Automation in Construction 107 (2019) 102910

Contents lists available at ScienceDirect

Automation in Construction
journal homepage: www.elsevier.com/locate/autcon

Direct digital construction: Technology-based operations management T

practice for continuous improvement of construction industry performance
⁎
Müge Tetika, , Antti Peltokorpia, Olli Seppänena, Jan Holmströmb
a
School of Engineering, Department of Civil Engineering, Aalto University, Espoo, Finland
b
School of Science, Department of Industrial Engineering and Management, Aalto University, Espoo, Finland

A R T I C LE I N FO A B S T R A C T

Keywords: Construction operations suffer from fragmented structures and loose coupling among project actors. Building on
Construction operations direct digital manufacturing (DDM), we describe direct digital construction (DDC) as a technology-based op-
Direct digital construction erations management practice aiming to improve construction performance through design reuse and dimin-
Direct digital manufacturing ished human interpretation. We develop the operational principles, describe the operations in practice, and
Virtual design and construction
identify alternative implementation paths based on case examples of partial implementations. The core principle
Cyber-physical systems
of the practice is to direct the value-adding operations over the building's lifecycle through the digital design
model, resulting in increased reusability, project-specific differentiation, and automation of designs and pro-
cesses across projects. Three implementation paths are identified: (a) as-built modeling-driven path, (b) modular
product architecture-driven path, and (c) algorithmic and parametric design-driven path for the incremental
transformation of the construction industry.

1. Introduction designs differently, leading to improvised alignment actions and out-

comes that are different from the original intention.
The fragmented structure of construction projects has inhibited the When designs allow interpretation, the actors concerned are re-
construction industry's development in terms of efficiency and pro- quired to have special skills and commit efforts to interpret what has
ductivity [1]. Construction projects are loosely connected to other been designed correctly [5]. Even when human interpretation and the
projects, leading their own life, with limited past or future coordination resulting improvisation are considered appropriate [8], relying on ex-
and learning between projects [2]. As a result, the reuse of project pert and tacit knowledge often creates problems in the future [9].
designs and plans remains an untapped source in performance im- Merschbrock and Wahid [10] document a case where an improvisation
provement. Furthermore, construction tasks within individual projects in the use of building information modeling (BIM) technology in a
are also loosely coupled with one another, undermining the perfor- loosely coupled system later resulted in errors, costly production, and
mance of many projects [3]. This loose coupling of tasks and partici- additional working hours. By contrast, industrialized construction is
pants from different domains is partially due to the fragmented struc- characterized by rigorous planning and control of the construction
ture of the construction industry on the one hand [4] and the process, increasing upfront costs, and requiring tightly coupled design
operational practice of time-limited involvement of actors in task-spe- and management of the operations [11]. Well-prepared operations with
cific work on the other. Each of these actors relies on his or her own complete design from the beginning lead to the smooth execution of
interpretation of how to move the project forward from one stage to operations and decreased defects [12]. Thus, there is an opportunity to
another [5]. Improvisations in the construction process are frequent, improve the efficiency and productivity of construction through direct
emanating from unplanned, emergent situations [6]. Improvisations digital control of the construction operations based on the complete
occur because limited time is allocated to secure additional planning design.
resources when something unexpected happens [7]. However, im- The purpose of this paper is to develop a new operations manage-
provisations also occur because certain tasks and task interfaces are ment practice—DDC—based on a combination of technology-based
intentionally left un-designed. Moreover, because of their different operations practices from the construction industry and manufacturing.
backgrounds, the actors involved in the project can interpret the We elaborate on the application of this practice in the construction

⁎
Corresponding author.
E-mail addresses: muge.tetik@aalto.fi (M. Tetik), antti.peltokorpi@aalto.fi (A. Peltokorpi), olli.seppanen@aalto.fi (O. Seppänen),
jan.holmstrom@aalto.fi (J. Holmström).

https://doi.org/10.1016/j.autcon.2019.102910
Received 25 October 2018; Received in revised form 4 July 2019; Accepted 11 July 2019
Available online 16 September 2019
0926-5805/ © 2019 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/BY-NC-ND/4.0/).
M. Tetik, et al. Automation in Construction 107 (2019) 102910

Table 1
Forming DDC from existing technology-based practices—commonalities and differences of existing practices and DDC.
Operational practice elements BIM VDC DDM Required new operations management practice

Digital design Yes Yes Yes Yes

Design-based production Yes, but limited Yes Yes Yes
Cyber-physical control of production No Yes, but limited Yes Yes
Design-based operations beyond production No Yes No Yes
Cyber-physical control of operations beyond production No No No Yes
Reuse and improvement of digital designs and cyber-physical controls No No No Yes

industry through different implementation paths to enable continuous industry relying on BIM suggest that this has been achieved. BIM can be
improvement. The remainder of this paper is structured as follows. developed for the as-designed status; however, it will not reflect the
First, we establish the requirements for the new operations manage- details of the as-built reality [21].
ment practice by analyzing the commonalities and differences of the Similar to the fragmented construction industry characteristics, BIM
existing technology-based operations management practices. Next, we is a multifaceted approach that involves multiple stages of im-
justify and describe our research approach, design science research plementation and many different capabilities; it is also affected by
[13]. We then present our case examples to introduce the specific project networks and supply chain interdependencies [23]. The chal-
problems by the new solution development, and describe the elements lenges include problems with information exchange and integration,
of the operations practice design already implemented in the con- data interoperability, and computability [24]. Construction projects
struction industry. We use design theory [14] to structure and report often undergo changes during their lifecycles, resulting in differences
the design of the novel operations practice and its alternative im- between the intended design and end product [24]. Information about
plementation paths. The conclusion section will present the theoretical the project is typically recreated between five and eight times during
contribution, explain the limitations of the research, and recommend the various phases of the project [25]. This indicates that simply in-
directions for further research. creasing design resources is hardly a solution to human interpretation
and reuse challenges. Without design-based production and cyber-
2. The need for DDC: commonalities and differences of physical control, the design models are not kept up to date while pro-
technology-based operations management practices duction and operations are being carried out. Therefore, there is a need
for an operations management practice where operations are both
Before we describe DDC in detail, we review the technology-based planned and carried out based on the digital design model.
practices on which the novel operations management practice builds VDC uses the digital design model to model and plan the multi-
and investigate their commonalities and differences (Table 1). As a disciplinary execution of design-construction projects, enabling
technology-enabled operations management practice, DDC builds on achieving explicit and public business objectives [26]. Luth [27] states
Building Information Modeling (BIM), Virtual Design and Construction that VDC is the process of utilizing accurate three-dimensional (3D)
(VDC), and Direct Digital Manufacturing (DDM). The gap in current building information models for facilitating visualization, simulation,
research concerns the absence of an operations management practice communication, coordination, estimation, purchasing, and scheduling.
that combines the unique practice elements of VDC and DDM. DDC This concept allows all actors to access a project's shared data. On the
relies on digital design and design-based production, which are advanced maturity levels of VDC, cyber-physical control of fabrication
common for both VDC and DDM. Besides, DDC draws on the unique and building of subassemblies in a factory have been added [26].
practice element of cyber-physical control of production from DDM and However, in the construction industry, project orientation of design and
the extension of design-based operations beyond production from VDC. delivery limits the benefits from investments in design-based produc-
Cyber-physical control refers to cyber-physical systems (CPS), which tion and operations to individual projects. Moreover, BIM focusing on
are physical systems whose operations are monitored and controlled by building elements of VDC model can bring disadvantages due to man-
a computing and communication core [15]. Attempts have been made agement issues requiring process interactions (see Table 1) [26]. Fur-
to apply the CPS approach in construction to real-time bi-directional thermore, VDC as an operational practice does not explicitly aim to
coordination and communication between virtual models and physical reuse the designs and process controls to prevent improvisations and
construction [16]. We extend this research by focusing on a technology- improve productivity across projects. Therefore, integrated project de-
based operations management practice that not only coordinates the livery (IPD) and partnering have been suggested as appropriate con-
virtual and physical worlds but in which a digital design prevents tractual forms to incentivize actors to invest in VDC [28]. VDC together
harmful interpretations and improvisations in future operations. Also, with IPD would increase the use of construction knowledge upstream in
the new elements of DDC missing in both VDC and DDM are cyber- the design process and development of detailed design at an earlier
physical controls on operations beyond production, which include lo- stage [29], both of which are needed for more automated design and
gistics and lifecycle operations in use phase and the purposeful reuse construction. The challenge of introducing VDC is that it requires firms
and improvement of digital designs and controls of operational pro- to invest in the operational practice more than the value they can
cesses. capture in a single project [26]. Therefore, we argue that new, tech-
Digital building models can be used as information stores in the nology-based operations practices that focus on the continuous im-
design and construction phases and in the use and maintenance phases provement of the firm- and supply chain-level operational performance
without using either design-based production or cyber-physical control. need to be developed; this calls for the reuse of design and operating
However, when an operation is carried out in an improvised manner models beyond a single project lifecycle.
without been specified in the digital design model, the succeeding op- In DDM, the focus is on using the digital design model for the effi-
erations in the use and maintenance phases begin with inaccurate in- cient production or fabrication of customized design of parts. The direct
formation and inadequate initial conditions that had been left un- cyber-physical control of manufacturing offers a potential solution
documented. A digitalized operation relying on BIM leads to differences element to minimize improvisation while improving efficiency and
between as-designed [17,18], as-planned [19], and as-built [20–22]. customization in the construction value chain. In manufacturing in-
Creating virtual buildings through BIM promises that the design models dustries, the concept of DDM is popular because of its ability to resolve
can replicate real buildings. However, very few examples in the customization and reusability issues by combining digital design with

2
M. Tetik, et al. Automation in Construction 107 (2019) 102910

cyber-physical control of production [30]. DDM is a technology-based case examples.

operations management practice where the 3D computer-aided design In our research process, we utilized the four phases of research for
(CAD) models are used directly to manufacture end-use parts with ad- design exploration and theory building in operations management re-
ditive manufacturing [31], for example. DDM uses additive manu- search [13]. These phases (see below) represent the movement from
facturing, computer-numerical-controlled (CNC) machines, computer- new to tested ideas, mid-range theory, and formal theory:
aided manufacturing (CAM), and computer-aided engineering (CAE)
[32]. DDM reduces the time and cost required for the design and 1) We first framed the problem and collected fragments from practice
manufacturing of individual items, eliminates the need for tooling, and for a potential solution design;
offers customization possibilities to satisfy different customers' re- 2) The initial proposal for a solution design was then subjected to
quirements effectively [30,33]. further development using inputs from cases which adopt required
DDM is an important operational practice in implementing the vi- operational elements of the new operations management practice;
sion of Industry 4.0 and Construction 4.0, as its potential counterpart in 3) The findings were generalized, and a theoretical contribution was
the construction industry [34]. However, because of the highly variable demonstrated in terms of novel insights in the research literature;
maturity level of digital manufacturing in the construction industry, and
adapting the practice poses several challenges. The cyber-physical 4) The findings were presented more formally as design theory.
control and reuse of designs need to be supported by operational
practice elements from both BIM and VDC. Actors in construction need Fig. 1 illustrates our research process more closely: In the first
to address many issues, such as data security, implementation costs, and phase, two case examples of the state-of-the-art BIM practices from the
process changes [34], the resolution of which takes time. Here, BIM is field are analyzed to frame the problem, determine the issues requiring
available to address the collaboration, integration, and interoperability new solutions, and propose the purpose for a new practice. In the
issues [35]. VDC is available to align the design model with different second phase of our research process, six partial case examples are
operations and actors within the project and building lifecycles [36]. explored to develop a solution proposal, namely the DDC operational
As DDM has been developed in manufacturing industries with stable practice. The six cases partially implement the required operational
product design and supply chains, the development of a new practice practice elements (Table 1) beyond the mere digital design and design-
such as DDC should allow handling higher levels of customization and based production in specific firms or supply chains. Our research pro-
assembly activities on customer sites in construction. Partial im- cess investigates what these partial solutions are; it also investigates
plementations of DDC involving combining digital design and BIM- what the approaches to solve the issues identified in the problem-
based automated construction for structures, i.e. [37], are available. framing phase are. In the third and fourth phases, design theory [14] is
However, the holistic design of the technology-based operational utilized to frame the opportunities and problems in the six cases; this is
practice of DDC construction is needed to support the paradigm shift followed by formalizing the proposed operational practice. The last part
from a project orientation to solution orientation in architectural design of our research process corresponds to generalizing the findings, de-
and construction [38]. For this task, we will adopt a design science monstrating its theoretical contribution, and presenting the design
approach [39]. theory for the DDC concept.
According to Denscombe [41], case studies are suitable for under-
3. Methods standing the complex relationship between various factors that operate
in a specific social setting. To understand the current level of im-
Building on the existing technology-based operations management plementation of technology-based practices in the industry and observe
practices, this research aims to design and describe a new operations our rudimentary solution design in a real-life context, we chose a case
management practice for construction. Closely working with the prac- study method as part of our design science research approach, con-
tice in steering the design toward providing a solution to real-world ducting eight case studies in total.
problems, design science research is a suitable approach to develop and First, we analyzed two different leading-edge attempts to have the
evaluate solutions to organizational and operational problems [39]. We match between as-designed models and as-built models; we also de-
chose the design science approach because design science identifies the scribed the benefits and shortcomings of each. This was done to gain
problems that practitioners usually face, defines the objectives of a insights into the state-of-the-art attempts to create a constructible de-
solution by asking what would a better artifact accomplish, and then sign model that can be followed in construction without allowing im-
designs and develops the artifact to use it to solve the problem in a provisations intending to achieve an accurate as-built model in the end.
proper context [39]. It then observes the effectiveness of the artifact in One of the authors conducted a field study, which included interviews
solving practitioners' initial problems. We adopted this approach to and site visits to the two case companies.
solve the fundamental issues concerning limited productivity im- Following these two in-depth case studies, we analyzed six case
provement, excessive improvisation, and lack of learning in the con- examples to elaborate on DDC's design theory components. The six
struction industry. cases were selected based on their partial physical implementation of
Designs in design research can be operational concepts and prac- the required six operational practice elements (Table 1) required for the
tices, implementation methods, and example artifacts [39]. In our re- new technology-based operations management practices. We used an
search, DDC, as a new technology-based operations management exploratory approach, selecting cases with different settings via theo-
practice, proposed solutions to the problems that construction industry retical sampling [42]. The cases do not represent any countries or
practitioners face, such as common interpretation, improvisation, and building types; however, they all exhibit some exceptions from the
lack of reuse. The development of the DDC concept requires describing traditional fragmented practices and are selected purely based on their
its design and implementation components based on the design science elements related to the required new operations management practice,
approach. While design science seeks to generate new knowledge on such as the cyber-physical control of production and the reuse and
solving organizational problems, design theory offers to codify and continuous improvement of digital designs. The analysis of the six cases
generalize aspects of the created knowledge [40]. Gregor and Jones' was performed to understand how the identified weaknesses of the
[14] design theory posits that the design of a new concept or practice leading-edge as-built modeling practices were solved in the cases and to
can be structured using the components listed and described in Table 2. identify implementation principles for the DDC elements.
We apply this model by focusing on the purpose and scope of the new We used multiple data sources in our cases, including interviews,
practice, its constructs, principles of its form and function, im- observations, official documents, and public sources concerning the
plementation principles, and expository instantiations, which are the case companies and their solutions. In the case studies, semi-structured

3
M. Tetik, et al. Automation in Construction 107 (2019) 102910

Table 2
Design theory components and their descriptions [14].
Component Description

Purpose and scope Describe what the system is for; state the requirements and specify the type of entity to which the design theory applies along with various
scope elements
Constructs Comprise representations of the object of interest
Principles of form and function Present the abstract blueprint that describes the structure and functionality
Principles of implementation Describe the processes used to implement the theory in a certain context
Expository instantiation A physical implementation that helps illustrate the theory as an example or for testing purposes

Phase 1: Problem
Phase 2: Soluon Phases 3 and 4:
framing
development Contribuon and
Analysis of the formalizaon
Analysis of case firms
current best model-
and supply chains, DDC concept
based pracces in
parally applying formalizaon using
projects: Aempts
required operaonal design theory (Table
to have as-designed
pracce elements (Table 2) to convey design
models correspond
1) knowledge
to as-built models

Fig. 1. Research process.

interviews were conducted to collect data on solutions as well as their the contractor makes the models buildable with his or her detailing
development, features, effects, and challenges. According to Brinkmann team. This yields a construction model that is accurate, coordinated,
[43], semi-structured interviews allow data to contain more sponta- and leaves no room for improvisation. The contractor can build the
neous descriptions. Thus, the interviews with the company managers of design model using a systematic controlling process, modify the designs
the state-of-the-art practices included semi-structured questions to ob- immediately if there are deviations from the design, and follow these
tain a profound understanding of the level of application of the concept modified designs closely. Not only does this approach to securing ac-
and suggesting further implementation possibilities. In the six partial curate modeling at the outset of the process minimize the chances of
solution cases, we interviewed solution developers and managers. We interpretation in later designs and operations, but it also decreases
also utilized the available video materials, webinars, documents, and iterations, which often arise when the design is not sufficiently com-
news articles in professional journals to obtain a comprehensive view of plete at the beginning.
the solutions and validate our interview data. Finally, the concept was To ensure that the construction model is adhered to in the field,
validated along with its novelty and usability in a focus group discus- drawings are extracted from the 3D model, and the quality of the on-
sion with 12 professionals from the architecture, engineering, and site work is controlled every day. A full-time surveyor checks the field
construction (AEC) industry. More information about the collated data using a robotic total station with up to 600 measurement points per day
from the cases is presented in Appendix A. Finally, the design principles and compares it back to the model before and after the installation. Any
and implementation paths of the DDC concept and inherent operational deviations are checked against the drawings, and corrections are made
elements were presented by discussing the findings related to the best either in the model or in the field, depending on the severity of the
as-built modeling practices and partial solutions of the six cases in light issue. All subcontractors must control the quality of their work. If they
of the existing literature. lack surveying capability (most do), they can purchase surveying ser-
vices from the case company. Since the case company does not seek
4. Problem framing: state-of-the-art practices in matching the as- profit on providing this service, it is the most affordable solution for the
designed model with the as-built reality subcontractors. In the end, surveying does not incur costs for the case
company.
The exploration to match the as-designed model with the as-built The case company does not change the design model if there is only
reality started with interviewing the managers from the two state-of- a minor deviation in the as-built reality that would not matter in sub-
the-art projects. To achieve a design-based production, these companies sequent processes, such as centimeter differences in areas where there is
followed two different sophisticated processes, as described below. no requirement for small tolerances. The company shifted considerable
design scope from the designers to the case company's detailers and
4.1. Case 1: As-built modeling contractor shifted work from the subcontractors to the detailers. In this way, the
case company could assure a high-quality construction model to ensure
Our first case company is a general contractor. This contractor up- that it could be built as designed. To raise the design to the desired LOD
dates models during projects so that the design models can be built and quality before construction, the case company's detailers spent
without improvisation and the final updated as-designed models cor- roughly around 10,000 h on a project worth $150 million; the project
respond to the as-built reality. The company invests several thousands was a 300,000 square-foot residential apartment building. By commit-
of man-hours in the design and construction phases to achieve this goal. ting these additional hours to the design phase, the case company re-
It has a sophisticated process of documenting the as-built and control- ported that work in the field had reduced, the quality had increased,
ling the as-built reality against the design model. The operations are as and the number of requests for information had decreased significantly.
follows. The external design companies develop the models to a level of The results of the practice included $2.5 million savings for the owner
detail (LOD) of 290, meaning that each element is designed with an and a 20% increase in profit for the company. Moreover, the sub-
approximate shape, location, size, and orientation without precise in- contractors also profited from the practice. Furthermore, the process led
formation, such as the model number, supplier, or exact location. Thus, to increased quality and saved hours in the subsequent phases of the
the designers do not have to specify the exact dimensions and location project. The owner paid for the 10,000 h because it would be saved
information—precise detailing is the contractor's responsibility. Then, later in the process due to the lower number of change orders, as the

4
M. Tetik, et al. Automation in Construction 107 (2019) 102910

model could be used to produce dimensionally accurate shop drawings. that design-based production is possible in the fragmented and project-
As an additional deliverable, an accurate as-built model was obtained. based practice of the current construction industry, although some
However, the company experienced challenges in convincing the weaknesses persist. For as-designed models to match the as-built rea-
owners and subcontractors to invest more money earlier in the process, lity, companies need to invest substantial time and labor in the con-
hoping that the money would be saved later. Owners were convinced by struction project's early phases. Our analyses suggest that both case
learning about the benefits of having an accurate as-built model. The companies found it useful to have a detailed and complete as-built
savings, profits, and quality enhancements of the previous projects in- model as these investments can be recouped in future operations.
crease the credibility of the practice and hence convince the prospective The weaknesses identified in these two current best BIM practices
project stakeholders. mainly relate to the high need for labor and training. In the as-built
modeling contractor's case, each project had a new group of stake-
4.2. Case 2: the school developer holders: different owners, designers, and subcontractors. Although the
process is highly transferable across different project types, transferring
Another state-of-the-art case example is a school construction pro- the process through different project teams requires training. For in-
ject. In this project, the as-designed model and built reality were ulti- stance, the contractor's detailers had a different starting point each time
mately matched. This requirement was driven by the owner developer. in case examples. They could standardize some of the details, although
The analysis of the interview data shows that the motivation behind this they still had to carry out substantial custom work. In the school de-
complete match was the desire to omit any inaccuracies in the re- veloper case, the public owner developer had to select the design teams
novation and maintenance operations of the buildings. The owner de- and contractors competitively. Therefore, they needed considerable
veloper of the school wanted the designs to exactly represent the training for each new team and project. It is unclear whether any
building since it was important to avoid information loss; besides, up- transferable knowledge other than the knowledge related to the
to-date information was desired when renovation and annual repairs training materials had been developed in this organization. Design so-
were carried out in the building. lutions remain with the design teams; however, each project had a new
In practice, the engagement of the designer was increased on site, design team. Therefore, process and design models that could be uti-
and the project did not proceed to the construction phase until the lized in other school projects were not purposefully reused.
design model had been completed; this meant that there was no room Fig. 2 presents the DDC logic in the state-of-the-art cases compared
for improvisation during the installation phase. The designers su- to the more traditional construction practice where the design model is
pervised the process to ensure that everything was constructed ac- not kept up to date. The iterations between the design and construction
cording to the design model. They also participated in the construction phases preclude improvising and having to interpret the design models.
management meetings and actively collaborated with the workers on In practice, both case examples had expanded the role of the design
site. The designers were always available in case of a need for a design phase and design resources to design the building and assembly work
change. This presence ensured that the contractor did not have to im- more comprehensively and keep the design model up to date. We call
provise and that the as-built model always followed the as-designed this “as-built modeling-driven DDC,” whereby the workload increases
model. The design contract comprised a target/ceiling price and was in the design phase but decreases in the construction phase. The most
paid according to an hourly rate. The developer also partnered with notable benefits may be achieved during the use and maintenance
several design companies with agreed hourly rates and often asked the phases because the models are only useful in these phases provided they
partners for estimates. Sometimes, the designers issued change orders are accurate. However, even if the as-built model could be utilized
to follow the process. The contractor was responsible for assessing the during the lifecycle of the building, it is not clear whether it could be
BIM model and giving feedback on its constructability if any issues reused when designing new buildings. Thus, the challenges of the state-
arose or improvement suggestions were offered. The designers subse- of-the-art practices include the additional labor time required to keep
quently performed the necessary design changes. They modeled the the designs up to date and the limited opportunities for reusing the
building to the required LOD so that everything could be installed ac- models between projects.
cording to the design model.
The owner developer of the school kept the models up to date and 5. Solution development: partial DDC solutions
was able to utilize the same design model (from 10 years ago) in the
design of a new renovation operation with no deviations from the de- We next analyze the six case examples to investigate the ways of
sign model on the construction site. Based on the owner developer's solving the detected problems with the current best practices explained
experience, in ordinary projects, many subcontractors do not even ask in the previous section. These six case examples were selected to il-
for the design model in advance, as it is not usually updated; instead, it lustrate the adoption of the required operational practice elements
has already become outdated in the construction phase. (Table 1) in technology-based construction operations; they were also
The analysis of the interview data indicates that no other companies used to address the issues of the need for labor and training and lack of
invest in this practice because it is easier to stop updating models. opportunities for reusing the model. The case examples emphasize the
Considerable staff training is necessary to achieve the desired match direct use of digital design models in the operations following the de-
between the design model and the actual building, which can incur sign phase, such as manufacturing of parts and maintenance.
enormous costs. The owner developer of the school emphasized the The first case is a log house design and construction solution. The
significance of interaction in the team for keeping the models up to flow of operations begins with the architectural design phase, which is
date. Moreover, in comparison with other companies' traditional prac- followed by the planning and manufacturing of log house parts. Fig. 3
tices, the developer allocated substantial additional resources to the illustrates the workflow of construction operations for log house pro-
design phase. However, it was considered that a lack of investment in duction. In this case example, the software developer developed custom
better design would increase the workload at the subsequent stages on libraries for wooden log house design and fabrication. The log house
the worksite. Thus, investing in a detailed and complete design model architects use the program's custom libraries to model log houses in a
pays off in the construction and maintenance phases. BIM environment according to the customer requirements. The pro-
gram automatically decomposes the log house design into building
4.3. Summary of the state-of-the-art solutions components and their respective locations. In this solution, the set of
tasks to be performed on the building is fully specified by the design
Table 3 summarizes the practices and their purpose and scope ele- eliminating the possibility of misinterpretation and improvisation. The
ments in the state-of-the-art cases. The state-of-the-art examples show design model is directly transformed into a machine-readable file to be

5
M. Tetik, et al. Automation in Construction 107 (2019) 102910

Table 3
Purpose and scope elements of the state-of-the-art cases.
Practice Purpose and scope

Surveying up to 600 points at the construction site every day to ensure as-designed models + Substantial savings on time and quality error costs
corresponded to as-built models + Accurate as-built models
−/+ Investment in planning to obtain a buildable model that can be
constructed without improvisations
– Additional man-hours spent on surveying
– Difficult to convince owners and subcontractors
Designers involved in the construction phase to ensure as-designed models corresponded to as- + Accurate as-built models
built models + Savings in construction, maintenance and renovation phases
– Increased scope of work for the designers
– Need for training

Note: + denotes an advantage; − denotes a disadvantage.

used in CNC production for manufacturing the logs. The design model contractor's solution, the prefabricated modular components fit into
also defines loads and logistic kits for site deliveries and provides on- one another. The fitting wall system can create different types of room
site assembly instructions for the parts. The instructions (i.e., the pro- arrangements or partitions, and these can be changed if required. The
cess information) are generated using the digital design model. Hence, contractor stresses the importance of the early planning phase and
in log house production projects, it is possible to use the digital design precise service descriptions. Moreover, the contractor standardizes the
models directly in many operations and minimize human interpretation invisible building components, such as the shell and supporting struc-
in all process interfaces. The solution also automates the work of ture, although it leaves room for individual choices when it comes to
structural designers and production planners. Moreover, the design the visible components of the building (e.g., the façade). The models for
models, in whole or in part, can be reused in different log house fab- the invisible building components are reused and continuously im-
rication projects. proved in new projects. Computer-controlled operations do not di-
The second partial DDC case focuses on producing steel structures. minish human interpretation. Instead, standard interfaces between the
In this case example, a model-based software solution for steel struc- components rule out improvisation in assembly, and additional design
tures is used for automatic welding operations. The position of the work can be directed toward improving and optimizing single compo-
welds and welding robots, relative to the materials and all other in- nents.
formation on the welding operations, is stored in the digital model. This The fourth partial DDC case is a direct digital timber and steel
information is used directly during the manufacturing operation. The manufacturer. This manufacturer emphasizes an integrated approach in
robots produce welds in an automated manner and based on the digital the early design phase. They use 3D CAM/CAD software, which in-
model. Thus, the practice enables the digital designs to be used in cludes production and logistics information in the design model. This
different projects in whole or in part. The accuracy level of these op- practice reduces the need for costly redesigns and changes. Hence, they
erations is extremely high. With this practice, the design model be- insert material properties, such as strength, into the design model only
comes the built reality. Moreover, the same models can be used during once. The virtual building process used identifies potential geometric
maintenance operations. For example, when the sensors detect a mal- and installation issues before construction work, so no material or labor
function in a structure, the software displays the parts. These parts are is wasted. With this solution, customized and complex structures are
used in different areas of the structure in the updated digital design built according to the design model with a high level of accuracy.
model allowing other malfunctions to be checked and prevented. A similar approach is also adopted for the customized interior ar-
The third partial DDC case is a modular building contractor. In this chitecture area. The fifth case is a company that produces customized

Fig. 2. DDC logic in the state-of-the-art modeling cases in comparison with the traditional practice—the first version of as-built modeling-driven DDC.

6
M. Tetik, et al. Automation in Construction 107 (2019) 102910

Fig. 3. Construction workflow of a log house company.

interiors via modular and prefabricated solutions. The company uses 3D work. In the last one, the designer was driving the process; it works in a
design and manufacturing software, which automatically updates the project environment and creates unique products with a standard
information used in the factory for production operations, thereby platform. In contrast, the state-of-the-art cases were driven by the
leaving no room for interpretation. It then increases the control over owner or general contractor in a project environment with multiple
what is being built mainly because no worker is required to interpret stakeholders. Furthermore, the cases illustrate that the DDC concept
the installation drawings. The used software platform enables the can be utilized by companies with vertically integrated supply chains
company to build exactly what has been designed. Moreover, the de- that drive their own product range and designs for controlling con-
sign, engineering, manufacturing, and installation operations are struction operations directly from a digital design model, such as in the
tracked on the same platform. modular building case.
The sixth case is a platform-based building designer who works as
an integrated design and operations consultant and uses earlier com-
6. Concept formalization: the DDC components
ponents and models for guiding the construction of future buildings.
The company has developed three open design platforms for different
Building on the findings from the eight case examples, the concept
building types, which include detailed designs for standardized com-
of DDC and its principles are presented in Table 5 using design theory
ponents. While each project remains unique, significant aspects of the
[14]. The purpose and scope elements of DDC suggest making the as-
design from previous schemes are retained. For the sake of standardi-
designed model correspond to the as-built model. This leads to in-
zation, the production can be automated on the site. The company aims
creased efficiency over the lifecycle of the buildings or building sub-
to refine and improve the common elements by reusing the common
systems. All operations including maintenance and renovation can be
parts in the projects to enhance the quality and efficiency of the pro-
conducted based on the same, accurate and complete model. Based on
duction of design information. The company standardizes the manu-
the case studies presented in this paper, these complete design models
facturing process and connection interfaces. The same components and
would also increase the opportunity for reusing the entire or partial
manufacturing process can be used for different project types and sizes.
product and process models in different construction projects. Design
Thus, the company develops assets that include repeating components
patterns in the architecture field involve different ways of building.
that fit into specific locations and architectural designs. Having a de-
Although they are different in detail, they are similar in their general
tailed design model that comprises component designs makes it possible
outline [44]. Thus, although the buildings are unique, the processes
to achieve a complete match between as-designed and as-built models.
needed to build them can be similar, partially the same or the same.
In summary, solution opportunities from real-world examples show
Therefore, the design and process models can be reused between the
that by integrating the product and operations information into the
projects. Similarly, standardized components, component interfaces,
design model, design-based operations and cyber-physical control of
and sub-product structures can be reused between the projects. In ad-
operations become possible with different kinds of materials, con-
dition, by repetitively reusing and improving the previous solutions
struction practices, and product offerings. The reuse and improvement
that have been proven to be efficient, the quality and efficiency of the
of digital design models in whole or in part are also possible between
products and processes can be improved over time with the refined
projects. Table 4 shows the advantages and disadvantages of partial
design model in each iteration. Since with DDC, everything is digital
case examples. These practices illustrate how the operational practice
and up-to-date, it is possible to scale up the benefits from reused de-
elements of the DDC concept are used in different environments.
signs. Continuous elimination of problematic designs and reuse and
Based on the six case examples, we propose that the required op-
improvement of best designs enable continuous improvement over
erational practice elements are applicable to the construction environ-
time.
ment and that the partial solutions have addressed the problems of the
The constructs represent the elements that are utilized to achieve
state-of-the-art cases. Reuse of the designs and process information are
the targets of the proposed concept. Considering the state-of-the-art
possible in most of the case example. Increased specifications of design
case examples and partial cases, having complete, detailed, and up-to-
models result in increased constructability of models and ensure high
date digital design models with embedded operation instructions re-
accuracy levels. However, albeit for a few of the aforementioned state-
present the most significant constructs of DDC design theory. To
of-the-art projects, the technology-based elements require investments
achieve this goal, design models must have a high LOD. The design
in the early stages of the parallel design and development of the pro-
models should be complete before the start of the operations and should
ducts and operations, as the design model must be detailed to ensure
be kept up to date. Operation instructions should also be embedded in
design-based and computer-controlled operations. However, because
the design model. Product information (e.g., LOD 300 level) and in-
the same models in both building design and the cyber-physical system
structions on how to produce the product or its sub-part (e.g., LOD 350)
can be reused, the additional investments in design and operations and
should be part of the digital design model. To accurately describe and
controlling technology are shared by multiple projects. Thus, return on
define the processes that are directed and controlled by the design
investment can be evaluated based on a series of projects reusing the
object, specific investments are needed in process interfaces and in-
design and utilizing the same operations technology.
herent technology interfaces. Table 5 also presents the principles of
The first five partial solution examples were driven by the company
function and implementation elements for DDC practice. A fully im-
responsible for the delivery of building components and installation
plemented DDC enables all value-adding operations over the lifecycle of

7
M. Tetik, et al. Automation in Construction 107 (2019) 102910

Table 4
Advantages and disadvantages of partial case solutions.
Case companies Advantages/disadvantages

Log house design and construction + Computing-based automated engineering and manufacturing
+ Design-based loading, delivery, and assembly
+ Investment in the design phase can be recouped in future operations
– Constrained by manufacturing technologies
Model-based software solution for steel structures + Cyber-physical control of welding and maintenance operations
+ High level of accuracy due to model-based automated process
+ Investment in the design phase is paid back through reuse
– Constrained by manufacturing technologies
Modular building contractor + Design-based configuration and assembly
+ Continuously optimizing design through reuse and adaptation of component designs
+ Efficiency improves as investments are directed to component development, not interface management
– Limited cyber-physical control but likely implementation in future
Digital timber and steel manufacturer + Cyber-physical control of engineering and manufacturing
+ Design-based delivery and assembly information
+ Investment in the design phase is paid back through reuse
– Constrained by manufacturing technologies
Interior manufacturer + Cyber-physical control of interior product manufacturing
+ Design-based installation
+ Investment in the design phase is paid back through reuse
– Constrained by manufacturing technologies
Platform-based building designer + Cyber-physical control of site-operations through automation and robotics in assembly
+ Reuse of design components, faster design process, and unique products
– Limited reuse of designs to date; requires committed owners or developers to become scalable

Note: + denotes advantage; − denotes disadvantage.

a building, from its design to its use, to be directed and controlled by Effective customization makes the whole design and engineering pro-
the digital design model. This increases efficiency in the planning and cess faster and more affordable. If the design is not shared openly or
control of the operations and promotes the industrialization of the initial designers or design owners decide not to reuse the design in
construction operations [12]. further projects, the value of reusability may be lost. On the other hand,
Fig. 4 compares the operations of traditional construction practice if the owner, contractor, or designer has access to design model and
and DDC concept. It shows that in the case of DDC with a detailed and they utilize it in further operations, this would lead to increased effi-
complete design model, there is no distinction between the as-designed, ciency over the lifecycle of both building and supply chain when po-
planned, and built models since the design model is identical to the tential mistakes are prevented by the up-to-date design model in future
built model. To minimize interpretation and improvisation in opera- operations. Continuous improvement can eventually lead to automation
tions, not only is the connection between design model and operations in all the operation phases, including engineering, manufacturing, lo-
bi-directional [16] but complete design models from previous projects, gistics, installation, and maintenance.
including operations instructions, are also reused to reduce the need for The effectiveness and applicability of the concept were validated by
project-specific changes. Models can be reused both at the level of a focus group session with 12 design, development, and business
product designs and process control designs by utilizing cyber-physical managers from AEC companies. In the session, the components of the
systems in operations. Designers can customize designs at project level DDC concept (Table 5), benefits and disadvantages of the partial case
either through component configuration or through changing para- solutions (Table 4), and comparisons between the traditional con-
metric values. Effective customization is enabled by open design plat- struction practice and the DDC concept (Fig. 4) were first presented to
forms as in the cases of the platform-based building developer and the the participants. After that, the following three questions were debated:
log house developer whereby designers can utilize detailed component
designs from the platforms to come up with unique building design. • Are you familiar with similar concepts? (the novelty of the concept)
Table 5
The components of the DDC concept.
Design theory components DDC concept Evidence from case studies

Purpose and scope • As-designed corresponds to as-built • Based on state-of-the-art case analyses, as-designed
• Increased
subsystems
efficiency over the lifecycle of buildings and building corresponding to as-built leads to improvements in the later
phases of buildings' lifecycles.
• Increased
projects
reuse of designs and process control models between • Based on the state-of-the-art and partial case examples, utilizing
the same design in whole or in part increases quality and
decreases project cost
Constructs • Complete, detailed, and up-to-date digital design model • Limit the improvisations and enable the use of the same model
• Embedded operation instructions in the digital design model in every operation
• Increase
systems
control over later operations through cyber-physical

Principles of function and

implementation
• Direct use of the design model when carrying out operations:
Engineering, manufacturing, logistics, installation, use, and
• Utilizing the same digital model in design, logistics, and
assembly operations in partial DDC examples and state-of-the-
maintenance art case examples
• Design model, including operation instructions, can be reused,
adapted, and improved continuously at both the building and
• The original investment in the design model and cyber-physical
control systems are recouped through reuse between projects.
building subsystem levels • Alternative implementation paths toward DDC are available
• Operations and buildings incrementally develop toward DDC
8
M. Tetik, et al. Automation in Construction 107 (2019) 102910

Fig. 4. Comparison of traditional construction practice and the DDC concept.

• In which situations can the DDC concept be useful? (application another project will help in refining the model, leading to continuous
area) improvements, both in design and subsequent operations. The re-
• How can the current practices be transformed into the DDC way? presentatives emphasized that every actor has the model in his or her
(implementation) own server; they added that he or she works like silos while the owners
do not value the digital model. They raised questions such as who owns
The participants found it useful to conduct construction operations the model and how every actor can access it. They also suggested that
with methods drawing on the DDC concept. Some discussed industry the value of having updated and complete model should be commu-
practices that partially resemble the DDC concept, such as laser sur- nicated to the actors.
veying for MEP installation to match the model with reality. For the
application area, one representative mentioned that in tandem with the
7. Alternative paths of implementation for DDC
development of design automation algorithms, the concept could be
utilized in solutions that involve prefabrication mainly because it is
After developing the design theory elements for DDC, we now dis-
currently suffering from human errors. They also cited several examples
of how DDC could improve construction operations: The concept can cuss the case examples to illustrate their DDC elements and identify
alternative paths of implementation. Table 6 presents three identified
enable continuous improvement in design and building phases because
it requires performing the installations precisely according to the design implementation paths, their relevant elements and constructs, and case
examples following each path. Achieving similarity between the design
model. Thus, when the operations closely follow the design model,
hardly any error will occur during the installation phase, and thus, no model and the built reality (design-based production) is the common
rework will be required. Since the design represents the exact building, element among the paths. However, different initiators of the partial
further operations such as use and maintenance tend to begin with solutions, different application areas, and different emphases in the
more accurate information. The representatives mentioned that during purposes may cause variations in the ways that the DDC is im-
plemented. Based on the various practices (both common and different)
the facility management phase, updating the models is currently very
complicated. Utilizing the same model in every operation and then in of the case examples and literature, we argue that DDC can be ap-
proached using three alternative implementation paths:

9
M. Tetik, et al. Automation in Construction 107 (2019) 102910

Table 6
DDC elements of case examples and implementation paths.
Implementation path DDC purpose and scope elements DDC constructs, functions, and implementation Case examples

As-built modeling-driven DDC As-designed corresponds to as-built Complete, detailed, and up-to-date digital design model As-built modeling contractor
Increased efficiency over the lifecycle School developer
of a building

Modular product architecture- As-designed corresponds to as-built Complete, detailed, and up-to-date digital design model Modular building contractor
driven DDC Increased reuse of design and process Design-based operations: Embedded operation instructions in the Platform-based building
models between projects in digital design model designer (partially)
construction
Increased efficiency over the lifecycle
of buildings and building subsystems

Algorithmic and parametric As-designed corresponds to as-built Complete, detailed, and up-to-date digital design model Log house design and
design-driven DDC Increased reuse of design and process Embedded operation instructions in the digital design model construction
models between projects in Model-based software solution
construction Cyber-physical control of operations: Direct use of the design in for steel structures
Increased efficiency over the lifecycle carrying out automated operations (e.g., engineering,
of buildings and building subsystems manufacturing, and installation) Digital timber manufacturer

Interiors manufacturer

Platform-based building
designer (partially)

1. As-built modeling-driven DDC generating construction drawings saves a significant amount of time in
2. Modular product architecture-driven DDC the design phase [49]. These parametric designs on either the compo-
3. Algorithmic and parametric design-driven DDC nent or building level can be reused; this supports the reuse principle of
function and implementation of DDC. Moreover, designers can auto-
The as-built modeling-driven DDC implementation path enables the matically generate and assess design solutions via parametric design
development and maintenance of unique engineered-to-order buildings. and genetic algorithms [50]. This method would generate unique and
BIM objects can be used for engineered-to-order component production, efficient buildings; however, it would also require investments in sys-
supporting the workflow between design and manufacturing operations tems integration and algorithm development. Moderating the expected
and generating automated activities, as suggested in [45]. This ap- investments requires beginning implementation with rather simple
proach validates the intended design before passing it to the manu- buildings or the sub-products of those buildings. The partial case ex-
facturer. This approach supports DDC by highlighting the DDC con- amples of log houses, steel structures, furniture, and timber products
struct on having complete and detailed digital design models. represent approaches whereby focusing on certain materials helps in
Modular product architecture-driven DDC can be implemented by simplifying the development of algorithms. The platform-based
utilizing make-to-stock or assemble-to-order items and modular building developer, in contrast, adopted a whole building solution in
building blocks that are designed, fabricated, and installed according to which component designs can be used partially or fully in other projects
the specifications of the design model. When producing a design that is on the one hand and modified according to the specific needs in the
based on a digital design model, the control information needed to future projects on the other. In summary, it is possible to benefit fully
assemble the product from its components can be embedded in the from this practice once the required investments are made in manu-
design model [46]. With product modularity-driven DDC, the interfaces facturing technologies, such as assembly and welding robots or auto-
between product subsystems are standardized, thereby increasing mated CNC operations.
continuous improvement and control and minimizing interpretation The implementation paths differ from each other according to the
and improvisation. This enables the use of permanent design rules in required investments. As-built modeling-driven DDC requires major
the building design and engineering phases. The focus shifts from de- early investments of time and labor at the project level. At the product
signing and planning the project to the continuous development of level, the required early investment is still present, although it is re-
scalable modules and module interfaces. These interfaces are then uti- duced. Since DDC includes constant updates on the design model ac-
lized from project to project and even when updating the existing cording to the changes made during construction, it will also increase
building. In the case of the modular building contractor, the contractor the overall quality due to the regular checking of the design model's
owned the component designs and interface solutions and developed compatibility and built reality. As-built modeling-driven DDC enables a
them throughout his or her project portfolio. Modular solutions can be single building to be continuously improved over time. However,
an appropriate way of developing innovative design solutions and im- benefits across buildings are limited if the number of embedded op-
proving the cost and quality of the buildings [47]. Thus, although a erations instructions remains low.
modular product architecture may limit some design features, as the As-built modeling-based DDC seems to emerge in companies that
number of module and interface variants are finite, it would also pro- integrate other companies' products, while modular product archi-
vide a cost-efficient path for the DDC implementation because the tecture-driven DDC and algorithmic design-driven DDC are typically
modular product architecture would simplify the project management scaled up by companies with independent manufacturing capacities.
even before it is embedded into the digital design model. Fully modular product architecture-driven DDC in complex products
Algorithmic and parametric design-driven DDC relies on the in- can be achieved with either a vertical integration strategy or a network/
tegration between the early design phase and the engineering and supply chain management strategy, where the construction processes
production phases up to the level at which some operations—previously are flexible enough to accommodate the modular DDC approaches.
requiring human interpretation—are also automated. Parametric design As-built modeling-driven DDC is sufficiently flexible to be applied to
is an algorithm-aided process where various parameters are used to various environments. The state-of-the-art practices show that DDC is
control the design properties [48]. Automating the process of possible with both owner-driven and general contractor-driven

10
M. Tetik, et al. Automation in Construction 107 (2019) 102910

applications. In general, contractor-driven applications, newly added toward other paths to achieve automation and cyber-physical control
subcontractors, and the need for training can be seen as disadvantages. possibilities. From as-built modeling-driven logic, a company can de-
Moreover, continuous improvement can be constrained due to limited velop toward modular product architecture-driven DDC by re-
reuse opportunities. The modular building contractor is an example developing its best as-built solutions to achieve standardized solution
illustrating that the concept is also applicable by a single integrated components and their interfaces. Alternatively, the company can
actor. Modular product architecture-driven DDC is feasible in terms of choose to utilize its as-built designs to better direct and control en-
reuse and continuous improvement; however, problems can arise in gineering, manufacturing, and assembly operations, thus moving to-
terms of the uniqueness of the product when considering the end-user ward algorithmic and parametric DDC logic with computer-based
perspective. Algorithmic and parametric design-driven DDC, in con- control of operations. Similarly, the modular product-driven path can
trast, focuses more on investments in processes and their cyber-physical be extended to cyber-physical operations by developing assembly au-
control instead of product architecture. It is a powerful approach to tomation of modules or by utilizing sensors, which are embedded in the
reduce human interpretation and increase opportunities for reuse. Yet, physical product when controlling the use phase of the building. This
technically, it can limit the scope of the product and become con- combination requires further development in certain processes and
strained by the available direct manufacturing technologies. technologies, although it will enable the full use of DDC in future op-
erations.
8. Discussion Regarding the most advanced level of DDC application, this study
indicates that the problem of large investments of additional labor in
The developed design solution for DDC and the identified alter- the design phase in state-of-the-art BIM practices can be compensated
native paths to implement it suggest that the construction industry can for with algorithmic and parametric designs. When using parametric
increase the quality and efficiency of operations by adopting various systems [54], updating digital design models becomes less labor in-
elements of technology-based operations management practices. The tensive. Besides, algorithmic architecture [55] can be used in the design
practices from the field show that having a complete, up-to-date, and phase to create model drawings in a time-efficient way. Design models
continuously improving design model with embedded operational in- that are created algorithmically can compensate for labor costs by
formation and cyber-physical control in operations is feasible in the shortening the design phase and eventually, the project time. Moreover,
construction industry. platforms provide opportunities for reusing designs of components or
DDC should not be understood as an extreme solution that focuses buildings, which can also bring about savings in the design phase while
on technologies only, such as additive manufacturing, which are ap- still providing constructible design models. Thus, the drawbacks of
plicable only in a limited range of subsystems and operations. Some of investing in the design phase can be eliminated by exploring the less
the components of DDC are highly general and can be utilized in var- costly designs created via such technical solutions while implementing
ious building investments and products. Moreover, the industry or DDC. These solutions can be driven by the owners, manufacturers, and
supply chain can begin by building a subsystem and gradually moving general contractors, making sure that the subcontractors and manu-
toward completely design model-directed building investments. The facturers have the required production capabilities.
inefficient construction industry can be disrupted, although it is unclear Augmenting the current best practices of partial solutions opens up
how this can be achieved (i.e., through a technological revolution, with future automation opportunities for even more complete DDC solutions.
innovations in other fields, market disruption, or another means of Most of the cases represent opportunities for cyber-physical control of
change). Thus, this paper explored the potential approaches on how building part production based on the detailed design and even the
buildings may be built and used more efficiently in the meantime i.e. establishment of direct or automated logistics, installation, and main-
without any improvisations, rework, and interpretation, and with more tenance operations. Identifying a malfunction and using that informa-
reuse activities. tion in quality control and preventive maintenance of all steel products
The research identified three alternative paths to implement the is a notable example of those opportunities. Similarly, the log house
concept of DDC, and they can be utilized in sequence. The first level of design and construction solution are automated until the assembly
DDC is the as-built modeling-driven path. Reusing the as-built model in phase, and manual assembly instructions can be extracted from the
new projects is an appropriate way of obtaining a return on the in- program. The solutions can be augmented by directing future invest-
vestment made initially by utilizing the concept. The AEC industry has ments to areas in which potential benefits are the most promising. For
experienced problems with reusing digital data [51]. Increased reuse example, the log house solution can be extended to an automated de-
through the DDC concept in projects that are partially similar or sign model-directed installation and collection of status information of
identical [52] enables the required investment of time and effort to be log structures in the use phase.
recouped over many individual project lifecycles. The analysis of the Additive manufacturing and 3D printing present specific opportu-
interviews conducted with the state-of-the-art case companies revealed nities for the expansion of DDC. The 3D printing [56] that uses digital
that the process is the same among the projects; however, the content design models, such as 3D CAD models [57], enables manufacturing
changes from project to project. Thus, solution opportunities must seek operations to be directly controlled by the design model. Hence, 3D
the recouping of the initial investment inside a single project along with printing of buildings has potential mainly because the as-built model
the reusability of the solution in future projects in whole or in part. and reality are the same, as DDC suggests. To further improve this
The more advanced level of DDC favors offerings with more pro- practice, it is necessary to conduct the manufacturing operation and
ductized ways of manufacturing mainly because DDC can be used with other operations (e.g., logistics and maintenance) based on the same
modularized product offerings. Introducing modularization in manu- digital design model. This would ensure direct and digitally controlled
facturing when the design requirements are incomplete and fluctuate operations. The same 3D design model can be reused in another project.
from project to project is problematic [53]. Our analysis of the case Thus, using 3D printing technology in construction operations con-
examples yielded similar results. Having a completed representation of stitutes the potential for benefiting from DDC.
the construction product is one of the constructs of DDC. Moreover, The theoretical contribution of this research involves combining
similar to the modular production offerings, prefabricated structures previous knowledge of as-built BIM, VDC, and DDM with the operations
can also be used when DDC is applied. The prefabrication of building management of construction projects by developing the concept of
parts leads to continuous improvements and the industrialization of the DDC. The research broadens the possibilities of using as-built BIM
construction [11]. Thus, DDC creates opportunities for industrialized models, VDC practices, and novel design methods (e.g., parametric,
construction with the offerings of modular and prefabricated products. algorithm, and modular design) by better connecting the design con-
Owners and contractors can use the as-built driven path to gravitate cepts with all other operations of the building lifecycle. The

11
M. Tetik, et al. Automation in Construction 107 (2019) 102910

contribution and novelty of this study do not lie in the technology but or product part of construction industry operations but also in the entire
instead in how the proposed technologies, such as Construction 4.0 construction supply chain over its lifecycle. Evidence from the industry
[27], can be used to improve construction operations management in examples that already use the operational practice elements illustrate
supply chains and firms' processes that are not project specific. The the solution potential and feasibility of the concept. The DDC fills an
connection between design and operations can be formed by embed- existing gap in the technology-based construction operations practices
ding the operation instructions into the digital design model. In addi- and creates additional value by removing inefficiencies and establishing
tion, the research shows that novel design methods, such as algorithmic a continuously improving way of designing, engineering, producing,
design, can fix labor problems encountered in the state-of-the-art as- and maintaining buildings.
built BIM practices.
For practitioners, this research provides avenues to develop pro- Acknowledgements
duction strategies and inherent technological capabilities toward higher
quality and more efficient operations. The research underlines the need This research received funding from Academy of Finland with de-
for integration between design, production, and maintenance opera- cision number 275839 and Tekes with grant number 2758/31/2015.
tions in such development efforts. The DDC concept and its im-
plementation paths suggest how construction practitioners can benefit References
from DDM principles and cyber-physical operations originally devel-
oped in the manufacturing industry. This study describes the logic of [1] S. Xu, H. Luo, The information-related time loss on construction sites: a case study
how investments in design-based operations can pay back in buildings' on two sites, Int. J. Adv. Robot. Syst. 11 (8) (2014) 128 https://doi.org/10.5772/
58444.
lifecycle and investments of future projects. The study also highlights [2] A. Dubois, L.E. Gadde, The construction industry as a loosely coupled system: im-
the role of diminished human interpretation, decreased improvisation plications for productivity and innovation, Constr. Manag. Econ. 20 (7) (2002)
and reuse, and continuous improvement of designs and control systems 621–631, https://doi.org/10.1080/01446190210163543.
[3] T. Wegelius-Lehtonen, Performance measurement in construction logistics, Int. J.
as key mechanisms to quality and cost benefits of the DDC concept. Prod. Econ. 69 (1) (2001) 107–116, https://doi.org/10.1016/S0925-5273(00)
00034-7.
9. Limitations and further research [4] J. Cheng, K. Law, H. Bjornsson, A. Jones, R. Sriram, A service oriented framework
for construction supply chain integration, Autom. Constr. 19 (2) (2010) 245–260,
https://doi.org/10.1016/j.autcon.2009.10.003.
Despite delving into the existing literature on technology-based [5] J. Underwood (Ed.), Handbook of Research on Building Information Modeling and
practices and using multiple cases to increase the validity of the in- Construction Informatics: Concepts and Technologies: Concepts and Technologies,
IGI Global, 2009.
vented concept, this research has several limitations, which can be
[6] F.R. Hamzeh, H. Alhussein, F. Faek, Investigating the practice of improvisation in
addressed in further research. First, our concept validation is mostly construction, J. Manag. Eng. 34 (6) (2018) 4018039, https://doi.org/10.1061/
conceptual and theoretical. In terms of design science research, the (ASCE)ME.1943-5479.0000652.
practice is a design proposal and hardly a field-tested and evidence- [7] F.R. Hamzeh, F. Abi Morshed, H. Jalwan, I. Saab, Is improvisation compatible with
lookahead planning? An exploratory study, Proc. 20th Annual Conference
based design as yet. We acknowledge this limitation and appreciate the International Group for Lean Construction, 20 IGLC, 2012, https://www.
need for in-depth research into the field-testing and empirical evidence researchgate.net/profile/Farook_Hamzeh/publication/242346673_Is_
of the invented DDC concept. Improvisation_Compatible_with_Lookahead_Planning_An_Exploratory_Study/links/
00b7d51cd9a35ba1f4000000.pdf.
During the development of this paper, we did not have much in- [8] J.G. Geraldi, L. Lee-Kelley, E. Kutsch, The Titanic sunk, so what? Project manager
formation about the actual time and cost savings from the case com- response to unexpected events, Int. J. Proj. Manag. 28 (6) (2010) 547–558, https://
panies. The application and measurement of the benefits of the DDC doi.org/10.1016/j.ijproman.2009.10.008.
[9] R.B. Johnston, M. Brennan, Planning or organizing: the implications of theories of
concept should be researched in the future. The scope of the partial case activity for management of operations, Omega 24 (4) (1996) 367–384, https://doi.
examples is limited due to various material and technical factors. org/10.1016/0305-0483(96)00016-3.
Moreover, the customer value and architectural perspectives should be [10] C. Merschbrock, F. Wahid, Actors’ freedom of enactment in a loosely coupled
system: The use of building information modelling in construction projects, ECIS
considered in terms of the reuse of the same designs in different projects European Conference on Information Systems, 2013, p. 107 http://aisel.aisnet.org/
as customers may prefer original designs. Also, the ownership issues ecis2013_cr/107.
regarding the product designs and control system in DDC require fur- [11] J. Lessing, L. Stehn, A. Ekholm, Industrialised house-building–development and
conceptual orientation of the field, Constr. Innov. 15 (3) (2015) 378–399, https://
ther research. On a different albeit related note, our case studies in-
doi.org/10.1108/CI-06-2014-0032.
cluded cases from different countries. This approach may have in- [12] J. Lessing, Industrialised house-building, Concept and Processes, 2006 978-91-631-
troduced country- and culture-specific biases into our findings. 9254-8.
The construction industry's adaptation of the DDC concept requires [13] J. Holmström, M. Ketokivi, A.P. Hameri, Bridging practice and theory: a design
science approach, Decis. Sci. 40 (1) (2009) 65–87, https://doi.org/10.1111/j.1540-
increased effort from individual organizations and existing company 5915.2008.00221.x.
networks in terms of overcoming shortsightedness and recognizing the [14] S. Gregor, D. Jones, The anatomy of a design theory, J. Assoc. Inf. Syst. 8 (5) (2007)
lifecycle benefits of its implementation. Future research should further 312–323 http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.232.743&
rep=rep1&type=pdf.
analyze the existing and emerging examples of DDC applications by [15] R. Rajkumar, I. Lee, L. Sha, J. Stankovic, Cyber-physical systems: The next com-
utilizing objective data to evaluate the long-term impact of DDC in puting revolution, Design Automation Conference, IEEE, 2010, pp. 731–736, ,
construction operations. More knowledge is also required concerning https://doi.org/10.1145/1837274.1837461.
[16] A. Akanmu, C.J. Anumba, Cyber-physical systems integration of building in-
the endeavors of traditional AEC companies in moving toward DDC formation models and the physical construction, Eng. Constr. Archit. Manag. 22 (5)
principles in their operations. In general, further research is advised to (2015) 516–535, https://doi.org/10.1108/ECAM-07-2014-0097.
concentrate on measuring the benefits gained from the DDC concept. [17] B. Akinci, F. Boukamp, C. Gordon, D. Huber, C. Lyons, K. Park, A formalism f or
utilization of sensor systems and integrated project models for active construction
quality control, Autom. Constr. 15 (2) (2006) 124–138, https://doi.org/10.1016/j.
10. Conclusion autcon.2005.01.008.
[18] L. Chen, H. Luo, A BIM-based construction quality management model and its ap-
plications, Autom. Constr. 46 (2014) 64–73, https://doi.org/10.1016/j.autcon.
In this research, a new practice for technology-based management
2014.05.009.
of construction operations and supply chains was developed through [19] M. Golparvar-Fard, F. Peña-Mora, S. Savarese, D4AR–a 4-dimensional augmented
design science research. The core principle of the practice is to direct reality model for automating construction progress monitoring data collection,
the value-adding operations over the building's lifecycle through the processing and communication, J. Inf. Technol. Construct. 14 (13) (2009) 129–153
https://pdfs.semanticscholar.org/338e/
digital design model, resulting in increased reusability, project-specific 04a889d11a6c6d46827dada3e67dda1833ce.pdf.
differentiation, and automation of designs and processes across pro- [20] F. Bosché, Automated recognition of 3D CAD model objects in laser scans and
jects. The DDC concept improves efficiency, not just in a limited project calculation of as-built dimensions for dimensional compliance control in

12
M. Tetik, et al. Automation in Construction 107 (2019) 102910

construction, Adv. Eng. Inform. 24 (1) (2010) 107–118, https://doi.org/10.1016/j. (2007) 45–77, https://doi.org/10.2753/MIS0742-1222240302.
aei.2009.08.006. [40] K.A. Piirainen, R.O. Briggs, Design theory in practice–making design science re-
[21] P. Tang, D. Huber, B. Akinci, R. Lipman, A. Lytle, Automatic reconstruction of as- search more transparent, International Conference on Design Science Research in
built building information models from laser-scanned point clouds: a review of Information Systems, Springer, Berlin, Heidelberg, 2011, pp. 47–61, , https://doi.
related techniques, Autom. Constr. 19 (7) (2010) 829–843, https://doi.org/10. org/10.1007/978-3-642-20633-7_4.
1016/j.autcon.2010.06.007. [41] M. Denscombe, The Good Research Guide: For Small-Scale Social Research Projects,
[22] K.K. Han, M. Golparvar-Fard, Potential of big visual data and building information 1st ed., McGraw-Hill, 2014 (ISBN-13: 978-0-335-26470-4).
modeling for construction performance analytics: an exploratory study, Autom. [42] D.M. McCutcheon, J.R. Meredith, Conducting case study research in operations
Constr. 73 (2017) 184–198, https://doi.org/10.1016/j.autcon.2016.11.004. management, J. Oper. Manag. 11 (3) (1993) 239–256, https://doi.org/10.1016/
[23] A.L. Ahmed, M. Kassem, A unified BIM adoption taxonomy: conceptual develop- 0272-6963(93)90002-7.
ment, empirical validation and application, Autom. Constr. 96 (2018) 103–127, [43] S. Brinkmann, Interview, Encyclopedia of Critical Psychology, Springer New York,
https://doi.org/10.1016/j.autcon.2018.08.017. 2014, pp. 1008–1010, , https://doi.org/10.1007/978-1-4614-5583-7_161.
[24] A. Azhar, Building information modeling (BIM): trends, benefits, risks, and chal- [44] C. Alexander, The Timeless Way of Building, 1 Oxford University Press, New York,
lenges for the AEC industry, Leadersh. Manag. Eng. 11 (3) (2011) 241–252, https:// 1979.
doi.org/10.1061/(ASCE)LM.1943-5630.0000127. [45] M. Hamid, O. Tolba, A. El Antably, BIM semantics for digital fabrication: a
[25] D. Davis, LEAN, Green and Seen, J. Build. Inf. Mod. 1 (1) (2007) 16–18 https:// knowledge-based approach, Autom. Constr. 91 (2018) 62–82, https://doi.org/10.
www.1stpricing.com/pdf/jbim_fall07.pdf. 1016/j.autcon.2018.02.031.
[26] J. Kunz, M. Fischer, Virtual Design and Construction: Themes, Case Studies and [46] K. Främling, T. Ala-Risku, M. Kärkkäinen, J. Holmström, Design patterns for
Implementation Suggestions, Center for Integrated Facility Engineering (CIFE), managing product life cycle information, Commun. Assoc. Comp. Mach. 50 (6)
Stanford University, 2009, https://stacks.stanford.edu/file/druid:gg301vb3551/ (2007) 75–79 http://delivery.acm.org/10.1145/1250000/1247009/p75-
WP097_0.pdf. karkkainen.pdf?ip=130.233.138.62&id=1247009&acc=ACTIVE%20SERVICE&
[27] G.P. Luth, VDC and the engineering continuum, J. Constr. Eng. Manag. 137 (10) key=74A0E95D84AAE420%2E1D2200B0A299939C%2E4D4702B0C3E38B35%
(2011) 906–915, https://doi.org/10.1061/(ASCE)CO.1943-7862.0000359. 2E4D4702B0C3E38B35&__acm__=1548327527_
[28] Y. Zhang, G. Wang, Cooperation between building information modeling and in- 7887bbca2cfe9a935681269b3f3b0967.
tegrated project delivery method leads to paradigm shift of AEC industry, [47] A. Peltokorpi, H. Olivieri, A.D. Granja, O. Seppänen, Categorizing modularization
International Conference on Management and Service Science, IEEE, 2009, , strategies to achieve various objectives of building investments, Constr. Manag.
https://doi.org/10.1109/ICMSS.2009.5305337. Econ. 36 (1) (2018) 32–48, https://doi.org/10.1080/01446193.2017.1353119.
[29] R. Sacks, L. Koskela, B.A. Dave, R. Owen, Interaction of lean and building in- [48] A. Lalla, I. Pirhonen, Parametric design of structural frames – a leap towards a fully
formation modeling in construction, J. Constr. Eng. Manag. 136 (9) (2010) automated structural design process, 17th International Conference on Computing
968–980. in Civil and Building Engineering, 2018 Tampere, Finland, June 2018 https://
[30] J. Holmström, M. Holweg, S. Khajavi, J. Partanen, The direct digital manufacturing icccbe2018.exordo.com/files/papers/106/final_draft/20180321_paper.pdf.
(r)evolution: definition of a research agenda, Oper. Manag. Res. 9 (1–2) (2016) [49] J. Manrique, M. Al-Hussein, A. Bouferguene, R. Nasseri, Automated generation of
1–10, https://doi.org/10.1007/s12063-016-0106-z. shop drawings in residential construction, Autom. Constr. 55 (2015) 15–24,
[31] I. Gibson, D.W. Rosen, B. Stucker, Additive Manufacturing Technologies, 238 https://doi.org/10.1016/j.autcon.2015.03.004.
Springer, New York, 2010, https://doi.org/10.1007/978-1-4419-1120-9. [50] A. Turrin, P. Von Buelow, R. Stouffs, Design explorations of performance driven
[32] D. Chen, S. Heyer, S. Ibbotson, K. Salonitis, J. Steingrímsson, S. Thiede, Direct di- geometry in architectural design using parametric modeling and genetic algorithms,
gital manufacturing: definition, evolution, and sustainability implications, J. Clean. Adv. Eng. Inform. 25 (4) (2011) 656–675, https://doi.org/10.1016/j.aei.2011.07.
Prod. 107 (2015) 615–625, https://doi.org/10.1016/j.jclepro.2015.05.009. 009.
[33] S. Khajavi, J. Partanen, J. Holmström, Additive manufacturing in the spare parts [51] C. Merschbrock, T. Tollnes, C. Nordahl-Rolfsen, Solution selection in digital con-
supply chain, Comput. Ind. 65 (1) (2014) 50–63, https://doi.org/10.1016/j. struction design – a lazy user theory perspective, Proc. Eng. 123 (2015) 316–324,
compind.2013.07.008. https://doi.org/10.1016/j.proeng.2015.10.097.
[34] T.D. Oesterreich, F. Teuteberg, Understanding the implications of digitisation and [52] V. Popov, V. Juocevicius, D. Migilinskas, L. Ustinovichius, S. Mikalauskas, The use
automation in the context of industry 4.0: a triangulation approach and elements of of a virtual building design and construction model for developing an effective
a research agenda for the construction industry, Comput. Ind. 83 (2016) 121–139, project concept in 5D environment, Autom. Constr. 19 (3) (2010) 357–367, https://
https://doi.org/10.1016/j.compind.2016.09.006. doi.org/10.1016/j.autcon.2009.12.005.
[35] B. Ilhan, H. Yaman, Green building assessment tool (GBAT) for integrated BIM- [53] P. Jensen, T. Olofsson, M. Sandberg, L. Malmgren, Reducing complexity of custo-
based design decisions, Autom. Constr. 70 (2016) 26–37, https://doi.org/10.1016/ mized prefabricated buildings through modularization and IT support, International
j.autcon.2016.05.001. Conference on Informations Technology in Construction, Universidad de Talca,
[36] B. Dave, S. Kubler, E. Pikas, J. Holmström, V. Singh, K. Främling, L. Koskela, 2008, pp. 429–437 http://www.diva-portal.org/smash/get/diva2:1000974/
Intelligent products: Shifting the production control logic in construction (with lean FULLTEXT01.pdf.
and BIM), Proceedings of the 23rd Annual Conference of the International Group for [54] J.J. Shah, M. Mäntylä, Parametric and Feature-Based CAD/CAM: Concepts,
Lean Construction, International Group for Lean Construction, Perth, Australia, Techniques, and Applications, John Wiley & Sons, 1995.
201529-31 July 2015 https://aaltodoc.aalto.fi/handle/123456789/17643. [55] K. Terzidis, Algorithmic Architecture, Routledge, 2006, https://doi.org/10.4324/
[37] L. Ding, R. Wei, H. Che, Development of a BIM-based automated construction 9780080461298.
system, Proc. Eng. 85 (2014) 123–131, https://doi.org/10.1016/j.proeng.2014.10. [56] H. Lipson, M. Kurman, Fabricated: The New World of 3D Printing, John Wiley &
536. Sons, 2013.
[38] N. Labonnote, A. Rønnquist, B. Manum, P. Rüther, Additive construction: state-of- [57] F. Rengier, A. Mehndiratta, H. Von Tengg-Kobligk, C.M. Zechmann,
the-art, challenges and opportunities, Autom. Constr. 72 (2016) 347–366, https:// R. Unterhinninghofen, H.U. Kauczor, F.L. Giesel, 3D printing based on imaging
doi.org/10.1016/j.autcon.2016.08.026. data: review of medical applications, Int. J. Comput. Assist. Radiol. Surg. 5 (4)
[39] K. Peffers, T. Tuunanen, M.A. Rothenberger, S. Chatterjee, A design science re- (2010) 335–341, https://doi.org/10.1007/s11548-010-0476-x.
search methodology for information systems research, J. Manag. Inf. Syst. 24 (3)

13