Computers in Industry 82 (2016) 57–68

Contents lists available at ScienceDirect

Computers in Industry
journal homepage: www.elsevier.com/locate/compind

Establishing a maturity model for design automation in sales-delivery

processes of ETO products
Olga Willnera,* , Jonathan Goslingb , Paul Schönslebena
a
ETH Zurich, Department of Management, Technology, and Economics, BWI Center for Industrial Management, Weinbergstrasse 56/58, 8092 Zurich,
Switzerland
b
Cardiff University, Cardiff Business School, Logistics Systems Dynamics Group, Aberconway Building, Cardiff, CF10 3EU, UK

A R T I C L E I N F O A B S T R A C T

Article history:
Received 7 September 2015 Short delivery times are considered a competitive advantage in the engineer-to-order (ETO) sector.
Received in revised form 24 April 2016 Design-related tasks contribute to a substantial amount of delivery times and costs since ETO products
Accepted 3 May 2016 have to be either fully developed or adapted to customer specifications within tendering or order
Available online 10 June 2016 fulfillment. Approaches aiming at a computerised automation of tasks related to the design process, often
termed design automation or knowledge-based engineering, are generally regarded as an effective
Keywords: means to achieve lead time and cost reductions while maintaining, or even improving product quality. In
Engineer-to-order this study we propose a maturity model as a framework for analyzing and improving such activities in
Design automation
ETO companies. We contribute to the literature in being the first to investigate design automation in the
Knowledge-based engineering
ETO sector from a maturity perspective. Beyond that, we extend the extant literature on design
Product configuration
Maturity model automation, which is of a highly technical nature, by providing a framework considering organizational
and managerial aspects. The findings indicate that five different levels of maturity can be achieved across
the dimensions strategies, processes, systems, and people. Empirical cases give insight into these
different levels. Our investigation draws from extant literature and a comparative case study involving
four companies over two years.
ã 2016 Elsevier B.V. All rights reserved.

1. Introduction introduction of sales configurators, which constitutes as an

element of design automation, contributes to better on-time
Fast and cost-efficient tendering and order execution processes delivery, a decrease in personnel efforts and quality improvements
are considered as sources of competitive advantage in the along both product and process dimensions [11,12].
engineer-to-order (ETO) sector [1–3]. Since ETO products either While technical aspects of design automation (e.g. system
have to be fully developed or adapted to customer specifications architecture, product modelling) are well researched [9,13–15],
within tendering or order fulfillment [4,5], design-related tasks studies related to organizational and managerial requirements of
contribute to a substantial amount of delivery lead times and costs. design automation are hardly available [16,17]. Researchers
Approaches aiming at computerised automation of tasks related to particularly emphasize the need for a framework guiding the
the design process, often termed design automation or knowledge- design automation process and supporting the identification of
based engineering (KBE), are generally regarded as an effective design automation opportunities [7,16]. More specifically,
means to achieve lead time and cost reductions while maintaining, Cederfeldt and Elgh [16] in a sample of eleven ETO manufacturers
or even improving product quality [6–8]. For example, case studies identified scope of implementation (e.g. implementation of sales
conducted by Raffaeli et al. [9] and Frank et al. [10] found that configurators, engineering configurators, CAD systems, or spread-
design automation based on integrating product configurators and sheet macros) and how far to push the automation level as topics
CAD systems may result in a reduction of the engineering time by requiring additional research. Well-established concepts associat-
up to 90%. Empirical evidence further suggests that the ed with maturity models are relevant to these issues, but the
review presented later in the paper shows that these have not been
adequately adapted to either design automation or ETO situations.
Beyond the shortcomings identified in the literature, discussions
* Corresponding author.
with company representatives brought to light that managers are
E-mail addresses: owillner@ethz.ch, olga.willner@gmx.de (O. Willner),
GoslingJ@cardiff.ac.uk (J. Gosling), pschoensleben@ethz.ch (P. Schönsleben).

http://dx.doi.org/10.1016/j.compind.2016.05.003
0166-3615/ã 2016 Elsevier B.V. All rights reserved.
58 O. Willner et al. / Computers in Industry 82 (2016) 57–68

often uncertain which steps to take in approaching design Design automation is predominantly seen as an approach for
automation. minimizing the effort required for repetitive design tasks [7,13,16].
To fill this gap, the present paper examines the following However, engineering ETO products encompasses the execution of
research question: What stages do ETO companies undergo in both repetitive and creative design tasks. Consequently, product
automating their design processes and how can we describe them? structures distinguishing between components that already exist
We base our investigation on a comparative case study with four and therefore can be reused in a repetitive manner and
ETO manufacturers from the mechanical engineering sector. The components that have to be engineered for a particular order
concept of the maturity model was selected to guide the are a prerequisite [28,29]. A review of the literature shows that
investigation due to its suitability for describing organizational various terminology has been applied to break down the structures
development paths [18,19] and supporting transformation of ETO products. A proliferation of terms from the design literature
processes [20]. seek to describe ways of responding to the challenge of configuring
This study contributes to the literature in being the first to and designing to customer order. Examples include modular
investigate design automation in the ETO sector from a maturity design [30], platform designs [31], and configuration design [32].
perspective. Beyond that, it extends the extant literature on design Jiao et al. [33] show the considerable range of terms that have
automation, which is of a highly technical nature, by providing a emerged. Further, the terms ‘common features’, ‘base product’
framework considering organizational and managerial aspects. It [28], ‘fixed components’ [34] and ‘standard parts and modules’
further provides companies with a step-wise guideline on how to [35] have been proposed to describe the standard components of
approach design automation in sales-delivery processes as a an ETO product. The terms ‘parameterized features’, ‘reused
means to foster a competitive advantage. Following Verhagen’s [7] variants’ [28], ‘configurable components’ [34] and ‘generic product
call for research, we further suggest that our maturity model can be structure’ [35] all describe its configurable components. To
used as an instrument for assessing design automation oppor- describe the components that are truly engineered for a specific
tunities. customer order the terms ‘special features’, ‘new components’
This paper is structured as follows. Section 2 provides an [28], ‘special components’ [34], ‘parts which are developed based
overview of the related work and state of the art. Section 3 on norms and standards’ [35] and ‘white spots’ [36] can all be
describes our methodological approach and introduces the found. In this paper, we use the terms standard components,
empirical setting in which we conducted our research. In Section 4, configurable components and special components to distinguish
a maturity model for design automation is conceptually drafted, between the different components of ETO products.
thereafter empirically refined through a comparative case study
and finally validated. Lastly, the conclusion section highlights the 2.2. Design automation
theoretical and managerial implications and proposes opportu-
nities for further research. The term design automation has its origins in the electronics
sector where it has been used since the early 1970s to describe the
2. Related work and state of the art automated design of circuits and electronics chips [37,38]. More
recently, the term has increasingly been applied when referring to
2.1. Engineer-to-order the automation of design-related tasks in the field of mechanical
engineering [6,7,10,13,16,39]. There exists no general consensus on
A number of papers have sought to define and categorize ETO the definition of design automation in the literature (see
situations, as well as give insight into their complex nature. Gosling [7,10,16,29,40]). In this paper we apply the definition of design
and Naim [5] define an ETO supply chain where production is automation by Cederfeldt and Elgh [16] as ‘computerised automa-
customized for each order and where the customer penetrates into tion of tasks that are related to the design process through the
the design phase, often operating in project specific environments. implementation of information and knowledge in tools or systems.’
Since ETO products either have to be fully developed or adapted to A broad range of literature related to the techical aspects of
customer specifications [4,5], engineering tasks have to be design automation exists (see Elgh [6] for a detailed review),
conducted within tendering or order execution. This can lead to whereas literature discussing the organizational and managerial
a range of co-ordination issues in terms of integrating engineering aspects is scarce. Both Elgh [6] and Cederfeldt [16] give
and production [21]. recommendations for planning design automation in ETO compa-
The ETO sector encompasses a broad range of industries, nies. While Elgh [6] proposes an information model for design
including mechanical engineering, construction, and ship-build- automation in quotation preparation, Cederfeldt [16] conducts a
ing. A number of ETO archetypes may also be identified, based on study with ETO manufacturers on the need and perceived potential
volume and the amount of order specific engineering work to be for a design automation framework. In describing the move from
performed [22]. Customers in this challenging sector often wish for ETO to mass customization, Haug et al. [41] identify five
lead times to be short and are not willing to pay high price dimensions (product variety, customer view, manufacturing costs,
premiums [23–25]. Hence, companies that operate in an ETO business purpose, configurator challenge) which they regard as
environment face the difficult prospect of undertaking order- relevant for deciding to what extent to standardize and automate.
driven design and engineering activities while customers wait Scholars in our field of study regard design automation for ETO
impatiently, often making last minute requests for changes. This as highly similar to KBE [7,42,43] or respectively regard KBE as one
leads to unpredictable work flows, ‘rush jobs’, out-of-date of its core sub-disciplines [10]. Typically, the automation of design
information, and distorted delivery dates [26]. processes for highly customized products is seen to encompass
From an engineering design perspective, ETO might be developing and implementing the following IT applications: sales
considered as the extent to which orders penetrate the scientif- configurators [11,44,45], engineering or technical configurators
ic-technical flow of design activities [27]. Hence, we might [10,41,44–46], as well as the linking of those with CAD systems
consider a spectrum between pure ‘engineer-to-stock’, where [9,10]. Although product lifecycle management (PLM) systems are
designs are held in stock, to pure ETO, where new designs must be generally regarded as enablers for sharing product data along
developed [2]. Despite this continuum being well recognized, the entire supply chains or product lifecycles [47], there exists no
appropriate design approach along it has not been addressed consensus on how well these systems are equiped to cope with the
comprehensively in the ETO literature. challenges the ETO environment presents. While Hicks and
 O. Willner et al. / Computers in Industry 82 (2016) 57–68 59

McGovern [48] found that some functionalities of PLM systems are have been made (see [18,61–63]). The proposed guidelines for MM
applicable for ETO products, it still remains to be determined how development include a problem identification phase in which
big their overall value is when lifecycles are short and volumes low. purpose and scope of the model are determined, a model
An empirical study conducted in the shipbuilding sector confirms development phase in which model and assessment instruments
that PLM systems have been designed with predominantly are defined, and an implementation and validation phase in which
assemble-to-order (ATO) and make-to-order (MTO) enivronments the model is evaluated based on empirical cases.
in mind and require adaptions for a successful implementation in We present an overview of the extant maturity-related
ETO environments [49]. Additionally, Hani et al. [50] report that literature in the realms of ETO and design automation in Table 1.
PLM systems do not sufficiently suport the reuse of design process There exists general consensus that MMs can contribute to an
knowledge through identifying appropriate workflows within analysis of the ETO environment but require some tailoring to
previous projects. unlock their full potential [48,64,65]. In none of the papers did we
Literature describing how standard and configurable compo- find such a tailoring. Tiihonen and Soininen [66] conducted a
nents, which are characteristic for MTO products, can be stored in survey on methods, practices, and tools supporting product
IT applications and later retrieved for reuse abounds (see Zhang configuration tasks. They conclude that companies can be at
[51] for a review). However, these approaches neglect the special different stages regarding the use of product configurators and
requirements of the ETO environment, such as the execution of propose the MM as instrument for understanding and improving
creative design tasks for the development of order-specific configuration processes. Cederfeldt and Elgh [16] and Cederfeldt
solutions. Silventoinen et al. [52] conducted an entire study [17] conducted empirical studies in the field of design automation.
exploring and classifying the factors hindering an information They associate potential for design automation with a company’s
reuse in ETO companies. In describing the ETO situation, McGovern degree of product and process maturity. Making the link between
et al. [53] state that a limited reuse of engineering designs is not ETO and design automation is not within the scope of any of the
uncommon. They further refer to anecdotal evidence highlighting reviewed papers.
that designers appreciate the task of developing new designs.
Further, Brière-Côté et al. [28] report that project-specific data 3. Methodology
tends to be regarded as transient and is therefore often not linked
to the lifecycle of the product family. In our literature review, we As outlined in the previous section, the literature proposes a
could identify first attempts targeting the design automation variety of guidelines and frameworks for developing maturity
challenges characteristic for the ETO environment. Brière-Côté models. We decided to use the four-step guideline for MM
et al. [28] propose a product structure concept systematically development introduced in Neff et al. [20], which is rooted in the
promoting the reuse of order-specific solutions. Kristianto et al. procedure model developed by Becker et al. [18]. As presented in
[54] develop a system level configurator that processes incomplete Fig. 1, we slightly adjusted the guideline to make it more applicable
configurations and engineering changes. to our specific research setting.
In the ETO environment, design automation can be applied Prior to MM development, the relevance of the problem that the
either for the generation of conceptual new designs as part of new model is meant to address has to be demonstrated, and the target
product development or in later project stages, such as tendering group of the model should be defined (step 1: problem identifica-
and order execution, for the development of detailed designs tion). As presented in Sections 1 and 2, both empirical evidence
linked to specific customer projects. In the following, we refer to gained from preliminary interviews with company representatives
design activites conducted within tendering and order execution as well as an initial literature review revealed that the automation
that are linked to customer projects as ‘order-specific engineering’. of design processes is crucial to enhancing the competitiveness of
This paper investigates design automation in sales-delivery ETO manufacturers. Yet both the extant literature as well as
processes while design automation in new product development empirical insights obtained from company representatives con-
is not within its scope. Our main rationale for excluding design firmed that there is a lack of established frameworks or guidelines
automation in new product development from this investigation is assisting ETO companies in automating their design processes.
that design automation in this phase encompasses very similar According to Becker et al. [18], the need for a new MM must be
challenges for a broad variety of product types. On the other hand, confirmed by an analysis of the existing models (step 2: comparison
the ETO environment has very unique requirements for design of existing MMs). We conducted a structured literature review to
automation within tendering and order execution. identify the MMs predominant in our field of research. As search
terms we used ‘maturity model’ combined with ‘engineer-to-
2.3. Maturity models order’, ‘design-to-order’, ‘design automation’ or ‘product configu-
ration’. Major databases, such as Science Direct, Emerald, Pro
The Oxford English Dictionary describes maturity as the state of Quest, and Google Scholar, were used to search for related works.
being complete, perfect, or ready [55]. Maturity models (MMs) are Since we were unable to identify any domain-specific MMs within
widely applied tools for assessing the maturity of organizations our field of research, we choose to broaden our research scope to
and provide a framework for process improvements or bench- maturity-related literature within the realms of ETO and design
marks [19]. They usually consist of a series of stages representing automation. Based on a content check, we determined which
an anticipated, desired, or logical organizational evolution path publications to consider relevant with respect to our research
[18] with the bottom stage describing a very low degree of interest. Within the relevant papers, we conducted backward and
maturity and the highest degree of maturity located at the top. forward searches with the objective of detecting additional
Besides generic MMs, which are suitable for a very broad field of material. In total, we identified six publications (see Section 2,
applications, such as the Quality Management Maturity Grid [56], Table 1) that we analyzed in detail.
the Capability Maturity Model [57], or the Capability Maturity MMs should be developed iteratively (step 3: iterative model
Model Integration (CMMI) [58], models explicitly focusing on development). Our approach consisted of two iterations. In the first
narrower defined domains can be found in the literature (e.g. iteration, we conceptually developed our a-priori model based on
[20,59,60]). the requirements we had previously derived from both the
In recent years, efforts to generalize the MM development literature review and preliminary interviews with company
process aiming at a theoretically sound and replicable MM design representatives. In the second iteration, we empirically refined
60 O. Willner et al. / Computers in Industry 82 (2016) 57–68

Table 1
Maturity-related literature in the realms of ETO and design automation.

Author Title Research design and method Contents Contribution to maturity-

related aspects
ETO
Veldmann & Klingenberg Applicability of the Empirical (single case study)  evaluation of applica-  concludes that CMMI
[64] capability maturity model bility of CMMI on ETO has to be enhanced to
for engineer-to-order companies become applicable for
firms ETO companies (e.g. lo-
gistics, construction
and maintenance are
not sufficiently covered)

Hicks & McGovern [48] Product life cycle Conceptual  analysis of characteris-  states that MM is a
management in engineer- tics of ETO companies suitable tool for man-
to-order industries (e.g. markets, products, aging the ETO life cycle
internal processes and
supply chains)

Kärkkäinen & Myllärniemi Maturity assessment for Empirical (multiple case study)  analysis of the potential  PLM maturity assess-
[65] implementing and using of a PLM maturity as- ment based on an
product lifecycle sessment in ETO com- existing PLM MM is
management in project- panies conducted
oriented engineering
companies
 outlines that generic
MMs require tailoring
to become applicable in
ETO settings

Design automation
Tiihonen et al. [66] State-of-the-practice in Empirical (survey with 10 companies)  empirical study on  states that companies
product configuration – a methods, practices and are at different levels of
survey of 10 cases in the tools that support maturity in respect to
Finnish industry product configuration product configuration
tasks

 proposes MM as (1)
suitable tool for asses-
sing configuration pro-
cesses and product data
management and (2)
improvement roadmap
for product configura-
tion

Cederfeldt & Elgh [16] Design automation in Empirical (questionnaire with 11 companies)  empirical study identi-  focus on process matu-
SMEs – current state, fying the perceived po- rity (see Cederfeldt
potential, need and tential for, current state 2007)
requirements of, and requirements for
design automation from
a SME perspective

Cederfeldt [17] Planning design Empirical (questionnaire/multiple case study)  development of a  focus on process and
automation – a structured structured method for product maturity:
method and supporting planning design auto- attributes the potential
tools mation for design automation
to a company’s degree
of product and process
maturity

 definition of product
maturity as ‘known
rules in relation to all
rules’

 definition of process
maturity as ‘level of task
and knowledge formal-
ization’

the model by means of a comparative case study with four ETO priori model to illustrate the study scope and to provide our case
manufacturers (see Table 2). At each of the companies, we study partners with a framework that allowed them to describe
conducted targeted interviews following an interview guideline their path towards design automation in a structured and
(see Appendix A). As part of the interviews, we introduced our a- comparable manner. We recorded all interviews and later reduced
 O. Willner et al. / Computers in Industry 82 (2016) 57–68 61

1 2 3 4
Problem Comparing of Iterative model
Model validation
identification existing MMs development

• Problem identification and • Identification and evaluation of • Iteration 1: Conceptual • Model validation with design
motivation existing MMs development of a-priori model automation experts
Performed
activities

• Definition of target group • Iteration 2: Empirical refinement • Self-assessment of ETO

of MM by the means of a manufacturer
comparative case study
Conceptual

• Initial literature review (ETO, • Structured literature review on • Triangulation of existing literature
Used techniques

design automation, maturity maturity-related literature within (ETO, design automation,

models) the realms of ETO and design maturity models) with empirical
automation insights

• Preliminary interviews with • Iteration 1: Preliminary interviews • Focus group workshops

Empirical

company representatives with company representatives • Self-assessment

• Iteration 2: Targeted interviews
with 4 ETO manufacturers

• Section 1 (Introduction) • Section 2 (Related work and • Section 4.1 (Development of the • Section 4.3 (Model validation)
Output & relation

• Section 2 (Related work and state of the art) a-priori model) & Section 4.2 • Figure 4 (Maturity model)
to paper

state of the art) • Table 1 (Maturity-related (Model elaboration and

literature in the realms of ETO refinement)
and design automation) • Figure 3 (A-priori maturity model)
• Figure 4 (Maturity model)

Fig. 1. Guideline for maturity model development (based on [18,20]).

their contents into categories along our five-level analysis frame, structures are always established before sales or even engineering
which contributes to both within-case and cross-case analysis [67]. configurators are introduced.
By doing so, we were able to identify common maturity paths As emphasized in Wendler [19], the development of a
across the companies. For example, our data showed that product meaningful and useful MM should conclude with model validation

Table 2
Case study companies.

Company Corporate Number of Turnover Number Department in Preliminary interviewsc Targeted Supplementary data
division employeesa in million of units charge of order- (1st round of interviews) interviewsd
sa sold specific engineering (2nd round of
(incl. number and interviews)
qualification of
employees)
ALPHA Environmental 2000 >250 200b Engineering (45 full- – 2 interview –
simulation time employees: participants
engineering degree (Technical
from universities of Director; Head
applied sciences or of Control
vocational training) Engineering);
5 h in total
BETA Turbomachine 160 >100 10a Engineering (75 full- 1 interview participant 1 interview Participation in
time employees: 90% (Director of participant company meetings
with university Engineering); 3 h in total (Director of (>50 h in total),
degree in Engineering); process mappings,
mechanical/ 3 h in total company data
electrical
engineering; 10%
with vocational
training)
GAMMA Asphalt mixing n/a n/a 2001 Product 4 interview participants 1 interview Participation in
plant development and (Director of participant company meetings
engineering (n/a; Development Core Parts, (Product with various
engineering degrees Technical Manager Paver, Manager); 3 h company
from universities/ Technical Director China, in total representatives
universities of Technical Director Italy); (>90 h in total,
applied sciences or 9 h in total process mappings,
vocational training) company data
DELTA High-rise 250 n/a 2000a Application 4 interview participants 1 interview Participation in
elevator Engineering (n/a; (Manager Engineering participant company meetings
bachelor degree in Switzerland, Manager (Director with various
mechanical/ Engineering China, Product Line company
electrical Director Product Line Management); representatives
engineering or Management, 3 h in total (>150 h in total),
vocational training) Engineering Director); process mappings,
12 h in total company data
a
Figures of 2014 for the division.
b
Figures of 2014 for the site.
c
Conducted in the period from 07/2013–03/2014.
d
Conducted in the period from 09/2014–12/2014.
62 O. Willner et al. / Computers in Industry 82 (2016) 57–68

Table 3
Participants of focus group and self-assessment workshops.

Company Corporate Number of Turnover Self-assessment workshopsb Focus group workshopsb

division employeesa in sa
EPSILON Design 9 n/a – 3 participants (Managing Director, Head of Design
(industry-oriented research Automation Automation Division; Research Engineer); 4 h in total
firm specialized in design
automation)
ZETA Corporate 6000 n/a – 3 participants (Program Manager Modularization,
(large corporation offering a Technology Program Manager Product Portfolio Management,
broad range of ETO products) Researcher); 2 h in total
ETA Elevator 200 40 1 participant (Director Product –
Development and Engineering);
3 h in total
a
Figures of 2014 for the division.
b
Conducted in 05/2015.

(step 4: model validation). As shown in Table 3, our approach for purposes. The three categories ‘strategies’, ‘processes’ and
model validation was twofold: First, we conducted focus group ‘systems’ proposed in Österle [69] were initially applied as
workshops with design automation experts. Second, we requested dimensions. We opted for developing a multi-dimensional instead
a company that had not participated in the model development to of a one-dimensional model. The results obtained from multi-
conduct a self-assessment with our model. Based on the workshop dimensional models are much more suited to letting organizations
results, we further adjusted and refined the model. gain awareness of their strengths and weaknesses and providing
guidance for improvements [63]. Later, the model was extended by
4. maturity model for design automation the ‘people’ dimension following De Bruin and Rosemann [70]
since empirical evidence gained in the first round of interviews
As described in the methodology section, we selected an revealed that the mindset and abilities of employees have a strong
iterative approach for developing the maturity model. This section impact on the level of design automation a company can achieve.
describes how initially the a-priori model was designed, thereafter To communicate our understanding of design automation to the
empirically refined with multiple case studies and finally case study partners, we predefined the two extremes of the model.
validated. We believe that an alternative could have been the As shown in Fig. 2, level 1 implies that effectively no standardiza-
development of a stage gate model [68] for design automation. tion and design automation has been put into practice. The
However, stage gate models are mainly applied in the context of customer is free to define the specifications of his order since the
new product development, and the conventionally used stages are solution space is completely open. Processes are ad-hoc, and
not entirely suitable for describing sales and order execution hardly any systems supporting tendering and order execution are
processes in the ETO environment. available. Level 5 is characterized by specified and implemented
processes and systems that allow full automation of the tendering
4.1. Development of the a-priori model and order execution processes. Since a fixed solution space is
regarded as a prerequisite for a full automation [46], we argue that
As a starting point, we developed a rough a-priori model (see in practice only fully configurable products (MTO) can reach level
Fig. 2). For the a-priori model, we drew from concepts underlying 5. By definition, the solution space of an ETO product has to remain
CMMI [58] to define the different levels of maturity. As the at least partially open and therefore the maturity of an ETO
literature shows the CMMI is a very popular foundation for the organization can at most converge towards level 5.
development of new maturity models (according to Wendler’s
mapping study [19] 75% of established maturity models are based 4.2. Model elaboration and refinement
on the CMMI). An alternative would have been the use of the stages
proposed in the Quality Management Maturity Grid [56]. However, To empirically elaborate and refine the a-priori model into a
we considered the terms used to describe the stages in that model full-scale maturity model, a comparative case study involving four
such as “awakening” or “enlightening” not as appropriate for our ETO manufacturers was conducted. The investigated products

Fig. 2. A-priori maturity model.

 O. Willner et al. / Computers in Industry 82 (2016) 57–68 63

(testing chamber, turbomachine, asphalt mixing plant, high-rise components and the task of executing the engineering for
elevator) of all four participating manufacturers have been on the individual orders.
market for more than 30 years and can therefore be considered
mature and well-established. All four companies serve both 4.2.2. Company BETA
developed, mainly Central Europe, as well as emerging markets, BETA is a large multinational corporation that participated in
particularly China. Since our cases demonstrate very similar the case study with one of its turbomachine divisions. The division
degrees in product and market maturity, we believe that they are was founded less than 10 years ago and shows characteristics of a
not suitable for investigating the impact of product and market start-up (e.g. high growth rate, low formalization and routinization
maturity on design automation. Instead, our unit of analysis is the of processes, no established product portfolio). In 2014, the
corporate division and our study investigates ‘what stages ETO division received orders for ten turbomachines, each requiring
companies undergo in automating their design processes’. First, we 11,500 h of order-specific development and engineering on
present the four empirical cases individually. Second, we aggregate average.
our findings by means of a cross-case analysis and from there Level 1—ultimate freedom
elaborate and refine the model. The turbomachine R&D department was founded in 2008. In its
beginnings, very limited customer intelligence that could be used
4.2.1. Company ALPHA for a delimitation of the solution space was available. Product
ALPHA participated in the case study with its site producing structures were not fully defined and processes were ad hoc,
special testing chambers, part of the environmental simulation partially inefficient, and redundant. A large number of design
division. The division develops and produces testing chambers in iterations and subsequent design reviews were required for each
five countries at seven different locations. In 2014, the site order.
participating in our study built 200 special testing chambers, each Level 2—product standardization
requiring 500 h of engineering on average. Initially, BETA structured its machine types into different
Level 1—ultimate freedom performance clusters and defined standardized components
For a long time, the management at ALPHA regarded testing covering the clusters. When asked for his motivation for product
chambers as one-of-a-kind products and made no efforts towards standardization, the Director of Engineering at BETA explained:
standardization and automation. A consistent product structure ‘Beyond a reduction in costs and lead times, standardized product
did not exist, and both engineering and production departments structures allows us to compare the prices of purchased parts and
frequently customized products during order execution. Most bundle orders for parts of a similar or identical design. Plus, I believe
employees had the mindset of craftsmen and enjoyed following that consistent product structures are a prerequisite for automation.’
their own processes and ideas when engineering products. He also reported: ‘Even today, our product portfolio is by far not
Engineers generally preferred to design everything from scratch complete. Our current strategy is to participate in tenders for a large
instead of using existing solutions. Plus, they were often not array of different machine sizes and application types. Obviously, it
aware of the order-specific solutions their colleagues have takes more time to engineer a “first-of-its-kind” since the number of
developed in the past since no proper database with search engineering hours required decrease with experience. However, it
functionalities existed. A systematic retrieval and reuse of similar helps us in broadening our knowledge and product base. If you have
projects and/or components tends was almost impossible. seen many different variants of a product, it becomes easier to develop
Consequently, the company had problems with costs, quality, modular product structures allowing a reuse of components for many
and lead times. different orders.’
Level 2—product standardization (today) Level 3—automation of tendering (today)
In 2010, the top management at ALPHA changed and it became a In its third year of business, BETA introduced sales configurators
core objective of the new management team to increase the to support an automated generation of tender documents. Most
profitability of the division. The Technical Director reported that an recently, the commercial product structures stored in the
essential step towards this objective was the definition of a configurators were remodeled to allow cost calculations for
consistent product structure. He explained: ‘Many of our projects different production stages instead of only the final turbomachine.
did not really require order-specific engineering. Instead, a well- As a manager of BETA explained: ‘I believe the remodeling of the
elaborated, modular product structure would have allowed a frequent product structures considerably increased our data quality. The newly
reuse of components.’ available data improves the accuracy and speed of the cost
When asked for the expected benefits of product standardiza- calculations that we execute in tendering.’
tion, he explained: ‘We expected a standardization to result in cost Outlook
and lead time reductions as well as quality improvements. It was also As a result of the standardization and automation, the
supposed to allow us to build the exact same products at different management at BETA expects revenues to grow disproportionately
locations.’ He then continued: ‘Today, we still have some difficulties to the number of people employed in the future. The management
with the new product structures. It takes our engineers more time to considers it key to further improve the product structures and
combine our new templates for standard components instead of extend the product portfolio. As a manager explained: ‘If we
simply using old projects and adapting them. However, this should not manage to improve our product structures, the use of pre-engineered
be an issue anymore once our product structures have been properly solutions will become feasible and we will be able to advance our level
implemented in a configurator.’ of design automation. Today, by far too many calculations have to be
Outlook done for each order. A major advantage would be to have more design
At the time of investigation, ALPHA stored its product guidelines. They would avoid that calculations have to be repeated for
structures in an ERP system. It is expected that sales might need every order to confirm the feasibility of the design.’
a configurator to support tendering in the future. The Technical
Director further reported that some departments might require 4.2.3. Company GAMMA
restructuring due to the product standardization. While today GAMMA is a construction equipment producer that participat-
ALPHA has a large department solely responsible for the order- ed in the case study with its division developing and producing
specific engineering, in the future ALPHA will have to asphalt mixing plants. In recent years, the division expanded its
distinguish between the task of defining standard/configurable global operations by opening new development and production
64 O. Willner et al. / Computers in Industry 82 (2016) 57–68

sites abroad. In 2014, the company sold 200 asphalt mixing plants, tendering and order execution, and processes were only roughly
each requiring 1400 h of order-specific engineering on average. defined.
Level 1—ultimate freedom Level 2—product standardization
Initially, processes were only roughly defined and bill-of- Faced with growing competition, the management at DELTA
materials were often incomplete or not fully specified. Tenders and came to realize that customers regarded their products as very
orders were handled according to the understanding and expensive and the delivery times as too long. A manager of DELTA
knowledge of individuals. A product manager of the division stated: ‘That is why we defined our first product lines. We started with
described the level of automation at that time as follows: ‘I believe the very top segments and then slowly worked our way down. Initially,
that automation only happened in the mind of people. Some of us product lines were noted down on paper. We also defined index price
automated processes for ourselves.’ lists.’
Level 2—product standardization Level 3—automation of tendering
In 2009, GAMMA launched ‘Project Optima’, which aimed at In 2005, DELTA introduced the first sales configurators to speed
reducing costs and lead times. The reductions were to be achieved up tendering. A manager of DELTA emphasized: ‘The introduction of
by a concise definition of the technical product structure, sales configurators led to new processes and the organization required
accompanied by a guideline explaining how the new product restructuring. For example, we split up the responsibilities between
structure was to be used. As a manager explained: ‘As a result of new product development and order-specific engineering. Further, we
Optima it wasn’t possible to order parts by simply describing them pushed sales to sell the pre-engineered solutions specified in the
anymore. Instead, material numbers had to be specified. Before configurator.’ The manager also expressed: ‘Sales configurators
Optima our engineering had to confirm every single order. Optima helped collect and prepare data that helped us decide what else we
achieved that orders not requiring special parts could go straight into could standardize. Another advantage of the configurator was that
work preparation.’ everybody started doing everything right or wrong in the exact same
Level 3—automation of tendering (today) way.’
GAMMA uses sales configurators for the generation of tender Level 4—automation of order execution (today)
documents. However, the commercial product structures stored in In the next step, it was decided that product specifications
the sales configurators are not coherently linked with the technical should no longer be copied manually from tendering documents
product structures stored in the ERP system and used for order after an order had been won. Instead, the configurators, originally
execution. To date, no interface between the two systems exists. conceived for the generation of tendering documents were to be
Component groups are manually copied into the ERP system after extended for use in order execution. Parameters selected within
an order has been won. Custom-built software for the configura- tendering were to be used to automatically generate engineering
tion of core parts is scattered throughout the engineering drawings and purchase orders later on. Just the special
department. Since most of the solutions are complex and require components not included in the fixed solution space should be
a certain expertise, the tool developers and their close peers calculated and designed manually by the department in charge of
primarily use them. order-specific engineering. Additionally, a database for storing
Outlook order-specific engineering requests with search functions allow-
In its quest for global market presence, GAMMA seeks to ing the retrieval and reuse of engineering solutions from previous
advance its current level of standardization and automation to projects was introduced. As a director expressed when discussing
improve operations efficiency. A major challenge related this the changes: ‘Processes had to be redesigned again, and calculation
aspect is the fact that the division conducts the order-specific rules had to be validated. In the beginning, it was difficult for some of
engineering at five different locations and that each location stores our engineers to trust in the automated order process. Previously, our
their order-specific solutions locally. In the past, the engineering engineers had calculated safety margins based on their individual
sites in China and India have already worked on highly similar experiences. Now, we had intense debates if the tolerances and rules
order-specific solutions simultaneously and only realized this after proposed by the systems were correct.’
project end. Company representatives unanimously expressed that Outlook
they regard further automation of order execution as the next step. DELTA does not intend to advance its current level of design
At time of the investigation, the division faced the challenge of automation in the future. The division considers the capability to
identifying the product families for which automation promised deliver products that are partly engineered to customer specifi-
the highest savings. cations as a core order winner. A new release of the configurators
expected to go-live in 2017 primarily targets performance
4.2.4. Company DELTA improvements and a simplification of the solution space.
DELTA participated in the case study with its division Fig. 3 illustrates the design automation paths of the four case
delivering high-rise elevators. The division, which designs and companies with the key milestones.
produces elevators for particularly high and often extremely
challenging buildings, is known for its innovativeness and strong 4.3. Model validation
global market presence. In 2014, the division sold 2000 elevators,
each requiring eleven hours of order-specific engineering on Model validation was based on two focus group workshops and
average. In merely requiring eleven hours of order-specific a self-assessment. The participants of all three validation rounds
engineering on average, elevators are not the most extreme type generally confirmed the selected levels and dimensions and agreed
of ETO (see Willner et al. [22] for an analysis of different ETO upon the proposed design automation paths.
types). We gained the following insights from the focus group
Level 1—ultimate freedom workshops. First, workshop participants at EPSILON expressed
Until the early 1990s, DELTA engineered every high-rise doubts that the tendering phase necessarily has to be automated
elevator basically from scratch. As a director pointed out: ‘At that before automation of the order execution can take place. We came
time, every single order required engineering. We had not yet to the conclusion that certain engineering subtasks (e.g. related to
discussed which components could be pre-engineered and which particular modules or components) can be automated without
should be engineered-to-order. We simply accepted orders the way having automated tendering but not the full order execution.
they came in.’ The division hardly used supporting IT systems for Therefore, we slightly altered the wording used to describe level 3
 O. Willner et al. / Computers in Industry 82 (2016) 57–68 65

Automated generation of drawings &

Sales Configurators purchase orders
Product lines & index price lists Split-up PD & ENG Engineering database
DELTA
Level 1 Level 2 Level 3 Level 4

Sales configurators
Custom-built software for the
GAMMA Project Optima engineering of core parts

Level 1 Level 2
Level Level 3

Standard components
for performance
BETA clusters Sales configurators

Level 2 Level 3

Modular product structure

ALPHA
Level 1 Level 2
Year
1995 2000 2005 2010
PD: Product development
ENG: Engineering

Fig. 3. Design automation paths of case companies..

and 4 in the model. Second, workshop participants at ZETA in the specific projects. Formalized knowledge sharing processes and
proposed to incorporate industry-specific factors as stage indica- systems are not yet fully developed in our company.’
tors in the model. While we generally agree that this might Our study participants unanimously confirmed that the model
increase the usefulness of the model for managers, we regard an delivers meaningful and applicable insights. A participant
elaboration of this issue as out of scope for our research question. expressed that he intends use to the maturity model to discuss
When discussing the maturity models at the focus group work- the next steps required for automation with the upper manage-
shops, it also emerged that managers should not necessarily ment. The managing director of one of the validation partners
attempt to advance all their products to Level 5, in which case they intends to apply the model in design automation projects at
would become MTO products. In line with Willner et al. [22], we customer sites.
argue that it depends on the product type which degree of design
automation is most appropriate. 4.4. Summary and discussion
As part of the self-assessment, the Engineering Director at ETA
noted: ‘I consider my division to be currently located at level 2 aiming Fig. 4 presents the maturity model that we derived from within-
towards moving on to level 3. In that respect, I regard it as a major case analysis combined with cross-case comparisons. It comprises
obstacle that the information and knowledge gathered in previous five distinct maturity levels (ultimate freedom, product standardi-
projects is primarily accessible to the engineers having been involved zation, automation of tendering, automation of order execution, full

Fig. 4. Maturity model for design automation.

66 O. Willner et al. / Computers in Industry 82 (2016) 57–68

automation) that are delimited by the criteria that a change of process (called product line management at ALPHA), the ones
activities has taken place through all four dimensions (e.g. an handling the order execution (called work preparation at BETA),
overall level 3 is achieved only when a level 3 or higher is achieved and the ones who improvise the ETO (called application engineer-
across all four dimensions). We used a bottom-up approach for ing at ALPHA). In level 4, emergent routines (defined by Nelson and
developing the distinct maturity levels in determining the required Winter [72] as patterns of action that store tacit knowledge and
activities first and then recorded the appropriate names that reflect function as organizational memory) contribute to automated order
these. According to De Bruin et al. [63] such a bottom-up approach execution for standard/configurable components. We use the term
should be used for the development of maturity models in more routinized improvisation (defined by Tan [73] as repeated
established domains. improvisation that entails simultaneous planning and execution)
Along the strategies dimension, the four cases supported us in to describe how special components, which are often character-
identifying the steps required to develop a solution space ized by a high degree of novelty and complexity, are handled
promoting design automation. In that context, our case studies efficiently and consistently.
brought to light that mature product structures are an important
prerequisite for successful design automation. Companies have to 5. Conclusion
distinguish between standard, configurable, and special compo-
nents to reach level 2 in this dimension. Advancing to level 3 and 4 This paper proposes a maturity model as a framework for
entails formalizing the solution space through the implementation analyzing and improving design automation activities in ETO
of product structures in configurators. Level 5 requires a fixed companies. Through integrating evidence from literature, case
solution space, meaning that a product is fully configurable and studies, and focus group workshops, we identified five distinct
does not contain any special components. maturity stages across the dimensions strategies, processes, systems
Along the processes dimension, we observed that processes and people. Empirical cases gave insight in the activities happening
evolve together with strategies and systems. In level 2, companies at the different stages and allowed us to describe them in detail.
start to develop nascent processes and replicate these across Our investigation makes a number of contributions to the
locations. Distinct processes for standard/configurable and special literature. First, we bring together several literature streams, which
components are required for advancing to level 3. Processes for have formerly been disconnected, in investigating design automa-
standard/configurable components are fully defined in level 4 tion in the ETO sector from a maturity perspective. Beyond that, we
while meta-processes (higher-order processes used to construct extend the extant literature on design automation by providing a
other processes [71]) exist for special components. In our view, the framework that takes organizational and managerial aspects into
concept of the meta-process is closely linked to the ETO-enabling account. Second, our cases revealed that design automation is not
process introduced in Schönsleben [24] and based on the exclusively applicable to repetitive design tasks but also supports
capability of routinized improvisation (see people dimension). In creative tasks. Through identifying this additional opportunity for
level 5, all processes are fully defined and coordinated. design automation, we augment previous research in our field. Third,
Along the systems dimension, the case studies helped to we adopted the concepts of routines and routinized improvisation
determine which IT systems to implement in which order for from the field of organizational studies to understand how tacit
design automation. In level 2, product structures are stored in a knowledge can be incorporated in ETO processes. We believe that
large variety of IT applications, which are not necessarily suitable additional studies applying these concepts on the operational
for handling complex and hierarchical product structures challenges of the ETO sector might yield promising results.
coherently. Beyond serving as data repositories for both part Managers can apply the model as a guideline on how to
numbers as well as bill-of-materials, PDM/PLM systems do not approach design automation in sales-delivery processes. This
play a big role in the sales-delivery process of our case should help them reduce the time and effort required for design-
companies. Some of them use PLM systems in product related tasks leading to competitive advantage. We argue that the
development but we could not identify a single case where a model also supports the assessment of design automation
PLM system is used as leading system along the entire product opportunities. In its current form, managers can use the model
lifecycle. In level 3 and 4, configurators with interfaces to CAD to determine where they stand today and what the next steps
systems are implemented to enable the automation of repetitive should be. As the validation rounds brought up, future research
design tasks for standard/configurable components. Correspond- could seek to develop stage indicators that help assess which
ingly, we noticed that engineering databases are set up to degree of design automation should ultimately be targeted in a
facilitate the reuse of special components and order-specific particular line of business.
solutions. Contrary to the common notion that design automa- This study has only begun to explore the organizational and
tion is mainly applicable for repetitive design tasks (e.g. managerial requirements of design automation in the ETO sector.
[7,13,16]), the cases studies demonstrated that creative design The maturity model for design automation was developed with
tasks can also benefit from design automation. Company cases from the mechanical engineering industry. Future inves-
representatives at DELTA reported how their engineers deliber- tigations may wish to assess the applicability of the model in a
ately retrieve former projects stored in an engineering database broader range of industries and identify industry-specific adap-
and use them as inspiration for creating new order-specific tions the model might require. For example, we believe that an
solutions. In level 5, fully integrated IT systems for tendering and application in the construction industry might make a particularly
order execution are in place. interesting case allowing a comparison of the similarities in
Along the people dimension, we found that the required skill requirements between design automation and building informa-
sets and behaviors of people change with automation. While tion modelling.
success initially depends on individual skills and ‘heroic’ perfor-
mance, the importance of collective effort and a comprehensive Acknowledgments
integration of tasks and roles later on gains momentum. As
demonstrated by the cases, moving to level 3 requires the This research is funded by the research project FastETO (CTI no.
formation of groups and specialization. The empirical cases 15021.2 PFES-ES). The authors would like to thank all organizations
demonstrates how it is distinguished between the people in participating in the case studies and validation rounds for sharing
charge of developing the solution space and defining the MTO their insights in the field of design automation in the ETO sector.
 O. Willner et al. / Computers in Industry 82 (2016) 57–68 67

Appendix A. Interview guideline [8] A. Haug, L. Hvam, N.H. Mortensen, Definition and evaluation of product
configurator development strategies, Comput. Ind. 63 (2012) 471–481, doi:
http://dx.doi.org/10.1016/j.compind.2012.02.001.
General information [9] R. Raffaeli, M. Mengoni, M. Germani, Improving the link between computer-
assisted design and configuration tools for the design of mechanical products,
1. Interviewee Information (name, position, in the position since Artif. Intell. Eng. Des. Anal. Manuf. 27 (2013) 51–64, doi:http://dx.doi.org/
10.1017/S0890060412000388.
when). [10] G. Frank, D. Entner, T. Prante, M. Schwarz, Towards a generic framework of
2. Division information (name, main products, # employees, engineering design automation for creating complex CAD models, Int. J. Adv.
annual revenue). Syst. Meas. 7 (2014) 179–192.
[11] L. Hvam, S. Pape, M.K. Nielsen, Improving the quotation process with product
configuration, Comput. Ind. 57 (2006) 607–621, doi:http://dx.doi.org/10.1016/
j.compind.2005.10.001.
Engineer-to-order [12] L. Hvam, N. Mortensen, J. Riis, Product Customization, Springer, Berlin, 2008.
[13] K. Amadori, M. Tarkian, J. Ölvander, P. Krus, Flexible and robust CAD models for
design automation, Adv. Eng. Inf. 26 (2012) 180–195, doi:http://dx.doi.org/
1. How does your division define ETO? 10.1016/j.aei.2012.01.004.
2. Name the different ETO products of your division. Estimate how [14] I. Rask, Rule-based Product Development—Report 1, Industrial Research and
many units of each product are sold annually and how many Development Corporation, Mölndal, 1998.
[15] M. Stokes, Managing Engineering Knowledge: MOKA—Methodology for
order-specific engineering hours are required per unit on the Knowledge Based Engineering Applications, ASME Press, London, 2001.
average. [16] M. Cederfeldt, F. Elgh, Design automation in SMEs-Current state, potential,
3. Describe the ETO processes of your division (product develop- need and requirements, ICED 05 15th Int. Conf. Eng. Des. Eng. Des. Glob. Econ.,
Melbourne, 2005. (accessed July 8, 2014) http://search.informit.com.au/
ment, sales, customer-specific engineering, production & documentSummary;dn=390095181685561;res=IELENG.
logistics, delivery). Which departments are involved in each [17] M. Cederfeldt, Planning Design Automation, PhD Dissertation, Jönköping
of the process phases? University, Sweden, 2007.
[18] J. Becker, R. Knackstedt, J. Pöppelbuß, Developing maturity models for IT
4. How do you expect your share of ETO products to develop management, Bus. Inf. Syst. Eng. 1 (2009) 213–222, doi:http://dx.doi.org/
within the next 10 years? 10.1007/s12599-009-0044-5.
[19] R. Wendler, The maturity of maturity model research: a systematic mapping
study, Inf. Softw. Technol. 54 (2012) 1317–1339.
[20] A. Neff, F. Hamel, T. Herz, F. Uebernickel, W. Brenner, J. vom Brocke, Developing
Design automation a maturity model for service systems in heavy equipment manufacturing
enterprises, Inf. Manage. (2014) 1–17, doi:http://dx.doi.org/10.1016/j.
im.2014.05.001.
1. What does a standardization and automation of design
[21] M.H. Mello, J.O. Strandhagen, E. Alfnes, Analyzing the factors affecting
processes imply for your division? coordination in engineer-to-order supply chain, Int. J. Oper. Prod. Manage. 35
2. What are the main drivers for design automation in your (2015) 1005–1031, doi:http://dx.doi.org/10.1108/02683940010305270.
division? [22] O. Willner, D. Powell, M. Gerschberger, P. Schönsleben, Exploring the
archetypes of engineer-to-order: an empirical analysis, Int. J. Oper. Prod.
3. Does your division attempt to achieve different degrees of Manage. 3 (2016), doi:http://dx.doi.org/10.1108/ijopm-07-2014-0339.
automation for different types of ETO products? If this is the [23] O. Willner, D. Powell, A. Duchi, P. Schönsleben, Globally distributed
case, which criteria do you apply to decide to which degree to engineering processes: making the distinction between engineer-to-order and
make-to-order, Procedia CIRP 17 (2014) 663–668, doi:http://dx.doi.org/
automate for which product type? 10.1016/j.procir.2014.02.054.
4. Please describe the current status of design automation in your [24] P. Schönsleben, Methods and tools that support a fast and efficient design-to-
division. Which elements of your design processes are order process for parameterized product families, CIRP Ann. Technol. 61 (2012)
179–182.
automated? [25] Y. Tu, Production planning and control in a virtual One-of-a-Kind Production
5. Describe the pathway of your design automation along the company, Comput. Ind. 34 (1997) 271–283, doi:http://dx.doi.org/10.1016/
dimensions ‘strategies’, ‘processes’, ‘systems’ and ‘people’. S0166-3615(97)00046-8.
[26] J. Gosling, D.R. Towill, M.M. Naim, A.R.J. Dainty, Principles for the design and
6. Which data did you require for design automation? operation of engineer-to-order supply chains in the construction sector, Prod.
7. How far are your product structures currently developed? Plan. Control. (2014) 1–16, doi:http://dx.doi.org/10.1080/
8. Which challenges did you encounter during design automation? 09537287.2014.880816.
[27] J.R. Dixon, Design Engineering: Inventiveness, Analysis, and Decision Making,
9. What is the future design automation path of your division?
McGraw-Hill, New York NY, 1966.
[28] A. Brière-Côté, L. Rivest, A. Desrochers, Adaptive generic product structure
modelling for design reuse in engineer-to-order products, Comput. Ind. 61
(2010) 53–65, doi:http://dx.doi.org/10.1016/j.compind.2009.07.005.
[29] W.J.C. Verhagen, B. de Vrught, J. Schut, R. Curran, A method for identification of
References automation potential through modelling of engineering processes and
quantification of information waste, Adv. Eng. Inf. (2015), doi:http://dx.doi.
[1] C. Bozarth, S. Chapman, A contingency view of time-based competition for org/10.1016/j.aei.2015.03.003.
manufacturers, Int. J. Oper. Prod. Manage. 16 (1996) 56–67, doi:http://dx.doi. [30] C.Y. Baldwin, K.B. Clark, Design Rules: the Power of Modularity, MIT, Press,
org/10.1108/01443579610119090. Cambridge MA, 2000.
[2] J. Wikner, M. Rudberg, Integrating production and engineering perspectives on [31] T.W. Simpson, J.R.A. Maier, F. Mistree, Produt platform design: method and
the customer order decoupling point, Int. J. Oper. Prod. Manage. 25 (2005) application, Res. Eng. Des. 13 (2001) 2–22.
623–641, doi:http://dx.doi.org/10.1108/01443570510605072. [32] B.J. Wielinga, A.T. Schreiber, Configuration design problem solving, IEEE Expert
[3] A. Trentin, E. Perin, C. Forza, Overcoming the customization-responsiveness 12 (1997) 49–56. http://citeseerx.ist.psu.edu/viewdoc/summary?
squeeze by using product configurators: beyond anecdotal evidence, Comput. doi=10.1.1.1.2118.
Ind. 62 (2011) 260–268. [33] J. Jiao, T.W. Simpson, Z. Siddique, Product family design and platform-based
[4] G. Amaro, L. Hendry, B. Kingsman, Competitive advantage, customisation and a product development: a state-of-the-art review, J. Intell. Manuf. 18 (2007) 5–
new taxonomy for non make-to-stock companies, Int. J. Oper. Prod. Manage. 19 29, doi:http://dx.doi.org/10.1007/s10845-007-0003-2.
(1999) 349–371, doi:http://dx.doi.org/10.1108/01443579910254213. [34] C. Lutz, Rechnergestütztes Konfigurieren Und Auslegen Individualisierter
[5] J. Gosling, M.M. Naim, Engineer-to-order supply chain management: a Produkte, PhD Dissertation, Vienna University of Technology, Austria, 2011.
literature review and research agenda, Int. J. Prod. Econ. 122 (2009) 741–754, [35] F. Elgh, Supporting management and maintenance of manufacturing
doi:http://dx.doi.org/10.1016/j.ijpe.2009.07.002. knowledge in design automation systems, Adv. Eng. Inf. 22 (2008) 445–456,
[6] F. Elgh, Decision support in the quotation process of engineered-to-order doi:http://dx.doi.org/10.1016/j.aei.2008.05.004.
products, Adv. Eng. Inf. 26 (2012) 66–79, doi:http://dx.doi.org/10.1016/j. [36] H. Wortmann, Comparison of information systems for engineer-to-order and
aei.2011.07.001. make-to-stock situations, Comput. Ind. 26 (1995) 261–271, doi:http://dx.doi.
[7] W.J.C. Verhagen, P. Bermell-Garcia, R.E.C. van Dijk, R. Curran, A critical review org/10.1016/0166-3615(95)00047-8.
of knowledge-based engineering: an identification of research challenges, Adv. [37] D. Jansen, The Electronic Design Automation Handbook, Kluwer Academic
Eng. Inf. 26 (2012) 5–15, doi:http://dx.doi.org/10.1016/j.aei.2011.06.004. Publishers, Boston MA, 2003.
68 O. Willner et al. / Computers in Industry 82 (2016) 57–68

[38] D. Macmillen, R. Camposano, D. Hill, T. Williams, An industrial view of on Information Systems (2005) 8–19. (accessed 03.03.15) http://eprints.qut.
electronic design automation, IEEE Trans. Comput. Des. Integr. Circuits Syst. 19 edu.au/25152/.
(2000) 1428–1448. [64] J. Veldman, W. Klingenberg, Applicability of the capability maturity model for
[39] T. Tomiyama, Intelligent computer-aided design systems: past 20 years and engineer-to-order firms, Int. J. Technol. Manage. 48 (2009) 219–239.
future 20 years, AI EDAM Artif. Intell. Eng. Des. Anal. Manuf. 21 (2007) 27–29, [65] H. Kärkkäinen, J. Myllärniemi, Maturity assessment for implementing and
doi:http://dx.doi.org/10.1017/S0890060407070114. using product lifecycle management in project–oriented engineering
[40] F. Elgh, Supporting management and analysis of quotations in a design companies, Int. J. Electron. Bus. 11 (2014) 176–198 http://www.
automation approach to customization, New World Situation: New Directions inderscienceonline.com/doi/abs/10.1504/IJEB.2014.060218?journalCode=ijeb
in Concurrent Engineering, vol. 46, Springer, London, 2010, pp. 361–368, doi: (accessed July 17, 2014).
http://dx.doi.org/10.1007/978-0-85729-024-3. [66] J. Tiihonen, T. Soininen, State of the practice in product configuration—a survey
[41] A. Haug, K. Ladeby, K. Edwards, From engineer-to-order to mass customization, of 10 cases in the finnish industry, in: S.F.T. Tomiyama, M. Mäntylä (Eds.),
Manage. Res. News 32 (2009) 633–644, doi:http://dx.doi.org/10.1108/ Knowledge Intensive CAD—1, Chapman-Hall, London, 1996.
01409170910965233. [67] M.B. Miles, M. Huberman, Qualitative Data Analysis: an Expanded Sourcebook,
[42] P. Bermell-Garcia, W.J.C. Verhagen, S. Astwood, K. Krishnamurthy, J.L. Johnson, Sage, Thousand Oaks, CA, 1994.
D. Ruiz, et al., A framework for management of knowledge-based engineering [68] R.G. Cooper, Stage-gate systems: a new tool for managing new products, Bus.
applications as software services: enabling personalization and codification, Horiz. 33 (1990) 44–54, doi:http://dx.doi.org/10.1016/0007-6813(90)90040-I.
Adv. Eng. Inf. 26 (2012) 219–230, doi:http://dx.doi.org/10.1016/j. [69] H. Österle, Business in the Information Age: Heading for New Processes,
aei.2012.01.006. Springer, Heidelberg, 2010.
[43] C. van der Velden, C. Bil, X. Xu, Adaptable methodology for automation [70] T. De Bruin, M. Rosemann, Towards a business process management maturity
application development, Adv. Eng. Inf. 26 (2012) 231–250, doi:http://dx.doi. model, ECIS 2005 Proceeding Thirteen European Conference on Information
org/10.1016/j.aei.2012.02.007. Systems (2005). (accessed 18.08.14) http://eprints.qut.edu.au/25194/.
[44] T. Olofsson, P. Jensen, A. Rönneblad, Configuration and design automation of [71] Collette Rolland, A comprehensive view of process engineering, Advanced
industrialised building systems, Proceeding CIB W78 27th International Information Systems Engineering, Springer, Berlin, 1998, pp. 1–24.
Conference (2010). [72] R.R. Nelson, S.G. Winter, An Evolutionary Theory of Economic Change, Harvard
[45] C. Forza, F. Salvador, Managing for variety in the order acquisition and University Press, Cambridge, MA, 1982, doi:http://dx.doi.org/10.2307/
fulfilment process: the contribution of product configuration systems, Int. J. 2232409.
Prod. Econ. 76 (2002) 87–98. [73] B. Tan, S.L. Pan, T.-C. Chou, J.-Y. Huang, Enabling agility through routinized
[46] M. Mäkipää, P. Paunu, T. Ingalsuo, Utilization of design configurators in order improvisation in IT deployment: the case of chang chun petrochemicals, ICIS
engineering, Int. J. Ind. Eng. Manage. 3 (2012) 223–231. 2010 Proceeding (2010). http://aisel.aisnet.org/icis2010_submissions/225.
[47] A. Silventoinen, H.J. Pels, H. Kärkkäinen, H. Lampela, Towards future PLM
maturity assessment dimensions, Proceeding 8th Intenational Conference Dr Olga Willner is a Senior Researcher at the BWI Center
Conference on Product Lifecycle Management (2011) 480–492. for Industrial Management at ETH Zurich. Her research
[48] C. Hicks, T. McGovern, Product life cycle management in engineer-to-order agenda centers on the engineer-to-order environment
industries, Int. J. Technol. Manage. 48 (2009) 153–167, doi:http://dx.doi.org/ with a particular focus on design automation and the
10.1504/IJTM.2009.024913. organization of a global engineering. During her Ph.D.
[49] J.H. Lee, J. Lee, Features of data management in PLM customised for ship design studies, she was founder and project leader of the KTI-
adopting engineering to order strategy, Int. J. Prod. Lifecycle Manage. 7 (2014) financed project FastETO, which was conducted jointly
292–317, doi:http://dx.doi.org/10.1504/IJPLM.2014.066818. between ETH Zurich and three major Swiss industrial
[50] M. El Hani, L. Rivest, R. Maranzana, Product data reuse in product firms. Prior to her Ph.D., Olga worked as a management
development: a practitioner’s perspective, Product Lifecycle Management. consultant in the field of supply chain management and
Towards Knowledge-rich Enterprises, Springer, Berlin Heidelberg, 2016. logistics. Olga holds a master’s degree in Industrial
[51] L.L. Zhang, Product configuration: a review of the state-of-the-art and future Engineering and Management with majors in logistics
research, Int. J. Prod. Res. 52 (2014) 6381–6398, doi:http://dx.doi.org/10.1080/ and strategic management from the Technical University
00207543.2014.942012. of Berlin as well as a Postgraduate Certificate in International Studies from the
[52] A. Silventoinen, A. Denger, H. Lampela, J. Papinniemi, Challenges of University of Sydney.
information reuse in customer-oriented engineering networks, Int. J. Inf.
Manage. 34 (2014) 720–732, doi:http://dx.doi.org/10.1016/j.
ijinfomgt.2014.07.001. Dr Jon Gosling is a Senior Lecturer in Supply Chain
[53] T. McGovern, C. Hicks, C. Earl, Modelling supply chain management processes Management in the Logistics and Operations Manage-
in engineer-to-order companies, Int. J. Logist. Res. Appl. A Lead. J. Supply Chain ment Section at Cardiff Business School, teaching
Manage. 2 (1999) 147–159. purchasing, supply chain management and operations
[54] Y. Kristianto, P. Helo, R.J. Jiao, A system level product configurator for engineer- management. Prior to becoming an academic, he worked
to-order supply chains, Comput. Ind. 72 (2015) 82–91, doi:http://dx.doi.org/ as a supply chain analyst for Unipart/Jaguar. He works
10.1016/j.compind.2015.04.004. closely with a range of organizations to better understand
[55] J.A. Simpson, E.S.C. Weiner, The Oxford English Dictionary, University Press, appropriate strategies for engineer-to-order supply
Oxford UK, 1989. chains, and his work has been published in a range of
[56] P.B. Crosby, Quality is Free: the Art of Making Quality Certain, McGraw-Hill, operations management and engineering journals. He is
New York NY, 1979. currently leading a funded research project in the area of
[57] M.C. Paulk, B. Curtis, M.B. Chrissis, C. Weber, Capability Maturity Model for ‘appropriate contracting for complex engineering proj-
Software, Software Engineering Institute, 1993. ects’ and supervising a Knowledge Transfer Partnership to
[58] CMMI Product Team, CMMI for Development, Version 1.3, Software deliver a new and reconfigured product range for a manufacturer of prefabricated
Engineering Institute, 2010. building products.
[59] V. Introna, V. Cesarotti, M. Benedetti, S. Biagiotti, R. Rotunno, Energy
management maturity model: an organizational tool to foster the continuous Paul Schönsleben is Professor for Industrial Engineering
reduction of energy consumption in companies, J. Clean. Prod. (2014) 1–10, and Management at ETH Zurich, Switzerland. Paul
doi:http://dx.doi.org/10.1016/j.jclepro.2014.07.001. Schönsleben studied Mathematics and Operation Re-
[60] A. Sen, K. Ramamurthy, A. Sinha, A model of data warehousing process search at ETH Zurich. He obtained a Ph.D. degree in 1980.
maturity, IEEE Trans. Softw. Eng. (2012) 38. Several years of responsibilities for computer and
[61] M. Van Steenbergen, R. Bos, S. Brinkkemper, I. Van De Weerd, W. Bekkers, The organization departments of industrial companies fol-
design of focus area maturity models, Proceeding 5th International Conference lowed. His research and teaching areas are logistics,
on Design Science Research in Information Systems and Technology (DESRIST operations and supply chain management, global service
2010), Springer, Heidelberg,, 2010, pp. 17–19, doi:http://dx.doi.org/10.1007/ management and service innovation. He is member of
978-3-642-13335-0_22. several boards of directors, as well as advisory boards. He
[62] T. Mettler, P. Rohner, Situational maturity models as instrumental artifacts for wrote several books, as well as numerous scientific and
organizational design, Proceeding 4th International Conference on Design popular articles. The most important book “Integral
Science Research in Information Systems and Technology-DESRIST’09 (2009), Logistics Management – Operations and Supply Chain
doi:http://dx.doi.org/10.1145/1555619.1555649. Management Within and Across Companies” was published in German by Springer,
[63] T. De Bruin, R. Freeze, U. Kaulkarni, M. Rosemann, Understanding the main Germany and in English by CRC/Taylor & Francis Group/Auerbach.
phases of developing a maturity assessment model, Australasian Conference