08.14).
Gebhardt, Andreas, 2003. Rapid Prototyping. Hanser Gardner Publications, Cincin-
nati, OH.
Geroski, Paul A., Pomroy, Richard, 1990. Innovation and the evolution of market
structure. J. Ind. Econ. 38 (3), 299–314.
Gershenfeld, Neil A., 2005. Fab. The Coming Revolution on Your Desktop—From
Personal Computers to Personal Fabrication. Basic Books, New York.
Gerwin, Donald, 1993. Manufacturing flexibility: a strategic perspective. Manag. Sci.
39 (4), 395–410.
Gibson, Ian, Rosen, David W., Stucker, Brent, 2010. Additive Manufacturing
Technologies. Rapid Prototyping to Direct Digital manufacturing. Springer,
New York, London.
Hotelling, Harold, 1929. Stability in competition. Econ. J. 39 (153), 41–57.
Jopson, Barney, 2013. New stamping ground for Nike and Adidas as 3D shoes kick
off. Financial Times June 09, 2013. Available online at 〈http://www.ft.com/cms/
s/0/1d09a66e-d097-11e2-a050-00144feab7de.html〉(accessed 18.06.13).
Kakati, M., 1997. Strategic evaluation of advanced manufacturing technology. Int.
J. Prod. Econ. 53 (2), 141–156.
Khanna, Tarun, 1995. Racing behavior technological evolution in the high-end
computer industry. Res. Policy 24 (6), 933–958.
Kleer, Robin, Piller, Frank T., 2013, Modeling benefits of local production by users:
welfare effects of radical innovation in flexible manufacturing utilizing additive
manufacturing and 3D printing. In: Presented at the 73rd Annual Meeting of
the Academy of Management 2013, Orlando, FL.
Koren, Yoram, 2006. General RMS characteristics. comparison with dedicated and
flexible systems. In: Dashchenko, A.I (Ed.), Reconfigurable Manufacturing
Systems: 21st Century Technologies. Springer, London, pp. 27–45.
Kota, Sridhar, Sethuraman, Kannan, Miller, Raymond, 2000. A metric for evaluating
design commonality in product families. J. Mech. Des. 122 (4), 403.
Kurfess, Thomas, Cass, William J., 2014. Rethinking additive manufacturing and
intellectual property protection. Res. Manag. 57 (5), 35–42.
Li, Phoebe, Mellor, Stephen, Griffin, James, Waelde, Charlotte, Hao, Liang, Everson,
Richard, 2014. Intellectual property and 3D printing: a case study on 3D
chocolate printing. J. Intellect. Prop. Law Pract. 9 (4), 322–332.
Lipson, Hod, Kurman, Melba, 2010, Factory@home: The Emerging Economy of
Personal Fabrication. Report Commissioned by the US Office of Science and
Technology Policy.
Lipson, Hod, Kurman, Melba, 2013. Fabricated. The New World of 3D Printing.
Wiley, Indianapolis, IN.
Lott, Philipp, Schleifenbaum, Henrich, Meiners, Wilhelm, Wissenbach, Konrad,
Hinke, Christian, Bültmann, Jan, 2011. Design of an optical system for the
in situ process monitoring of selective laser melting (SLM). In: Proceedings of
the 6th International WLT Conference on Lasers in Manufacturing, vol. 12(0),
pp. 683–690.
Lux Research, 2013. Led by Auto, Medical and Aerospace, 3D Printing to Grow into
$8.4 Billion Market in 2025. Available online at 〈http://www.luxresearchinc.
com/news-and-events/press-releases/164.html〉(accessed 30.04.13).
MacCarthy, Bart, Brabazon, Philip G., Bramham, Johanna, 2003. Fundamental modes
of operation for mass customization. Int. J. Prod. Econ. 85 (3), 289–304.
Mellor, Stephen, Hao, Liang, Zhang, David, 2014. Additive manufacturing: a frame-
work for implementation. Int. J. Prod. Econ. 149, 194–201.
Mellor, Stephen, 2014. An Implementation Framework for Additive Manufacturing
(Dissertation thesis), University of Exeter.
Merle, Aurélie, Chandon, Jean-Louis, Roux, Elyette, Alizon, Fabrice, 2010. Perceived
value of the mass-customized product and mass customization experience for
individual consumers. Prod. Oper. Manag. 19 (5), 503–514.
Milgrom, Paul, Roberts, John, 1990. The economics of modern manufacturing:
technology, strategy, and organization. Am. Econ. Rev. 80 (3), 511–528.
Mills, David E., Schumann, Laurence, 1985. Industry structure with fluctuating
demand. Am. Econ. Rev. 75 (4), 758–767.
Nicholls, Dale, 2013. 3D printing could revolutionise mining in Australia. 3D Printing
System June 2013. Available online at 〈http://3dprintingsystems.com/3d-prin
ting-could-revolutionise-mining-in-australia〉(accessed 18.06.13).
Norman, George, Thisse, Jacques-François, 1999. Technology choice and market
structure: strategic aspects of flexible manufacturing. J. Ind. Econ. 47 (3),
345–372.
Obama, Barack, 2013. State of the Union 2013 |The White House. Available online at
〈http://www.whitehouse.gov/state-of-the-union-2013〉(accessed 3.03.13).
Pei, Eujin, 2014. 4D printing –revolution or fad? Assembly Autom. 34 (2), 123–127.
Petrovic, Vojislav, Vicente Haro Gonzalez, Juan, Jordá Ferrando, Olga, Delgado
Gordillo, Javier, Ramón Blasco Puchades, Jose, Portolés Griñan, Luis, 2010.
Additive layered manufacturing: sectors of industrial application shown
through case studies. Int. J. Prod. Res. 49 (4), 1061–1079.
Reeves, Philip, 2009. Additive Manufacturing–A Supply Chain Wide Response to
Economic Uncertainty and Environmental Sustainability. Econolyst Ltd, Wirks-
worth, UK.
Rogalski, Sven, 2012. Factory design and process optimisation with flexibility
measurements in industrial production. Int. J. Prod. Res. 50 (21), 6060–6071.
Röller, Lars-Hendrik, Tombak, Mihkel M., 1993. Competition and investment in
flexible technologies. Manag. Sci. 39 (1), 107–114.
Salop, Steven C., 1979. Monopolistic competition with outside goods. Bell J. Econ. 10
(1), 141–156.
Schuh, Günther, Schubert, Johannes, Orilski, Simon, Aghassi, Susanne, 2011.
Technology roadmapping for production in high-wage countries. In: Christian
Brecher (Ed.), Integrative Production Technology for High-Wage Countries.
Springer, New York, pp. 50–64.
Sethi, Andrea Krasa, Sethi, Suresh Pal, 1990. Flexibility in manufacturing: a survey.
Int. J. Flex. Manuf. Syst. 2 (4), 289–328.
Stigler, George, 1939. Production and distribution in the short run. J. Polit. Econ. 47
(3), 305–327.
Stratasys, 2013. Press release: stratasys to acquire MakerBot, Merging Two Global
3D Printing Industry Leaders. Available online at 〈http://investors.stratasys.
com/releasedetail.cfm?ReleaseID=772534〉(accessed 15.07.13).
Suarez, Fernando F., Cusumano, Michael A., Fine, Charles H., 1996. An empirical
study of manufacturing flexibility in printed circuit board assembly. Oper. Res.
44 (1), 223–240.
The Economist (2011): Print me a Stradivarius: how a new manufacturing
technology will change the world. The Economist February 12, 2011, p. 7.
The Economist, 2012. A Third Industrial Revolution. Special Report: Manufacturing
and Innovation. The Economist April 21, 2012.
Tseng, Mei-Chiun, 2004. Strategic choice of flexible manufacturing technologies.
Int. J. Prod. Econ. 91 (3), 223–227.
Tseng, Mitchell M., Jiao, Jianxin, Merchant, M. Eugene, 1996. Design for mass
customization. CIRP Ann. - Manuf. Technol. 45 (1), 153–156 .
Vickers, John, 1986. The evolution of market structure when there is a sequence of
innovations. J. Ind. Econ. 35 (1), 1–12.
Von Hippel, Eric, 2005. Democratizing Innovation. MIT Press, Cambridge, MA.
Wilbanks, Kelsey B., 2013. The challenges of 3D printing to the repair-
reconstruction doctrine in patent law. George Mason Law Rev. 20, 4.
Wohlers, Terry T., 2013. Wohlers Report 2013. Additive Manufacturing and 3D
printing State of the Industry: Annual Worldwide Progress Report. Wohlers
Associates, Fort Collins, CO.
Womack, James P., Jones, Daniel T., Roos, Daniel (1991): The Machine that Changed
the World. New York, NY, HarperPerennial.
C. Weller et al. / Int. J. Production Economics 164 (2015) 43–5656
