Automation in rammed earth construction for industry 4.0: Precedent work, current progress and future prospect

Abstract

The increasing awareness of the undesirable environmental impact of cement-based products has led to rethinking earth construction within the digitally mechanised framework of Industry 4.0. The attempts to mechanise earth construction started nearly a decade ago, yet, the past four years have seen a surge in the intensity of research on the advanced manufacturing of earth construction. Additive manufacturing of clay-rich soil, like in cob and adobe methods, has been attracting the primary attention, while on the contrary, considerably limited research has been conducted to mechanise other methods such as rammed earth. This paper collected, reviewed, and analysed the state-of-the-art research on the advanced manufacturing of rammed earth construction, with a focus on process design aspects and machine development. Two case studies of recent digitally manufactured rammed earth projects in Australia and Switzerland were comprehensively analysed to provide a realistic feasibility assessment of the potential and technical challenges. The insights from this research provide a holistic and tangible roadmap to the digital manufacturing of rammed earth construction, which aids in making it more desirable and capable of meeting contemporary demands of the architecture and construction of Industry 4.0.

Full text

Journal of Cleaner Production 398 (2023) 136569
Available online 2 March 2023
0959-6526/© 2023 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
Review
Automation in rammed earth construction for industry 4.0: Precedent work,
current progress and future prospect
Mohamed Gomaa
a
, Sascha Schade
b
, Ding Wen Bao
a
,
c
, Yi Min Xie
a
,
*
a
Centre for Innovative Structures and Materials, School of Engineering, RMIT University, Melbourne, 3001, Australia
b
ERNE AG Holzbau, Laufenburg, 5080, Switzerland
c
School of Architecture and Urban Design, RMIT University, Melbourne, 3001, Australia
ARTICLE INFO
Handling Editor: Mingzhou Jin
Keywords:
Rammed earth
Mechanisation
Digital manufacturing
Industry 4.0
Earth construction
ABSTRACT
The increasing awareness of the undesirable environmental impact of cement-based products has led to
rethinking earth construction within the digitally mechanised framework of Industry 4.0. The attempts to
mechanise earth construction started nearly a decade ago, yet, the past four years have seen a surge in the in-
tensity of research on the advanced manufacturing of earth construction. Additive manufacturing of clay-rich
soil, like in cob and adobe methods, has been attracting the primary attention, while on the contrary, consid-
erably limited research has been conducted to mechanise other methods such as rammed earth. This paper
collected, reviewed, and analysed the state-of-the-art research on the advanced manufacturing of rammed earth
construction, with a focus on process design aspects and machine development. Two case studies of recent
digitally manufactured rammed earth projects in Australia and Switzerland were comprehensively analysed to
provide a realistic feasibility assessment of the potential and technical challenges. The insights from this research
provide a holistic and tangible roadmap to the digital manufacturing of rammed earth construction, which aids in
making it more desirable and capable of meeting contemporary demands of the architecture and construction of
Industry 4.0.
1. Introduction
The construction industry has been undergoing a radical change with
a faster pace than ever before as mankind leaps into the fourth industrial
revolution (Industry 4.0). The term construction 4.0 has been intro-
duced recently to reect the eagerness for using digital manufacturing,
machine learning and articial intelligence (AI) technologies in con-
struction industry (Kozlovska et al., 2021). The demand for this tech-
nology has especially grown upon empirical proofs in various studies
over the past decade, which demonstrated that automating the con-
struction process can provide substantial benets to productivity, ef-
ciency, and quality, while it also provides high freedom in design
(Zareiyan and Khoshnevis, 2017). It has become obvious that the uptake
of modern manufacturing technologies in the construction sector is
increasing rapidly, where the number of companies that provide
advanced manufacturing services is rocketing worldwide, in response to
the evolving market demand. Additive manufacturing (AM) methods,
and especially 3D printing of building components, are receiving most of
the attention in research and practice (Feng et al., 2015; Gomaa et al.,
2021a).
Nowadays, the race to develop a fully automated construction pro-
cess is well under way. Several companies and institutions worldwide
have been rushing to build the full-scale structures and houses using
digital and robotic manufacturing methods. This frantic race is heavily
dependent on employing cement-based materials due to their relatively
high productivity and structural efciency. Moreover, the research and
development of cement and concrete products has also become quite
well-established over the past few decades (Geneidy et al., 2019), while
other sustainable materials receive a considerably lower attention in
comparison (Alhumayani et al., 2020). This surging interest in cemen-
titious materials is triggering further concerns over the possible increase
in environmental impact of the construction industry, which is already
responsible for almost 40% of the energy consumption and greenhouse
gas emissions globally. Furthermore, cement products alone are gener-
ating 5–8% of the global CO
2
emissions (International Energy Agency,
2021).
With the increased awareness of the global warming crisis and re-
sources depletion, researchers and other stakeholders in building and
* Corresponding author
E-mail address: mike.xie@rmit.edu.au (Y.M. Xie).
Contents lists available at ScienceDirect
Journal of Cleaner Production
journal homepage: www.elsevier.com/locate/jclepro
https://doi.org/10.1016/j.jclepro.2023.136569
Received 17 November 2022; Received in revised form 30 January 2023; Accepted 20 February 2023

Journal of Cleaner Production 398 (2023) 136569
2
construction elds have become highly conscious about minimising the
environmental footprint of construction processes and materials. Since
nearly 50% of the world’s raw materials are consumed in construction
industry (Weiβenberger et al., 2014), it has become a necessity to
rethink conventional construction materials and seek more sustainable
substitutes. In this context, earth-based materials re-emerge as potential
alternatives to the cement-based counterparts, where it can provide a
more affordable and eco-friendly construction solution for several con-
struction sectors (e.g., residential projects) (Gomaa et al., 2022). For
centuries, earth (or subsoil) has been used as a core ingredient for
traditional construction mixtures, where local resources and manual
skills are efciently utilised to build various structures and buildings. A
traditional earth mixture includes subsoil, water, and occasional addi-
tion of bre to improve the mixture’s mechanical strength. Air lime was
also added sometimes for increased durability and to enhance the
structural stability when the clay content is naturally low in the used raw
earth (Parracha et al., 2020; Silva et al., 2022). Earth buildings are
historically well-known for their sustainability, durability, and adequate
thermal efciency when complemented with good passive design stra-
tegies. However, since the discovery of modern concrete, there has been
a dramatic retreat from building with earth materials, especially when
considering the lengthy and labour-intense construction process (Alhu-
mayani et al., 2020; Veliz Reyes et al., 2019).
Modern construction and architecture industries have beneted
signicantly from the mechanisation of several building materials such
as concrete, masonry, and timber. Nowadays, digital manufacturing
technologies like 3D printing and robotic manufacturing have been
employed intensively in several elds of construction research, with a
primary attention given to 3D concrete printing. However, since 2016,
the interest in mechanising earth construction has been increasing
noticeably. Several studies worldwide have emphasised the importance
of leveraging digital manufacturing technologies to promote the re-use
of earth construction in a contemporary context (Perrot et al., 2018;
Veliz Reyes et al., 2019). Adopting digital manufacturing for earth
construction (DMEC) proved its ability to harness the sustainability
benets of earth materials, while also providing a potential alternative
for modern concrete construction (Gomaa et al., 2022).
Despite the increase in the number of studies on DMEC, the current
experiments and applications are largely focused on additive
manufacturing of wet-based earth methods (e.g., cob and adobe)
(Gomaa et al., 2022). On the contrary, there is considerably limited
research on digital manufacturing of dry-based earth methods (i.e.,
rammed earth). From a modern perspective, rammed earth presents a
unique form of additive manufacturing, where the structures are
generated through the addition and compaction of successive layers of
earth. To date, there are only seven recorded projects that utilise
advanced manufacturing methods for rammed earth construction.
Furthermore, most of these research projects focus on mechanising the
ramming process through replacing manual tampers with automated
ones, while little has been done to revolutionise the design of rammed
earth structures, through changing the conventional at or
single-curved walls into more intricate, free-form ones. In addition,
there is still a knowledge gap between the different studies and a lack of
denitive feasibility evidence, which both hinder adopting advanced
manufacturing for rammed earth (AMRE) in modern construction by
stakeholders and the regulating authorities (Bick, 2016; Gomaa et al.,
2021b; Schweiker et al., 2021).
Within this context, this paper aims to bridge the gaps in knowledge
between the separate research silos on AMRE. A critical review for the
state-of-the-art research on AMRE is conducted to establish a wider
perception for their potential and limitations. In addition, two case
studies from recent projects in Australia and Switzerland are compre-
hensively analysed to provide a realistic feasibility assessment of the
technological development and material processing. The insights from
this research provide a holistic and tangible roadmap to the digital
manufacturing of rammed earth construction, which aids in making it
more desirable and capable of meeting contemporary demands of the
architecture and construction of Industry 4.0.
2. Rammed earth construction
Rammed earth is a type of earth-based construction, where dry soil
and water are mixed, then compacted in consecutive lifts within a
formwork to create structures. Out of the several styles of earth con-
struction, rammed earth presents the highest mechanical and structural
strength due to the near-dry nature of mixture and the formation
through compaction process. Rammed earth buildings are traditionally
known for their strength and durability, which make them one of the
most desirable types of earth construction worldwide (Hall et al., 2012;
Walker, 2005). Stabilised Rammed earth compares favourably with
masonry or concrete blocks in compression strength and resistance to
erosions, while the cost of its raw material is cheaper. The overall cost of
rammed earth construction varies based on region and the associated
labour price.
The soil used in rammed earth must be excavated at a depth of more
than 1 m below the ground surface; therefore it is called subsoil. Topsoil
from a lower depth is good for plantation and not suitable for rammed
earth. Subsoil typically contains four main grades of materials: gravel,
sand, silt, and clay. Each grade is classied according to particle sizes. A
consistent variety of particle sizes is ideal, with a good balance between
coarse and ne aggregates (Gomes et al., 2014). Conducting an on-site
testing or lab analysis is essential to guarantee the quality of rammed
earth mixture. Typically, good subsoil for rammed earth should have a
high sand content (30–50%) with no more than 8% clay. Clay is a nat-
ural binder in a rammed earth mixture, which enhances the compaction
process and the mechanical performance. Binding agents, or binders, are
core ingredients of rammed earth construction. However, excessive clay
in the subsoil may cause heaving and cracking problems. Subsoil with
more than 8% clay must be modied through adding washed, sharp sand
(Hall et al., 2012; Keable, 1996).
Subsoil quality can be enhanced to reach the optimal balance be-
tween the different grades of materials through addition of sand or clay.
The rammed earth that uses clay only as the main binder is commonly
known as unstabilised rammed earth (URE). Other binding agents have
been used over the time such as lime and more recently cement to
substantially enhance the mechanical properties of the mixture, leading
to what is known as stabilised rammed earth (SRE). Cement has become
the most widely used binder for rammed earth stabilisation as it
signicantly enhances the mechanical strength and durability, while it is
also abundant worldwide. However, the increased awareness of the
environmental impact of cement production has spurred increasing
research on other sustainable binders (Hall et al., 2012; Keable, 1996).
Several recent studies have highlighted the strong potential of using
bio-based additives such as xanthan gum, guar gum and linseed oil for
rammed earth stabilising and earth-based structures in general (Chang
et al., 2015; Fatehi et al., 2021; Perrot and Rangeard, 2021; Schweiker
et al., 2021).
Rammed earth walls act as the main structural system in earth
construction. The wall thickness varies according to the expected loads
and the number of storeys, with an average of 60 cm (Quagliarini et al.,
2010; Weismann and Bryce, 2006). The wall thickness increases pro-
portionally with the number of storeys or the height of the building and
may also taper to be larger at the bottom and smaller at the top. The
purpose of rammed earth construction also inuences the wall thickness.
For example, defensive military structures usually requires much thicker
walls (Parracha et al., 2020). In general, the mechanical properties of
earth mixtures depend on several factors: subsoil properties, water
content, the use of bre, and the quality of craftsmanship.
To date, most of rammed earth construction methods worldwide still
employ conventional approaches for production, which are heavily
dependent on labour and conventional wooden or metal formwork.
There have been a few transitions during the past few decades from
M. Gomaa et al.

Journal of Cleaner Production 398 (2023) 136569
3
using manual tampers to pneumatic assisted tampers (Fig. 1), while
more advanced formwork systems have been developed to enhance the
reusability, versatility and time efciency (Fig. 2)(PERI, 2022; Rammed
Earth Consulting, 2022). These enhanced techniques of production have
helped modernise rammed earth industry through the increased pro-
ductivity and quality.
The denition of the term ‘modern rammed earth’ construction de-
pends on several factors. Hall et al. (2012) demonstrated that any con-
struction must meet one or more of the following criteria to be classied
as modern:
•To leverage new technological advancement in tools and machinery
in construction process.
•To reect a high level of quality in detailing, accuracy, nishing, and
reproducibility.
•To comply with modern building regulations for structural, thermal,
and environmental performance.
•To meet the contemporary expectations and needs of occupants, in
terms of design, comfort and well-being.
Following the previous criteria, modern rammed earth construction
can be classied into two categories according to the technological
advancement level of the construction process. The rst category is
where electric and/or fuel powered tools are utilised to enhance the
efciency, productivity, and quality of rammed earth construction. Such
processes are highly dependent on humans to actively operate and
observe the machines. Machines in this context are not programmed to
be functioning in fully automated modes. There are several rms around
the world that provide modern rammed earth buildings that meet the
previously described standards (Fig. 3-right).
The second category of modern rammed earth, which is the focus of
this study, is when advanced manufacturing techniques are leveraged in
the construction process. This approach utilises various levels of auto-
mation or mechanisation in the production of either the formwork or the
compaction process. The recent decade has witnessed an increase in
adopting advanced manufacturing techniques in the rammed earth in-
dustry as it offers more productivity, higher quality, and less labour
intensity. However, the number of projects that have explored and
developed this approach are very low, while there are still considerable
limitations in the versatility of design and process efciency.
3. Materials and methods
This study aims to investigate new approaches to rammed earth
manufacturing that can leverage the advantages of the precedent pro-
jects and propose solutions to overcome the current challenges. The
study comes as a part of a larger project that investigates the potential of
using advanced manufacturing for different types of earth construction.
The objectives of this study are twofold:
1) To critically review state-of-the-art projects on utilising advanced
manufacturing methods for rammed earth construction.
2) To present comprehensive details on two recent case studies that
utilise digital manufacturing methods in rammed earth construction.
The ndings from this study provide a holistic understanding of the
advantages and limitations of the existing approaches of modern ram-
med earth construction. This improved understanding provides neces-
sary guidance for the prototyping and experimentation of new
approaches to implementing digital manufacturing in both processes of
formwork production and earth ramming.
3.1. Review method
The ndings are based on reviewing the literature/projects found
through several online scholarly databases (e.g., ScienceDirect and
Google Scholar), media releases, and social media platforms (e.g.,
LinkedIn, Instagram). The inclusion criteria for work/data extracted
from the literature are as follows:
•The projects must be utilising advanced manufacturing technologies
(e.g., robotic arms, 3D printing, automated/autonomous machines)
in rammed earth construction, either in the formwork production or
in the compaction process.
•The project must address actual building components/prototypes,
either on a small scale (e.g., modular components, bricks/blocks) or
on large scale (e.g., full-size walls). Purely artistic and non-functional
elements are disregarded.
The search terminologies included combinations between digital
fabrication and rammed earth construction. Search terms such as “dig-
ital rammed earth” “mechanised rammed earth” “robotic rammed earth”
have been used. The search using the inclusion criteria resulted in six
recorded works/approaches associated with digital fabrication and/or
mechanisation of rammed earth (Table 1). Also, the recorded works
included one published paper and one book. The rest of the works are
available online as showcases and student projects, which are all pre-
sented using online platforms such as institutional websites, online
magazines, online videos, and social media.
3.2. Physical experiments method
This study provides comprehensive details on two recent projects in
Australia and Switzerland respectively. Both projects leverage robotic
fabrication at different levels of the production process. The project in
Australia was conducted by a research team at the Centre for Innovative
Structures and Materials (CISM) at RMIT University, which focused on
producing free-form formwork using robotic 3D printing of recycled
plastic. The project in Switzerland, conducted by ERNE AG Holzbau,
utilised industrial robotic arms for material compaction, as well as using
Fig. 1. Traditional tampers (left)(Rammed Earth Consulting, 2022); pneumatic tamper (right).
M. Gomaa et al.

Journal of Cleaner Production 398 (2023) 136569
4
semi-autonomous material feeding and formwork assembly. Both pro-
jects followed a rigorous procedure for material testing and quality
assurance. Basic material testing was conducted in laboratories to verify
the suitable particle size distribution, clay content, and compression
strength test. Since the focus of this paper is the manufacturing process,
information on specic material testing will not be detailed. Both pro-
jects used CAD tools for geometry production and robotic code gener-
ation. The project at RMIT University used Rhinoceros via Grasshopper
and KUKA PRC® for robotic programming. ERNE used their own
developed interface to control the robot directly from any high-level
language (e.g., Python).
4. Advanced manufacturing for rammed earth: review of works
4.1. Mechanised manufacturing of rammed earth
4.1.1. Lehm Ton Erde LLC © (Martin Rauch)
In 2015, Martin Rauch through his company in Austria (Lehm Ton
Erde LLC) introduced one of the earliest mechanised approaches for
rammed earth production. Rauch’s machine “Roberta” resembles an
industrial production line, which consists of four main components:
formwork, material conveyer, compaction machine, and cutting ma-
chine (Fig. 4). Each linear section of the formwork has a length of 20 m
and a height up to 2.8 m. Assembly and disassembly of formwork is
executed semi-automatically with machine assistance. The fresh,
unstabilised earth mixture is then delivered to the formwork section
through the conveyor. The conveyor moves automatically along the
length of the mould to assure the consistent distribution of earth mixture
per each layer inside the mould. Meanwhile, a special compaction ma-
chine is inserted within the formwork and follows the material conveyor
to conduct the ramming process. This process is then repeated until
reaching the desired high of the wall. Each compressed layer is roughly
12 cm high. A thin layer of cement mortar is added after every third
layer of rammed earth at the outer edge of the wall to enhance abrasion
resistance and improve durability. The usual wall thickness is 60 cm.
After disassembling the formwork, a semi-automatic vertical cutting
machine is used to slice each rammed earth linear section into smaller
panels according to the envisaged design requirement. These panels are
then packed and transported to construction sites (Howe et al., 2019;
Rauch, 2020; Sauer and Kapnger, 2015).
Rauch’s approach is considered highly efcient and productive,
while it provides adequate quality and durability. However, it is still
dependant on human workers to operate and direct the machines. The
process is not considered autonomous. Moreover, the produced elements
are at wall only, which constrains the design freedom when a more
intricate design is required by the customer.
4.1.2. Freeform by Form Earth©
In 2020, the Australian start-up company Form Earth© introduced a
prototype for an automated system for rammed earth wall production.
The system, called ‘Freeform’, utilises a compact unit consisting of a
material hopper, a compaction tool, and a slip formwork. The whole
system is portable and can operate either off-site or on-site. The standard
system creates a 2.0 m ×2.7 m wall panel, and it can be adjusted to suit
Fig. 2. Traditional timber formwork (left)(Rammed Earth Consulting, 2022); Modern formwork system by PERI (right)(PERI, 2022).
Fig. 3. Traditional rammed earth house (left)(Rammed Earth Consulting, 2022); modern rammed earth house by Olnee Construction (right)(Olnee Rammed Earth,
n.d.).
Table 1
List of recorded projects on advanced manufacturing of rammed earth.
Referenced Project Design Formwork Location
Lehm Ton Erde (2015) Prefabricated
panels
Timber +metal Austria
Form Earth (2022) Prefabricated
panels
Metal (slip form) Australia
Hurtado and Eloualid
(2016)
Modular Timber +metal USA
Akipek and Yazar
(2017)
Free form module Timber +
Styrofoam
Turkey
Kloft et al. (2019) Prefabricated
panels
Timber (slip
form)
Germany
ERNE AG Holzbau
(2022)
Prefabricated
panels
Timber +metal Switzerland
RMIT University (2022) Free form 3D printed plastic Australia
M. Gomaa et al.

Journal of Cleaner Production 398 (2023) 136569
5
the desired requirements of different projects. The material hopper is
designed to cover the full length of the slip form. This way, a full layer of
rammed earth mixture is added per every load discharge from the
hopper. The process starts by locating the slip formwork at the
designated height of the wall. The fully loaded hopper ascends auto-
matically to discharge the rammed earth mixture into the formwork. The
compaction tool, which also ascends simultaneously with the formwork,
starts moving within the formwork to conduct the ramming process. The
Fig. 4. Roberta machine for rammed earth production by Martin Rauch. (1) assembled production line; (2) the compaction machine; (3) fresh product after
disassembly; (4) stored wall section after cut; (5) constructed wall on-site (Rauch, 2020).
Fig. 5. Form earth system for prefabricated rammed earth panels (Form Earth, 2022).
M. Gomaa et al.

Journal of Cleaner Production 398 (2023) 136569
6
compaction tool is equipped with sensors to determine the optimum
compaction level. Upon every nished lift, the form work slides up-
wards, and the process repeats to create the successive lifts (Fig. 5).
This system by Form Earth is considered similar in concept to
Rauch’s system. Despite the smaller size, it presents a step forward from
Rauch’s machine, where the new system is ‘nearly’ autonomous and
portable. Also, this approach can eliminate the need for cutting, storage,
and shipping as it can operate directly on-site. The slip form mechanism
also means a reduction in raw material consumption for formwork.
However, the system requires signicant further development in several
aspects. Although Form Earth claims that the process is fully automated,
it seems from the released media that the system is still dependant on
human direction to some extent. Also, there seems to be a lack of quality
in the surfaces and edges of the walls. This could be resulting from the
slip form which may jeopardise the nishing quality of fresh rammed
earth. On the other hand, the produced elements are straight-shaped
panels, which constrains the freedom of producing intricate designs.
4.2. Digitally manufactured rammed earth
The previous attempts, while innovative, mimic a conventional in-
dustrial production line, where the process is mechanised, but not yet
computational or digital. This means that the tools are made to mass
produce the same shape and geometry. This is efcient in increasing the
productivity, yet contemporary architecture and construction industry
requires achieving more versatile and bespoke geometries that are
aesthetically attractive and more intricate. Reaching this goal requires
utilisation of advanced manufacturing tools such as computer numerical
control (CNC) machines or robotic arms, either for the ramming process,
the mould making, or both combined.
4.2.1. Engineered cast earth
One of the earliest attempts to introduce digital fabrication into
mould production was recorded in 2016 by Hurtado and Eloualid
(2016). The project uses CNC laser cutting to fabricate the formwork
parts out of metal and plywood. The metal mould is extracted from a
single sheet, where an unfolded shape is generated to on a planar surface
then folded and tack-welded into the desired geometry. This metal
element is then inserted within a plywood box to create the nal brick
mould. The soil mixture is then added and compacted to create a ram-
med earth brick. The bricks are then assembled into the desired wall size
(Fig. 6). The design provides a modular way to create self-shaded wall
that provides better thermal efciency and a more intriguing look. While
the idea itself of using inner mould with rammed earth is not new, this
project differs as it presents an early approach to utilising computational
design and manufacturing methods in the production process of the
mould. However, the project is completely dependent on humans for
material feeding, compaction, and assembly.
4.2.2. Common action rammed earth wall by Akipek and Yazar (2017)
Common action walls is a short project that was rst presented in
2017 by a team from Istanbul Bilgi University in 2017 (Akipek and
Yazar, 2017). This project marks the rst recorded attempt to utilise
robotic manufacturing in formwork production for modular rammed
earth. The project also uses computational methods for form generation
of intricate formwork. Robotic hot-wire cut method was utilised to carve
rigid styrofoam blocks into the desired mould. These carved styrofoam
elements are then slotted into a rectangular metal skeleton. Fresh earth
mixture is added and then rammed manually. These produced rammed
earth blocks are then assembled to create wall panels (Fig. 7).
This project presents an innovative approach to produce a compu-
tationally designed rammed earth walls with embedded cavities that
improves air circulation and self-shading, thus enhancing its environ-
mental efciency. Moreover, the modular approach enhances the pro-
ductivity and reduces the raw material consumption for formwork. The
use of robotic arms in hot wire cutting also offers high accuracy and
quality. However, the ramming process was totally dependent on
humans and manual tools, which was deemed essential considering the
complex nature of the mould at that time. Also, there are no released
data on the structural performance of the produced geometry, which is a
concern considering that conventional rammed earth gains its structural
strength from its mass.
4.2.3. Robotic rammed earth components by Kloft et al. (2019)
Although the previous examples exhibited innovative ways for the
production and formation of rammed earth, they involved human labour
at different stages of the production process. Consequently, these ex-
amples would be classied as semi-automated approaches. Kloft et al.
(2019) introduced a fully automated approach that took a step further
into industrialising rammed earth construction. The new approach
leveraged a robotic system that provides an augmented ramming pro-
cess with highly optimised material consumption and labour. The sys-
tem replaces the manually controlled compaction and material feeding
with robotically controlled ones. Moreover, the system adopts a slipform
technique which aids in minimising the need for formwork. The slipform
size is adjustable to suit different design requirements. Both the tamper
and formwork create a semi-enclosed cell that is controlled by the ro-
botic arm. A collaborative robot performs deposition of earth mixture
layer per layer along the wall path, then the robotic tampering cell
follows the same path to apply the compaction. The process repeats until
the full height of the wall is achieved (Fig. 8).
This automated approach provides innovative solutions for several
limitations in the previous projects. The use of synchronised collabora-
tive robotic systems for material feeding and ramming offers highly
economical and environmental benets due to the considerable reduc-
tion in labour, construction time, and formwork. However, despite the
use of advanced robotic arms, the produced elements are still the stan-
dard linear wall panels. The suggested approach does not unleash the
full capabilities of robotic arms to move in multi-directions and thus
achieve more intricate geometries. The current mechanism of the slip
form still considerably limits the design freedom, while it may inuence
detrimentally the nishing quality of the walls.
Fig. 6. Rammed earth blocks by Hurtado and Eloualid (2016).
M. Gomaa et al.

Journal of Cleaner Production 398 (2023) 136569
7
5. Case study 1: free-form rammed earth, RMIT university,
Australia
The observations from previous projects show a consistent short-
coming either in the mechanised ramming process or in the formwork
production. It is clearly noticed that each project focuses mostly on one
aspect of development, which is mostly the ramming tools and process,
while very little has been done to develop a formwork system with free-
form formation capabilities. Hence, a team from the Centre for Inno-
vative Structures and Materials at RMIT University has been investi-
gating the possibility of leveraging advanced manufacturing methods
for both formwork production and compaction/ramming since 2021.
The project consists of two phases: the rst phase focuses on exploring
the potential of using 3D printing for free-form formwork production,
while the second phase will focus on developing a robotically controlled
ramming system that offers multi-directional ramming technique. This
paper only focuses on presenting the early outcome of the ongoing phase
one, while phase two experimentations will be conducted in the future.
5.1. Formwork and ramming tool
The selected material for 3D printing is Polyethylene terephthalate
glycol (PETG), which is a type of thermoplastic that offers adequate ri-
gidity and high recyclability. PETG has been used for years in different
projects within the school of Architecture at RMIT University, which
made it an excellent candidate material for 3D printing of rammed earth
formwork. The PETG used in this project was all recycled from other past
projects to further reduce the carbon footprint of the experiments. Prior
to starting the full-scale test, a pilot experiment was conducted on a
small scale to examine the feasibility and workability of using 3DP
formwork for rammed earth. An existing 3D printed plastic formation
was repurposed to create a small, rammed earth brick (Fig. 9). The pilot
study was successful, and provided early indications for process
improvement on a large scale.
The full-scale project reused an existing 3D printed formwork from
another recent project by the team in RMIT called NerviPrint (Ma et al.,
2022). The selected design features a 2.5 m tall morphing column that
mimics one of the designs by the famous engineer Pier Luigi Nervi
(Fig. 10). Using this design also provides an opportunity to observe
several prospective challenges that are critical for the feasibility
assessment of this proposed 3D printed formwork system. These pro-
spective challenges are:
1) The impact of ramming process on the 3D printed PETG formwork,
and the correlation between the cross-section design of the formwork
and the sufcient rigidity
Fig. 7. Digital fabrication for rammed earth formwork, Common action wall project by Akipek and Yazar (2017).
Fig. 8. Robotic rammed earth components by Kloft et al. (2019).
M. Gomaa et al.

Journal of Cleaner Production 398 (2023) 136569
8
2) The impact of the corrugated nature of 3D printed formwork on
demoulding process and rammed earth components’ surface nish
3) The ability to produce complex geometries with protruding parts
and/or grooves
4) The connection rigidity between the formwork sections.
The 3D printing facility at RMIT University leverages a robotically
controlled plastic extruder that uses plastic pellets instead of laments
(Fig. 11). The robotic platform enables 3D printing of large parts within
a bounding box of 2 m ×1 m ×1 m. The printing nozzle diameter is 3
mm, which facilitates printing large and rigid plastic components in a
signicantly shorter time as compared to other conventional gantry
printers. Two different approaches for the cross-section design were
examined to determine the suitable rigidity that better withstand the
compaction forces. The full formwork is 2.5 m tall, with a footprint of 50
cm ×50 cm at the bottom, and 40 cm ×40 cm at the top. It consists of
four main parts with a height of 1.25 m each (Fig. 12). The parts are
connected using metal clamps. The inner surfaces of the formwork were
lightly greased prior to the compaction process to improve the
demoulding. The project uses a Kawasaki KPT 2L pneumatic rammer
with a rubber butt of 50 mm diameter. The rammer compresses a slim,
light-weighted design (4.0 kg) with an adequate compaction power
delivers 90 psi over 1000 blows per minutes (BPM), making it effective
for operations within narrow spaces.
5.2. Material mixture preparation
The project uses two different rammed earth mixtures to examine the
inuence of the soil particle sizes on the nal shape, while also exhib-
iting an interesting colour blend in the nal nish (Fig. 13). The subsoil
sourcing location is the major factor that controls the properties of the
mixture. Cream crushed rocks (CCR) and red crushed rocks (RCR)
sourced from local quarries in Victoria, Australia, are used as subsoil in
this experiment. CCR in Australia naturally contains low clay content
(<5%), which makes it favourable for stabilised rammed earth methods.
Cement stabilisation is a very common practice in the rammed earth
industry in Australia, where 5–10% cement (by weight) is added to the
mixture to improve its mechanical performance and durability. RCR, on
the other hand, naturally contains high clay content (average 12%),
which makes it more suitable for the unstabilised rammed earth method.
However, the good clay content for unstabilised rammed earth is around
8% (Hall et al., 2012), therefore, washed sand was added to the RCR
patch. In general, sand is also added to subsoil to control/enhance the
particle size distribution, where subsoil with high sand content
(30–50%) is also preferred. The CCR type in this experiment is 14 mm
minus, which means the maximum particle size is 14 mm in diameter,
while the RCR is 20 mm minus. RCR can be sourced at lower particle size
range; however, this study intended to examine the inuence of the
particle size on the ramming process and surface quality. Both CCR and
RCR mixtures were stabilised by 5% cement in this study.
Fig. 9. Pilot test for a small scale 3D printed PETG formwork (left) and the produced rammed earth element (right).
Fig. 10. Original design of the morphing column by Nervi (left); the PrintNervi column replica (middle) and 3D printed plastic formwork by Ma et al.(right) (Ma
et al., 2022).
M. Gomaa et al.

Journal of Cleaner Production 398 (2023) 136569
9
It is worth mentioning that, the major aim of this project was to focus
on the feasibility and material processing aspects of the 3D printed
formwork for rammed earth. Hence, the study used standard practice for
rammed earth mixture preparation by the industrial partner in Mel-
bourne, Olnee Construction (Olnee Rammed Earth, n.d.).Specic ma-
terial properties and testing of alternative binders is the subject of an
ongoing study by the team. The reproduction of this experiment can be
conducted anywhere according to standard local practice for rammed
earth mixture preparations, whether it is stabilised or not.
5.3. Production process
After assembling the bottom section of the formwork, the two ram-
med earth mixtures are deposited alternately into successive lifts. Each
lift is nearly 20 cm high before compaction, which turns into roughly 15
cm after compaction. The compaction is conducted using a pneumatic
rammer, starting from the outside circumference and gradually shifting
inwards. After completing the ramming of the bottom section, the upper
section is assembled and connected to the bottom section with clamps.
The same compaction process is repeated (Fig. 14). The full process of
assembly and ramming of the column took nearly 70 min. The column
was left to harden for 24 h before demoulding as recommended by the
industry partner. The demoulding process started by releasing the
clamps from top to bottom, then gave few minutes to the formwork to
relax before removing the formwork sections. Due to the corrugated
nature of 3DP form work, it was essential to pull the formwork outwards,
and try to limit any movement upwards or downwards to preserve the
quality of the surface texture and reduce peeling. The demoulding
process was completed successfully with very minor peels and/or
Fig. 11. Robotic system for 3D printing from plastic pellet.
Fig. 12. Diagram of the formwork parts.
Fig. 13. Different piles of CCR and RCR subsoil-based mixtures for ram-
med earth.
Fig. 14. Assembled formwork and ramming process of the column.
M. Gomaa et al.

Journal of Cleaner Production 398 (2023) 136569
10
surface damages.
5.4. Remarks and recommendations
The nal outcome shows strong potential for using 3D printed plastic
formwork to produce free-form rammed earth structures (Fig. 15). Using
robotic 3D printing of recycled plastic presents an efcient and sus-
tainable way for formwork production, while it empowers designers to
achieve high complexity of rammed earth formations that meet the
demands of contemporary construction. The proposed method, howev-
er, still requires essential development on different levels. The experi-
ment revealed several challenges that must be tackled to improve the
quality and efciency of this new approach. The following are key re-
marks for the development of future work:
1) The maximum particle size in the soil mixture noticeably affects the
nishing quality. It was clearly observed that the lifts where CCR has
been used exhibited better compaction and a smoother nish as
compared to the RCR lifts. This happens due to the difference in the
maximum particle size between the two blends, where the CRC has a
maximum size of 14 mm, while it is 20 mm in the RCR. The smaller
range of particle size distribution presents better workability during
compaction as it better lls the micro grooves of the 3D printed
formwork.
2) Intricate design of rammed earth structures means intricate form-
work. High complexity can be easily achieved with modern tech-
nologies, yet it is strongly recommended to avoid sharp corners and
deep grooves as they complicate the demoulding process and in-
crease the risk of damaging the structure surfaces. Using smoother
lines and lleted corners should aid in overcoming these problems.
Furthermore, the corrugated nature of the 3D printed mould surface
generates micro interlocking between the compacted soil and the
formwork. Therefore, demoulding should be performed through
pulling the formwork sections outwards to avoid peeling and also to
maintain the unique corrugation effect on the rammed earth
components.
3) The formwork sections in this experiment have a height of 125 cm.
However, shorter heights can be used when necessary to further
facilitate the assembly and disassembly of the formwork. It is worth
mentioning that increasing the number of sections is not always
favourable, as the increased number of connections can affect the
rigidity of the overall formwork while also increasing the time
needed for assembly and demoulding.
4) Since using 3D printed PETG formwork for rammed earth is a new
approach, there were no clear expectations on how the formwork
will handle the compaction forces during the ramming process.
Hence, the project team initially decided to adopt a warren-truss
design for the formwork cross section to enhance its rigidity. The
formwork exhibited high strength and adequate workability during
the ramming process. However, the dense cross-sectional design
resulted in heavy formwork with higher material consumption and
longer printing time. The observations from the experiment show
that a less dense formwork will still be sufcient. Therefore, an
optimised cross section design is currently being developed by the
team, which will have nearly 40% reduction in material consump-
tion and printing time. Other types of thermoplastic or blends of
recycled materials can be used as well instead of PETG in the future.
5) Phase 1 of the project depended on manually guided pneumatic
tampers to perform the ramming process. The project team is
currently developing a multidirectional robotic ramming system to
be employed in phase 2 of the project, which will deliver higher
productivity and efciency in producing complex rammed earth el-
ements like double curved walls (Fig. 16).
Overall, the approach of using 3D printed thermoplastic formwork
for rammed earth construction is promising. Leveraging robotically
assisted ramming will further elevate this approach into a highly
desirable, environmentally friendly, and efcient construction tech-
nique that is capable of competing with other modern construction
methods.
6. Case study 2: robotic rammed earth walls, ERNE AG Holzbau
ERNE AG Holzbau is a Swiss company that specialises in techno-
logical solutions for modular and hybrid construction. ERNE has expe-
rience with gantry robots and industrial robots in timber construction
for wall and facade production. In 2022, they presented a robotically
controlled ramming system for industrialised, prefabricated rammed
earth, which forms a unique merge between the two earlier approaches
by Rauch (2020) and Kloft et al. (2019). ERNE system uses a six-axis
high-load industrial robotic arm with pneumatic rammer on its end
effector (Fig. 17). The company’s focus is to leverage advanced
manufacturing methods with robots to set up an industrial process for
mass-producing individual rammed earth elements directly from a
CAD/BIM model. The development project was carried out in close
cooperation with an ongoing design-and-build project by ERNE.
Fig. 15. Final look of the rammed earth morphing column after demoulding.
M. Gomaa et al.

Journal of Cleaner Production 398 (2023) 136569
11
6.1. Formwork and ramming tool
As requested by the architect of this project, the shapes of the wall
elements are rectangular cuboids with at surfaces. Elements had
different sizes, layer count, and layer height. For this purpose, a rein-
forced and adjustable timber formwork with a special mechanism for
quick assembly and disassembly was used. The dried rammed earth el-
ements are stacked with a clay mortar in the nal building. To
compensate for the accumulative thicknesses of the mortar, the lowest
layer in the bottom rammed earth element in the complete wall has a
reduced height. This technique assures that all layers from different el-
ements are equal in height and aligning perfectly when stacked, which
then gives a seamless look to the nished wall.
The robotic ramming system has seven degrees of freedom and can
work on any feasible shape given by a CAD/BIM model. The robotic arm
sits on linear track which enables moving within a working envelope of
about 15 m ×6 m ×6 m. The end-effector is a pneumatic rammer from
Mannesmann Demag© that was modied to be attached to the robot.
The ramming plate is square shaped and made of abrasion resistant steel
with a size of 80 mm ×80 mm. To protect the robot from mechanical
vibrations and excessive forces excreted by the rammer, a tuned mass
damper was engineered and installed. The rammer is self-oscillating
when placed at the right distance to the surface of the uncompacted
RE. The observed resonance frequency depends on achieved compact-
ness and can be used to estimate the level of compaction during the
process. To maximise quality control and reduce the risk of failure, the
project utilised a pneumatic rammer that has proven reliability in con-
ventional rammed earth production. Other methods of compaction such
as vibrating plate compactors and rotary electric vibrators will be
considered for future projects.
6.2. Material mixture preparation
This project uses unstabilised rammed earth. The raw material was
harvested from the excavation site where the nal building is located.
The subsoil consisted of different fractions and were analysed. The
initial mixture was directly done at the excavation site to reach the
required average clay content. The material was stored outside of the
production hall for further processing. No binding agent of any kind was
added. Only gravel was added to reach the required particle size dis-
tribution. The water content was adjusted for best compaction proper-
ties. For preparation of the production run, multiple cubic specimens
were specially manufactured to undergo compressive strength testing in
the company’s material testing laboratory to optimise the mixture for
the required load bearing capabilities and desired visual appearance.
6.3. Production process
Initial studies of the production process showed that the actual earth
ramming stage is the most time-consuming part of the process. However,
logistics play an important role in the overall efciency of the process.
Therefore, ERNE’s approach is to holistically design all stages of the
rammed earth construction process, including material handling and
logistics, where the robotic rammer is in constant operation, while all
secondary processes are complementing this primary process of ram-
ming. It was envisioned that this leads to the highest possible produc-
tivity of the whole production facility. The production cycle consists of
ve steps: 1) formwork assembly, 2) material deposition, 3) robotic
ramming, 4) demoulding, 5) transportation of completed elements to
storage. Each cycle consists of interlaced steps and operates simulta-
neously throughout three parallel zones (Fig. 18). It is important to note
Fig. 16. Envisaged process and outcome of the ongoing development of robotic ramming and 3D formwork systems by the team at RMIT University.
Fig. 17. Robotic ramming system by ERNE (Note: righ-side picture was taken during the robot installation. Safety fences were not yet placed).
M. Gomaa et al.

Journal of Cleaner Production 398 (2023) 136569
12
that formwork assembly, demoulding and transportation (steps 1, 4 and
5) occur only once per element, whereas material deposition and ram-
ming (step 2 and 3) repeats multiple times per element.
Ramming earth with a pneumatic rammer by hand is a very uner-
gonomic and repetitive task which puts heavy mechanical stress on the
workers. Thus, the goal of the project was to avoid any manual ramming
process to create attractive and safe workplaces. Given the limited re-
sources to develop a new pilot production facility, this task was chosen
to be transferred to a robotic system without losing the exibility of a
manual process. In this project, the formwork preparation and
demoulding were still designed to be carried out by workers as this is a
delicate process. However, future projects will utilise automated depo-
sition of material and formworks. The amount of deposited material in
the formworks plays a very crucial role in achieving a consistent layer
height, constant compaction, and high overall quality. Hence, a special
adjustable hopper was designed and manufactured to control precisely
the volume that is deposited for each layer.
This project required prefabricated elements of lengths between 0.5
m and up to 3.0 m. Therefore, only half of the full working envelope was
used (one side of the track) and divided into three equal zones for a
parallel production in three formworks. The production line can be
adjusted easily to accommodate more zones and more robots on the
same, or multiple, linear tracks. The process runs sequentially
throughout the zones, where every zone hosts a single step at a time. In
the balanced state, the full process is performed in the following
sequence:
1. Formwork is assembled in zone 1 while the robot arm is ramming in
zone 2 and new material is deposited in zone 3 (formwork is already
assembled in zone 2 and 3).
2. Material deposition and quality check in zone 3 is nished. The zone
is conrmed empty of any personnel and ready for ramming.
3. Formwork assembly continues in zone 1 while the robotic arm n-
ishes ramming in zone 2. The robot automatically folds to a safe
position and moves from zone 2 to zone 3.
4. Robot nishes ramming in zone 3, folds, then returns to zone 2 to
resume ramming, the next deposition of rammed earth mixture re-
starts in zone 3.
5. The sequence between zone 2 and 3 continues as long as the form-
work assembly in zone 1 takes place.
6. Once the formwork is nished in zone 1, the rst layer of rammed
earth mixture is deposited.
7. After the last layer of zone 3 is rammed and thus the element
completed, the robot continues its cycle between zone 1 and zone 2.
Meanwhile, the formwork in zone 3 is disassembled immediately,
and the element is relocated for quality inspection and then to
storage area to dry and prepare for shipping.
The dynamic assignment of the zone and element order is crucial for
an uninterrupted production. In a balanced production, elements are
nished in the three zones in a round-robin fashion. In addition, the
operation path and ramming strategies were rmly planned and
experimented to create a collision-free path, and a highly safe and
efcient operation. The path is planned independently of the robot
manufacturer (but still considering different capabilities of models),
checked in a simulation, and transferred to the robot automatically over
a communication network. No manual programming of the robot takes
place. On the other hand, CAD/BIM model was developed to hold all
necessary geometric and semantic information of the rammed earth
element (e.g., material, dimension, required qualities) as well as infor-
mation about the production process. A full description of a similar kind
of data model can be found in Kosse et al. (2022).
6.4. Quality management and quality control
As these rammed earth elements were used as load bearing elements
in this project, a rigorous quality management system and quality con-
trol had to be implemented. During production, multiple specimens of
rammed earth cubes with dimensions of 20 cm ×20 cm ×20 cm were
produced, dried and taken on a daily basis for lab-testing of multiple
properties, including load bearing capabilities. The raw material was
inspected twice per day and all results were documented. Engineers
supervised the production process regularly. Experiments showed a
strong correlation between compaction (roughly equal to density) and
load bearing capabilities. The compaction of each element was visually
checked and the required weight of each element was directly calculated
from CAD data. Each element was tagged with a serial number and the
weight was measured before storage. If any insufcient compaction was
discovered, the element was directly recycled and rebuilt. Although
rammed earth is based on natural ingredients that can vary in properties
and quality according to sourcing conditions, it is a very reproducible
product that meets given quality standards. An automated production
benets greatly from stable and consistent processes.
6.5. Safety
The robotic arm and the pneumatic rammer can pose signicant risk
to humans. Thus, strict safety measures and standards were carefully
considered in the early stages of designing the production line to ensure
compliance with the essential health and safety requirements. The nal
safety concept includes three independent zones in which the robot and
rammer operate within safe distances from the workers. Access to the
zones is restricted by safety fences. The safe working boundaries of the
robot and the entrance to each zone are monitored by several safety
sensors and light curtains respectively. Light curtains were chosen to
facilitate the process of formwork assembly and disassembly. Using a
separate light curtain per zone allows safe access to the other zones
where the robot is not operating, which aids in achieving a sequential
collaboration between the different zones. The transfer of the robot from
one zone to another is always supervised by authorised personnel to
ensure that other zones (i.e., source and target zone) are free of any
person. It is worth mentioning that following these safety measures
helps maximising the productivity, where the robotic ramming system
can operate autonomously 12 h per day and ve days per week with
minimal interruptions in a real production.
Fig. 18. Robotic working envelope with 3 zones (left); produced rammed earth elements (right).
M. Gomaa et al.

Journal of Cleaner Production 398 (2023) 136569
13
6.6. Remarks and recommendation
For a three-storey ofce building in Switzerland, digitally assisted,
robotically fabricated rammed earth was used as load-bearing interior
walls to replace concrete in several zones and to create a unique archi-
tecture. The production was done at an industrial scale (350+wall el-
ements) and could be scaled to any required output. The robotic
ramming system by ERNE is undergoing continued development to offer
more versatility through handling different layer heights, material
compositions (e.g., for colouring) and shapes within the layer in the
future. Furthermore, a more intricate formwork system could be paired
with the existing robotic ramming system. Currently, there are ongoing
testing for arch-shaped elements that will be suitable as ceiling. Pilot
testing was conducted successfully, and the details will be released in the
future.
Nevertheless, the demonstrated method still requires essential
development and adoption:
•Outreach and education of architects and their clients is urgently
necessary to demonstrate that rammed earth is now very compatible
with modern construction methods and can be implemented in large-
scale building projects. The expertise for planning with rammed
earth is not yet widespread among architects. Special care must be
taken to come up with a design appropriate to the material and its
specic mechanical and structural properties.
•The efciency of the process could be further increased by replacing
the mechanised hopper with an automated, robotically controlled
feeding mechanism.
•When improvements on the standard procedure are implemented
and tested, further developments could be incorporated as newly
developed framework systems, coloured layers, free-form layers and
so on, expanding the design space and design freedom for the ar-
chitect even more.
7. Discussion and concluding remarks
Research on modernising earth construction, including rammed
earth, has undergone a signicant development in the past decade.
Based on the precedent work, it has become evident that rethinking
rammed earth construction within a mechanised and/or automated
framework presents huge potential, where the new approaches stand out
as a more economic and environmentally friendly construction method
that is highly capable of competing with other modern counterparts (e.
g., 3D printed concrete). Furthermore, the incorporation of rammed
earth in the contemporary construction industry promises signicant
benets to the circular economy (Schweiker et al., 2021). However,
there are still several technological aspects that require substantial
development to help accelerate the development process, in terms of the
employed automation level, design freedom and productivity. This
study observed that most of the work to date exhibits an inversely
proportional relation between the automation level and the design
freedom, where increasing the automation/mechanising level usually
reects a simple, at wall, while using manual ramming enables more
intricate designs. Table 2 demonstrates an analysis for the correlation
between the three aforementioned factors, and whether the
manufacturing process is digitally assisted or not.
Generally, the analysis of the recorded projects in this study revealed
two target domains that require improvement. Identifying the chal-
lenges in each domain will aid in establishing clear roadmap for future
development. These domains are:
1) Ramming and material feeding process.
2) Formwork system.
7.1. Ramming and material feeding process
To date, there is still considerable dependence on humans to
participate in one or more stages of the construction process. At the same
time, the robotic systems that are used are not fully leveraged to their
maximum capabilities, where they are only used to perform 3-axis
movement during the ramming process, which is in a very similar
manner to conventional gantry systems. Fortunately, the teams in ERNE
and RMIT University are currently working on developing a robotic
ramming system that can perform compaction in multi-directional
manner, and thus delivering more intricate rammed earth geometries
(e.g., double-curved walls). The development of the robotic systems is
expected to happen collaboratively with the development of advanced
formwork systems.
7.2. Form-work system
Despite the noticeable progress in the developing the ramming pro-
cess for rammed earth, the formwork system, which is an integral part of
rammed earth construction, still requires signicant development. Most
of the existing formwork systems to date are made either from timber or
metal, which are designed to produce basic rectangular-shaped walls.
Leveraging robotic 3D printing or milling/subtraction can expand
widely the possibilities for creating more innovative and intricate
formwork, which was evident in the projects at RMIT University and
Akipek and Yazar (2017). Monolithic or modular formwork approaches
can provide tailored solution as desired by the designer. The team at
RMIT University is currently investigating the feasibility of the two
approaches, and the ndings will be presented in the future.
Funding
This work was funded by the Australian Research Council
(FL190100014).
Declaration of competing interest
The authors declare that they have no known competing nancial
interests or personal relationships that could have appeared to inuence
the work reported in this paper.
Data availability
Data will be made available on request.
Acknowledgments
We would like to thank Ahmed Abdelaal, Jiaming Ma and Jeff Lee
from the Centre for Innovative Structures and Materials in RMIT
Table 2
Demonstration of the correlation between the different technical aspects of the
recorded projects.
Recorded project Automation
level
Design
freedom
Productivity Digitally
assisted
Lehm Ton Erde
(2015)
High Low High No
Form Earth
(2022)
High Low Medium No
Hurtado and
Eloualid (2016)
Low Medium Medium Yes
Akipek and Yazar
(2017)
Medium Medium Medium Yes
Kloft et al. (2019) High Low Medium Yes
ERNE AG Holzbau
(2022)
High Medium High Yes
RMIT University Medium High Medium Yes
M. Gomaa et al.

Journal of Cleaner Production 398 (2023) 136569
14
University for their help and support during the manufacturing process
of the morphing column. Also, we would like to extend our gratitude to
Olnee Construction team in Melbourne for their guidance and technical
support.
References
Akipek, F.¨
O., Yazar, T., 2017. Common action walls. POT+Des. Res. Gr. URL 8.25.21. htt
p://potplus.org/common-action-walls/.
Alhumayani, H., Gomaa, M., Soebarto, V., Jabi, W., 2020. Environmental assessment of
large-scale 3D printing in construction: a comparative study between cob and
concrete. J. Clean. Prod. 270, 122463 https://doi.org/10.1016/j.
jclepro.2020.122463.
Bick, Z.R., 2016. Mechanising Rammed Earth. Pennsylvania State Univ. URL. https
://etda.libraries.psu.edu/catalog/29163.
Chang, I., Im, J., Prasidhi, A.K., Cho, G.C., 2015. Effects of Xanthan gum biopolymer on
soil strengthening. Construct. Build. Mater. 74, 65–72. https://doi.org/10.1016/j.
conbuildmat.2014.10.026.
Form Earth, 2022. Free form rammed earth. URL 10.26.22. https://www.youtube.com/
c/Formearth.
Fatehi, H., Ong, D.E.L., Yu, J., Chang, I., 2021. Biopolymers as green binders for soil
improvement in geotechnical applications: a review. Geosci. 11, 1–39. https://doi.
org/10.3390/geosciences11070291.
Feng, P., Meng, X., Chen, J.-F., Ye, L., 2015. Mechanical properties of structures 3D
printed with cementitious powders. Construct. Build. Mater. 93, 486–497. https://
doi.org/10.1016/j.conbuildmat.2015.05.132.
Geneidy, O., Ismaeel, W.S.E., Abbas, A., 2019. A critical review for applying three-
dimensional concrete wall printing technology in Egypt. Architect. Sci. Rev. 1–15.
https://doi.org/10.1080/00038628.2019.1596066, 0.
Gomaa, M., Jabi, W., Veliz Reyes, A., Soebarto, V., 2021a. 3D printing system for earth-
based construction: case study of cob walls. Autom. ConStruct. 124, 103577 https://
doi.org/10.1016/j.autcon.2021.103577.
Gomaa, M., Vaculik, J., Soebarto, V., Grifth, M., Jabi, W., 2021b. Feasibility of 3DP cob
walls under compression loads in low-rise construction. Construct. Build. Mater. 301,
124079 https://doi.org/10.1016/j.conbuildmat.2021.124079.
Gomaa, M., Jabi, W., Soebarto, V., Xie, Y.M., 2022. Digital manufacturing for earth
construction: a critical review. J. Clean. Prod. 338, 130630 https://doi.org/
10.1016/j.jclepro.2022.130630.
Gomes, M.I., Gonçalves, T.D., Faria, P., 2014. Unstabilized rammed earth:
characterization of material collected from old constructions in south Portugal and
comparison to normative requirements. Int. J. Architect. Herit. 8, 185–212. https://
doi.org/10.1080/15583058.2012.683133.
Hall, M.R., Lindsay, R., Krayenhoff, M., 2012. Modern Earth Buildings : Materials,
Engineering, Constructions and Applications, 2012th ed (Cambridge).
ERNE AG Holzbau, 2022. Robotic rammed earth. URL 10.26.22. https://www.linkedin.
com/feed/update/urn:li:activity:6990574421816045568/.
Howe, L.B., Heringer, A., Rauch, M., 2019. Upscaling Earth: Material, Process, Catalyst,
rst ed. gta publishers, Zurich.
Hurtado, E., Eloualid, I., 2016. Engineered cast earth. URL 10.19.22. https://doi.org/10
.13140/RG.2.2.23253.12008.
International Energy Agency, 2021. Global Energy Review : CO2 Emissions in 2021
(Paris).
Keable, J., 1996. Rammed Earth Structures: A Code of Practice. Intermediate Technology
Publications, London.
Kloft, H., Oechsler, J., Loccarini, F., Gosslar, J., Delille, C., 2019. Robotic fabrication of
components from rammed earth (translated from German). Dtsch. BauZeitschrift
54–59.
Kosse, S., Vogt, O., Wolf, M., K¨
onig, M., Gerhard, D., 2022. Digital twin framework for
enabling serial construction. Front. Built Environ. 8, 1–20. https://doi.org/10.3389/
fbuil.2022.864722.
Kozlovska, M., Klosova, D., Strukova, Z., 2021. Impact of industry 4.0 platform on the
formation of construction 4.0 concept: a literature review. Sustain. Times 13, 1–15.
https://doi.org/10.3390/su13052683.
Ma, J., Gomaa, M., Bao, D.W., Javan, A.R., Xie, Y.M., 2022. PrintNervi – Design and
Construction of a Ribbed Floor System in the Digital Era. in: Innovation,
Sustainability and Legacy: Proceedings of the IASS Annual Symposium and the 13th
Asian-Pacic Conference on Shell and Spatial Structures, Beijing.
n.d. Olnee Rammed Earth. Rammed earth houses: Olnee Constructions’ residential
gallery, URL 10.25.22. https://olneerammedearth.com.au/galleries/completed-pro
jects-residential/.
Parracha, J.L., Silva, A.S., Cotrim, M., Faria, P., 2020. Mineralogical and microstructural
characterisation of rammed earth and earthen mortars from 12th century Paderne
Castle. J. Cult. Herit. 42, 226–239. https://doi.org/10.1016/j.culher.2019.07.021.
PERI, 2022. DUO system formwork. URL 10.25.22. https://www.peri.com.au/products/f
ormwork/wall-formwork/duo-formwork.html.
Perrot, A., Rangeard, D., 2021. LINSEED OIL AND XANTHAN GUM : PROMISING
STABILISERS FOR EARTHEN BUILDING MATERIALS, pp. 1–8.
Perrot, A., Rangeard, D., Courteille, E., 2018. 3D printing of earth-based materials:
processing aspects. Construct. Build. Mater. 172, 670–676. https://doi.org/10.1016/
j.conbuildmat.2018.04.017.
Quagliarini, E., Stazi, A., Pasqualini, E., Fratalocchi, E., 2010. Cob construction in Italy:
some lessons from the past. Sustainability 2, 3291–3308. https://doi.org/10.3390/
su2103291.
Rammed Earth Consulting, 2022. Traditional rammed earth construction. URL 10.25.22.
http://rammedearthconsulting.com/?fbclid=IwAR1yd_AkZ58Ttdu_xbKk
XSKajyFH2q1VJGuLuO3P2g1EbgyTnTzwBv6r5CE.
Rauch, M., 2020. Unstabilised Rammed Earth – 100% EARTH. Why Is it So Hard to Trust
in 100%? in: LEHM 2020- 8th International Conference on Building with Earth,
Weimar, pp. 1–8.
Sauer, M., Kapnger, O., 2015. Martin Rauch: Rened Earth Construction and Design
with Rammed Earth. DETAIL – Institut für internationale Architektur-
Dokumentation GmbH & Co., Munich.
Schweiker, M., Endres, E., Gosslar, J., Hack, N., Hildebrand, L., Creutz, M., Klinge, A.,
Kloft, H., Knaack, U., Mehnert, J., Roswag-Klinge, E., 2021. Ten questions
concerning the potential of digital production and new technologies for
contemporary earthen constructions. Build. Environ., 108240 https://doi.org/
10.1016/j.buildenv.2021.108240.
Silva, A., Oliveira, I., Silva, V., Mir˜
ao, J., Faria, P., 2022. Vernacular Caramel’s adobe
masonry dwellings–material characterization. Int. J. Architect. Herit. 16, 67–84.
https://doi.org/10.1080/15583058.2020.1751343.
Veliz Reyes, A., Jabi, W., Gomaa, M., Chatzivasileiadi, A., Ahmad, L., Wardhana, N.M.,
2019. Negotiated matter: a robotic exploration of craft-driven innovation. Architect.
Sci. Rev. 62, 1–11. https://doi.org/10.1080/00038628.2019.1651688.
Walker, P., 2005. Rammed Earth : Design and Construction Guidline, rst ed. IHS BRE
Press, London.
Weismann, A., Bryce, K., 2006. Building with Cob: a Step-by-step Guide. Green Books ltd,
Devon.
Weißenberger, M., Jensch, W., Lang, W., 2014. The convergence of life cycle assessment
and nearly zero-energy buildings: the case of Germany. Energy Build. 76, 551–557.
https://doi.org/10.1016/j.enbuild.2014.03.028.
Zareiyan, B., Khoshnevis, B., 2017. Interlayer adhesion and strength of structures in
contour crafting - effects of aggregate size, extrusion rate, and layer thickness.
Autom. ConStruct. 81, 112–121. https://doi.org/10.1016/j.autcon.2017.06.013.
M. Gomaa et al.