Automation in Construction 72 (2016) 347–366

Contents lists available at ScienceDirect

Automation in Construction

journal homepage: www.elsevier.com/locate/autcon

Review

Additive construction: State-of-the-art, challenges and opportunities

Nathalie Labonnote a,⁎, Anders Rønnquist b, Bendik Manum c, Petra Rüther a
a
SINTEF Building and Infrastructure, PO Box 4760, Sluppen, NO-7465 Trondheim, Norway
b
Structural Engineering Department, Norwegian University of Science and Technology, Richard Birkelands vei 1a, Trondheim, N-7491, Norway
c
Department of Architectural Design, History and Technology, Norwegian University of Science and Technology, Alfred Getz vei 3, N-7491, Norway

a r t i c l e i n f o a b s t r a c t

Article history: The present study has investigated to what extent additive manufacturing technologies can be successfully applied
Received 18 December 2015 to the construction of large-scale structures. The central concept of additive construction was defined, and a system-
Received in revised form 23 June 2016 atic mapping study was performed in order to assemble relevant publications selected according to a well-defined
Accepted 17 August 2016
set of criteria. Knowledge gathered from the relevant publications was organised into four main categories: material
Available online 21 September 2016
science, engineering, building design and market analysis. The lack of focus of material science research towards the
Keywords:
characterisation and potential improvement of construction-related material properties has been emphasised. The
3D printing evolution of technological solutions to deposit the construction materials from gantry solutions to more lightweight
Additive manufacturing systems has been described. The governing parameters for deciding on the most appropriate solutions have been
Building design identified as the type of building component, the location for production, and the assembly technique. Benefits of
Structural design additive construction for building design were shown to mostly address the perspective of end-users, but should in-
Buildings stead be understood as the emergence of new opportunities and new constraints that will necessitate a greater de-
Civil engineering gree of rational decision-making in the design phase. The relevant markets for additive construction were shown to
Architectural design
be closely related to the inherent specificities of the project in question. This implies that additive construction can
Building materials
be successfully applied in connection with general housing projects only if housing in general changes to become
Additive construction
Parametric design more optimised and more individualised. It was concluded that additive construction has the potential to revolu-
Multiconstraint design tionise the construction industry, its success depending on how the whole building industry is ready to tackle
LCA three challenges: the need for an architectural paradigm shift, the need for a holistic design process, and the need
for rational designs. A list of suggestions for further research is provided, among them the development of tools
for assessing the disruptive potential of additive construction in an objective and scientific way.
© 2016 Elsevier B.V. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 348
1.1. General background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 348
1.2. Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 348
1.3. Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 348
2. Methodology for the systematic mapping study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 349
2.1. Search overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 349
2.2. Screening of publications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 349
2.3. Keywording and the mapping process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 349
3. Material science challenges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 350
3.1. Relevant materials and processes for additive construction: state-of-the-art . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 350
3.2. Process-related material requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351
3.3. Construction-related material requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352
3.4. How non-homogeneous materials should be used . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353
3.5. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353
4. Relevant technological solutions for additive construction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354
4.1. Gantry solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354
4.2. Cable-suspended platforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354

⁎ Corresponding author.
E-mail addresses: nathalie.labonnote@sintef.no (N. Labonnote), anders.ronnquist@ntnu.no (A. Rønnquist), bendik.manum@ntnu.no (B. Manum), petra.ruther@sintef.no (P. Rüther).

http://dx.doi.org/10.1016/j.autcon.2016.08.026
0926-5805/© 2016 Elsevier B.V. All rights reserved.
348 N. Labonnote et al. / Automation in Construction 72 (2016) 347–366

4.3. Swarm approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 355

4.4. Multi-purpose robotics and automated assembly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 356
4.5. Folding and self-assembly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 356
4.6. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 357
5. Opportunities for additive construction in connection with building design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 357
5.1. Complex geometries at a structural scale . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 357
5.2. Optimised topologies provides potential individualisation and contextualisation at no cost . . . . . . . . . . . . . . . . . . . . . . . 358
5.3. “Smart” constructions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 358
5.4. Additive construction – the continued digital workflow from design process to building process? . . . . . . . . . . . . . . . . . . . . 359
5.5. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 359
6. Relevant fields of application for additive construction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 360
6.1. Additive construction for harsh environments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 360
6.2. Additive construction for “common” environments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 360
6.3. Limitations and open questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 361
6.4. Specific projects and demonstrations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 361
6.5. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 362
7. General discussion: challenges and opportunities for additive construction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 363
7.1. The need for an architectural paradigm shift . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 363
7.2. The need for a more holistic design process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 363
7.3. The need for (even more) rational designs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 363
8. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 364
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 364

1. Introduction ○ a digital 3D model is created by dedicated software, or by the scan-

1.1. General background
Gartner [1], one of the world's leading information technology
research companies, identified 3D printing both as “a rare example
of a single technology that has become truly disruptive by itself”,
and as one of the five emerging technology trends that are believed
will significantly impact business during the period 2014–2019. 3D
printers are currently claimed to be able to shorten design and de-
velopment cycles, to improve communication and collaboration,
and to resolve issues in the realm between of design and engineer-
ing [2]. Although 3D printing has discretely been applied in
prototyping for many years, the term “3D printing” has now be-
come a widely-used general term for additive methods, indepen-
dent of specific technology, material and intended application [3].
New clothes, innovative medical implants, better-tasting pizzas,
and futuristic houses are just some of the numerous media-
heralded “3D printing achievements”.
In contrast, labour productivity in the construction industry has ex-
hibited either a decline or stagnation over the past fifty years [4]. The
causes are numerous and include factors such as labourers' resistance
to change (due to excessively long-term payback), poor data interop-
erability, declining real labour costs (which tend to reduce investment
in capital equipment), and high levels of turnover at the bottom end of
the industry (which make it more difficult to implement new
methods) [4]. Up until a few years ago, 3D printing applications in
the construction industry were largely confined to the production of
affordable architectural models. However, more recently, several spec-
tacular attempts to 3D print complete houses have been the subject of
much publicity.
The primary aim of this study is to investigate the extent to which 3D
printing technologies can be successfully applied to the construction of
large-scale structures. US President Barack Obama was reported to have
said [5] that 3D printing is a technology with “the potential to revolu-
tionise the way we make almost everything”. So, is 3D printing also ca-
pable of revolutionising the construction industry?
1.2. Definitions
In the present study the following definitions are used:
• “3D printing” refers to the various processes used to synthesise a
three-dimensional object. The general process is as follows:
ning of an existing object,
○ an algorithm cuts the digital model into 2D slices, and
○ a “printer” prints the object, slice by slice, according to the dimen-
sions of digital 3D model.
In 2009, “3D printing” was defined more specifically by the Ameri-
can Society for Testing and Materials (ASTM) as “the fabrication of
objects through the deposition of a material using a print head, noz-
zle, or another printer technology” [6].
• The term “additive manufacturing” is preferred in this study because
this concept is more generally defined by the ASTM as: “the process
of joining materials to make objects from 3D model data, usually
layer upon layer, as opposed to subtractive manufacturing methodol-
ogies” [6].
• “Construction” is here defined as a general term describing the act of
generating large-scale structures whose purpose is linked to structural
engineering. Such structures may include large civil engineering prod-
ucts such as houses, bridges, architectural pavilions and other types of
building.
• “Additive construction” has a similar definition to “additive manu-
facturing”. It is described as “the process of joining materials to create
constructions from 3D model data”. This means that the design, produc-
tion and/or assembly processes should be digitally controlled at least to
some extent.
1.3. Objectives
Additive construction represents the entire process of building a dig-
ital form (building design) from materials produced on-site (material
science), which are then deposited according to a digital model (engi-
neering). As is the case for any other new technology, the success of ad-
ditive construction depends on both technological progress and
commercial relevance.
The following questions are thus of major interest:
• What construction-specific material science challenges do we face?
• What structural engineering challenges come into play when scaling-
up additive manufacturing?
• What building design opportunities emerge when using additive
construction?
 N. Labonnote et al. / Automation in Construction 72 (2016) 347–366 349

• What are the requirements for a successful (marketable) concept for Table 1
use in the building industry? Keywords and Boolean operators used for the initial search.

How? AND What? AND NOT To exclude

“3D printing” “Building construction” Food
In order to answer these questions, this study is based on the results “4D printing” “Civil engineering” Organ
“Additive manufacturing” “Building technology” Medical
of a systematic mapping study, the methodology for which is described “Contour crafting” “Building fabrication” DNA
in Section 2. The states-of-the-art for each of the research fields current- “Swarm printing” Skyscraper Tissue
ly associated with additive construction are summarised in Section 3 “CNC weaving” “Large scale”
(Material Science), Section 4 (Engineering) and Section 5 (Building De- “3D weaving” “Building components”
“Buildings”
sign). An analysis of the market is provided in Section 6. A short discus-
House
sion is provided at the end of each section. The paper concludes with Dwelling
general discussion (Section 7) that argues for a more holistic design pro- Bridge
cess. Fig. 1 summarises the structure of the present study. Construction
Architecture

2. Methodology for the systematic mapping study

The present study is based on a systematic mapping study designed
to carry out an objective selection of literature sources relevant to the
topic in question. Petersen et al. [7] described how to conduct a system-
atic mapping study in connection with software engineering, and their
proposed framework has since been implemented in other fields [8].
The same methodology is adhered to in the present study.
2.2. Screening of publications
The screening of publications was performed on the basis of the fol-
lowing inclusion and exclusion criteria with the aim of excluding publi-
cations that are not relevant to the research questions set out in Section
1.3. The qualitative inclusion criterion was as follows:

- English language
2.1. Search overview

In the present study, the following five main abstract and citation
databases of peer-reviewed literature were considered: Scopus, Web
of Science, Direct Science, Engineering Village and IEEE. The search
string was structured using Boolean operators to refine the search pro-
cess in terms of action description (“how”) and object (“what”), and
corresponds to the keywords set out in Table 1.
The search was performed on 7 September 2015, and produced the
following results:
The scientific inclusion criteria were as follows:
- The technology under study should correspond to the definition of
additive manufacturing provided in Section 1.2, and
- The object to which the study in question is applied should corre-
spond to any construction within the scope of the construction in-
dustry as defined in Section 1.2. This excludes publications related
to prototyping.

- Scopus: 674 publications

- Web of Science: 276 publications After a screening process based on titles and abstracts, a total of 165
- Direct Science: 200 publications publications were deemed to be relevant to the research questions. This
- Engineering Village: 359 publications number is surprisingly low, particularly in the light of the extensive
- IEEE: 483 publications media coverage given to additive construction. Fig. 2 illustrates the re-
cent marked increase in publication numbers. Prior to 2012, the subject
was confined to somewhat confidential sources and readerships. Only
Seven additional publications from academic sources and 42 web very few scientific publications appeared each year. However, since
pages were also included in the results. After the removal of duplicates, 2012, the number of publications addressing the topic has increased sig-
a total of 1560 publications were considered for screening. nificantly both in scientific and peer-reviewed works, and also in the
popular science media (web pages and magazines). For 2015, the latter
is more important than the former.

2.3. Keywording and the mapping process

The 165 publications were then subjected to keywording and classi-

fication based on their full text content. The keywording was performed
in two steps:

- Firstly, abstracts were read in order to identify keywords and con-

cepts that reflect the contribution of the publication in question.
- Secondly, all identified keywords were combined in order to develop
a high-level understanding of the field of additive manufacturing as
it applies to the construction industry.

Table 2 provides an overview of the main categories identified and

Fig. 1. Structure of the present study.
their sub-levels. The classification scheme used here reflects the objec-
tives set out in Section 1.3 (see Fig. 1).
Each publication was mapped into the relevant categories using
both manual and automated methods, including text queries. All the re-
sults were then reviewed manually in order to ensure consistency and
to remove “false” or biased occurrences.
350 N. Labonnote et al. / Automation in Construction 72 (2016) 347–366

Fig. 2. Evolution of number and type of publications during the last twenty years.

The following sections of this paper adhere to the classification given
in Table 2
- Section 3 – state-of-the-art and challenges related to material sci-
ence,
- Section 4 – state-of-the-art and challenges related to engineering,
- Section 5 – opportunities linked to building design,
- Section 6 – opportunities for the construction market.
found in Barnatt's book [9]. Nevertheless, a brief overview of the differ-
ent types of additive manufacturing processes is given here, focusing on
the state of the material before deposition, and according to its rele-
vance to the additive construction process; a) solid-like (Fig. 3a), b) vis-
cous-like (Fig. 3b), c) powder-like (Fig. 3c), and d) liquid-like (Fig. 3d).
The potential relevance and possible applications of each additive
manufacturing process in relation to additive construction are briefly
discussed in the following.

• Additive manufacturing processes based on solid-like materials. These

A short discussion is provided at the end of each section, and the involve the layer-wise assembly of material in solid form. Bonding be-
paper is concluded with a general discussion (Section 7) and conclu- tween the different layers is achieved by a number of different
sions (Section 8). methods, such as by the use of a glue-like material, or as an integral
part of the assembly process itself e.g. weaving. Masonry is probably
3. Material science challenges one of the oldest additive construction processes based on solid-like
materials. Building a brick wall using mortar to bond the layers was
3.1. Relevant materials and processes for additive construction: state-of- an additive manufacturing process well before the development of
the-art 3D printers. Other innovative processes such as the 3D weaving of car-
bon fibres have been used in applications from vehicle body frames to
A detailed review of all additive manufacturing processes is clearly critical aircraft parts [10], and can be envisaged as a means of produc-
beyond of the scope of this paper. However, such a review can be ing highly optimised building components.
• Additive manufacturing processes based on viscous-like mate-
rials. These involve the deposition of material in a viscous liquid
Table 2 form via a printing nozzle. They are thus also referred to as “extru-
Classification scheme for categories identified during the mapping study. sion-based processes”. Solidification of the material is achieved by
Discipline Most investigated topics
curing following extrusion. Well known and related additive
manufacturing processes include Fused Deposition Modelling, Inkjet
Material Science Production process
and PolyJet.
Process-related material requirements
Construction-related material requirements
Non-homogeneous materials
Engineering Gantry solutions This is by far the most investigated additive construction process.
Cable-suspended solutions Relevant materials commonly comprise a combination of paste and
Swarm solutions bulk materials. Bulk materials may be composed of:
Robotics solutions
Solutions combined with folding - natural aggregates such as soil, sand, natural gravel, crushed stone,
Building Design Freeform capacity
clay or mud
Optimisation at no cost
Embedded functionalities - recycled aggregates such as those from construction, demolition or
Digitalisation excavation waste
Market Analysis Potential applications - manufactured aggregates such as air-cooled blast furnace slag and
Current or planned projects bottom ash
Current limitations
- natural fibres, such as cellulose and recycled wood fibre
 N. Labonnote et al. / Automation in Construction 72 (2016) 347–366
Fig. 3. The different states of a material before deposition; a) solid-like, b) viscous-like, c) powder-like and d) liquid-like
Bulk material is combined with a binding paste that may be com-
posed of:
- cements consisting of mixtures of oxides of calcium, silicon and alu-
minium
- polymer blend
In the rest of this paper, materials created by means of an extrusion-
based process using a cement-like paste will be referred to as “concrete-
type” materials. Similarly, materials using a polymer-like paste will be
referred to as “plastic-type” materials.
Both “Contour Crafting” [11] and “Freeform construction” [12] (also
known as “Concrete Printing”) have been using this process extensively
for concrete-type materials. The Freeform construction technique is re-
ported to enable the additional deposition of fine details using the same
process [13], while Contour Crafting can achieve lateral smoothing of a
surface by means of a trowel [14]. Polymer blend was used by Hunt et
al. [15] to extrude thermoplastic from a heatable extrusion head to facil-
itate the construction of larger structures.
• Additive manufacturing processes based on powder-like materials.
This involves transforming a material from a powder to a solid state.
This is achieved either by melting or sintering, enabled by an energy
source (usually a laser or electron beam), or curing by means of a
chemical reaction. Once the object is finished, the remaining powder
can be removed and recycled for re-use.
The powder-bed process builds up a solid object within a pile of
powder. Well-known and related additive manufacturing processes in-
clude Laser Sintering, Power Binding Printing, Selective Laser Melting
and Selective Laser Sintering.
The powder-feed process uses an energy source to generate a melt-
ing bath on a surface. The powder is then sprayed into the melting bath
through nozzles. This method has been used mostly for metals and the
repair of tools. Some combination of materials is possible using this ap-
proach. Well-known and related additive manufacturing processes in-
clude Laser Powder Deposition, Direct Metal Deposition, Laser Metal
Deposition, Direct Laser Deposition, Direct Metal Laser Deposition, Di-
rect Laser Deposition, and Laser Engineered Net Shape.
Mostly, this process has been applied in additive construction by de-
veloping a powdered concrete mix, subsequently cured by hydration
using an ink jet spray [16]. Dini developed his D-shape printer based
on the same process using sand, stone powder and an inorganic binder
[17]. In 2007, he built what is generally considered to be the first large-
scale additively manufactured structure – Radiolaria (see also Section
6.4). Sintering, melting and exothermic chemical reactions have also
been investigated for additive construction applications in extra-terres-
trial environments [18–20].
• Additive manufacturing processes based on liquid-like materials. This
involves transforming a material from a liquid to a solid state. Curing
is achieved by means of a light source. A well-known and related
351
additive manufacturing process is Stereolithography. To the best of
the authors' knowledge, no application of this form of additive con-
struction has been investigated to date.
Materials considered for additive construction are required to meet
distinct and sometimes contradictory requirements related both to the
desirable properties essential for an efficient additive manufacturing
process, and those demanded by their efficient application in the con-
struction project in question.
3.2. Process-related material requirements
This paper has investigated process-related requirements primarily
in the case of extrusion-based processes (as defined in Section 3.1).
Lim et al. [21] discussed large-scale additive manufacturing processes
that have been applied in the fields of construction and architecture,
and identified the following four key extrusion-based process-related
characteristics:
• Pumpability: defined as the ease and reliability with which material is
moved through the delivery system (Fig. 4a),
• Printability (sometimes also reported as extrudability [22] for mate-
rials used in extrusion-based processes): defined as the ease and reli-
ability of depositing material through a deposition device (Fig. 4b),
• Buildability: defined as the resistance of deposited wet material to de-
formation under loads. Le at al. [22] suggested quantifying buildability
as the number of filament layers which can be added without signifi-
cant deformation of the lower layers (Fig. 4c),
• Open time: defined as the period during which the aforementioned
properties remain consistently within acceptable tolerances (Fig. 4d).
To some extent, both concrete-type and plastic materials meet the
requirements set out in the above. With a few exceptions, most large-
scale additive manufacturing technologies involving concrete do not
use frameworks to confine the concrete, but instead deposit it through
a nozzle to form successive layers along a pre-defined fabrication
path. In the absence of a skeleton, freshly-layered concrete must be
load-bearing immediately upon placement [23], and this demands un-
precedented structural requirements.
For this reason, the trade-off between printability and buildability
has often been reported as the most critical property [3,22–26] because
of the risk of development of potential flaws between extrusions that
can induce stress concentrations. Bond strength is highly dependent
on adhesion, which is a function of the time interval between extru-
sions. For this reason, the base material should be formulated in such
a way that individual layers are able to bond with one another (an
open time that is short enough), and yet which still have time to cure
(an open time that is long enough). This enables the creation of a single
monolithic form with sufficient rigidity to support its own weight [3].
Different approaches have been employed to solve this problem:
352 N. Labonnote et al. / Automation in Construction 72 (2016) 347–366
Fig. 4. Extrusion-based process-related material requirements. a) Pumpability, b) printability, c) buildability and d) open time.
- Le et al. [22] investigated the rheology of their material, and ob-
served it to be significantly influenced by the mix proportions and
the presence of superplasticizer, retarder and accelerator. Di Carlo,
Khoshnevis and Chen [23] have proposed a material maturity model.
- Pegna [27] investigated the effect of building rate and reported on
the difficulty of increasing it. More recently, Perrot, Rangeard and
Pierre [26] proposed an optimisation method, and validated it ex-
perimentally using a column built layer-wise.
- Bukkanatman and Clark [28] developed a two degrees-of-freedom
model to capture the salient features of the process dynamics under-
lying extrusion-based processes. Their experimental investigations
showed that their model could enable the effective monitoring of
process conditions, including the overflow and underflow of materi-
al from the extrusion nozzle, as well as sub-optimal (fast and slow)
feed rates from the extrusion head.
- In a more holistic approach, Khoshnevis et al. [29] attempted to cal-
ibrate input parameters such as extrusion rate, linear speed, layer
thickness, part diameter and number of layers in relation to output
parameters such as layer width, part height, vertical profile and sur-
face roughness.
More general process-related requirements include:
- Heath, safety and the environment. Here, concerns in relation to ce-
ment-based binding agents were resolved by Dini who preferred to
shift from potentially flammable and toxic polymer binding mate-
rials to sand, and oxides and chlorides derived from sea water [30].
- Maintenance. Issues here were underlined by Valkenaers et al. [31],
who noted the high levels of abrasiveness of concrete and its effect
on essential pump maintenance.
- Material storage. Large-scale construction involves large volumes of
material. For this reason, material considered for additive construc-
tion should be affordable and preferably lightweight in order to facil-
itate storage. This issue has generated a major challenge to the
development of large-scale additive manufacturing processes and
has been regarded as a significant obstacle [32]. Such constraints
can be addressed by favouring locally-derived materials that require
no transport or storage. Kayser was reported [33] to have proposed
an “off-grid” 3D printer that could function entirely autonomously
regardless of location, using sand as the raw material for building
melted silicate components. Rael and San Fratello [16] designed an
installation entitled Earthscrapers, in which environments where de-
sertification, erosion, mining and dredging had shaped the land-
scape constituted theoretical sources of material.
- Control of material: Large-scale construction processes are particu-
larly vulnerable to the accumulation of error – a factor usually pro-
portional to part height. This problem was recognised as being a
substantial cause of rejection of many material types [32], and
leads us in to the final process-related requirement – that material
should be controllable. For this reason, plastic-type materials were
also considered for additive construction [15] since they combine
both low cost and low density, while enabling storage in a ready-
to-be-deposited-and-controllable state, unlike their concrete-type
counterparts.
3.3. Construction-related material requirements
The ability to scale-up process-related material requirements is in-
adequate as a basis for the development of rational additive construc-
tion processes. The material produced should also exhibit material
properties that are relevant to the constructions in question (as defined
in Section 1.2).
Mechanical material properties have become the most investigated
construction-related properties. Unfortunately, investigations have fo-
cused almost entirely on concrete-based processes, and here mainly in
relation to the load-bearing capacity parameter and, to some extent,
early strength and long-term strength [34]. Jeon et al. [34] obtained a
maximum compressive strength of approximately 55 MPa after
28 days, as well as values for the most suitable setting times (initial
set 60 min, and final set 185 min) using a mortar mix comprising 30%
fly ash and 10% silica fume. For comparison, Le et al. [22] concluded
that the mix 70% cement, 20% flu ash and 10% silica fume, together
with an additive consisting of 1.2 kg/m3 micro-polypropylene fibres
corresponded to a maximum compressive strength of 110 MPa after
28 days, and an optimum open time of up to 100 min.
However, even well-established concrete materials had to be re-
characterised due to significant differences in production process as
compared with current practices. Measurement characterisations
were performed at an early stage by Pegna [27] in 1997 in order to ob-
tain approximate values of tensile and compressive yield strength, as
well as Young's modulus, for the concrete-type material in question.
His findings revealed some rather novel anisotropic properties, but his
process, which involved rapid curing using steam, was not developed
further. More recently, Feng et al. [35] employed microscope observa-
tions to study the characteristics of a 3D printed, layered, concrete-
type material, a mechanical test of cubes and small beams, and FE
modelling on the effect of the construction process on structural behav-
iour. These indicated that the layer-on-layer printing process led to an
apparent orthotropic behaviour that was relevant to compression
strength and elastic modulus, but not to failure mode. Their results con-
firmed that due to the anisotropic nature of material distribution, all ex-
trusion-based processes are likely to create components that are
strongly anisotropic [3], and that this will have a significant effect on
the load-bearing capacity of the construction in question. The same con-
clusions were reached by Le et al. [25] who measured the effects of voids
that appeared between deposited filaments on the orthotropic com-
pressive and flexural strengths of extruded concrete.
Although the property of durability is well-known for concrete-type
materials produced by traditional processes, it has also had to be re-
assessed. Internal curing was regarded as problematic for “graceful age-
ing” because in general, unconfined extrudate was reported to be ex-
posed to less-than-optimal curing conditions [23].
 N. Labonnote et al. / Automation in Construction 72 (2016) 347–366
Very few studies are available concerning other critical properties of
building materials, most notably those linked to the field of building
physics. These include insulation [12] and water vapour diffusion resis-
tance properties, etc. Exceptions include studies investigating the feasi-
bility of additive construction in extra-terrestrial environments.
Wilhelm and Curbach [18] provided an overview of the benefits and dis-
advantages of selected materials and processes that might be applied in
lunar environments. The following properties were assessed:
- durability
- ductility
- vapour imperviousness
- high tensile and compressive strength
- low coefficient of thermal expansion
- resistance to UV light
3.4. How non-homogeneous materials should be used
To date, most research has investigated graded materials. Strauss
[3] defined “Functionally Graded Materials” as those consisting of
different material types that interpenetrate one another along a con-
trolled gradient, thus enabling the creation of gradients from hard to
soft or from rigid to flexible. This is of special interest in connection
with processes such as the construction of articulated joints or earth-
quake-prone buildings in which “programming flexibility” can be in-
troduced for critical components [3], for special foldable structures
necessitating creases [36] and, not least, in potential mass optimisa-
tion processes in which the required amount of material is placed
only where it is needed.
The Netherlands Organisation for Applied Scientific Research was re-
ported [3] to have developed a prototype system that made it possible
to print a small spiral of powder-filled polymer paste with a density of
100% at one end and zero at the other. In 2013, Craveiro et al. [37] devel-
oped extrusion-based technologies designed to fabricate functional,
graded structural components with different material compositions.
The principal motivation here was to facilitate the construction of build-
ings that were eco-efficient in terms of their thermal, acoustic and struc-
tural properties, and which commonly incorporated potentially
complex forms and geometries.
In recent years, Neri Oxman and the research group Mediated Matter
at the Massachusetts Institute of Technology (MIT) [38] have been fo-
cusing on improving strength, weight, and materials consumption by
carrying out design using functionally graded materials. For the purpose
of manufacturing graded material objects, they are working to construct
a 3D printer capable of dynamic mixing of the composition material,
starting with concrete and UV-curable polymers. This proof-of-concept
3D printer can be used for the production of skeletal bone-inspired
beams as well as constructions that reflect real-time structural and en-
vironmental constraints. The same team is also promoting heterogene-
ity and differentiation in connection with material properties [38] in
order to demonstrate the concept of a structural skin using a pattern
in which the skin's density responds to multi-scalar loading conditions.
The distribution of shear-stress lines and surface pressure is embodied
in the allocation and relative thickness of the vein-like elements built
into the skin. A final application from the MIT research group to materi-
al-property gradients is the fabrication of fully recyclable products or
temporary architectural components, such as tent structures exhibiting
graded mechanical and optical properties [38] resulting from the use of
biodegradable hydrogel composites.
To date, most of the materials used in construction are either a mix of
binding and fibres/aggregates that are regarded as isotropic for large
volumes, or isotropic metals and plastics. No study has been published
on materials that were successful in benefiting from controlled anisot-
ropy, for example by controlling the orientation of added fibres.
353
The building of a multi-material or heterogeneous object by means
of an additive manufacturing system is fairly new, but concepts are
emerging that would make such systems more attractive and more effi-
cient. The principal challenges are related to variations in melting tem-
peratures, curing behaviour and curing times, all of which have to be
considered when combining different materials into a single compo-
nent [3]. Khoda and Bashir [39] decomposed a heterogeneous material
into two-dimensional heterogeneous layers containing islands in
which each island exhibits specific feature's properties. The authors
underlined the fact that such multi-feature/multi-contour layers require
more resources to be incorporated into the material deposition path-
plan, and presented a framework for determining optimum build direc-
tions for heterogeneous objects.
At the extreme end of the non-homogeneous materials scale are ma-
terials made out of small-dimensioned objects. Yoshida et al. [40] inves-
tigated an original approach by building an architectural pavilion out of
a mix of chopsticks and glue delivered by a hand-held dispenser. Al-
though the deposition process was not automated, guidance was pro-
vided to the human operator by means of a projection mapping
system which enabled constant scanning and evaluation of the prevail-
ing shape of the pavilion, and real-time feedback to the operator. This
approach is interesting because the authors modelled the aggregations
of sticks as homogeneous volumes following mechanical testing. Fur-
ther work includes optimising the structural properties of the material
by means of using stronger binders and a closer investigation of aggre-
gation behaviours. The 3.8 metre-high pavilion consumed about one
million 200 millimetre-long chopsticks and demonstrated the benefits
of the use of macro-scale composite materials in relation to both costs
and the porosity of the construction – a desirable parameter in connec-
tion with material optimisation.
3.5. Discussion
The state-of-the-art of additive construction is largely dominated by
the concrete-extrusion research field (see Section 3.1). Although the
process faced important technical challenges in its early phases (see
Section 3.2), systems are now being employed that successfully ensure
correct speed, resolution and load-bearing capacity parameters as part
of the construction of full-scale real-world structures. However, a
good building material cannot rely on sound load-bearing capacity
alone. Very little research has been conducted into the load-bearing ca-
pacities of other materials, and/or other essential qualities of additive
construction building materials (see Section 3.3) such as durability,
water vapour diffusion resistance, thermal properties, or fire-resistant
properties. This is most probably due to the way in which constructions
are organised today, in that most are composed of a combination of ma-
terials, each of which exhibits specific material properties. For example,
a wall may be composed of timber studs with adequate load-bearing
properties, combined with a mineral wool exhibiting good insulation
properties, and with a vapour barrier exhibiting good water vapour dif-
fusion resistance properties. This means that the application of additive
construction processes are nowadays limited mostly to the production
of (concrete-based) load-bearing components. The future opportunities
for the application of material science in additive construction are thus
two-fold (see Section 3.4):
- research should investigate additive construction processes for non-
homogeneous materials in order to facilitate the production of ver-
satile building materials. This research direction has already been in-
vestigated for additive manufacturing, from Pegna et al. [41] in 1999
to Chianrabutra et al. [42] in 2014, and should as well be considered
for additive construction.
- research should develop new innovative materials that exhibit a
good combination of all essential material properties. A definition
of specifications, followed by optimisation, and possibly involving
354 N. Labonnote et al. / Automation in Construction 72 (2016) 347–366
multiscale analysis, would then become an important component of
the development phase of tailored building materials.
In both cases, it is likely that today's standard use of extruded con-
crete will be disqualified in favour of other materials and/or processes.
A parameter such as the mass of building materials, among others, is
an important factor which makes other materials, such as the cellu-
lose-based, particularly promising. More generally, the replication of
existing building materials and production processes, such as the cur-
rently prevailing concrete extrusion processes, is unlikely significantly
to favour additive construction approaches ahead of traditional con-
struction methods.
4. Relevant technological solutions for additive construction
Existing additive manufacturing technologies were originally devel-
oped for, and are thus currently appropriately adapted to, a given “prod-
uct” size. It has often been emphasised that the greatest challenge facing
additive construction is the scaling-up of existing additive manufactur-
ing technologies. In this section, the challenges related to the structural
engineering aspects of depositing material are investigated. The techno-
logical solutions are described in chronological order starting with a)
gantry solutions, which were the first to be developed (Fig. 5a), b)
cable-suspended solutions (Fig. 5b), c) swarm solutions (Fig. 5c), d)
multi-purpose robotics (Fig. 5d), and e) combination with folding (Fig.
5e).
4.1. Gantry solutions
The gantry solution simply represents a direct scaling-up of additive
manufacturing to additive construction – in short, a giant 3D printer. In
gantry solutions, an actuator is controlled in translation in any direction
defined by along the X, Y and Z-axes in Cartesian coordinates.
Gantry solutions were first developed for concrete extrusion in 2001,
and Khoshnevis from the University of South California in the USA de-
posed a patent for the combination of this solution with the material
process under the name “Contour Crafting” [29]. At first the process ap-
peared as a hybrid method that combined an extrusion process for
shaping the object surfaces, and a filling process (pouring or injection)
that built the object core [11]. The “exceptionally smooth and accurate”
exterior surfaces were obtained by constraining the extruded flow in
the vertical and horizontal directions using trowels [14]. In a companion
paper, Khoshnevis [11] considered the possibility of constructing domes
and vault structures by the direct deposition of concrete-like materials
without the need to first produce a mould. It seems that as time went
by, this latter option was more strategically developed [43] than the
former.
The “Freeform Construction” process, later also referred to as “Con-
crete Printing”, emerged from a research group at Loughborough Uni-
versity in the UK in 2007 [12]. Unlike Contour Crafting, where the
focus had always been directed on entire constructions fabricated in
one piece, Freeform Construction was dedicated to the fabrication of
full-scale construction components such as panels and walls [13]. Lim
Fig. 5. Different technological solutions. a) Gantry, b) cable-suspended, c) swarm, d) robotics, e) combined with folding.
et al. presented a practical example by designing and manufacturing a
concrete component called “Wonder Bench”, that included service
voids and reinforcement [24]. The gantry solution was reported to oper-
ate within a steel frame (5.4 m in length by 4.4 m in width by 5.4 m in
height) [22]. This system adhered to the same principle as Contour
Crafting and comprised a printing head digitally controlled by a CNC
machine to move in the X, Y and Z directions along three chain-driven
tubular steel beams. A material container was mounted on top of the
printing head and was connected to a pump that conveyed the material
to the printing nozzle which was activated by the CNC machine.
Following the examples of Contour Crafting and Freeform Construc-
tion, a cost-effective solution involving a mechanical control system for
the construction of a free curved structure without a mould was de-
scribed from South Korea. The entire system consists of triaxial linear
actuators mounted on a box-type support frame. Depending on the re-
quirements appropriate to the axis in question, the linear actuator used
is selected from among either the belt-driven or ball-screw types [34].
The dimensions of the usable workspace were reported to be
1.3 m × 0.95 m × 0.8 m. The maximum speed of the X and Y-axis was
set at 500 mm/s, and that for the Z-axis at 300 mm/s.
This type of gantry solution has been reported by the mass media to
have been developed in China by WinSun for the construction of resi-
dential homes (see Section 6.4). However, to date, and to the authors'
best knowledge, no peer-reviewed study of this solution has been
published.
To date, the only non-extrusion-based process tool using a gantry
solution is Dini's D-shape printer. The dimensions of the process cham-
ber have been reported as being as large as 6 m × 6 m × 6 m, although in
principle, it can be expanded to any required size [3]. The printing ma-
chine is a form of gigantic plotter, equipped with a spraying head
which moves along two frames in X–Y axis space and selectively sprays
a binding liquid onto predefined areas of the sand layer. The core of the
system is the printing head, which is mounted on a 6-metre aluminium
beam that spans the printing area horizontally and moves in response to
a predetermined vertical pitch. The printing head is composed of 300
nozzles separated by an inter-axis distance of 20 mm. In order to fill
the 20 mm gaps between the nozzles, and to ensure that the entire
area is printed and uniformly covered by the liquid ink the printing
head also shifts along an auxiliary Y-axis [44].
Programming gantry solutions for extrusion-based processes, in-
volving several cranes and nozzles, can rapidly develop into a complex
task. Zhang and Khoshnevis [43] presented an approach designed to
identify the optimal tool path for a single-nozzle Contour Crafting sys-
tem incorporating both the physical constraints of the technology com-
bined with construction considerations. They also provided several
algorithms developed to identify the collision-free tool path for a multi-
ple nozzle system based on the single-nozzle approach.
4.2. Cable-suspended platforms
At first sight, Cartesian coordinates may appear to be more appropri-
ate for additive construction. However, due to the large size of the ob-
jects being produced (from large-scale building components to entire
 N. Labonnote et al. / Automation in Construction 72 (2016) 347–366
houses and completed buildings), their use also requires a considerable
amount of energy for transport and installation, as well as the transla-
tion of heavy cranes, even when the tool paths are optimised. A more ra-
tional solution, that improves the transportability of the system, has
been developed under the generic label of “cable-suspended platform”.
The main benefits of cable-suspended platforms are the provision of
larger workspaces and the fact that they are relatively inexpensive
[32,45]. They are also easy to transport, disassemble and reassemble
[45] and are more easily reconfigurable [32].
A cable suspended platform consists of an end-effector attached to
an external frame using multiple cables. The end-effector is manipulat-
ed by motors that can extend or retract the cables in a fully automated
way. A cable-suspended platform is fully constrained when the cables
are attached to the end-effector from below [32].
In 2007, Bosscher et al. [45] presented a cable-suspended platform
named “C4 robot”, that consisted of a rigid frame combined with an
end-effector suspended from twelve cables (four upper and eight
lower). In addition, the eight lower cables were divided into four paral-
lel pairs. The authors used the robot's reachable workspace, as well as
the corresponding cable tensions, to approximate the maximum achiev-
able size dimensions – 44 m × 44 m × 40 m for a given cube frame struc-
ture of 50 m. They studied both kinematics and statics and provided a
proof of translation-only motion [45]. Pseudostatics, dynamics and con-
troller design were also addressed in a companion paper [46].
A second, improved, version of the cable-suspended platform was
presented in 2008 [47] involving both crossed and parallel cables
mounted from top, instead of a base, frame. The workspace and stiffness
parameter (defined as the ability to resist unwanted distributions) were
both considered to be better in the later version.
Barnett and Gosselin [32] recently described the performance of
a cable-suspended platform in terms of the kinematics, statics and
dynamics of their own six-degrees-of-freedom example. This plat-
form was not fully constrained and was composed of an end-effec-
tor attached to six cables controlled by six overhead motors.
Although they did not build a construction as such (as defined in
Section 1.2), they created a 2.16 metre-high polyurethane foam
statue of the seventh Prime Minister of Canada in the space of
38 h, to an accuracy of approximately 1 cm. They also reported
that they were planning to enlarge their workspace and investigate
different printing materials.
Neri Oxman at MIT also investigated cable-suspended platforms with
the aim of developing an easily deployable platform capable of printing
large structures [38]. This cable-suspended platform was called the “Spi-
der Bot”, and was intended to be connected to stable points suspended
high in the external environment, such as large trees or buildings. The sys-
tem was originally designed to use extruded expanding foam, but the ex-
trusion or spinning of tension elements such as cable or rope were also
considered for wrapping bridges or webs [48]. Similar ideas led students
Monroy and Chen from the California College of the Arts in San Francisco
to design the “Sky Printer” – a cable-suspended platform designed to “hi-
jack existing structures and topography to construct its own coordinates
system” [49].
WASP (the “World's Advanced Saving Project”) is a company based in
Italy that started investigating low-cost additive construction in 2012.
Their solution involves a delta robot, but it exhibits some similarities to
cable-suspended platforms. Instead of cables, the end-effector is attached
to a lightweight frame by means of parallel-oriented arms. The key con-
cept of the delta robot is the use of parallelograms which restrict the
movement of the end platform to pure translation, i.e. movement is per-
mitted only in the X, Y or Z directions – no rotation is permitted. They con-
structed three cable-suspended platforms which they called “BigDelta”
[50]. The first platform was 4 m high, the second 8 m, and the third
12 m. Due to its shape, the third platform is claimed to be “the world's
largest delta 3D printer”. To date, no housing units have been constructed
using BigDelta, but potential solutions involve the use of locally available
and inexpensive materials such as clay, earth and mud.
355
4.3. Swarm approach
A dramatically different approach rejects the use of a single giant
fixed frame in favour of several smaller, mobile “robots”. This solution
is commonly referred as the “swarm approach”, referring to the concept
of “swarm behaviour”, in order words a collective behaviour exhibited
by entities. The approach had already been described in 1997 by
Pegna [27], who envisaged “the construction of a large structure by an
army of mechanical ‘ants’, one grain of sand at a time”.
The swarm approach is particularly attractive for additive con-
struction in extra-terrestrial environments, for which payload and
transportation considerations are critical. Ceccanti et al. [51]
recognised that “a small printer (with a printing width of 1 or 2 m),
running on wheels and aided by another rover performing the
function of collecting and depositing the regolith, could do the job
in a more efficient way” than the originally envisaged D-Shape
alternative.
Another benefit of the swarm approach is that smaller robots can
find their own way to the construction site without human interven-
tion, which is not possible in the case of fixed frame installations.
Again, this is highly beneficial in extra-terrestrial environments, and
also in locations where human exposure to the building process may
be deemed hazardous [52].
Oxman from MIT followed this approach by introducing the concept
of building architectural structures using a swarm of single, autono-
mous robots [38]. In Oxman's vision, the robots would be able to build
and then climb tubular structures which in turn become single threads
within a larger construction network constituting the emerging archi-
tecture. Oxman considered communication between the individual ro-
bots within the swarm as the key to additive construction, requiring
both embedded hierarchies and sensors to support an algorithmic deci-
sion-making process.
Students Muhleman, Cation and Satish from the California College of
the Arts in San Francisco shared the same aim of having autonomous ro-
bots carry out construction work in hostile environments, and designed
units that were able to perform additive construction from sawdust
[53].
The ability to climb is key to the application of the swarm approach,
since this characteristic enables the solution to adapt to all construction
dimensions, and height in particular. Two leading researchers from the
Institute of Advanced Architecture of Catalonia, Sasa Jokic and Petr
Novikov, developed a team of three small robots (“Minibuilders”)
with a volume of well under 0.05 m3 for the construction of large-
scale structures [54]. They demonstrated the approach by constructing
an architectural pavilion, and the main features of their method are as
follows;
- a “Foundation Robot”, which employs tracks for mobility, and a line-
follower sensor mounted in the front of the robot for positioning.
The sensor recognises undulations on the ground and enables the
robot to steer. The Foundation Robot is equipped with a vertical lin-
ear actuator to position the nozzle relative to the height of printed
layer. One or more Foundation Robots make the construction's foot-
print – the first twenty layers of the structure – using an extrusion-
based process.
- a “Grip Robot” that attaches itself to the structure by clamping itself
between four rollers. As well as a rotational actuator, each roller is
connected to a steering actuator that enables the robot to position it-
self precisely within and above the structure. After the base structure
is completed, Grip Robots are clamped onto the footprint. They ex-
tend the structure upwards by printing layers of material while
holding onto the layers they have just printed. The grip of these ro-
bots is enough strong, and the curing speed of the material fast
enough, to enable the robots to carry out the horizontal printing of
components such as ceilings and window/door lintels. Their nozzles
can be moved laterally to enable greater control over the printed
356 N. Labonnote et al. / Automation in Construction 72 (2016) 347–366
form, thus enabling the printing of curved walls. Heaters are used to
accelerate material curing times.
- “Vacuum Robots” are used to reinforce the shell printed by the first
two robot types. Each robot is held onto the surfaces by means of a
vacuum generator and a suction cup that enables the robot to
move freely around it. The robot moves and steers itself along two
tracks. Vacuum robots reinforce the shell first by attaching them-
selves to it, and then printing additional layers over it. These layers
don't have to be parallel to each other, but can be free-form and ap-
plied only where structure requires them. Vacuum robots can travel
over surfaces at any angle.
Overcoming height restrictions can also be accomplished by other
means, such as by the use of flying robots. This solution also enables a
significant amount of flexibility in terms of the location where additive
construction is taking place. The feasibility of this approach has been ex-
amined by Hunt et al. [15], who also presented the design and charac-
terisation of an “aerial 3D printer” using robots called “quadcopters”.
Further developments involve improving flight stability in order to
allow more accurate deposition and thus facilitate printing at more pre-
cise tolerances. The main restrictions to this approach are related to low
materials availability due to the relatively small payload capacity and
limited battery life of the quadcopter, as well as an inability to regulate
and stop a printing cycle halfway through before moving on to print at
another location.
4.4. Multi-purpose robotics and automated assembly
A number of solutions involve the use of robotic arms. These can ei-
ther extrude materials by themselves [55–57], or perform ancillary con-
struction-related tasks [11] such as painting or the spreading of tile
adhesives prior to picking the tiles and accurately placing them over
the treated area. These robots are particularly well suited to extra-ter-
restrial environments, and a full description of a conceptual robotic mo-
bility platform named “ATHLETE” was proposed by Scott Howe et al.
[58]. The ATHLETE is a six-limbed wheel-on robotic mobility platform
that can operate without human intervention. Each limb is a six- or
seven degrees-of-freedom robotic arm that can perform various tasks,
including additive construction.
This type of multi-purpose robotics shares many similarities with
Oxman's concept of a “Digital Construction Platform” [38]. The aim of
the team at MIT was to replicate the biological construction of the
human shoulder and hand by combining a 5-axis mobile boom arm at-
tached to a 6 axis robotic arm. The entire platform is based on a mobile
truck vehicle. The team envisages that potential applications include the
fabrication of non-standard architectural forms, the integration of real-
time on-site sensing data, improvements in construction efficiency, en-
hanced resolution, lower error rates, and increased safety.
Automated assembly solutions have been developed independently
of additive manufacturing and/or additive construction systems, but
there is no reason why a combination of the two could not be consid-
ered for multi-purpose robotics solutions.
ROB technology AG [59] is a spin-off company from the Swiss Feder-
al Institute of Technology in Zurich (ETH). The firm Gramazio & Kohler
worked in close cooperation with ROB Technology to develop a mobile
fabrication unit that can fit into a modified freight container and per-
form several construction tasks. The enormous robotic arm employs al-
gorithms to enable autonomous operation, and is particularly suited to
bricklaying. Famous applications include art installations exhibited in
Venice and New York, as well as the construction of a non-standard
brick façade [60] for a service building at the Gantenbein Winery in
Switzerland in 2006. The robotic arm was used to lay each one of the
20,000 bricks precisely according to programmed parameters – at the
desired angle and at exactly prescribed intervals. In this way, each
wall could be designed and constructed to deliver the desired lighting
and air permeability, while creating a pattern that covers all of the
building's façades.
Another bricklaying solution, also investigated by ROB in collabora-
tion with ETH, is what they called “Flight assembled Architecture”.
This was demonstrated during the construction of a large-scale art in-
stallation exhibited at the FRAC in Orléans, France [61]. The installation
consists of over 1500 modules which are positioned by 50 quadrotor he-
licopters (drones), collaborating in response to mathematical algo-
rithms that translate the digital design data into the behaviour of the
drones. The algorithms determine the efficient deployment of the
drones, ensuring that collisions are avoided and best-case flight paths
are selected for fast payload pickup and release.
4.5. Folding and self-assembly
Appropriate folding operations can significantly improve fabrication
speed [36]. To date, research has been limited to medium-sized objects,
but the concept promises major benefits for the additive construction of
thin-shell structures.
Deng and Chen [36] observed that the additive manufacturing of a
large thin-shell feature usually requires many layers using a small vol-
umes of material. They decided to adopt the principles of origami design
and to integrate the layer-based fabrication approach with appropriate
folding operations with the aim of significantly improving fabrication
speed. The critical issues they encountered when designing folding
structures included:
- the development of an algorithm to convert a 3D thin-shell structure
into a foldable 2D structure
- an additive manufacturing process able to produce either a multi-
material design crease design or a rigid material hinge design
- development of actuation mechanisms to unfold a 2D structure to
produce a designed 3D structure.
Deng and Chen presented a cellular-based decomposition method,
and proposed both internal (using a thermally sensitive polystyrene
plastic film) and external actuation mechanisms. They also achieved a
controlled bending angle by developing an angle lock design. Ge et al.
[62] also investigated self-assembly by folding using heat, and intro-
duced the paradigm of “printed active composites”. They demonstrated
a self-assembling structure comprising an elastomeric matrix reinforced
with glassy polymer fibres and exhibiting shape memory effect. They
used the example of self-folding and box-opening using heat as the ac-
tivation energy.
Water can also be used as activation energy. The Self-Assembly Lab
at MIT collaborated with Stratasys Ltd. to develop an additive
manufacturing process using a hydrophilic polymer that expands by
150% on contact with water. When printing a specific combination of
both active and rigid polymers, the printed object can be forced to fold
in response to a digital template. Actual demonstrations have been lim-
ited to medium-sized objects, such as single strands that transform into
the letters ‘MIT’ when submerged in water. However, the self-assembly
of building components could be envisaged, starting as flat-pack mate-
rials activated on delivery to achieve their full volume and functionality
[63].
Of course, electronic materials can provide the necessary activation
energy for folding, and Oxman from MIT [38] developed computational
origami in order “to create from a single sheet […] almost any 3D shape
that is already embedded with electronics”. As part of the project “Logic
Matter”, Tibbits [64], also from MIT, demonstrated the benefits of a
building block embodying physical logic for architectural assembly.
Finally, folding also provides a means of maximising what can be ad-
ditively manufactured from a fixed-size 3D printer by digital folding.
Tibbit's approach relies on breaking down a given structure into a linear
strand that is subsequently folded into a super-dense curve designed to
 N. Labonnote et al. / Automation in Construction 72 (2016) 347–366
fit within the 3D printer bed. Tibbits was reported to successfully
having printed a 15 meter-long strand as the basis for creating a
chandelier that was eight times larger than the 13 cm × 13 cm ×
15 cm print bed used.
4.6. Discussion
Contour-crafting and cable-suspended platforms are primarily
intended for building one-piece constructions, whereas Freeform Con-
struction and D-Shape were designed for making single building ele-
ments for subsequent assembly. Using the gantry (Section 4.1) or
cable-suspended approaches (Section 4.2), the size of the constructed
object is more or less limited by the size of the frame. Increasing the
size of the frame does not necessarily solve the problem. For example,
how high should a frame be to enable the additive construction of a sky-
scraper? Increased frame size introduces potential issues related to the
transportability and mobility of the solutions in question. The swarm
approach (Section 4.3) is attractive because of the use of climbing robots
and the resulting reduced size of the additive construction equipment.
However, swarm technologies currently remain confined to the field
of experimental application.
The existing solution that is perhaps the most easily adaptable
for additive construction in the short-term is the use of multi-pur-
pose robotic arms (Section 4.4), not least because they combine
the advantages of limited size and maturity of technology. This
also opens the field of additive construction to new methods for
“joining materials”, as defined in Section 1.2. Weaving and knitting
can be regarded as true “joining” processes, even if they do not cor-
respond exactly to the concept of “additive” layers. “Material” is
also a rather wide-ranging concept that theoretically can encom-
passes any type of substance, including small objects. Thus, if bricks
are regarded as “materials”, the automated assembly of bricks
based on a digital template is theoretically relevant to additive con-
struction according to the definition given in Section 1.2. Robotics,
although well-known and commonly applied, are thus able to gen-
erate exciting opportunities for additive construction by providing
new and innovative ways of “joining materials”.
Folding technologies (Section 4.5) also challenge the definition of
additive construction in that the resulting structure is no longer neces-
sarily built as a stack of layers. Although the technology is currently in
the experimental phase, its relevance is key in the case of thin shell
structures. The corresponding issue for thick structures is assembly.
We choose not to include this issue in the definition given in Section 1.2.
The relevance of all the solutions mentioned in the present section
should thus be evaluated in relation to the different types of additive
construction reviewed so far:
- additive construction of specific building elements (but not neces-
sarily all of the elements required for a given structure). In this
case, additive construction takes place in the controlled environ-
ment of a factory, and can be compared to high-tech prefabrication.
Gantry and cable-suspended solutions are fully relevant since mobil-
ity and transportability are insignificant concerns. Size will still rep-
resent a limitation to some extent.
- additive construction of the majority of building elements com-
prising a structure – later assembled on-site. If assembly is carried
out by traditional methods, this case is similar to that described
above. On the other hand, if assembly is automated and seamlessly
integrated with the design process, it falls under the definition of ad-
ditive construction as given in Section 1.2 (a process carried out to
“create constructions from 3D model data”). Relevant solutions
here include robotic arms and swarm robots. Robotic equipment
can either be multi-purpose (by which the same robot is able not
only to produce the building element but also to assemble it), or
task-specific (by which a single robot specialises either in the pro-
duction or the assembly of the element in question).
357
- additive construction of the entire structure in one piece. Al-
though gantry and cable-suspended solutions were initially devel-
oped for this purpose, their relevance is declining in favour of the
swarm robotic approach which offers greater mobility and which
is less limited in terms of the size of the construction. However,
the combination of folding technologies with gantry, cable-
suspended or robotic arm approaches may still be appropriate
here, but more research is needed to fully evaluate the potential of
compacting the production of building material into a fixed and lim-
ited space, and then subsequently deploying the product outside
these limits.
5. Opportunities for additive construction in connection with
building design
The development of additive construction opens a new era for build-
ing design. It provides opportunities for designs that would not have
been technologically or commercially feasible in the past. This section
provides an insight into new building design opportunities resulting
from advances in additive construction approaches, as compared with
traditional design process. These include an almost unlimited potential
for complex geometries (Fig. 6a), more optimised topologies for no ad-
ditional cost (Fig. 6b), the design with multi-functional building compo-
nents (Fig. 6c), and union of digital design process and digital building
process (Fig. 6d).
5.1. Complex geometries at a structural scale
Strauss [3] used the concept of “design for production” (by which
building materials dictate a predetermined range of fixed measure-
ments and hence repeating products) to explain the limited opportuni-
ties when it came to generating new designs and constructions. This
applies in particular to constructions made out of concrete, for which
the creativity of architects was constrained to adhere to certain pre-de-
fined geometries in order to reduce costs and enable the reuse of form-
works [26,65]. Thus the building of freeform constructions, i.e. those
that exhibit complex forms that cannot be defined by elementary geo-
metrical concepts [66], was achieved only for a limited number of se-
lected projects.
However, once Anatasiou et al. [67] had demonstrated mathemati-
cally that any 3D structure can be produced by means of a layer-by-
layer approach, structures can now assume “virtually any shape” [51],
including those that would have been difficult, expensive (freeform),
or impossible (fully-enclosed voids [27,36,68]) to create using standard
construction procedures [69].
Anatasiou et al. [67] also noted the practical limitations of slicing res-
olution and the achievement of stability during layering. While methods
for building vaulting and domes without external supports have been
suggested by Khoshnevis [11], the issue of the capability of additive con-
struction to build structures exhibiting overhangs, which are essential
for the maintenance of high degrees of geometrical freedom, is a recur-
ring theme [33]. This capability is also essential in the case of more tra-
ditional architectural designs, e.g. roof constructions [56]. In general, the
capability to construct major overhanging structures is dependent on
both the presence of support materials and/or the speed at which the
material transforms, cf. Gardiner [30].
Additive construction processes that are capable of building over-
hangs without the need for supporting structures are:
- those involving powder-based processes in which the unconsolidat-
ed material actually provides support [21]
- those involving materials made out of small-dimensioned objects
bound together by adhesives [40]
- those involving liquid-based processes combined with electromag-
netic confinement (to some extent) [70]
358 N. Labonnote et al. / Automation in Construction 72 (2016) 347–366
Fig. 6. Architectural design opportunities provided by additive construction. a) Complex geometries, b) optimised topologies at no additional cost, c) multi-functional building
components, d) union of digital design process and digital building process.
Additive construction processes that require a supporting structure
when building overhangs are:
- Freeform Construction, which requires a secondary structure that
has to be removed once the processing operation is complete [21]
- those involving liquid-based processes, in situations where a liquid
metal at a boundary would flow downwards in response to gravity
due to lack of adequate support [70]
Supporting structures may either be a) inflatable elements that
serve as a reference for internal wall boundaries [51], b) peripheral
structures similar to scaffolding that are removed after the building pro-
cess is finished [21], or c) a second ancillary material that is easy to re-
move. However, the removal operation is likely to reduce the quality
of the surface of the element in question [3,33].
Freeform capacity is commonly thought to apply to bulk material
constructions, but in fact can also be applied to truss-like structures
that make use of the innovative technologies needed to build so-called
“anti-gravity” structures. “Mataerial” is the result of a collaborative re-
search effort between Petr Novikov and Saša Jokić from the Institute
for Advanced Architecture of Catalonia (IAAC), and the Joris Laarman
Studio [71]. By employing innovative extrusion technology, they
claimed to be able “to neutralize the effect of gravity during the course
of the printing process”. This provides the flexibility needed to create
truly natural-looking objects because the 3D curves can follow the pre-
cise stress lines inherent in a customised form. This is a special feature
that cannot be achieved using 2D layers.
Additive construction also enables freeform capability at material
scale. Auxetic materials are those that exhibit a negative Poisson's
ratio. Auxetic behaviour is of interest to architects, for example, in situ-
ations in which domes are designed by means of avoiding saddle-effect.
Park et al. [72] employed digital modelling and additive construction to
generate and test experimental auxetic architectural design proposals.
5.2. Optimised topologies provides potential individualisation and
contextualisation at no cost
Both structural and material scale designs can benefit from additive
construction as a means of developing freeform or innovative shapes.
Additive construction thus represents an opportunity for making the
link between well-developed, computational-based, optimisation and
the previously missing physical processes required to reproduce
optimised structures [73]. This is still the case even if some post-pro-
cessing is required on the products resulting from the optimisation
methods. Zegard and Paulino [74] addressed this issue by describing a
simple procedure using a topology optimisation output appropriately
adapted to additive manufacturing.
It is common for optimised components to yield material savings
and reductions in weight [31], but they can also contribute to other
types of optimisation, such as those linked to thermal or acoustic prop-
erties. Arup [75] has been reported as using additive manufacturing pro-
cesses for “significantly reducing weight and cost of future construction
materials” by optimising the amount of material required. Arup claimed
to have been able to reduce the weight of traditional steel nodes by 75%,
resulting in an overall weight reduction of more than 40% for the consid-
ered structure. Medeiros e Sá et al. [76] developed an automatic ap-
proach for the generation, within the constraints of a volume
boundary, of parameterised, adaptive-infill, primal or dual cellular
structures by means of additive manufacturing. The authors
emphasised the potential applications of additive construction for
building the honeycomb panels that optimise heat transfer. The me-
chanical properties of additively manufactured honeycomb sandwich
panels were studied by Bagsik et al. [68], who underlined their potential
for “load-adjusted” lightweight constructions.
The potential of additive construction processes to enhance true
“individualised” [3] and “contextualised” [16] constructions at “no addi-
tional cost” [33] was claimed to be “evident” by Kwon and Jang [77].
Contextualisation can be understood as a local optimisation process tai-
lored to specific environments, i.e. “making structures more efficient by
controlling the amount and the distribution of material in the structure
according to its need” (Bowen [78]). This applies both to building com-
ponents and one-piece structures built using additive construction pro-
cesses. Ceccanti et al. [51] examined the case of a single element
assembly lunar dwelling which required thicker walls facing the sun
in order to protect against hazardous radiation, while Ageieva and
Zhuikov [57] deconstructed a load-bearing concrete skeleton into indi-
vidual concrete elements whose geometries and material properties
were varied gradually in response to local structural demands.
Furthermore, according to Mahamood et al. [79], additive construc-
tion represents a significantly better approach than mass customisation
for meeting customer and/or architectural requirements for
individualised products. Mass customisation relies on either the use of
different combinations of pre-assembled modules or delayed differenti-
ation strategies but, according to Strauss [3], offers only a limited degree
of individualisation. Flexibility with respect to individualisation was
even considered as a potential future niche market for additive con-
struction by Gardiner [30].
5.3. “Smart” constructions
The potential to embed building utilities such as heaters, electrical
facilities and plumbing has been emphasised ever since the very early
days of additive construction by Khoshnevis [11]. For example, attempts
to produce optical glass for manufacturing façade elements have been
reported by Oxman [38] as part of the “Glass Printing” of objects.
However, according to Soar [73], the added value of additive
construction has been largely surpassed by the opportunity to embed
“increased functionality”. “Smart structures” is a general concept
encompassing a variety of properties such as solar responsivity [16],
technologies for the monitoring and maintenance of buildings, the dis-
sipation of seismic forces [16] and self-healing materials.
Strauss [3] investigated the potential of additive manufacturing for
façade construction, and foresaw futuristic smart applications such as
self-acting kinetic façades that automatically create shading, and guid-
ance systems that change colour or pattern in the case of fire. Tibbits
 N. Labonnote et al. / Automation in Construction 72 (2016) 347–366
[80] envisaged designing pipes that can expand or contract to control
flow rates.
So-called “4D printing” has been popularised largely by the numer-
ous presentations given by Tibbits in recent years. 4D printing was offi-
cially defined by Campbell et al. [81] as a fabrication process which
entails a change in form or function that occurs after the 3D printing
process is completed.
Transformation requires energy, which can be either harvested from
external sources or embedded into the building component during the
additive construction process. Applications will likely focus on building
environmentally-tailored and resilient constructions that can adapt to
environmental variation such as changes in moisture content, tempera-
ture, pressure, altitude, illumination, heat or sound.
5.4. Additive construction – the continued digital workflow from design
process to building process?
Digital building process is a vast concept encompassing everything
from design, engineering, production to possibly even assembly. To
date, design and engineering have been revolutionised by digitisation
[82], involving the ever-growing use of architectural- and engineer-
ing-dedicated software products [83]. Computer Numerical Control
(CNC) has contributed, to a somewhat lesser extent, to the development
of digital production. By merging digital production and assembly, addi-
tive construction has the potential to significantly simplify the process
from digital design to the digital production of finished structures and
buildings. File sharing enables both digital architecture and additive
manufacturing to facilitate collaborative design. Thus, additive con-
struction has the potential to “democratise” building design (Leblanc
[33]), and to significantly contribute to the architectural expansion of
the worldwide “Do-It-Yourself” (DIY) movement (Diez [84]). The po-
tential for an enlargement of the “open source building system” pro-
posed by WikiHouse [85] has also been emphasised. The success of
such an open source system has been reported [33] to be based on
three criteria:
- Repeatability: the requirement of replicability facilitated by provi-
sion of the specific hardware and its operating software (including
tested settings) for small entrepreneurs offering manufacturing ser-
vices,
- Transformability: the requirement of access that permits modifica-
tions to the current version and the repair of faults, and
- Publishability: the requirement that encourages a community of
users to share its results and discoveries.
5.5. Discussion
Among the publications examined in this study, the opportunity for
freeform applications is by far the most commonly used argument
supporting the benefits of additive construction. In theory, additive con-
struction can enable the production of almost any geometry (Section
5.1). However, freeform is not exactly obtained “for free” in all domains.
It is true that additive construction processes can produce both simple
and complex geometries for the same “equipment costs”, meaning
here that, contrary to traditional construction approaches, new and ir-
regular forms can be created without having to make large numbers
of new moulds. It is also true that this will provide architects with
new and creative opportunities that are unavailable using traditional
construction approaches. The flexibility of forms permitted by additive
construction is substantial in cases involving the horizontal curvatures
of wall-like elements but, due to fundamental physical conditions
constraining load-bearing and gravity, is far less in the case of horizontal
spans.
This is true in particular in the case of extruded, concrete-based, ma-
terials, where load-bearing capacity during the building phase (i.e.
359
before curing) is far lower than that of the completed structure. Con-
crete has been used successfully in construction mainly for two reasons.
The first is its suitability for compression-based structural forms like
arches, vaults and shells, and the second is its suitability for steel-rein-
forced structures that can also withstand tension and thus accommo-
date large horizontal spans by means of bending. To date, the load-
bearing performance of additive constructions based on extruded con-
crete-based materials is far from that that can be achieved by traditional
and advanced concrete structures. The future of extruded concrete-
based additive construction is likely to depend on improvements in
the aforementioned aspects, which must involve the development
both of construction methods that permit building forms that exploit
the load-bearing potential of the material, and of new materials that
are lighter than conventional concrete and can also address tension-re-
lated issues.
In summary, future freedom of form will only become a reality at the
cost of strengthened material in critical areas in order to counteract
gravity-induced load during the building phase. Freeform opportunities
should thus be regarded as rational design compromises that promote
greater creativity, but will not be unrestricted or entirely “free” in the
true sense of the word.
Optimisation commonly appears as the second most-often cited
benefit of additive construction (Section 5.2). Structural optimisation
is probably best suited to single building components, and preferably
those that are frequently repeatable (in order to achieve a significant
mass loss), such as connection joints. Once again, true structural optimi-
sation may not be the best option, since ending up with too many indi-
vidual components is likely to result in increased labour costs for correct
assembly. It is also desirable to avoid extreme optimisation since this
only results in the construction of structures exhibiting very low robust-
ness. Of greater interest to architects is the individualisation and
contextualisation of structures. This interest corresponds to some ex-
tent to the idea that structural design responds to architectural inten-
tions rather than to structural requirements alone. In that sense,
additive construction may represent a true game-changer for architec-
tural design and for future fruitful collaboration between engineers
and architects, and for architecture in general. Both from the architects'
and engineers' points of view, additive manufacturing is interesting
simply because its minimal use of materials might allow for the
utilisation of materials that are more expensive and of higher quality,
and will also create opportunities for the development of new materials.
The combination of additive construction and “Internet of the
Things” (IoT) technologies could significantly facilitate the development
of “smart” constructions (Section 5.3), although the technology must
first demonstrate some success in the production of non-homogeneous
materials (Section 3.4).
Perhaps more important to modernisation within the construction
industry is that the “smart” implementation of additive construction
may represent the last step towards full digitisation of the construction
process from design to erection (Section 5.4). In general, the prevailing
attitude is that the introduction of digital tools to the construction in-
dustry is resulting in safer and more reliable structures over their ser-
vice lifetime. However, some contradictory views are being expressed
[83]. The argument addressing the democratisation of design and con-
struction (Leblanc [33]) makes sense given that the tools available for
design and construction are open, easy to access, and apply in all phases
from design to construction. On the other hand, if the sector continues
to be dominated by large industrial corporations, the final result may
well be the opposite, in that the behaviour of monopolies within the
market leaves little room for the influence of small contractors or
individuals.
On the basis of the development of building materials and construc-
tion techniques specifically tailored to additive methods, additive con-
struction may introduce entirely new constraints as well as
opportunities for the building sector. From a building industry point of
view, additive construction represents new opportunities for fast and
360 N. Labonnote et al. / Automation in Construction 72 (2016) 347–366
efficient construction. From structural and architectural design points of
view, the greatest potential probably lies in the opportunities for
freeform design and improved conditions for unique, rather than repet-
itive, designs. This is likely to result in new kinds of efficient load-bear-
ing structures and new generations of site- and context-specific
architectural design.
6. Relevant fields of application for additive construction
In order to generate applications for their newly-developed additive
construction system, Buswell et al. [82] asked a representative group
within the building industry the following question: “If you could
have an [additive construction system] today, what would you use it
for?”. This section of the paper reports on the answers to this question
as extracted from the selected literature. This section considers poten-
tial future additive construction applications, describes built or soon-
to-be-built projects, and finally discusses the relevance of each of the
identified fields of potential application.
6.1. Additive construction for harsh environments
Additive construction is claimed to confer considerable benefits on
the construction of structures in harsh environments, where access for
humans is either difficult, impossible [53] or dangerous. Such applica-
tions will reduce stressful workloads and prevent fatigue and accidents
[52].
Most of the publications investigated in this study mention hazard-
ous situations resulting from either natural or man-made disasters,
such as may be the case in war zones. Several potential applications
were emphasised, and in particular those focusing on the construction
of first response shelters [15,58] and the repair of damaged infrastruc-
ture [15,48]. In such situations, the need for “quickly deployable” appli-
cations has been considered by Peter [48] as critical. The University of
Nantes in France has developed the INNOprint 3D printer for this pur-
pose. This printer is capable of building a small emergency facility in
just under 30 min that is insulated, sealed, and safe to live in [56]. In
such situations, additive construction can also be used to build infra-
structure in remote regions with the aim of facilitating access, such as
bridges, etc. in discontinuous terrain to assist humanitarian aid (Hunt
et al. [15]).
Also mentioned are aggressive environments such as the Poles, de-
serts, and environments exposed to chemical contamination [58,82] or
high levels of pollution [26].
Extra-terrestrial environments represent a combination of harsh en-
vironments where access is difficult (for example, the “landing pad” is
defined as “key infrastructure” by Chang et al. [86]), and where the con-
ditions for human workers are dangerous. Plans to erect constructions
on the Moon have already been proposed. Johann-Dietrich Wörner,
who is head of the European Space Agency (ESA), has indicated that
his organisation is intending to start building “Lunarville” by as early
as 2024 [87], and US President George W. Bush is reported to have an-
nounced that the National Aeronautics and Space Administration
(NASA) is aiming to construct a lunar base by 2020 as a jumping-off
point for future missions [18]. In addition, Mars is also claimed to be a
target for human colonisation before the end of the new century
(Koshnevis [29]).
Projects involving the construction of settlements on the moon have
been highly publicised. They include plans originating both from NASA
(in collaboration with the inventor of contour crafting, Boris Koshnevis
[11]), from SinterHab [88], and from the ESA, in partnership with Foster
and Partners [51].
It is also argued that extra-terrestrial environments will demand a
significant contribution from additive construction, not least when it
comes to facilitating the maximum exploitation of in-situ resources
[89,90] in order to minimise the enormous costs of transporting mate-
rials to construction sites using space shuttles [58]. Such applications
are assumed not only to reduce transportation costs but also to increase
the efficiency of extra-terrestrial operations (Fateri et al. [19]).
The use of additive construction for “space” applications has been in-
vestigated in general terms by Johnston et al. [91] in collaboration with
NASA. This study was primarily intended to develop a long-term strate-
gy for the on-demand fabrication of components and equipment for
manned missions. Additive manufacturing and (potentially) also addi-
tive construction were considered to be the best technologies offering
cost-effective, high-precision and low-unit on-demand manufacturing
as part of the so-called “space roadmap”. The authors' principal objec-
tive was to carry out a scientific technology demonstration of additive
manufacturing in a microgravity environment [91].
6.2. Additive construction for “common” environments
The frequently cited environmentally-friendly benefit of additive
construction throughout the literature we investigated is reduction in
waste. This has been assessed by Berman [92], among others, when
making comparisons with subtractive manufacturing technologies. It
is argued that additive construction produces less waste, and that it
also enables the recycling of most of the waste during the next round
of additive construction [27,92,93]. Rael and San Fratello [16] consid-
ered that this applied in particular in the case of excess cement and
aggregates.
Achillas et al. [93] considers that the carbon footprint left by additive
construction is significantly smaller than that left by traditional con-
struction techniques. In general, less material is used, firstly because of
the absence of a requirement for moulding and casting operations,
and secondly because additive construction enables highly-optimised
construction processes and the production of highly-optimised compo-
nents that naturally reduce the amount of material used. When com-
bined with technologies that favour in-situ resources, additive
construction is also described as resulting in a significant diminution
of carbon emissions related for the most part to transport [3,93].
Kwon and Kang [78] argued that transport-related emissions were fur-
ther reduced because of the need for less traffic to and from the con-
struction site as a result of limited labour requirements. Haymond and
Noble [94] used the 2008 LEED Homes system as a guide to isolate the
sustainable potential of additive construction (in the form of contour
crafting) as an alternative to conventional construction for single family
housing.
In the context of 4D printing (Section 4.5), Campbell et al. [81] pro-
posed that “by commanding the object to decompose into programma-
ble particles or components that then can be reused to form new objects
and perform new functions” the process would contribute towards the
better conservation of our planet's limited resources.
Additive manufacturing is often compared by the popular media to
disruptive technologies such as digital books. Currently, the construc-
tion industry is focusing on modular solutions, either fabricated directly
on-site or as systems pre-assembled off-site. Additive construction is a
“new” application of additive manufacturing that excites the media be-
cause it is argued that it requires lower volumes of materials and less in-
tensive human labour, while at the same time being faster. However,
Soar [73] prudently prefers to consider it as a “third way”. Additive con-
struction was considered to be particularly well-suited to complex
building projects by Thangavelu et al. [52] since it improves the overall
efficiency of the building process. Ehrnstöm and Demir [95], on the
other hand, recognised that the key is to know whether or not additive
construction can be used to replace conventional building methods in
the future.
The main benefits that additive construction provides when com-
pared to traditional construction techniques are known to be cost-effi-
ciency and process controllability. In 2007, Boscher et al. [45]
presented a comparative cost and productivity analysis designed to
evaluate the potential of their cable-suspended additive construction
technology against traditional concrete work. Although the cost savings
 N. Labonnote et al. / Automation in Construction 72 (2016) 347–366
were considered to be “tremendous” (for example by Rael and San
Fratello [16]), Boscher et al. nevertheless concluded that the cost of
their additive construction alternative was in fact very similar to a con-
ventional operation, although they did not take into account costs
linked to accident and safety training, which are assumed to be consid-
erably lower for additive construction.
In 2014, Skanska's Director of Innovation and Business Improvement
considered that additive construction using concrete had the potential
to revolutionise the entire process of construction, and anticipated
that it would “reduce the time needed to create complex elements of
buildings from weeks to hours” [96]. The Chinese company WinSun
probably shares this view, and has claimed in the public domain to
have saved between 30 and 60% of building materials and to have short-
ened production times by between 50 to even 70%, while at the same
time decreasing labour costs by 50 to even 80%. This took place while
the company was allegedly applying additive construction (a version
similar to contour crafting) to the construction of ten individual houses
and a six-story apartment building (see Section 6.4). However, Brandon
[97] was critical of WinSun's achievements, emphasising that no de-
scription was provided by the Chinese company as to what had in fact
been produced, and considered that its figures were highly doubtful.
Finally, the issues of low income housing and third-world housing
have been mentioned in the literature. Among other initiatives, Tridom
formed a strategic partnership with WASP in Italy in order to investigate
the additive construction of affordable and sustainable housing for “bot-
tom-of-the-pyramid” people [55].
6.3. Limitations and open questions
Intellectual property issues have been raised in particular in connec-
tion with concerns that digital objects (including a digital file describing
the additive construction of a given structure) can easily be copied and
re-sold [92]. The lack of standardisation and regulation governing 3D
printed objects, whether these are products or the construction itself,
Fig. 7. Additive constructions projects. a) “Radiolaria” ©D-Shape , b) “Landscape House” ©Universe Architecture, c) “Protohouse 2.0” ©Softkill Design, d) “3D-printed Canal House” ©DUS
Architects, e) Chinese houses ©WinSun, f) a Chinese building ©WinSun, g) a pedestrian bridge ©Joris Laarman Lab, h) an office building ©WinSun Global.
361
has also been noted by Strauss [3]. Validation should apply in this specif-
ic case to both hardware and the construction in question [33]. The reg-
ulation issue is a serious matter because of the consequences that will be
incurred if construction failures result in fatalities. This situation repre-
sents a serious challenge if additive construction is ever to compete with
traditional construction approaches [30].
The next issue is that of liability. Campbell et al. [81] speculated as to
who would be liable in the event of failure of a PM-fabricated compo-
nent of an aircraft wing. Would it be the original manufacturer, the pro-
grammer, or the manufacturer of the new design or smart material?
This problem is even more complicated in the case of building compo-
nents that may have two or more functions, such as a wall installed
with hot water transport and electrical distribution equipment. Buswell
et al. [82] ask who will be liable for the design in such situations? And in
the case of additive construction as a whole, the same authors wonder
“Who is responsible for the construction; the designer, the contractor
or those who built the machine?”.
4D printing and programmable matter face the additional challenges
linked to cyber security and risk of hacking. Campbell et al. [81] raised
the issue of securing embedded programmable capabilities into objects.
6.4. Specific projects and demonstrations
This section describes some additive constructions with the status
“built” or “projected with a reasonable likelihood to be built at some
point”.
• Italy: the “Radiolaria Pavilion”, a 3 m × 3 m × 3 m architectural pavil-
ion that was built in 2009.
○ Illustrated in Fig. 7a
○ Companies involved: D-shape and Shiro Studio
○ Material: an environmentally-friendly combination of an inorganic
binder with sand or mineral dust, without internal reinforcement.
The binder transforms any kind of sand or marble dust into a
362 N. Labonnote et al. / Automation in Construction 72 (2016) 347–366

stone-like material (i.e. a mineral with microcrystalline characteris- ○ Companies involved: Hejmans, MX3D's technology and Joris
tics) with a resistance and traction superior to Portland cement, at Laarman Lab
levels that require no iron reinforcement. ○ Material: steel
○ Production process: powder-based ○ Production process:
○ Technology: gantry solution ○ Technology: robotic arm
○ Product: an entire (one-piece) structure ○ Product: entire (one-piece) structure
○ Type of construction: traditional compression structure ○ Type of construction: Truss-like structure

• The Netherlands: the “Landscape house” was designed in 2013, esti- • United Arab Emirates/Dubai: A one-storey office building is projected
mated to be built in 2017. to be completed in 2017, with an estimated area of 185 m2.
○ Illustrated in Fig. 7b ○ Illustrated in Fig. 7h
○ Companies involved: D-shape and Universe Architecture ○ Companies involved: Gensler, Thornton Thomasetti and Syska
○ Material: a bio-plastic made out of 80% vegetable oil combined with Hennessy.
concrete paste ○ Material: specially reinforced concrete, fibre-reinforced plastic and
○ Production process: powder-based glass-fibre reinforced gypsum
○ Technology: gantry solution ○ Production process: extrusion
○ Product: hollow components ○ Technology: gantry solution, expected to be 6 m high
○ Type of construction: non-traditional compression structure ○ Product: building components
○ Type of construction: traditional compression structure
• UK: the “ProtoHouse 2.0” was designed in 2013 (approximate size
8 m × 5 m), and was built at a 1:33 scale the same year.
○ Illustrated in Fig. 7c
○ Company involved: Softkill Design 6.5. Discussion
○ Material: bio-plastic
○ Production process: laser-sintering The potential of additive construction is closely related to the inher-
○ Technology: likely gantry solution ent specificities that govern the project in question. These may include
○ Product: components of dimensions 2.5 m long and 1 m wide the environment (Section 6.1) and/or design premises, and control the
○ Type of construction: porous bulk material potential of the context-specific additive construction in question. This
is well illustrated by the numerous and heavily funded projects carried
• The Netherlands/Amsterdam: the “3D printed Canal House” has been out to investigate additive construction in lunar and martian environ-
under construction since 2014. No estimated date for completion (at ments. On the other hand, the more generally defined a construction
the time of writing in 2016). is, the less cost-effective additive construction will be. This factor
○ Illustrated in Fig. 7d makes additive construction much less attractive for general construc-
○ Companies involved: DUS Architects, Kamer Maker tion projects, at least when it is considered as a means of replicating
○ Material: polymer existing buildings. This is true in spite of the enormous mass media
○ Production process: extrusion coverage of certain rare projects (Section 6.4). The widespread use of
○ Technology: gantry solution additive construction for “general housing” projects in “normal environ-
○ Product: building components ments” will only become economically viable if housing itself changes to
○ Type of construction: traditional compression structure become more optimised and more individualised. This vision is shared
by some, but not all, of the projects presented in Section 6.4. In terms
• China: Ten individual houses were built in 2014 using additive of housing, the future of additive construction probably lies somewhere
construction. between the Chinese concrete houses (Fig. 7e), whose design is identi-
○ Illustrated in Fig. 7e cal to traditional concrete houses (although their detailing and finishing
○ Company involved: WinSun is of poorer quality), and the ProtoHouse project (Fig. 7c) that repre-
○ Material: combination of cement, glass fibre and construction waste sents a major step-up in terms of architectural design.
○ Production process: extrusion Additive construction may reduce the amount of human labour and
○ Technology: gantry solution (6.6 m high, 10 m wide and 150 m materials consumed to some extent. It may therefore be of benefit to
long) projects where:
○ Product: components
- material savings would be significant, or
○ Type of construction: traditional compression structure
- human labour would be costly.

• China: A five-storey residential complex and a 1100 m2 villa, built in

2015. In such cases, an evaluation of the cost-efficiency of additive con-
○ Illustrated in Fig. 7f struction must rely on knowledge of the allocation of costs involved in
○ Company involved: WinSun the design phase, material consumption, human labour for construction,
○ Material: combination of cement, glass fibre and construction waste and equipment. Unfortunately, overall knowledge of these factors is
○ Production process: extrusion often incomplete or missing. There is a need for comparisons between
○ Technology: gantry solution (6.6 m high, 10 m wide and 150 m existing approaches and new additive construction techniques that
long) are scientifically documented. And the question will still remain
○ Product: components concerning the potential of additive construction for “non-complex”
○ Type of construction: traditional compression structure building elements, for which neither material savings nor savings in
• The Netherlands/Amsterdam: A pedestrian bridge under construc- human labour would be significant.
tion in 2016. No estimated date of completion (at the time of writing In the case of metal-based materials, the current main challenges are
in 2016). the high costs and the limited sizes of units produced. If we assume that
○ Illustrated in Fig. 7g production costs will decrease rapidly as technology advances, the
 N. Labonnote et al. / Automation in Construction 72 (2016) 347–366
metal-based additive manufacturing of building components is likely to
increase significantly. In terms of both metal-based and other kinds of
additive manufacturing linked to building, the next few years are likely
to reveal a large number of as yet unknown, but interesting and useful
components and buildings.
It has been argued that additive construction will generate less mate-
rial waste (Section 6.2). If this is so, additive construction would definite-
ly show an important advantage over traditional construction processes
in light of Paris Agreement on Climate Change, which will unleash ac-
tions and investment towards a low carbon, resilient and sustainable fu-
ture [98]. This environmental aspect is even more governing since the
building sector has been recognised [99] to contribute up to 30% of global
annual greenhouse gas emissions and to consume up to 40% of all energy.
However, in order to achieve a recognisably environmentally-friendly
construction process, a full life-cycle analysis (LCA) of additive construc-
tion processes must be performed in the same scientific and objective
way this has been already performed for additive manufacturing pro-
cesses by for example Drizo et al. [100]. Additive construction is such a
wide-ranging concept, involving a large number of different materials
and technologies, that it is doubtful if all additive construction processes
will be able to reduce carbon footprints to the same extent when com-
pared to traditional construction techniques. Some are probably better
than others, whereas for concrete-based materials the reality is that it
is most likely that the opposite is true (Section 5.5). In this sense, “digital”
additive construction would also act as a catalyst for the integration of
LCAs as a basis for construction projects.
Finally, concerns about intellectual property and liability (Section
6.3) are relevant to additive construction, but do not constitute a
major obstacle to its development. However, they will have to be
taken into account in due course, and issues will probably have to be
specifically adapted to each technology in question.
7. General discussion: challenges and opportunities for additive
construction
7.1. The need for an architectural paradigm shift
It is striking to note that concrete-type materials and gantry solu-
tions have been the subject of much more discussion in the literature
than other materials and methods (Fig. 8a and b). The original concept
of “contour-crafting” was first developed by Khoshnevis [29], and this
paper has become a standard reference in the field of additive construc-
tion. The term “contour-crafting” is cited a total of 508 times in the pub-
lications selected for this study. The swarm approach is a more recent
concept and has been cited only 40 times. Both gantry solutions and
concrete-type materials were thought to be a viable means of replacing
conventional techniques in connection with traditional housing (Fig.
8d). However, this somewhat contradicts the frequently mentioned
ability to produce freeform using additive construction.
Fig. 8. The various study topics sorted according to number of publications for a) material science, b) engineering, c) architecture, d) relevant markets.
363
As described in Section 6, the houses of the future will have to free
themselves from traditional (and less efficient) designs if they are to ben-
efit from the potential inherent in additive construction. In fact, the fu-
ture of additive construction must rely on an architectural paradigm
shift. An important issue for architecture as a field will therefore be
whether or not architects are ready to make use of the “complex” [30]
potential and the “tremendous degree of design freedom” [96] provided
by additive construction, and whether or not they will be able to “re-
shape the way we think about architectural components” [16] (Fig. 8c).
7.2. The need for a more holistic design process
In order for the paradigm shift to take place, the production and as-
sembly stages of the construction process must be considered right
from the start of the design process. Ideally, the design process should
be thought of as collaboration between architects, engineers and con-
structors. The present study has highlighted some successful, but often
disjointed, research efforts in three different domains: a) material sci-
ence (Section 3), b) engineering (Section 4), and c) building design
(Section 5). All of these aspects must be far more incorporated into a
single design process if the true potential of additive construction is to
be realised.
A holistic design process should develop as a result of smaller re-
search steps than those described in this study. The development of
small research steps making up the design process provides an opportu-
nity to assemble them more easily at a later stage. Only then will inno-
vative engineers, architects and builders be given the opportunity to
implement the required paradigm shift.
7.3. The need for (even more) rational designs
Additive construction is often described by the mass media as the
new and “miraculous” construction technique. Although additive con-
struction is very likely to trigger the development of innovative con-
structions that were not technologically or economically possible only
some years ago, constructions will have to obey the same laws of me-
chanics. There is no “magic” involved in additive construction, and in
fact its development is likely to make the design process more complex
because of the exploitation of new opportunities within the frameworks
of new constraints. It will encourage more innovative designs, but these
designs will also have to be more rational.
Rational decision-making will involve trade-offs, and the task will be
made more complex by the huge volume of information made available
via the holistic design process. However, this will also mean that addi-
tive construction should become the next natural step in the evolution
of “smarter” construction. Designers should then be in a position to pro-
pose new types of construction, adapted to the various additive con-
struction methodologies available. For example, bulk materials with
adaptable densities could be used for structures working in
364 N. Labonnote et al. / Automation in Construction 72 (2016) 347–366
compression, whereas the construction of lightweight structures work-
ing in tension or membrane could now be facilitated as a result of
optimised truss techniques.
8. Conclusions
The present study has investigated to what extent 3D printing tech-
nologies can be successfully applied to the construction of large-scale
structures (Section 1). The central concept of additive construction
was defined, and a systematic mapping study (Section 2) performed
in order to assemble relevant publications selected according to a
well-defined set of criteria. Knowledge gathered from the relevant pub-
lications was organised into four main categories: material science, en-
gineering, architecture and market analysis.
Material science research (Section 3) was shown to be dominated
by the development of concrete-extrusion processes. Process-relat-
ed requirements were the subject of a large number of studies, but
there were relatively few investigations into construction-related
material requirements. The materials used for additive construction
are very rarely characterised according to said requirements, e.g.
fire resistance, durability, thermal properties, although exceptions
include load-bearing capacity. More focus should thus be directed
towards the characterisation and potential improvement of con-
struction-related material properties. This is likely to trigger the de-
velopment of new materials exhibiting an effective combination of
essential homogeneous or non-homogeneous properties. In the
case of the latter, this will necessitate improvements in current addi-
tive construction processes.
Research into the development of technologies designed to organise
material into a construction (Section 4) was originally directed towards
gantry, and subsequently cable-suspended, solutions. More recently,
more lightweight systems involving swarm robots and robotic arms
have been considered. It emerges that there is no single optimum sys-
tem, but instead a set of solutions, each of which should be considered
on the basis of the following criteria:
- the type of object to be produced by additive construction (building
component or one-piece construction),
- the location where additive construction is to take place (on-site or
in a factory)
- the assembly technique for the different components (fully para-
metric and automated, or not)
Building design has been often mentioned in the literature in con-
nection with additive construction (Section 5). However, these refer-
ences are almost always delivered from the perspective of end-user
benefits, as opposed to a research perspective considering the impact
of this factor on the research effort itself. The concept of “freeform”
capacity appeared in the literature as the most frequently mentioned
benefit for designers, followed by optimisation and embedded function-
alities. In fact, such concepts were often misinterpreted and should in-
stead be understood to represent a combination of the emergence of
new opportunities and new constraints that will necessitate a greater
degree of rational decision-making in the design phase. Decision-mak-
ing was also shown to be empowered by the digitisation of workflow.
The relevant markets for additive construction (Section 6) were
shown to be closely related to the inherent specificities of the project
in question: the more specific conditions governing the project, the
greater potential of the context-specific additive construction. This im-
plies that additive construction can be successfully applied in connec-
tion with general housing projects only if housing in general changes
to become more optimised and more individualised. However, this vi-
sion is not always obvious on the basis of projects already built or
about to be built.
The challenges facing, and opportunities available to, additive con-
struction were identified as:
- the need for an architectural paradigm shift. Current construction
design approaches must be upgraded in response to the new capa-
bilities offered by additive construction,
- the need for a holistic design process. The design process must pro-
mote synergy facilitated by real collaboration between architects,
engineers, and builders,
- the need for rational designs. The design process must adopt a ratio-
nal approach to the opportunities available and the new constraints
governing building design, structural engineering and the construc-
tion method.
More specifically, the research gaps identified along the present
study lead to the following list of recommendations for further
investigations:
- develop innovative additive construction processes based on solid-
like materials used in line-based (such as 3D weaving) or even sur-
face-based techniques,
- investigate the possibility of liquid-like materials for additive
construction,
- develop process modelling for additive construction,
- investigate additive construction processes for non-homogeneous
materials in order to facilitate the production of versatile building
materials,
- develop new innovative materials that exhibit a good combination
of all essential material properties,
- challenge the definition of “joining materials” by developing new
uses of multi-purpose robotic arms,
- develop decision support tools in order to find the most appropriate
(possibly hybrid) technological solution with respect to the topology
of the structure (truss, shell, etc.),
- develop multi-constraint parametric design to clearly assess the dis-
ruptive potential of additive construction for a given building project
with respect to different performance indicators: cost, carbon foot-
print, quality, health and safety, construction time, etc.
In conclusion, additive construction has the potential to revolution-
ise the construction industry. However, the industry itself, including its
architects, engineers and builders, have to be ready to innovate and
think in new ways. When these adaptations have been made, the
needs of additive construction will result in the development of new
materials. Projects will become mechanically feasible and economically
viable, and new, rational and uncomplicated production methods will
evolve. This study has been intended to trigger a movement towards
the additive construction of functional, environmentally-friendly, ele-
gant, interesting, and user-friendly structures and spaces.
References
[1] S. Prentice, The five SMART Technologies to Watch, 2014 Accessed: November
2015, Available from https://www.gartner.com/doc/2669320?ref=unauthreader.
[2] Stratasys, The power of prototyping. MaximizedAccessed: November 2015, Avail-
able from http://www.stratasys.com/3d-printers/design-series.
[3] H. Strauss, AM Envelope - The Potential of Additive Manufacturing for Facade Con-
struction, Faculty of Architecture, Delft University of Technology, 2013.
[4] P. Teicholz, Labor-productivity declines in the construction industry: causes and
remedis (another look), AECBytes Blog, Vol. 67, 2013http://www.aecbytes.com.
[5] www.3ders.org. President Obama called 3D Printing: “the potential to revolution-
ize” in State of the Union Address, 2013 Accessed: November 2015, Available from
http://www.3ders.org/articles/20130213-president-obama-called-3d-printing-
the-potential-to-revolutionize-in-state-of-the-union-address.html.
[6] The American Society for Material and Testing, Standard Terminology for Additive
Manufacturing Technologies, vol. F2792 - 12a, ASTM International, West
Consholhocken, United States, 2009.
[7] K. Petersen, R. Feldt, S. Mujtaba, M. Mattsson, Systematic Mapping Studies in Soft-
ware Engineering, Proceedings of the 12th International Conference on Evaluation
 N. Labonnote et al. / Automation in Construction 72 (2016) 347–366
and Assessment in Software Engineering, British Computer Society, Bari, Italy,
2008 68–77.
[8] N. Labonnote, K. Høyland, Smart home technologies that support independent liv-
ing: challenges and opportunities for the building industry – a systematic mapping
study, Intelligent Buildings International 2015, pp. 1–26.
[9] C. Barnatt, 3D Printing, Second Edition CreateSpace Independent Publishing Plat-
form, 2014.
[10] T. Shelley, Maximum material performance, Eureka 26 (9) (2006) 15–16.
[11] B. Khoshnevis, Automated construction by contour crafting - related robotics and
information technologies, Autom. Constr. 13 (1) (2004) 5–19.
[12] R.A. Buswell, R.C. Soar, A.G.F. Gibb, A. Thorpe, Freeform construction: mega-scale
rapid manufacturing for construction, Autom. Constr. 16 (2) (2007) 224–231.
[13] R.A. Buswell, A. Thorpe, R.C. Soar, A.G.F. Gibb, Design, data and process issues for
mega-scale rapid manufacturing machines used for construction, Autom. Constr.
17 (8) (2008) 923–929.
[14] D. Hwang, B. Khoshnevis, An innovative construction process-contour crafting
(CC), Proceedings of the 22nd International Symposium on Automation and Ro-
botics in Construction (ISARC 2005), The Fraunhofer-Informationszentrum Raum
und Bau IRB, Ferrara, Italy, 2005 p. IAARC.
[15] G. Hunt, F. Mitzalis, T. Alhinai, P.A. Hooper, M. Kovac, 3D printing with flying ro-
bots, Proceedings of the IEEE International Conference on Robotics & Automation
(ICRA), Hong-Kong, China 2014, pp. 4493–4499.
[16] R. Rael, V. San Fratello, Developing concrete polymer building components for 3D
printing, Proceedings of the 31st Annual Conference of the Association for Comput-
er Aided Design in Architecture: Integration Through Computation (ACADIA), As-
sociation for Computer Aided Design in Architecture, Calgary/Banff, Canada 2011,
pp. 152–157.
[17] E. Dini, D. Shape, Accessed: November 2015, Available from http://www.d-shape.
com/cose.htm.
[18] S. Wilhelm, M. Curbach, Review of possible mineral materials and production
techniques for a building material on the moon, Struct. Concr. 15 (3) (2014)
419–428.
[19] M. Fateri, A. Gebhardt, M. Khosravi, Experimental investigation of selective laser
melting of lunar regolith for in-situ applications, Proceedings of the ASME Interna-
tional Mechanical Engineering Congress and Exposition (IMECE), vol. 2 A, Ameri-
can Society of Mechanical Engineers (ASME), San Diego, USA, 2013 p. ASME.
[20] V.K. Balla, L.B. Roberson, G.W. O'Connor, S. Trigwell, S. Bose, A. Bandyopadhyay,
First demonstration on direct laser fabrication of lunar regolith parts, Rapid
Prototyp. J. 18 (6) (2012) 451–457.
[21] S. Lim, R.A. Buswell, T.T. Le, S.A. Austin, A.G.F. Gibb, T. Thorpe, Developments in
construction-scale additive manufacturing processes, Autom. Constr. 21 (1)
(2012) 262–268.
[22] T.T. Le, S.A. Austin, S. Lim, R.A. Buswell, A.G.F. Gibb, T. Thorpe, Mix design and fresh
properties for high-performance printing concrete, Mater. Struct. 45 (8) (2012)
1221–1232.
[23] T. Di Carlo, B. Khoshnevis, Y. Chen, Manufacturing additively, with fresh concrete,
Proceedings of the ASME International Mechanical Engineering Congress and Ex-
position (IMECE), Vol. 2 A, American Society of Mechanical Engineers (ASME),
San Diego, USA, 2013 p. ASME.
[24] S. Lim, R. Buswell, T. Le, R. Wackrow, S. Austin, A. Gibb, T. Thorpe, Development of a
viable concrete printing process, Proceedings of the 28th International Symposium
on Automation and Robotics in Construction (ISARC), IIT Madras, Seoul, Republic
of Korea 2011, pp. 665–670.
[25] T.T. Le, S.A. Austin, S. Lim, R.A. Buswell, R. Law, A.G.F. Gibb, T. Thorpe, Hardened
properties of high-performance printing concrete, Cem. Concr. Res. 42 (3)
(2012) 558–566.
[26] A. Perrot, D. Rangeard, A. Pierre, Structural built-up of cement-based materials
used for 3D-printing extrusion techniques, Mater. Struct. 49 (4) (2016)
1213–1220.
[27] J. Pegna, Exploratory investigation of solid freeform construction, Autom. Constr. 5
(5) (1997) 427–437.
[28] S. Bukkanatnam, B. Clark, Dynamic modeling and monitoring of contour crafting -
an extrusion-based layered manufacturing process, J. Manuf. Sci. Eng Trans. ASME
129 (1) (2007) 135–142.
[29] B. Khoshnevis, R. Russell, H. Kwon, S. Bukkapatnam, Crafting large prototypes, IEEE
Robot. Autom. Mag. 8 (3) (2001) 33–42.
[30] J.B. Gardiner, Exploring the Emerging Design Territory of Construction 3D Printing
– Project Led Architectural Research, School of Architecture and Design, RMIT Uni-
versity, 2011.
[31] H. Valkenaers, D. Jansen, A. Voet, A. Van Gysel, E. Ferraris, Additive manufacturing
for concrete: A 3D printing principle, Proceedings of the 14th International Confer-
ence of the European Society for Precision Engineering and Nanotechnology
(EUSPEN), vol. 1, European Society for Precision Engineering and Nanotechnology,
Dubrovnik, Croatia 2014, pp. 139–142.
[32] E. Barnett, C. Gosselin, Large-scale 3D printing with a cable-suspended robot, Addi-
tive Manuf. 7 (2015) 27–44.
[33] F. Leblanc, Anything, anyone, anywhere: towards a cloud-based 3D printing fabri-
cation in architecture, Proceedings of the 19th International Conference on Com-
puter-Aided Architectural Design Research in Asia - Rethinking Comprehensive
Design: Speculative Countercultur (CAADRIA), The Association for Computer-
Aided Architectural Design Research in Asia, Kyoto, Japan 2014, pp. 461–470.
[34] K.-H. Jeon, M.-B. Park, M.-K. Kang, J.-H. Kim, Development of an automated
freeform construction system and its construction materials, Proceedings of the
30th International Symposium on Automation and Robotics in Construction and
Mining (ISARC 2013), Canadian Institute of Mining, Metallurgy and Petroleum,
Montreal, Canada 2013, pp. 1359–1365.
365
[35] P. Feng, X. Meng, J.F. Chen, L. Ye, Mechanical properties of structures 3D printed
with cementitious powders, Constr. Build. Mater. 93 (2015) 486–497.
[36] D. Deng, Y. Chen, An origami inspired additive manufacturing process for building
thin-shell structures, Proceedings of the ASME International Mechanical Engineer-
ing Congress and Exposition (IMECE), Vol. 2 A, American Society of Mechanical En-
gineers (ASME), San Diego, USA, 2013.
[37] F. Craveiro, H. Bártolo, P. Bártolo, Functionally graded structures through building
manufacturing, Adv. Mater. Res. 683 (2013) 775–778.
[38] Mediated Matter, M. I. T. Media Lab, Designing for, with and by Nature, 2015
Accessed: November 2015, Available from http://www.media.mit.edu/research/
groups/mediated-matter.
[39] A.K.M.B. Khoda, Build direction for improved process plan in multi-material addi-
tive manufacturing, Proceedings of the IIE Annual Conference and Expo, Institute
of Industrial Engineers, Montreal, Canada 2014, pp. 662–669.
[40] H. Yoshida, T. Igarashi, Y. Obuchi, Y. Takami, J. Sato, M. Araki, M. Miki, K. Nagata, K.
Sakai, S. Igarashi, Architecture-scale human-assisted additive manufacturing, ACM
Trans. Graph. 34 (4) (2015) 88:1–88:8.
[41] J. Pegna, S. Pattofatto, R. Berge, C. Bangalan, H. Herring, M. LeSaux, J. Engler, The
SAND PAINTER: two-dimensional powder deposition, Proceedings of the Tenth
Solid Freeform Fabrication Symposium (SFF), Austin, USA, 1999.
[42] S. Chianrabutra, B.G. Mellor, S. Yang, A dry powder material delivery device for
multiple material additive manufacturing, Proceedings of the 25th Solid Freeform
Fabrication Symposium (SFF), Austin, USA 2014, pp. 36–48.
[43] J. Zhang, B. Khoshnevis, Contour crafting process planning and optimization, Pro-
ceedings of the 26th International Symposium on Automation and Robotics in
Construction (ISARC) IIT Madras, Austin, USA 2009, pp. 576–583.
[44] G. Cesaretti, E. Dini, X. De Kestelier, V. Colla, L. Pambaguian, Building components
for an outpost on the Lunar soil by means of a novel 3D printing technology,
Acta Astronaut. 93 (2014) 430–450.
[45] P. Bosscher, R.L. Williams Ii, L.S. Bryson, D. Castro-Lacouture, Cable-suspended ro-
botic contour crafting system, Autom. Constr. 17 (1) (2007) 45–55.
[46] R.L. Williams Ii, M. Xin, P. Bosscher, Contour-crafting-cartesian-cable robot system:
dynamics and controller design, Proceedings of the ASME International Design En-
gineering Technical Conferences and Computers and Information in Engineering
Conference (IDETC/CIE), Vol. 2, American Society of Mechanical Engineers, Brook-
lyn, USA 2008, pp. 39–45.
[47] R.L. Williams Ii, M. Xin, P. Bosscher, Contour-crafting-cartesian-cable robot system
concepts: workspace and stiffness comparisons, Proceedings of the ASME Interna-
tional Design Engineering Technical Conferences and Computers and Information
in Engineering Conference (IDETC/CIE), Vol. 2, American Society of Mechanical En-
gineers, Brooklyn, USA 2008, pp. 31–38.
[48] B.J. Peter, Spiderbot: Large Scale 3D Printer, 2015 Accessed: November 2015, Avail-
able from http://www.bnpeters.com/spiderbot.html.
[49] T. Monroy, T. Chen, Sky printer - prototyping an adaptable large-scale 3d-printer,
2015 Accessed: November 2015, Available from http://www.instructables.com/
id/Sky-Printer-Prototyping-An-Adaptable-Large-Scale-3/.
[50] World's Advanced Saving Project, BigDelta WASP experiments new materials in
Barcelona, 2015 Accessed: November 2015, Available from http://www.
wasproject.it/w/en/bigdelta-wasp-experiments-new-materials-at-construmat-in-
barcelona/.
[51] F. Ceccanti, E. Dini, X. De Kestelier, V. Colla, L. Pambaguian, 3D printing technology
for a moon outpost exploiting lunar soil, 61st International Astronautical Congress
2010, IAC 2010, September 27, 2010 - October 1, Vol. 11, International Astronauti-
cal Federation, Prague, Czech Republic 2010, pp. 8812–8820.
[52] M. Thangavelu, B. Khoshnevis, A. Carlson, N. Leach, Architectural concepts
employing co-robot strategy and contour crafting technologies for lunar settle-
ment infrastructure development, Proceedings of the AIAA SPACE Conference
and Exposition, American Institute of Aeronautics and Astronautics Inc., Pasadena,
USA, 2012 (p. American Institute for Aeronautics and Astronautics (AIAA).
[53] B.B. Millsaps, The swarmscapers: students create 3D printing robots capable of
building in outlying areas, 3DPrint.com2015.
[54] Institute of Advanced Architecture of Catalunya, Minibuilders: small robots print-
ing big structuresAccessed: November 2015, Available from http://robots.iaac.
net/#process.
[55] D. Sher, Tridom Joins the Race for 3D Printed Affordable Housing, 3D Printing In-
dustry, 2014.
[56] Alec. INNOprint 3D printer can build emergency housing in just under 30 minutes,
2015 Accessed: November 2015, Available from http://www.3ders.org/articles/
20150716-3d-print-emergency-housing-in-under-30-minutes-with-innoprint-
3d-printer.html.
[57] A. Ageieva, D. Zhuikov, Smart Masonry, Dessau Institute of Architecture, Anhalt
University of Applied Sciences, 2015.
[58] A.S. Howe, B. Wilcox, C. McQuin, D. Mittman, J. Townsend, R. Polit-Casillas, T.
Litwin, Modular additive construction using native materials, Proceedings of the
ASCE's Aerospace Division 14th Earth and Space Conference, American Society of
Civil Engineers (ASCE), St. Louis, USA 2014, pp. 301–312.
[59] ROB Technology AG, Flexible production of building elementsAccessed: November
2015, Available from http://www.rob-technologies.com/en/services.
[60] Gramazio Kohler Architects, Non-standardised brick facade, 2006 Accessed: No-
vember 2015, Available from http://www.gramaziokohler.com/web/e/bauten/52.
html.
[61] Gramazio Kohler Architects, Flight assembled architecture, 2011 Accessed: No-
vember 2015, Available from http://www.gramaziokohler.com/web/e/projekte/
209.html.
[62] Q. Ge, H.J. Qi, M.L. Dunn, Active materials by four-dimension printing, Appl. Phys.
Lett. 103 (13) (2013) 131901-1–131901-5.
366 N. Labonnote et al. / Automation in Construction 72 (2016) 347–366
[63] S. Tibbits, 4D printing: multi-material shape change, Archit. Des. 84 (1) (2014)
116–121.
[64] S. Tibbits, K. Cheung, Programmable materials for architectural assembly and auto-
mation, Assem. Autom. 32 (3) (2012) 216–225.
[65] C.M. Rouhana, M.S. Aoun, F.S. Faek, M.S. Eljazzar, F.R. Hamzeh, The reduction of
construction duration by implementing contour crafting (3D printing), Proceed-
ings of the 22nd Annual Conference of the International Group for Lean Construc-
tion: Understanding and Improving Project Based Production (IGLC), The
International Group for Lean Construction, Oslo, Norway 2014, pp. 1031–1042.
[66] I. Vizotto, Computational generation of free-form shells in architectural design and
civil engineering, Autom. Constr. 19 (8) (2010) 1087–1105.
[67] A. Anastasiou, C. Tsirmpas, A. Rompas, K. Giokas, D. Koutsouris, 3D printing: basic
concepts mathematics and technologies, Proceedings of the IEEE 13th Internation-
al Conference on Bioinformatics and Bioengineering (BIBE), (Chania, Greece) 2013,
pp. 1–4.
[68] A. Bagsik, S. Josupeit, V. Schoeppner, E. Klemp, Mechanical analysis of lightweight
constructions manufactured with fused deposition modeling, Proceedings of PPS-
29: The 29th International Conference of the Polymer AIP Conf. Proc. 1593 2014,
pp. 696–701.
[69] B. Sanz-Izquierdo, E.A. Parker, 3D printing technique for fabrication of frequency
selective structures for built environment, Electron. Lett. 49 (18) (2013)
1117–1118.
[70] X.W. Bai, H.O. Zhang, G.L. Wang, Electromagnetically confined weld-based additive
manufacturing, Procedia CIRP 6 (2013) 515–520.
[71] Mataerial. Anti-gravity object modeling, 2015 Accessed: November 2015, Available
from http://www.mataerial.com.
[72] D. Park, J. Lee, A. Romo, Poisson's ratio material distributions: variable Poisson's
ratio material systems for adaptive building applications, Proceedings of the 20th
International Conference on Computer-Aided Architectural Design Research in
Asia: Emerging Experiences in the Past, Present and Future of Digital Architecture
(CAADRIA), The Association for Computer-Aided Architectural Design Research in
Asia, Daegu, Republic of Korea 2015, pp. 735–744.
[73] R. Soar, Additive manufacturing technologies for the construction industry, in: N.
Hopkinson, R. Hague, P. Dickens (Eds.), Rapid Manufacturing: An Industrial Revo-
lution for the Digital Age, John Wiley & Sons, Inc. 2006, pp. 249–273.
[74] T. Zegard, G.H. Paulino, Bridging topology optimization and additive manufactur-
ing, Struct. Multidiscip. Optim. 53 (1) (2016) 175–192.
[75] Arup. 3D makeover for hyper-efficient metalwork, 2015 Accessed: November
2015, Available from http://www.arup.com/News/2015_05_May/11_May_3D_
makeover_for_hyper-efficient_metalwork.aspx.
[76] A. Medeiros e Sa, V.M. Mello, K.R. Echavarria, D. Covill, Adaptive voids: primal and
dual adaptive cellular structures for additive manufacturing, Vis. Comput. 31 (6–8)
(2015) 799–808.
[77] H.K. Kwon, M.K. Jang, 3D free form fabrication using the pivoting side trowel with
ceramic material, Mater. Sci. Forum 658 (2010) 268–271.
[78] T.S. Bowen, As prefabrication sheds its off-the-rack image, automation via 3D
printing threatens to transform conventional construction, Archit. Rec. 195 (4)
(2007) 166.
[79] R.M. Mahamood, E.T. Akinlabi, M. Shukla, S. Pityana, Revolutionary additive
manufacturing: an overview, Lasers Eng. 27 (3–4) (2014) 161–178.
[80] S. Tibbits, The emergence of “4D printing”, 2013 Accessed: November 2015, Avail-
able from https://www.ted.com/talks/skylar_tibbits_the_emergence_of_4d_print-
ing?language=en.
[81] T.A. Campbell, S. Tibbits, B. Garrett, The Next Wave: 4D Printing - Programming the
Material World, Washington DC, 2014.
[82] R.A. Buswell, R.C. Soar, A.G.F. Gibb, A. Thorpe, Freeform construction application re-
search, advances in engineering structures, mechanics and construction, Solid
Mech. Applic. 140 (8) (2006) 773–780.
[83] N. Labonnote, J.H. Mork, A. Gunleiksrud, A. Rønnquist, B. Manum, Trondheim
HolzBau pier: the elegance of efficiency, IABSE Symp. Rep. 104 (25) (2015) 1–8.
[84] T. Diez, Personal fabrication: fab labs as platforms for citizen-based innovation,
from microcontrollers to cities, Nexus Network Journal 14 (3) (2012) 457–468.
[85] WikiHouse. Architecture of the people by the peopleAccessed: November 2015,
Available from http://www.wikihouse.cc/about/.
[86] B.C. Chang, J. Lee, T.S. Lee, Experimental Study on Waterless Lunar Concrete for
Landing Pad Construction, 64th International Astronautical Congress, IAC, Septem-
ber 23–September 27, vol. 3, International Astronautical Federation, Beijing, China,
2013 1728–1731.
[87] European Space Agency, Lunarville! 2015 Accessed: November 2015, Available
from http://www.youtube.com/watch?v=BFr2WmeQu74.
[88] T. Rousek, K. Eriksson, O. Doule, SinterHab, Acta Astronaut. 74 (2012) 98–111.
[89] B. Khoshnevis, M.P. Bodiford, K.H. Burks, E. Ethridge, D. Tucker, W. Kim, H. Toutanji,
M.R. Fiske, Lunar contour crafting - a novel technique for ISRU-based habitat devel-
opment, Proceeedings of the 43rd AIAA Aerospace Sciences Meeting and Exhibit,
American Institute of Aeronautics and Astronautics Inc., Reno, USA 2005,
pp. 7397–7409.
[90] B. Kading, J. Straub, Utilizing in-situ resources and 3D printing structures for a
manned Mars mission, Acta Astronaut. 107 (2015) 317–326.
[91] M.M. Johnston, M.J. Werkheiser, M.P. Snyder, J.E. Edmunson, 3D printing in Zero-G
ISS technology demonstration, Proceedings of the AIAA SPACE 2014 Conference
and Exposition, San Diego, USA, 2014.
[92] B. Berman, 3-D printing: the new industrial revolution, Bus. Horiz. 55 (2) (2012)
155–162.
[93] C. Achillas, D. Aidonis, E. Iakovou, M. Thymianidis, D. Tzetzis, A methodological
framework for the inclusion of modern additive manufacturing into the produc-
tion portfolio of a focused factory, J. Manuf. Syst. 37 (1) (2015) 328–339.
[94] L. Haymond, D. Noble, Rapid contour crafting to create more sustainable housing,
Proceeedings of the SOLAR 2008: Catch the Clean Energy Wave, Vol. 8, American
Solar Energy Society, San Diego, USA 2008, pp. 5051–5057.
[95] E. Ehrnström, M. Demir, The Potential of Additive Manufacturing for Building Sin-
gle-Family Houses in the Near Future (in Swedish), School of Industrial Engineer-
ing and Management, KTH Royal Institute of Technology, 2013.
[96] www.3ders.org. Loughborough University teams up with Skanska to build
commerical 3D concrete printing robot, 2014 Accessed: November 2015, Available
from http://www.3ders.org/articles/20141121-loughborough-university-skanska-
to-build-commerical-3d-concrete-printing-robot.html.
[97] J.F. Brandon, Don't believe the hype: WinSun unveils bland multi-story 3D printed
buildings, SolidSmack, www.solidsmack.com2015.
[98] United Nations - framework convention on climate change. Historic Paris agree-
ment on climate change, 2015 Accessed: November 2015, Available from http://
newsroom.unfccc.int/unfccc-newsroom/finale-cop21/.
[99] United Nations Environment Programme, Buildings and Climate Change - Summa-
ry for Decision-Makers, Paris, France, 2009.
[100] A. Drizo, J. Pegna, Environmental impacts of rapid prototyping: an overview of re-
search to date, Rapid Prototyp. J. 12 (2) (2006) 64–71.