energies

Review
Artificial Intelligence in the Construction Industry: A Systematic
Review of the Entire Construction Value Chain Lifecycle
Christian Nnaemeka Egwim 1, * , Hafiz Alaka 1, *, Eren Demir 2 , Habeeb Balogun 1 , Razak Olu-Ajayi 1 ,
Ismail Sulaimon 1 , Godoyon Wusu 1 , Wasiu Yusuf 1 and Adegoke A. Muideen 1

1 Big Data Technologies and Innovation Laboratory, University of Hertfordshire, Hatfield AL10 9AB, UK
2 Decision Sciences Business Analysis and Statistics Group, Hertfordshire Business School, University of
Hertfordshire, Hatfield AL10 9AB, UK; e.demir@herts.ac.uk
* Correspondence: c.egwim@herts.ac.uk (C.N.E.); h.alaka@herts.ac.uk (H.A.);
Tel.: +44-01707-285426 (ext. 5426) (H.A.)

Abstract: In recent years, there has been a surge in the global digitization of corporate processes and con-
cepts such as digital technology development which is growing at such a quick pace that the construction
industry is struggling to catch up with latest developments. A formidable digital technology, artificial
intelligence (AI), is recognized as an essential element within the paradigm of digital transformation,
having been widely adopted across different industries. Also, AI is anticipated to open a slew of new
possibilities for how construction projects are designed and built. To obtain a better knowledge of the
trend and trajectory of research concerning AI technology application in the construction industry, this
research presents an exhaustive systematic review of seventy articles toward AI applicability to the entire
lifecycle of the construction value chain identified via the guidelines outlined by the Preferred Reporting
Items for Systematic Reviews and Meta-Analyses (PRISMA). The review’s findings show foremostly that
AI technologies are mostly used in facility management, creating a huge opportunity for the industry to
profit by allowing facility managers to take proactive action. Secondly, it shows the potential for design
expansion as a key benefit according to most of the selected literature. Finally, it found data augmentation
as one of the quickest prospects for technical improvement. This knowledge will assist construction
Citation: Egwim, C.N.; Alaka, H.; companies across the world in recognizing the efficiency and productivity advantages that AI technologies
Demir, E.; Balogun, H.; Olu-Ajayi, R.; can provide while helping them make smarter technology investment decisions.
Sulaimon, I.; Wusu, G.; Yusuf, W.;
Muideen, A.A. Artificial Intelligence Keywords: artificial intelligence; AI technologies; AI adoption; AI benefits; AI challenges; construction
in the Construction Industry: A industry
Systematic Review of the Entire
Construction Value Chain Lifecycle.
Energies 2024, 17, 182. https://
doi.org/10.3390/en17010182 1. Introduction
Academic Editor: Chi-Ming Lai On a regional, national, and global scale, construction is considered a large sector
with strategic importance [1]. It is also an industry that has been plagued by a slew
Received: 26 November 2023
of issues for decades, including low production, slim profit margins, waste, and safety
Revised: 22 December 2023
Accepted: 27 December 2023
concerns. Its projects are extremely complex, and the danger of inefficiency and risk, which
Published: 28 December 2023
eventually contribute to project costs and delays, grows geometrically with the project’s
scale [2,3]. In the past, to mitigate these issues, the construction industry traditionally
concentrated on generating operational benefits by employing technology to streamline
processes and procedures, but the data gathered as a result of this digitization trial is often
Copyright: © 2023 by the authors. overlooked [4,5]. Surprisingly, this industry is still on the edge of digitization, which is said
Licensee MDPI, Basel, Switzerland. to disrupt existing traditional procedures while also opening up a slew of new prospects [6].
This article is an open access article In recent years, there has been a surge in the global digitization of corporate processes
distributed under the terms and and paradigms, including industry 4.0, and digital twins and digital technology development
conditions of the Creative Commons
is growing at such a quick pace that the construction industry is struggling to catch up with
Attribution (CC BY) license (https://
the latest developments. A formidable digital technology, artificial intelligence (AI), is now a
creativecommons.org/licenses/by/
vital component of the digital shift (partly due to big data revolution), having gained broad
4.0/).

Energies 2024, 17, 182. https://doi.org/10.3390/en17010182 https://www.mdpi.com/journal/energies

Energies 2024, 17, 182 2 of 21

acceptance in diverse sectors, including healthcare, where it assists in patient diagnosis through
genetic data [7,8]; manufacturing, where it is utilized for workforce management, production
process optimization, and predictive maintenance [9]; education, where it facilitates virtual
lectures [10,11]; finance, particularly in fraud detection [12,13], and transportation, exemplified
by the development of self-driving autonomous cars [14,15], among many others.
The definition of AI has evolved throughout time, but its foundation has always been the
goal of creating machines that are capable of thinking like humans. It does this by trying to
imitate human intelligence through hardware and software solutions [16]. With more data being
generated every second, AI technologies encompassing robotics, machine learning, natural
language processing, speech recognition, expert systems, and computer vision, among others,
have aided the scientific community in harnessing the growth of big data [17]. On these massive
datasets, scientists can extract information that human eyes cannot interpret quickly enough
using AI.
As a result, it is clear that AI can help the construction industry improve decision-making,
drive project success, and deliver projects on time and on budget by proactively unlocking new
predictive insights from its ever-growing volume of project data, which was previously only
archived for future reference. For instance, data collected from smart devices, sensors within the
Internet of Things (IoT), Building Information Modeling (BIM), and various other data sources
can be analyzed by AI technologies to find patterns in the performance and usage of current
infrastructure assets and determine what sort of infrastructure is needed in the future and how
it should be supplied [5].
Furthermore, the number of incremental stages necessary to bring infrastructure designs
to operational status will most likely be reduced by AI. This will save time and money in
the manufacturing of construction materials, alongside the development and maintenance
of our infrastructure networks. In this regard, a vast body of international literature have
investigated the use of AI technologies to tackle concerns related to construction projects. For
example, machine learning has been applied to mitigate construction project delay risks [1,18,19],
occupational health and safety concerns within the construction sector [20–22], and construction
and demolition waste generation [23–25].
It is the viewpoint of Gamba, Balaguer, and Chu [26–28] that robotics can be used to
automate the assembly of building elements (e.g., masonry walls, steel structures etc.). Also,
Bruckmann and Wu [29–31] made an important point by adding that a robotic system, com-
prising a gripper connected to a frame via cables, demonstrates applicability in the domain of
bricklaying. Furthermore, an expert system has been devised for the computation of fault rates
in incidents related to falls in the construction domain [32–34] and natural language processing
has been applied to extract and exchange information, as well as a variety of downstream
applications to aid management and decision-making in smart construction projects [17,35,36].
More recently, some studies have conducted traditional narrative critical/literature reviews
for a specific AI technology in the construction industry (e.g., computer vision by Xu [37],
natural language processing by Wu [17], robotic system by Davila [38], etc.), while a few other
studies have conducted traditional narrative critical/literature reviews for adopting generic AI
technologies in the construction industry with a specific goal (e.g., Parveen [39] focused on AI’s
legal issues and regulatory challenges, Schia [40] focused on AI’s impact on human behavior,
and Abioye [41] focused on the current state, prospects, and forthcoming challenges in the field
of artificial intelligence).
However, to the best of the authors’ knowledge, there has been no exhaustive examination
of the application of AI technologies in the construction industry. Thus, this study is motivated
by the absence of a systematic review in this domain. In conducting a systematic review,
independent researcher(s) design a system, based on specific guidelines (protocol), and the
system then makes the decisions to determine the outcome of the research thus producing a
research outcome that is explicit, reproducible, and without prior assumptions [42]. Meanwhile,
in a typical traditional narrative literature review, the identification, selection, inclusion, and
extraction of research articles solely (all) depends on the judgement of the author(s) in order to
Energies 2024, 17, 182 3 of 21

support their model, hypothesis, and to identify the research gaps. This poses a great concern of
subjectivity, repeatability, and reproducibility of results from such research [43].
Secondly, and as a final rationale of this research, no study, to the best of our knowledge,
has conducted any kind of AI review toward its applicability to the entire lifecycle of the
construction value chain. In addressing this void, the aim of this study is to provide a
thorough systematic review of artificial intelligence and its implementation throughout the
entire construction value chain lifecycle—from building material manufacturing to design,
planning, and construction, as well as facilities management. The systematic review is
structured around the ensuing research questions:
1. What AI technologies have been documented in the literature so far?
2. What are the different stages of the construction project lifecycle wherein these AI
technologies are applied?
3. What potential benefits arise from the implementation of the identified AI technolo-
gies, and what are the existing challenges and deficiencies in their industry adoption?
This research makes a significant contribution to the body of knowledge by addressing the
knowledge gap in the field of AI in the construction industry, specifically by addressing several
imaginable application cases for AI across diverse phases of the construction project lifecycle
and the potential benefits of implementing AI technologies, as well as the present roadblocks
and gaps in their industrial adoption. This will immensely contribute to the advancement of the
Architecture, Engineering, and Construction (AEC) industry and the holistic built environment
ecosystem in identifying opportunities for technological advancement.

2. Methodology
This study adopts a philosophical paradigm rooted in pragmatism, emphasizing practical
applied research that employs multiple perspectives for data interpretation. Depending on the
research question, the study considers both observable occurrences and subjective meanings
as viable sources of knowledge [44]. This study employed a systematic review methodology.
A systematic review, distinct from a conventional literature review, utilizes a precise, compre-
hensive, replicable, and auditable methodology to evaluate and interpret all pertinent research
associated with a particular research question, subject, or domain [42]. Moreover, by scrutinizing
the holistic perspective and amalgamating discrete elements to synthesize results in a structured
manner, a systematic review can address the limitations inherent in a traditional narrative
literature review, which is commonly used in the vast body of literature.
To develop its systematic review protocol, the Preferred Reporting Items for Systematic
Reviews and Meta-Analyses (PRISMA) guidelines were applied. PRISMA serves as a protocol
for executing systematic reviews and meta-analyses, comprising a twenty-seven-item checklist
and a four-phase flow diagram. Conceived by a consortium of twenty-nine scholars in the
medical domain, PRISMA aims to enhance the lucidity and uniformity of literature reviews.
Consequently, our examination’s focal theme, exploration strategy, criteria for inclusion and ex-
clusion, eligibility standards, data extraction, and synthesis methodologies were all delineated in
adherence to this protocol which was chosen because of its comprehensiveness, wide acceptance,
and applicability in different fields of study, despite the fact that it was originally established
for the medical and health domain [43]. For a start, the review process was broken down
into four steps: article identification, article screening, critical assessment, and data extraction
and synthesis.
During the article identification process (step 1), a comprehensive exploration of the
literature was conducted to identify articles for this research. Specifically, articles available
in the Scopus electronic database up until 21 January 2022 were utilized as the principal
source of information for the search. This database was chosen over others like ScienceDirect
and Web of Science because it is the “largest abstract and citation database of peer-reviewed
literature” [45]. Furthermore, Scopus indexes practically the whole ScienceDirect database, and
Scopus offers a greater choice of journals than Web of Science, as well as quicker citation analysis
and coverage of more articles [46]. The abstract, title, and keywords of publications in this
database were searched using the following search terms: (“artificial intelligence” OR “machine
Energies 2024, 17, 182 4 of 21

learning” OR “deep learning” OR “reinforcement learning” OR “automation” OR “robotics”

OR “expert system” OR “natural language processing” OR “optimization” AND (“construction
industry” OR “building industry” OR “built environment” OR “Architecture Construction
and Engineering”)) with no date, language, or article type restrictions. The search terms were
divided into two major parts separated by the “AND” operator, namely, AI technologies and
the construction industry. The search terms also contained several interesting synonyms, word
variations, and exact phrase searching symbols, such as the usage of double quotation marks in
“machine learning”and “building industry”, among many others.
At the article screening process (step 2), the abstracts of 2716 articles were reviewed to see
whether they were related to the research questions and to make sure there were no duplicates.
As such, this led to the removal of 2306 items, leaving 410. At the critical assessment (step 3),
three distinct inclusion and exclusion criteria were implemented. Articles were deemed eligible
if their primary emphasis centered on the utilization of artificial intelligence technologies in
construction projects and excluded otherwise. Additionally, only articles presenting original
research data were included, while review articles were excluded. Finally, each article’s
relevance was determined using a previously developed rating scale by [42,43,47]. The scale
was adapted based on practical results of how artificial intelligence technology is used in
a construction project, with “1” indicating low relevance, “2” medium relevance, and “3”
indicating high relevance.
Consequently, the full text of all articles having information relating to genuine case studies
of AI technology implementation in construction projects or AI technology application proven
in a laboratory environment were extracted, exported to a file in Comma-Separated Values
(CSV) format, given a “3” rating, and included in the evaluation during the data extraction and
synthesis process (step 4). Finally, a comprehensive examination of the selected articles was
conducted for the purpose of data extraction, encompassing elements such as research aim,
project type, country/region, research method(s), and AI technology, among others. Figure 1
elucidates this process, presenting a flow diagram delineating the research article selection
procedure. The diagram outlines the total count of articles identified through the database
search, articles screened in accordance with eligibility criteria,
Energies 2024, 17, x FOR PEER REVIEW
articles fully accessed, and the
5 of 25
eventual count of articles utilized for analysis in this study.

Figure 1. Flow diagram of the systematic review process.

Figure 1. Flow diagram of the systematic review process.
3. Results
3.1. Summary of Selected Articles
A summary of the selected articles organized by their respective publication source
is provided. It was discovered that the journals Automation in Construction and Journal Con-
struction Engineering and Management, as well as conference proceedings from the Confer-
ence on Computer-Aided Architectural Design Research, CAADRIA, possessed the preeminent
quantity of articles, constituting 20% of the overall selected papers. In general, 75.71% of
 Energies 2024, 17, 182 5 of 21

3. Results
3.1. Summary of Selected Articles
A summary of the selected articles organized by their respective publication source
is provided. It was discovered that the journals Automation in Construction and Journal
Construction Engineering and Management, as well as conference proceedings from the Confer-
ence on Computer-Aided Architectural Design Research, CAADRIA, possessed the preeminent
quantity of articles, constituting 20% of the overall selected papers. In general, 75.71% of
Energies 2024, 17, x FOR PEER REVIEW the total papers (53 out of 70) were disseminated through peer-reviewed scholarly journals, 6 of 25
whereas 16 articles (22.86%) were presented at conferences, and just 1 paper was part of a
book series.
For two decades between 1993 and 2013, there were eight articles, each with a different
yearUpon scrutinizing
of publication (seethe geographic
Figure 2). Theredispersion of constant
is a notable academicincrease
papersinrelated to artificial
the quantity
of research publications throughout the AI in construction research community.
intelligence research applied in the construction/execution lifecycle of the construction More
specifically, between 2017 and 2021, there was a constant increase in the number
value chain, it was discovered that researchers in the Republic of Korea and China emerge of research
as publications published in the research community, with a total of 56 articles, indicating
primary contributors to scholarly inquiry within this domain (4 articles each), followed
a rising interest in the application of artificial intelligence technology to the construction
by researchers in GBR and Australia (3 articles each). Interestingly, researchers in China
industry. Figure 3 shows the number of publications according to the first author’s insti-
tend to have
tute’s devoted
location. the most attention
Predominantly, researcherstofrom
AI applications in the supply/facility
China (12 articles), the United Statesmanage-
of
ment lifecycle of the construction value chain, with seven publications
America (USA) (8 papers), Republic of Korea (6 papers), Italy, the United Kingdom dedicated to it, and
of Great
fiveBritain
articles dedicated
(GBR), to the (4
and Australia planning/design lifecycle
papers each) published of the
most construction
of the value
papers relevant chain as
to the
well.
research topic.

Figure 2. 2.
Figure Sequential
Sequentialdistribution of articles
distribution of articles(total
(totalnumber
number of articles
of articles is 70).
is 70).
 Energies 2024, 17, x FOR PEER REVIEW 7 of 25
Energies 2024, 17, 182 6 of 21

Geographical
Figure 3. Figure distribution
3. Geographical of articlesofbased
distribution on based
articles projecton
lifecycle.
project lifecycle.

Upon
3.2.scrutinizing the geographic
Types of AI Technologies dispersion of academic papers related to artificial
and Categorization
intelligence research applied in the construction/execution lifecycle of the construction
Table 1 shows the different types of AI technology and their distribution across dif-
value chain, it was discovered that researchers in the Republic of Korea and China emerge as
ferent application areas. In general, the seventy reviewed studies referenced seven AI
primary contributors to scholarly inquiry within this domain (4 articles each), followed by
technologies for use in the construction industry, with supervised learning, deep learning,
researchers in GBR and Australia (3 articles each). Interestingly, researchers in China tend
knowledge-based systems, robotics, and natural language processing being the most often
to have devoted the most attention to AI applications in the supply/facility management
mentioned. On the other hand, AI technologies such as optimization and reinforcement
lifecycle of the construction value chain, with seven publications dedicated to it, and five
learning garnered
articles dedicated less attention. In
to the planning/design terms of
lifecycle of the
theconstruction
distribution value
of AI chain
technologies
as well. to their
application areas in health and safety management, supervised learning was the most re-
searched
3.2. Types (4 articles),
of AI Technologies followed
and by deep learning and knowledge-based systems (3 articles
Categorization
each). Deep learning and supervised
Table 1 shows the different types of AI technology learning haveand shown to be effective
their distribution time
across dif-and cost
management
ferent application technologies
areas. In general,(2 articles
the seventyeach).reviewed
Also, robotics was
studies the most seven
referenced often mentioned
AI
technology for prefabrication (2 articles). Furthermore, the most promising
technologies for use in the construction industry, with supervised learning, deep learning, AI technolo-
gies for heating,
knowledge-based systems, ventilation, andnatural
robotics, and air conditioning (HVAC) optimum
language processing being thecontrol were identi-
most often
mentioned. On the other hand, AI technologies such as optimization and reinforcement papers
fied to be optimization and deep learning technologies (2 articles each), while most
learninghighlighted
garnered less theattention.
applicationInof natural
terms language
of the processing
distribution of AIin relation to sustainable
technologies to their con-
creteareas
application and regenerative
in health andsustainability (1 article).supervised learning was the most
safety management,
researched (4 articles), followed by deep learning and knowledge-based systems (3 articles
Energies 2024, 17, 182 7 of 21

each). Deep learning and supervised learning have shown to be effective time and cost
management technologies (2 articles each). Also, robotics was the most often mentioned
technology for prefabrication (2 articles). Furthermore, the most promising AI technologies
for heating, ventilation, and air conditioning (HVAC) optimum control were identified to
be optimization and deep learning technologies (2 articles each), while most papers high-
lighted the application of natural language processing in relation to sustainable concrete
and regenerative sustainability (1 article).

Table 1. Types of AI technology and their area of application.

AI Technology Description Subtype Application Area Reference

Support vector machine (6),
artificial neural network (6), Health and safety management (4), time and
bckpropagation (3), decision cost management (2), building structures (2),
tree (5), random forest (4), structural reliability (1), sustainable concrete
k-nearest neighbors (3), (1), demolition waste management (1),
A type of machine learning in
gradient boost machine (2), constructability analysis (1), construction
which a computer algorithm is
Supervised Learning (SL) adaptive boosting (2), naïve monitoring (1), construction equipment (1), [19,24,25,48–70]
trained on labeled input data
Bayes (2), extreme gradient occupant behavior (1), site layout (1),
for a certain output.
boosting (2), logistic cementitious composite (1), energy savings
regression (2), ensemble and demand response (1), project selection
method (2), light gradient (1), construction
boosting machine (1), extra negotiation and conflict resolution (1)
trees (1)
Health and safety management (3), time and
cost management (2), HVAC optimal control
A type of machine learning Convolutional neural network (1), construction monitoring (1), intelligent
that trains computers to (6), deep neural network (4), building design (1), 3D datasets (1), building
Deep Learning [66,71–88]
accomplish things that autoencoder (1), long information modeling (1), monument
humans do instinctively. short-term memory (1) recognition (1), surface defect detection (1),
building recognition (1), filing architectural
drawings (1) parametric design (1)
Health and safety management (3), building
automation (1), productivity estimation (1),
A computerized system
site layout (1), building diagnosis and
designed to capture and Expert system (5), Case-based
repairs (1), performance evaluation (1),
Knowledge-Based System imitate human intellect in reasoning (2), Fuzzy [51,57,89–97]
construction
symbolic form, often through logic (1)
negotiation and conflict resolution (1),
a series of if–then rules.
occupant behavior (1)
architectural innovation (1)
Additive
Robotic prefabrication (2), digital fabrication
A technology that deals with manufacturing (4), robotic
(1), collaborative
the creation, design, beam assembly (1), soft
Robotics robotics (1), block assembly (1), intelligent [26,71,98–104]
construction, and operation of robotics (1), robotic
hoisting (1), health and safety management
programmable machines. bricklaying (1), mobile
(1), environmental impact analysis (1)
robot (1)
An artificial intelligence
technology that utilizes Text clustering (1), word
Health and safety management (2), HVAC
computers to comprehend, segmentation (1), information
Natural Language Processing optimal control (1), sustainable concrete (1), [56,105–109]
generate, and analyze human retrieval and extraction (1),
regenerative sustainability (1)
languages known as natural text analysis (1)
language processing.
Genetic algorithm (2), gray
A technique that seeks to alter
wolf optimization algorithm
an existing process to enhance HVAC optimal control (2), health and safety
(1), genetic algorithm,
Optimization the occurrence of good results management (2), sustainable concrete (1), [60,68,85,110]
stochastic
and decrease the occurrence of building structures (1)
gradient descent (1), genetic
bad outcomes.
programming (1)
A form of machine learning
that enables an agent to
Energy savings and demand response (1),
acquire knowledge through
Reinforcement Learning Q-learning (4) HVAC optimal control (1), health and safety [110–113]
iterative experimentation and
management (1), look-ahead schedule (1)
experience while receiving
feedback from its actions.

Quite notably (see Table 1), the articles included in this study highlighted the emer-
gence of diverse subtypes within each AI technology. For instance, the most cited subtypes
of supervised learning AI technology were support vector machine and artificial neural
network (6 articles each), followed by the convolutional neural network subtype of deep
learning (6 articles), and the expert system subtype of knowledge-based systems (5 articles).
As Table 1 shows, most of the papers included cited adaptive manufacturing (4 articles)
Energies 2024, 17, 182 8 of 21

and Q-learning (4 articles) as the most widely used subtype of robotics and reinforcement
learning AI technology, respectively. In addition, genetic algorithm appeared to be the most
favorable subtype of optimization technique for the researchers (2 articles).
Energies 2024, 17, x FOR PEER REVIEW
3.3. Construction Project Types and Their Lifecycle Application Area 10 of 25

In terms of the types of construction projects in which AI technologies were used,

Figure 4 suggests that the majority of the scholarly articles included (58.60%) were related to
included
built papers on
environment commercial
and residentialbuildings
building (28(6 articles)
articles and bridge/roadway
in built environment projects (5 arti-
and 13 articles
cles)
in focused building).
residential on the three stages of that
Following the construction
were paperslifecycle
regardingin high-rise
near equaland
measure. Inter-
commercial
estingly, articles
buildings, pertaining
which made to power
for 18.60% of theplant projects
articles (2 (7
chosen articles)
articlesand retrofit building
in high-rise building(1and
ar-
6ticle)
articles in commercial
exhibit building).solely
a distinct emphasis Bridge/highway project and office building
on the construction/execution project
stage within thewere
con-
discovered
struction valuein 7.10%
chainand 4.30% Conversely,
lifecycle. of the total of 70 articles,
water treatmentrespectively. Moreover, power
projects predominantly con-
plant, timber construction, and architectural heritage projects all had the same
centrated on the planning/design stage of the construction value chain lifecycle. Addition- number of
articles (2 each),
ally, projects with onlytimber
involving one article (1.40%) proposing
construction the use ofheritage
and architectural AI technology
sharedinanretrofit
equal
building
number and water across
of articles treatment
theirplants.
respective stages in the construction value chain lifecycle.

Figure4.4.Distribution
Figure Distributionof
ofarticles
articlesbased
basedon
onproject
projecttype.
type.

Furthermore, Figure 5 depicts the distribution of articles by construction project type

and lifecycle stage. The majority of articles on built environment were focused on the
construction/execution stage (15 articles). However, there were only a few articles that
focused on the planning/design stage of the built environment (3 article). Likewise, the
planning/design and supply/facility management stages of the construction lifecycle for
residential buildings received the greatest attention (6 articles each), with a smaller number
of articles devoted to the construction/execution stage (1 article). Papers related to high-rise
buildings had a focus toward the construction/execution stage (3 articles). All the included
papers on commercial buildings (6 articles) and bridge/roadway projects (5 articles) focused
on the three stages of the construction lifecycle in near equal measure. Interestingly, articles
pertaining to power plant projects (2 articles) and retrofit building (1 article) exhibit a
Energies 2024, 17, 182 9 of 21

distinct emphasis solely on the construction/execution stage within the construction value
chain lifecycle. Conversely, water treatment projects predominantly concentrated on the
Energies 2024, 17, x FOR PEER REVIEW 11 of 25
planning/design stage of the construction value chain lifecycle. Additionally, projects
involving timber construction and architectural heritage shared an equal number of articles
across their respective stages in the construction value chain lifecycle.

Figure
Figure 5.
5. Distribution
Distribution of
of articles
articles based
based on
on project
project type
type and
and lifecycle.
lifecycle.

3.4. Benefits,
3.4. Benefits, Challenges,
Challenges, and
and Opportunities
Opportunities for
for Technological
Technological Advancement
Advancement
3.4.1. Benefits
3.4.1. Benefits
This subsection comprehensively addresses the manifold benefits derived from the
This subsection comprehensively addresses the manifold benefits derived from the
escalating prominence of artificial intelligence technologies in the construction industry.
escalating prominence of artificial intelligence technologies in the construction industry.
The principal benefits attributed to the integration of AI technology within the construction
The principal benefits attributed to the integration of AI technology within the construc-
sector, as delineated in twenty-six out of the seventy curated articles, are succinctly outlined.
tion sector, as delineated in twenty-six out of the seventy curated articles, are succinctly
outlined.
Potential for Design Expansion
Intelligent room layouts for better natural ventilation are one example of how AI
Potential for Design Expansion
technologies can lead to novel design aspects. As mentioned by Sonetti [107], who de-
Intelligent
veloped room for
AI solutions layouts for better natural
human-centered ventilation
regenerative design,areAIone example ofare
technologies how AI
strong
technologies
supporters ofcan lead to novel
human-centric design aspects.
regenerative designAswhen
mentioned
it comes bytoSonetti [107],technologies
developing who devel-
oped AI solutions
that improve for human-centered
interactions between buildingsregenerative
and theirdesign, AI technologies
occupants. are strong
It is the viewpoint of
supporters of human-centric regenerative design when it comes to
Lamio [82] that the application of AI technologies to automate building design processes developing technolo-
gies that improve
demonstrates thatinteractions
an image takenbetweenfrombuildings
a virtualand theircan
model occupants.
accurately It isdistinguish
the viewpoint the
of Lamio type.
building [82] that
Theythe application
developed an of
AIAItooltechnologies
using classicalto automate
and modern building
machine design pro-
learning
cesses demonstrates
techniques that an
to categorize image
images oftaken fromdesigns
building a virtual
intomodel
threecan accurately
classes. This isdistinguish
especially
the buildingconsidering
important type. They developed an AI tool
the large number of using classical and
BIM structures withmodern
missing machine learning
information or
techniques to categorize images of building designs into three classes. This is especially
incorrect labeling.
important considering the large number of BIM structures with missing information or
Possibility
incorrect for Big Data Analytics
labeling.
AI technologies are exposed to an unending quantity of data to learn from and im-
Possibility for Bigday
prove on every DataatAnalytics
a time when vast amounts of data are being produced in the
industry
AI technologies are exposedthe
[114,115]. For instance, research
to an unending led by Palmaof[81]
quantity dataexplored
to learnthe integration
from and im-
prove on every day at a time when vast amounts of data are being produced in theinto
of convolutional neural networks (CNNs), a subset of deep learning methods, the
indus-
realm of architectural heritage by developing a mobile app aimed at monument
try [114,115]. For instance, the research led by Palma [81] explored the integration of con- recognition,
pioneeringneural
volutional the usenetworks
of AI in this domain.
(CNNs), Palma’s
a subset [81] argument
of deep is compelling,
learning methods, especially
into the realm of in
terms of his pointing
architectural heritage out that the output
by developing of their
a mobile adoption
app aimed at of monument
AI technology resulted inpio-
recognition, the
neering the use of AI in this domain. Palma’s [81] argument is compelling, especially in
terms of his pointing out that the output of their adoption of AI technology resulted in the
creation of open datasets for testing and evaluating AI applications in the field of archi-
tecture and architectural heritage. In addition, Keshavarzi [86], who developed a genera-
tive system that addresses the challenge of limited 3D datasets for deep learning
Energies 2024, 17, 182 10 of 21

creation of open datasets for testing and evaluating AI applications in the field of architec-
ture and architectural heritage. In addition, Keshavarzi [86], who developed a generative
system that addresses the challenge of limited 3D datasets for deep learning methodologies
in the built environment stated that their AI technology has the potential to facilitate data
augmentation of parametric 3D scan datasets by taking an extant 3D scan as input and
generating alternative iterations of the architectural configuration, encompassing walls,
doors, and furnishings, accompanied by corresponding textures. This process extends the
prevailing 3D geometry datasets, which are conventionally constrained in their scope.

Workplace Health and Safety

AI technology can provide a project with precise job site safety best practices based on
learned knowledge. As one of the most hazardous industries to work for, this surveillance
keeps people safe and accidents to a limit. In the pursuit of minimizing accident occurrences
within construction sites, Zhang [56] employed a diverse set of AI technologies. Specifically,
an ensemble model was devised, integrating text mining, natural language processing
(NLP), and machine learning methodologies for the comprehensive analysis of construction
accident records. The objective was to discern and extract salient elements associated with
accidents, ultimately mitigating potential hazards. With reference to Yu [73], AI technolo-
gies can be used as non-invasive tools for workload monitoring and thorough ergonomic
assessment for various construction tasks, such as assessing risk factors for work-related
musculoskeletal disorders by developing an AI tool that employs a smartphone camera
with advanced deep learning algorithms to extract construction workers’ skeleton data,
complemented by smart insoles to quantify plantar pressures during various construction
activities. More so, Su [76] adopted an AI technology to predict the smoke motion and the
available safe egress time in a typical atrium.

Increase in Productivity
Some AI technologies can complete repetitive tasks swiftly and precisely while being
fatigue-free. For instance, Li [103] detailed the creation of a vision-based intelligent mobile
robot hoisting system to improve the hoisting process, including the process of hooks
identifying the hoist points and autonomously releasing the components without the
need of on-site construction employees. According to García de Soto [100], by offering a
process for evaluating productivity based on total cost and time per unit installed, it is
conceivable to obtain considerable economic advantage from the use of additive digital
fabrication to create complicated structures through the development of an AI-driven
robotic construction method as part of digital fabrication in the construction industry.
Furthermore, investigations by several researchers [71,98,99,101–103] have shown the
possibility of replacing risky and difficult manual construction work with automated robots.

Enhanced Risk Mitigation

All construction projects have a few risks, which can take numerous forms, including
quality, timeliness, and cost. A particular strength of Hong’s [105] argument is that AI
technologies can assist in assigning time and cost contingency to completing a construction
project through the development of natural language-related AI technologies including
clustering methods, including latent semantic analysis (LSA), latent Dirichlet allocation
(LDA), and word2vec, amongst many others for quantitative analysis in construction
scheduling. Varouqa [55] concurred and went on to say that AI technologies can be
employed as optimization strategies in prefabricated construction projects to save time and
money. Furthermore, Lee [108] adopted AI technology to perform a pre-emptive contract-
risk evaluation, which can offer stakeholders with contractual positions and rights based
on contract facts, minimizing the number of claims and conflict cases between participating
parties during construction.
Energies 2024, 17, 182 11 of 21

3.4.2. Challenges
There are several challenges described in the seventy selected articles about the im-
plementation of AI technology in the construction industry (see Table 2). In general, a low
accuracy level due to scarcity of available data was found to be the most frequently cited
challenge (41.43%) during the adoption of supervised learning AI technology (15 occur-
rences), followed by data transformation techniques not transferable to data from other
regions (12.86%) during the implementation of the same AI technology (5 occurrences),
lack of real-world applicability (11.43%) when using deep reinforcement learning AI tech-
nology (3 occurrences), and incorrect image classification of structures (4.29%) during deep
learning AI technology adoption (2 occurrences). However, 2.86% of the articles considered
the combination of industrial robot size and weight limits and high demand for sophisti-
cated algorithms and computing power owing to massive volumes of data being equally
troublesome when adopting deep learning and supervised learning AI technology in the
construction industry. Other notable challenges include difficulty in model calibration
and excessive modeling errors for heating demand prediction, long installation time for
robots, difficulties in developing inference rules for expert systems, and misclustering
of some project milestones into building works for natural language processing, among
many others.

3.4.3. Opportunities for Technological Advancement

More so, Table 2 emphasizes how AI technologies open a slew of opportunities for
technological advancement in the construction industry, giving it a competitive edge by
improving efficiency across the whole value chain from building materials manufacturing
to design/planning, construction/execution, and supply/facility management.

Data Augmentation
In concrete, it shows that 41.43% of the selected papers who experienced a low ac-
curacy level due to scarcity of available data suggested the need for future studies to
augment datasets for the development of more robust AI technologies in the industry. For
instance, Hu’s [71] automatic robotic disinfection framework was unable to investigate
the relationship between adequate UV light exposure and the effects of pathogen eradi-
cation due to low accuracy in segmenting the areas of potential contamination on small
objects such as doorknobs and cabinet handles in adverse conditions. Furthermore, Davila
Delgado [74] successfully demonstrated the application of undercomplete, sparse, deep,
and variational autoencoders as novel techniques for data augmentation and generation of
synthetic data in construction management, which can provide useful insights regarding
the underlying non-linear relationships among variables in the datasets amongst many
other selected studies.

Model Generalizability/Transferability
The opportunity for AI model generalizability and transferability became eminent as
the unique data transformation employed in 12.86% of the selected articles was not trans-
ferable to the data from other regions as typical of any data-driven model. Zhang’s [110]
argument is persuasive in this aspect, particularly when they pointed out that their building
energy AI model has better generalizability because it is based on fundamental scientific
laws. A building energy model, for example, can properly estimate the energy performance
of a new unseen control technique while a data-driven model may not. This is the case since
the data-driven model is built using a training dataset that contains no information about
the unseen control technique. In addition, Koc [60] raised awareness for future studies
to take advantage of model generalizability, since he encountered the difficulty of an un-
balanced dataset while using AI technology to assess construction workers’ post-accident
impairment status.
Energies 2024, 17, 182 12 of 21

Table 2. Cited challenges for integrating AI technologies and opportunities for technological advancement in the construction industry.

Challenges Opportunities AI Technology No. of Articles Reference %

Robotics, DL, expert system, [24,25,48,52–55,57,59–63,66,69,71–
Low accuracy level due to scarcity of available data Data augmentation 29 41.43
optimization, SL (15), NLP 73,75,80,84,88,89,91,94,96,97,104,108]
NLP, SL, optimization, expert
Data transformation not universally applicable Model generalizability and transferability 9 [19,50,51,58,70,74,92,95,109] 12.86
system, DL
Only experimental with lack of real-world applicability Real-world applicability SL, optimization, DL, RL 8 [49,68,76,78,79,85,112,113] 11.43
Incorrect image classification of structures Computer vision and AR DL, SL 3 [77,81,87] 4.29
Size and weight restrictions of industrial robots Design optimization Robot 2 [99,101], 2.86
High data demands, require advanced algorithms Cloud computing infrastructure DL, SL 2 [64,83] 2.86
Limited computing sSpeed Compiling Keras/TensorFlow RL 1 [111] 1.43
Model calibration challenges, errors in predictions Multi-objective RL RL, optimization 1 [110] 1.43
Programming safety and recalibration
Long installation time Robotics 1 [98] 1.43
of peripherals
Enhanced random neural network
Saturated neural network accuracy, optimization challenges SL, DL 1 [82] 1.43
structure generation.
Difficulties in developing inference rules High optimization techniques Expert system 1 [90] 1.43
Misclustering of some project milestones
Construction schedules analysis NLP 1 [105] 1.43
into building works
Inadequate network architecture Large (big) datasets SL, optimization 1 [65] 1.43
Omission of planning in robotic design Design optimization Robot 1 [100] 1.43
Limited scope in classification of building structures Computer vision and AR SL 1 [67] 1.43
Threshold limitations on Elasticsearch queried data Load balancing NLP 1 [106] 1.43
Limited customization of textures for walls 3D modeling and AR DL 1 [86] 1.43
Control approach for large hydraulic robots New control methods Robot 1 [26] 1.43
Robot response hindered by image quality Hgh-definition camera Robot 1 [103] 1.43
Natural language exhibits diverse expressions Advanced optimization NLP, SL 1 [56] 1.43
Inability to repair walls and columns Data-driven machine learning Expert system 1 [93] 1.43
Heterogeneous hardware and software integration AI knowledge experts SL 1 [64] 1.43
Social barriers to the adoption of AI AI education and trainings NLP, SL, DL 1 [107] 1.43
Energies 2024, 17, 182 13 of 21

Real-World Applicability
About 11.43% of the selected articles only tend to have simulated the adoption of
AI technologies in a controlled environment thus lacking the confidence to validate their
methods in a real-world setting.
According to Vázquez-Canteli [111], a useful graphical user interface (GUI) that
allows users to write machine learning code or set hyper-parameters of the algorithms
after the simulation environment is compiled is required for their fast AI-based building
energy simulator implemented in an integrated simulation environment to be tested in a
physical setting. Furthermore, as illustrated in Hong’s [105] AI framework for clustering
construction schedules in UK-based construction projects, Gondia’s [19] AI-based model
lacks the use of real construction project schedule information for their construction project
delay risk prediction implementation in Egypt.

Computer Vision and Augmented Reality

The application of the most recent computer vision techniques and augmented reality
functions followed suit, with 4.29% of selected articles recommending them as a means of
advancing technology in the construction industry. More precisely, in the implementation
of automatic image recognition of architectural heritage sites by Palma [81], for example,
items in the same cultural site appeared to be extremely similar, or could appear together in
the same view, making it impossible to distinguish one part from another. However, these
circumstances emphasize the practical value of AI rather than the recognition of landmarks
that are far apart. Thus, this implies the most up-to-date computer vision algorithms will
be key to obtaining more detailed information than previously or the use of augmented
reality functions to enhance their interaction.
Other notable opportunities for technological advancement that received less attention
from the selected papers include design optimization and cloud computing infrastructure
with 2.86%, followed by multi-objective reinforcement learning, safety programming and
re-calibration of peripheral modules enterprise AI knowledge experts, and AI education
and trainings, among many others (1.43% each).

4. Discussion
The findings of this systematic review primarily elucidate the dissemination pattern of
research articles based on their publication sources, underscoring a pronounced prevalence
of coverage within academic journals on the subject. This trend aligns cohesively with
the research conducted by [41,116] affirming the transformative surge in the integration of
artificial intelligence within the construction industry, largely attributed to the availability
of substantial funding in this domain. Additionally, a discernible and consistent upward
trajectory is observed in the quantity of scholarly research publications addressing the
application of AI technology in the construction industry across the broader research
community ecosystem. This arguably shows a significant improvement in the promotion
of research and development of trustworthy AI solutions by funding bodies and agencies
across the globe. Interestingly, this is in line with the findings of Rahkovsky [117], who
argued that artificial intelligence research clusters are experiencing extreme growth due
to great support from research funding organizations that are currently been led by the
National Natural Science Foundation of China (NNSFC). It is no surprise from the results
of this systematic review that researchers from China are dominating the research space of
the application on AI technologies in the construction industry since NNSFC is the largest
funder of AI technology research over other large funding bodies such as the National
Institutes of Health and National Science Foundation from the USA, European Commission
and European Research Council from Europe, and Japan Society for the Promotion of
Science from Japan, among others.
Secondly, the analysis of this study further provided answers to all the research
questions stated in the first section. More specifically, although seven major AI technology
types were found in the literature, supervised learning emerged as the most influential
Energies 2024, 17, 182 14 of 21

AI technology of choice for most researchers, especially toward its applicability in health
and safety management. Supervised learning is a branch of machine learning in which
computer algorithms are trained on a labeled input dataset for a certain output. From a
labeled training dataset (i.e., a dataset that already has a known value for each record’s
output variable), supervised machine learning algorithms can find insights, patterns, and
correlations. When proper answers for a given task during training are provided, the
machine learning algorithm can learn how the rest of the characteristics relate to the output,
allowing you to unlock insights and make predictions based on past data. This is extremely
crucial for the industry and consistent with the findings of [5], which argued that the
industry can derive key benefits from AI to drive further profitability only when it leverages
the amount of data produced from a backlog of project schedules, as-built drawings and
models, computer-aided designs, costs, and invoices, among many other sources.
Furthermore, the results showed that although AI technologies can be applied in
three major stages of the construction project lifecycle, more attention is drawn towards
the supply/facility management stage (see Section 3). It can be argued that this is owing
to the massive quantity of data collected over time (from the design stage all the way
through), making it ideal for the adoption of the most significant AI technology (supervised
learning). Thus, this creates a great opportunity for the industry to capitalize by allowing,
for example, facility managers to take proactive action. For instance, as argued by [110],
AI can recognize portions of buildings that are not being utilized and automatically turn
off the heating, ventilation, and air conditioning, substantially decreasing energy use.
Moreover, AI technologies in the construction industry were found to hold promise for
applications in many types of construction projects and their respective lifecycle application
area with about 20% of the literature reporting their implementation in any structures and
systems that are part of the built environment (urban area, pedestrian walkways, parks, etc.)
projects. Kılkış’s [118] argument about this is compelling, especially when they mentioned
that the built environment impacts all parts of our life, including the buildings we live in,
the distribution systems that provide us with water and energy, and the roads, bridges, and
transportation systems we use to move about.
Additionally, most articles acknowledged that artificial intelligence technologies’ grow-
ing popularity in the construction industry would offer a wide range of benefits with
potential for design expansion as a key benefit according to most of the selected literature.
As such, this study argues that given the substantial time investment by engineers and
architects in the architectural design process and their access to an extensive database
housing numerous pre-existing building plans, an artificial intelligence (AI) technology
system holds the capacity to generate diverse design alternatives based on the collective
information derived from the repository of designs. Consequently, designers can input
design objectives and parameters into the system, allowing it to systematically explore all
conceivable permutations of a solution. This process results in the creation of design alter-
natives that satisfy the predefined requirements, with the system progressively refining its
understanding of optimal design choices through iterative learning. This iterative learning,
in turn, enhances the system’s efficacy with each subsequent project. Beyond the potential
design benefits, the application of generative design holds promises in augmenting creativ-
ity. For instance, it can empower architects to unveil hitherto unimagined approaches to
designing forms and curves or guide them toward innovative design solutions that may
remain unexplored through conventional means.
However, most of the articles reported that it is challenging to apply AI technologies
in the construction industry because of a low accuracy level due to the scarcity of available
data. Data scarcity arises when there is a paucity of labeled training data or when there
is none at all. It might be a shortage of data for a particular label as compared to the
other labels (known as data imbalance). It was discovered from some of the selected
literature [52–55,80], which is also in line with research [1], that mega infrastructure projects
often have access to a lot of data; however, they may have data imbalances, whereas small-
sized projects typically have a limited amount of labeled training data. As a result, resolving
Energies 2024, 17, 182 15 of 21

this issue cannot be overstated, as reported articles (41.43%) universally agreed, citing “data
augmentation” as one of the quickest prospects for technical improvement in this area. For
instance, the research [74] successfully demonstrated the application of undercomplete,
sparse, deep, and variational autoencoders as novel techniques for data augmentation and
generation of synthetic data in construction management which can provide useful insights
regarding the underlying non-linear relationships among variables in the datasets amongst
many other selected studies.
Notwithstanding, practical obstacles beyond just data accuracy persist around cultural
readiness, ethical risks, skill shortages, and flaws in security posture for many construction
projects exploring AI solutions. As the research [119,120] assessed, construction has often
lagged significantly in digital transformation and technology assimilation compared to
other industries. Coupled with an aging workforce leaning on legacy methods, this exac-
erbates reluctance and barriers to AI change management. For instance, PwC [121] notes
generational shifts may gradually improve receptiveness, like modeling shows younger
workers are 67% more open to retraining on AI tools relative to senior staff. But broad
culture change inevitably remains for the long term. Customized change management
programs fitting construction realties are hence vital to align teams behind AI via strategic
internal communications campaigns and leadership vision as exemplified by firms such as
Bechtel. Additionally, the opaque decision-making of AI systems poses ethical dilemmas
around accountability as flagged by Parveen [39]. Lack of explainable outcomes or audit
trails can impede transparency and responsible oversight of automated systems. There is
also a dearth of standardized governance principles as highlighted in Egwim’s [1] delay
risk assessment.
The specialist expertise needed is another capacity challenge evidenced by widening
talent gaps globally per Johnson’s [122] labor market analysis. Most construction firms are
not staffed with multidisciplinary data scientists or algorithm auditors. As such, the shortage
of such AI and analytics roles may worsen for small- and mid-size construction companies
lacking resources to reskill staff or attract experts. Additionally, many construction industry
jobs also require on-site client coordination, hence diminishing flexibility that technology
candidates expect. Therefore, targeted training programs are crucial to developing well-
rounded internal capabilities. Finally, the vulnerability of connected tools or data handling
processes to malicious threats leaves unprepared adopters exposed to crippling breaches as
studied across industries [123,124]. Also, lack of transparency around data rights or algorith-
mic decision-making processes also introduces major ethical risks. Furthermore, antiquated
security postures coupled with failures to implement robust and resilient protections can
negate any assumed productivity gains. Thus, addressing these open socio-technical prob-
lems demands coordinated efforts across construction stakeholders to formulate frameworks,
standards, and cultural shifts guided by construction-specific nuances.

5. Conclusions
The systematic review study covers 70 studies that were judged to be rigorous, credible,
and relevant in their application of AI technology in the construction sector. The research
content of these 70 publications demonstrated that artificial intelligence research in the
construction sector has taken a quantum leap, with increased interest in academic journals,
particularly in the last few years, owing to the availability of funding in that area. Most
articles pertinent to the research topic in general were published by Chinese researchers.
More precisely, scholars from the Republic of Korea and China contributed the most
publications to the construction/execution lifecycle stage of the construction value chain.
Furthermore, China also published most of the related articles concerning AI applications
in the planning and facility management lifecycle stages of the construction value chain.
Construction AI technology was discovered to be a growing application field, with
supervised learning, deep learning, knowledge-based systems, robotics, natural language
processing, optimization, and reinforcement learning AI technologies all appearing to have
more potential to influence the development of AI research for increased efficiency and
Energies 2024, 17, 182 16 of 21

productivity. Regarding the construction AI technology categories given above, the bulk of
the featured publications used the supervised learning approach. A substantial number of
the articles were connected to the built environment and residential building in terms of the
construction project types in which such AI technologies are used. The papers on high-rise
and commercial buildings came after that. A few studies advocated the application of AI in
building retrofits and water treatment plants. Most publications on the built environment
concentrated on the construction/execution stage. Similarly, the planning and facility
management lifecycle stages of the residential building garnered the most attention.
According to the findings, the most significant number of studies across all build-
ing construction disciplines focused on the possibility for regenerative design expansion.
However, there are various obstacles to implementing AI technology in the construction
industry, with low accuracy owing to a lack of relevant data being the most commonly
mentioned issue. It is also worth noting that, despite being the most prevalent AI construc-
tion method, supervised learning has been the technology of choice for the most difficult
challenge to be solved in the industry. And as sparse, deep, and variational autoencoder
approaches show promise in providing meaningful insights into the underlying non-linear
correlations among variables in datasets, data augmentation was identified as one of the
most promising areas for technical advancement.
This study presents an all-inclusive systematic review of a vast body of knowledge
on artificial intelligence in the construction industry. The findings of this study present a
comprehensive assessment of the many types and categories of AI technologies, as well
as their application areas and the advantages of using them at the three lifecycle stages
of the construction value chain. This knowledge will assist construction organizations
across the world in recognizing the efficiency and productivity advantages that AI tech-
nologies can provide while helping them make smarter technology investment decisions.
It will point construction organizations in the right direction in terms of imagining the
construction problems that AI technology could solve. In addition, it is possible to integrate
evidence from the sorts of construction projects where AI technologies were used to address
technological difficulties and see what new AI technologies can accomplish in the future.
Evidently, the findings of this study are based on a systematic review methodology.
Given that the research article keywords were domain-specific, the principal drawback of
this study approach might be bias in publication selection. As a result, it is possible that
some important papers were overlooked throughout the search. Additionally, the PRISMA
protocol mandated the use of predetermined inclusion and exclusion criteria for article
selection, implying that important publications that did not meet these criteria may have
been overlooked as well. Furthermore, any breakthroughs in the field of AI technology in
construction are pushed by experts who are unable to publish in book series, conference
proceedings, or academic journals. Consequently, there is a chance that any important
research from the experts or somewhere else were overlooked throughout the search.
Although the study explored a variety of AI technologies for various construction
projects, further research is needed to figure out how to simplify these complicated systems
and processes to establish an integrated AI system for the construction sector. Therefore, an
implementation framework is crucial to soften the introduction of a system and bridge the
adoption gap by addressing low accuracy due to a scarcity of available data, model gener-
alizability, incorrect image classification of structures, high requirements for sophisticated
algorithms, and limited computing speed across the existing construction value chain.

Funding: The doctoral studies of the first author were financially supported by a University of
Hertfordshire Doctoral Scholarship, allowing for the completion of this research as part of the first
author’s doctoral degree training at the University of Hertfordshire in the United Kingdom.
Data Availability Statement: No new data were created or analyzed in this study. Data sharing is
not applicable to this article.
Conflicts of Interest: The authors declare no conflict of interest.
Energies 2024, 17, 182 17 of 21

References
1. Egwim, C.N.; Alaka, H.; Toriola-Coker, L.O.; Balogun, H.; Sunmola, F. Applied artificial intelligence for predicting construction
projects delay. Mach. Learn. Appl. 2021, 6, 100166. [CrossRef]
2. Kulkarni, P.; Londhe, S.; Deo, M. Artificial Neural Networks for Construction Management: A Review. J. Soft Comput. Civ. Eng.
2017, 1, 70–88. [CrossRef]
3. Egwim, C.N.; Egunjobi, O.O.; Gomes, A.; Alaka, H. A Comparative Study on Machine Learning Algorithms for Assessing
Energy Efficiency of Buildings. In Machine Learning and Principles and Practice of Knowledge Discovery in Databases. ECML PKDD
2021; Communications in Computer and Information Science; Springer: Cham, Switzerland, 2021; Volume 1525, pp. 546–566.
[CrossRef]
4. Zhang, R.; Li, D. Development of risk assessment model in construction project using fuzzy expert system. In Proceedings of the
2nd IEEE International Conference on Emergency Management and Management Sciences, Beijing, China, 8–10 August 2011;
pp. 866–869. [CrossRef]
5. Egwim, C.N.; Alaka, H.; Toriola-Coker, L.O.; Balogun, H.; Ajayi, S.; Oseghale, R. Extraction of underlying factors causing
construction projects delay in Nigeria. J. Eng. Des. Technol. 2021, 21, 1323–1342. [CrossRef]
6. Bajpai, A.; Misra, S.C. Identifying Critical Risk Factors for Use of Digitalization in Construction Industry: A Case Study. In
Proceedings of the 2020 IEEE India Council International Subsections Conference (INDISCON), Visakhapatnam, India, 3–4
October 2020; pp. 124–128. [CrossRef]
7. Malik, M.; Khatana, R.; Kaushik, A. Machine learning with health care: A perspective. J. Phys. Conf. Ser. 2021, 2040, 012022.
[CrossRef]
8. Huang, S.; Yang, J.; Fong, S.; Zhao, Q. Artificial intelligence in the diagnosis of COVID-19: Challenges and perspectives. Int. J.
Biol. Sci. 2021, 17, 1581–1587. [CrossRef] [PubMed]
9. Chen, J.; Lim, C.P.; Tan, K.H.; Govindan, K.; Kumar, A. Artificial intelligence-based human-centric decision support framework:
An application to predictive maintenance in asset management under pandemic environments. Ann. Oper. Res. 2021, 1–24.
[CrossRef] [PubMed]
10. Bajaj, R.; Sharma, V. Smart Education with artificial intelligence based determination of learning styles. Procedia Comput. Sci. 2018,
132, 834–842. [CrossRef]
11. Harmon, J.; Pitt, V.; Summons, P.; Inder, K.J. Use of artificial intelligence and virtual reality within clinical simulation for nursing
pain education: A scoping review. Nurse Educ. Today 2020, 97, 104700. [CrossRef]
12. Bao, Y.; Hilary, G.; Ke, B. Artificial Intelligence and Fraud Detection. In Innovative Technology at the Interface of Finance and
Operations; Springer: Cham, Switzerland, 2022; pp. 223–247. [CrossRef]
13. Chen, J.I.-Z.; Lai, K.-L. Deep Convolution Neural Network Model for Credit-Card Fraud Detection and Alert. J. Artif. Intell.
Capsul. Netw. 2021, 3, 101–112. [CrossRef]
14. Manoharan, S. An Improved Safety Algorithm for Artificial Intelligence Enabled Processors in Self Driving Cars. J. Artif. Intell.
Capsul. Netw. 2019, 1, 95–104. [CrossRef]
15. Ma, Y.; Wang, Z.; Yang, H.; Yang, L. Artificial intelligence applications in the development of autonomous vehicles: A survey.
IEEE/CAA J. Autom. Sin. 2020, 7, 315–329. [CrossRef]
16. Wang, P.; Monett, D.; Lewis, C.W.P.; Thórisson, K.R. On Defining Artificial Intelligence. J. Artif. Gen. Intell. 2019, 10, 2019–2022.
[CrossRef]
17. Wu, C.; Li, X.; Guo, Y.; Wang, J.; Ren, Z.; Wang, M.; Yang, Z. Natural language processing for smart construction: Current status
and future directions. Autom. Constr. 2022, 134, 104059. [CrossRef]
18. Yaseen, Z.M.; Ali, Z.H.; Salih, S.Q.; Al-Ansari, N. Prediction of risk delay in construction projects using a hybrid artificial
intelligence model. Sustainability 2020, 12, 1514. [CrossRef]
19. Gondia, A.; Siam, A.; El-Dakhakhni, W.; Nassar, A.H. Machine Learning Algorithms for Construction Projects Delay Risk
Prediction. J. Constr. Eng. Manag. 2020, 146, 04019085. [CrossRef]
20. Lee, Y.-C.; Shariatfar, M.; Rashidi, A.; Lee, H.W. Evidence-driven sound detection for prenotification and identification of
construction safety hazards and accidents. Autom. Constr. 2020, 113, 103127. [CrossRef]
21. George, M.R.; Nalluri, M.R.; Anand, K.B. Severity Prediction of Construction Site Accidents Using Simple and Ensemble
Decision Trees. In Proceedings of SECON’21; Lecture Notes in Civil Engineering; Springer: Cham, Switzerland, 2021; Volume 171,
pp. 599–608. [CrossRef]
22. Xie, Y.; Lee, Y.-C.; Shariatfar, M.; Zhang, Z.D.; Rashidi, A.; Lee, H.W. Historical Accident and Injury Database-Driven Audio-Based
Autonomous Construction Safety Surveillance. In Proceedings of the ASCE International Conference on Computing in Civil
Engineering 2019, Atlanta, GA, USA, 17–19 June 2019; pp. 105–113. [CrossRef]
23. Xiao, W.; Yang, J.; Fang, H.; Zhuang, J.; Ku, Y. A robust classification algorithm for separation of construction waste using NIR
hyperspectral system. Waste Manag. 2019, 90, 1–9. [CrossRef]
24. Cha, G.-W.; Moon, H.J.; Kim, Y.-M.; Hong, W.-H.; Hwang, J.-H.; Park, W.-J.; Kim, Y.-C. Development of a prediction model for
demolition waste generation using a random forest algorithm based on small datasets. Int. J. Environ. Res. Public Health 2020, 17,
6997. [CrossRef]
25. Cha, G.-W.; Moon, H.-J.; Kim, Y.-C. Comparison of random forest and gradient boosting machine models for predicting demolition
waste based on small datasets and categorical variables. Int. J. Environ. Res. Public Health 2021, 18, 8530. [CrossRef]
Energies 2024, 17, 182 18 of 21

26. Gambao, E.; Balaguer, C.; Gebhart, F. Robot assembly system for computer-integrated construction. Autom. Constr. 2000, 9,
479–487. [CrossRef]
27. Balaguer, C.; Abderrahim, M.; Navarro, J.; Boudjabeur, S.; Aromaa, P.; Kahkonen, K.; Slavenburg, S.; Seward, D.; Bock, T.; Wing,
R.; et al. FutureHome: An integrated construction automation approach. IEEE Robot. Autom. Mag. 2002, 9, 55–66. [CrossRef]
28. Chu, B.; Jung, K.; Lim, M.-T.; Hong, D. Robot-based construction automation: An application to steel beam assembly (Part I).
Autom. Constr. 2013, 32, 46–61. [CrossRef]
29. Bruckmann, T.; Reichert, C.; Meik, M.; Lemmen, P.; Spengler, A.; Mattern, H.; König, M. Concept Studies of Automated
Construction Using Cable-Driven Parallel Robots. In Cable-Driven Parallel Robots; Gosselin, C., Cardou, P., Bruckmann, T., Pott, A.,
Eds.; Mechanisms and Machine Science; Springer: Cham, Switzerland, 2018; Volume 53, pp. 364–375. [CrossRef]
30. Wu, Y.; Cheng, H.H.; Fingrut, A.; Crolla, K.; Yam, Y.; Lau, D. CU-brick cable-driven robot for automated construction of
complex brick structures: From simulation to hardware realisation. In Proceedings of the 2018 IEEE International Conference
on Simulation, Modeling, and Programming for Autonomous Robots (SIMPAR), Brisbane, QLD, Australia, 16–19 May 2018;
pp. 166–173. [CrossRef]
31. Bruckmann, T.; Spengler, A.J.; Karl, C.K.; Reichert, C.; König, M. Process Analysis of Cable-Driven Parallel Robots for Automated
Construction. In Mechatronics for Cultural Heritage and Civil Engineering; Ottaviano, E., Pelliccio, A., Gattulli, V., Eds.; Intelligent
Systems, Control and Automation: Science and Engineering; Springer: Cham, Switzerland, 2018; Volume 92, pp. 63–83. [CrossRef]
32. Birgonul, M.T.; Dikmen, I.; Budayan, C.; Demirel, T. An expert system for the quantification of fault rates in construction fall
accidents. Int. J. Occup. Saf. Ergon. 2016, 22, 20–31. [CrossRef] [PubMed]
33. Hadipriono, F.C. Expert System for Construction Safety. I: Fault-Tree Models. J. Perform. Constr. Facil. 1992, 6, 246–260. [CrossRef]
34. Imriyas, K. An expert system for strategic control of accidents and insurers’ risks in building construction projects. Expert Syst.
Appl. 2009, 36, 4021–4034. [CrossRef]
35. Jallan, Y.; Brogan, E.; Ashuri, B.; Clevenger, C.M. Application of Natural Language Processing and Text Mining to Identify
Patterns in Construction-Defect Litigation Cases. J. Leg. Aff. Disput. Resolut. Eng. Constr. 2019, 11, 04519024. [CrossRef]
36. Faraji, A.; Rashidi, M.; Perera, S. Text Mining Risk Assessment–Based Model to Conduct Uncertainty Analysis of the General
Conditions of Contract in Housing Construction Projects: Case Study of the NSW GC21. J. Arch. Eng. 2021, 27, 04021025.
[CrossRef]
37. Xu, S.; Wang, J.; Shou, W.; Ngo, T.; Sadick, A.-M.; Wang, X. Computer Vision Techniques in Construction: A Critical Review. Arch.
Comput. Methods Eng. 2020, 28, 3383–3397. [CrossRef]
38. Delgado, J.M.D.; Oyedele, L.; Ajayi, A.; Akanbi, L.; Akinade, O.; Bilal, M.; Owolabi, H. Robotics and automated systems in
construction: Understanding industry-specific challenges for adoption. J. Build. Eng. 2019, 26, 100868. [CrossRef]
39. Parveen, R. Legal Issues and Regulatory Challenges. Int. J. Civ. Eng. Technol. 2018, 9, 957–962. Available online: https:
//iaeme.com/MasterAdmin/Journal_uploads/IJCIET/VOLUME_9_ISSUE_13/IJCIET_09_13_096.pdf (accessed on 31 January
2022).
40. Schia, M.H.; Trollsås, B.C.; Fyhn, H.; Lædre, O. The Introduction of AI in the Construction Industry and Its Impact on Human
Behavior. In Proceedings of the 27th Annual Conference of the International Group for Lean Construction (IGLC), Dublin, Ireland,
6–12 July 2020; pp. 903–914. [CrossRef]
41. Abioye, S.O.; Oyedele, L.O.; Akanbi, L.; Ajayi, A.; Delgado, J.M.D.; Bilal, M.; Akinade, O.O.; Ahmed, A. Artificial intelligence
in the construction industry: A review of present status, opportunities and future challenges. J. Build. Eng. 2021, 44, 103299.
[CrossRef]
42. Sharif, S.P.; Mura, P.; Wijesinghe, S.N.R. Systematic Reviews in Asia: Introducing the ‘PRISMA’ Protocol to Tourism and
Hospitality Scholars. In Quantitative Tourism Research in Asia: Perspectives on Asian Tourism; Rezaei, S., Ed.; Springer: Singapore,
2019; pp. 13–33. [CrossRef]
43. Pahlevan-Sharif, S.; Mura, P.; Wijesinghe, S.N. A systematic review of systematic reviews in tourism. J. Hosp. Tour. Manag. 2019,
39, 158–165. [CrossRef]
44. Saunders, M.A.; Lewis, P.; Thornhill, A. Research Methods for Business Students, 6th ed.; Pearson Education: London, UK, 2019;
Available online: https://www.pearsoned.co.uk (accessed on 26 September 2021).
45. Cantú-Ortiz, F.J.; Fangmeyer, J. University performance in the age of research analytics. In Research Analytics: Boosting University
Productivity and Competitiveness through Scientometrics; CRC Press: Boca Raton, FL, USA, 2017; pp. 237–258. [CrossRef]
46. Falagas, M.E.; Pitsouni, E.I.; Malietzis, G.A.; Pappas, G.; Kouranos, V.D.; Arencibia-Jorge, R.; Karageorgopoulos, D.E.; Reagan-
Shaw, S.; Nihal, M.; Ahmad, N.; et al. Comparison of PubMed, Scopus, Web of Science, and Google Scholar: Strengths and
weaknesses. FASEB J. 2008, 22, 338–342. [CrossRef] [PubMed]
47. Gharbia, M.; Chang-Richards, A.; Lu, Y.; Zhong, R.Y.; Li, H. Robotic technologies for on-site building construction: A systematic
review. J. Build. Eng. 2020, 32, 101584. [CrossRef]
48. Pham, A.-D.; Ngo, N.-T.; Nguyen, Q.-T.; Truong, N.-S. Hybrid machine learning for predicting strength of sustainable concrete.
Soft Comput. 2020, 24, 14965–14980. [CrossRef]
49. Wang, D.; Li, J. Artificial Intelligence Aided Prediction of Building Structure Anti-seismic. In Proceedings of the 2021 5th International
Conference on Computing Methodologies and Communication (ICCMC), Erode, India, 8–10 April 2021; pp. 1403–1407. [CrossRef]
Energies 2024, 17, 182 19 of 21

50. Mahjoubi, S.; Barhemat, R.; Guo, P.; Meng, W.; Bao, Y. Prediction and multi-objective optimization of mechanical, economical,
and environmental properties for strain-hardening cementitious composites (SHCC) based on automated machine learning and
metaheuristic algorithms. J. Clean. Prod. 2021, 329, 129665. [CrossRef]
51. Zhang, J.P.; Liu, L.H.; Coble, R.J. Hybrid intelligence utilization for construction site layout. Autom. Constr. 2002, 11, 511–519.
[CrossRef]
52. Zhang, J. Potential energy saving estimation for retrofit building with ASHRAE-Great Energy Predictor III using machine learning.
In Proceedings of the 2021 1st International Conference on Control and Intelligent Robotics, Guangzhou, China, 18–20 June 2021;
pp. 425–429. [CrossRef]
53. Sanni-Anibire, M.O.; Zin, R.M.; Olatunji, S.O. Machine learning-based framework for construction delay mitigation. J. Inf. Technol.
Constr. 2021, 26, 303–318. [CrossRef]
54. Vahdani, B.; Mousavi, S.M.; Hashemi, H.; Mousakhani, M.; Ebrahimnejad, S. A New Hybrid Model Based on Least Squares
Support Vector Machine for Project Selection Problem in Construction Industry. Arab. J. Sci. Eng. 2014, 39, 4301–4314. [CrossRef]
55. Varouqa, I.F. Using Artificial Intelligence and computation Enhanced apply in neural network. J. Appl. Sci. Eng. 2021, 24, 763–770.
[CrossRef]
56. Zhang, F.; Fleyeh, H.; Wang, X.; Lu, M. Construction site accident analysis using text mining and natural language processing
techniques. Autom. Constr. 2019, 99, 238–248. [CrossRef]
57. Ayhan, B.U.; Tokdemir, O.B. Safety assessment in megaprojects using artificial intelligence. Saf. Sci. 2019, 118, 273–287. [CrossRef]
58. Xue, F.; Yao, E. Adopting a random forest approach to model household residential relocation behavior. Cities 2022, 125, 103625.
[CrossRef]
59. Toosi, H.A.; Lavagna, M.; Leonforte, F.; Del Pero, C.; Aste, N. A novel LCSA-Machine learning based optimization model for
sustainable building design—A case study of energy storage systems. Build. Environ. 2022, 209, 108656. [CrossRef]
60. Koc, K.; Ekmekcioğlu, Ö.; Gurgun, A.P. Integrating feature engineering, genetic algorithm and tree-based machine learning
methods to predict the post-accident disability status of construction workers. Autom. Constr. 2021, 131, 103896. [CrossRef]
61. Milošević, I.; Kovačević, M.; Petronijević, P. Estimating Residual Value of Heavy Construction Equipment Using Ensemble
Learning. J. Constr. Eng. Manag. 2021, 147, 04021073. [CrossRef]
62. Ayhan, M.; Dikmen, I.; Birgonul, M.T. Predicting the Occurrence of Construction Disputes Using Machine Learning Techniques. J.
Constr. Eng. Manag. 2021, 147, 04021022. [CrossRef]
63. Amin, M.N.; Iqtidar, A.; Khan, K.; Javed, M.F.; Shalabi, F.I.; Qadir, M.G. Comparison of Machine Learning Approaches with
Traditional Methods for Predicting the Compressive Strength of Rice Husk Ash Concrete. Crystals 2021, 11, 779. [CrossRef]
64. You, Z.; Feng, L. Integration of Industry 4.0 Related Technologies in Construction Industry: A Framework of Cyber-Physical
System. IEEE Access 2020, 8, 122908–122922. [CrossRef]
65. Barai, S.V.; Nair, R.S. Neuro-Fuzzy Models for Constructability Analysis. J. Inf. Technol. Constr. 2004, 9, 65–73. Available online:
http://www.itcon.org/2004/4/ (accessed on 22 March 2022).
66. Lee, Y.-C.; Scarpiniti, M.; Uncini, A. Advanced Sound Classifiers and Performance Analyses for Accurate Audio-Based Construc-
tion Project Monitoring. J. Comput. Civ. Eng. 2020, 34, 04020030. [CrossRef]
67. Zhou, P.; Chang, Y. Automated classification of building structures for urban built environment identification using machine
learning. J. Build. Eng. 2021, 43, 103008. [CrossRef]
68. Bagheri, A.; Nazari, A.; Sanjayan, J. The use of machine learning in boron-based geopolymers: Function approximation of
compressive strength by ANN and GP. Measurement 2019, 141, 241–249. [CrossRef]
69. Shehadeh, A.; Alshboul, O.; Al Mamlook, R.E.; Hamedat, O. Machine learning models for predicting the residual value of heavy
construction equipment: An evaluation of modified decision tree, LightGBM, and XGBoost regression. Autom. Constr. 2021, 129,
103827. [CrossRef]
70. Pereira, P.F.; Ramos, N.M.M.; Simões, M.L. Data-driven occupant actions prediction to achieve an intelligent building. Build. Res.
Inf. 2019, 48, 485–500. [CrossRef]
71. Hu, D.; Zhong, H.; Li, S.; Tan, J.; He, Q. Segmenting areas of potential contamination for adaptive robotic disinfection in built
environments. Build. Environ. 2020, 184, 107226. [CrossRef] [PubMed]
72. Kim, J.-M.; Bae, J.; Son, S.; Son, K.; Yum, S.-G. Development of Model to Predict Natural Disaster-Induced Financial Losses for
Construction Projects Using Deep Learning Techniques. Sustainability 2021, 13, 5304. [CrossRef]
73. Yu, Y.; Li, H.; Yang, X.; Umer, W. Estimating construction workers’ physical workload by fusing computer vision and smart insole
technologies. In Proceedings of the 35th International Symposium on Automation and Robotics in Construction and International
AEC/FM Hackathon: The Future of Building Things (ISARC 2018), Berlin, Germany, 20–25 July 2018. [CrossRef]
74. Delgado, J.M.D.; Oyedele, L. Deep learning with small datasets: Using autoencoders to address limited datasets in construction
management. Appl. Soft Comput. 2021, 112, 107836. [CrossRef]
75. Charoenkwan, P.; Homkong, N. CSDeep: A crushed stone image predictor based on deep learning and intelligently selected
features. In Proceedings of the 2017 2nd International Conference on Information Technology (INCIT), Nakhonpathom, Thailand,
2–3 November 2017; pp. 1–6. [CrossRef]
76. Su, L.-C.; Wu, X.; Zhang, X.; Huang, X. Smart performance-based design for building fire safety: Prediction of smoke motion via
AI. J. Build. Eng. 2021, 43, 102529. [CrossRef]
Energies 2024, 17, 182 20 of 21

77. Norrdine, A.; Motzko, C. An internet of things based transportation cart for smart construction site. In Proceedings of the 2020
International Conferences on Internet of Things (iThings) and IEEE Green Computing and Communications (GreenCom) and
IEEE Cyber, Physical and Social Computing (CPSCom) and IEEE Smart Data (SmartData) and IEEE Congress on Cybermatics
(Cybermatics), Rhodes, Greece, 2–6 November 2020; pp. 160–167. [CrossRef]
78. Fisher-Gewirtzman, D.; Polak, N. A learning automated 3D architecture synthesis model: Demonstrating a computer governed
design of minimal apartment units based on human perceptual and physical needs. Archit. Sci. Rev. 2019, 62, 301–312. [CrossRef]
79. Spallone, R.; Palma, V. Artificial intelligence and augmented reality: A possible continuum for the enhancement of built heritage.
Disegnarecon 2021, 14, 26. [CrossRef]
80. Bassier, M.; Vergauwen, M. Unsupervised reconstruction of Building Information Modeling wall objects from point cloud data.
Autom. Constr. 2020, 120, 103338. [CrossRef]
81. Palma, V. Towards Deep Learning for Architecture: A Monument Recognition Mobile App. ISPRS Ann. Photogramm. Remote Sens.
Spat. Inf. Sci. 2019, XLII-2/W9, 551–556. [CrossRef]
82. Lomio, F.; Farinha, R.; Laasonen, M.; Huttunen, H. Classification of Building Information Model (BIM) Structures with Deep
Learning. In Proceedings of the 2018 7th European Workshop on Visual Information Processing (EUVIP), Tampere, Finland,
26–28 November 2018; pp. 1–6. [CrossRef]
83. Long, R.; Li, Y. Research on Energy-efficiency Building Design Based on BIM and Artificial Intelligence. IOP Conf. Ser. Earth
Environ. Sci. 2021, 825, 012003. [CrossRef]
84. Šatrevičs, V.; Kul, ikovskis, G.; Ošs, O. Commercialization Potential for Deep Machine Learning Technology Using Line Scan
Camera. In Proceedings of the 24th World Multi-Conference on Systemics, Cybernetics and Informatics (WMSCI 2020), Virtual,
13–16 September 2020.
85. Ayadi, M.I.; Maizate, A.; Ouzzif, M.; Mahmoudi, C. Deep learning in building management systems over NDN: Use case of
forwarding and HVAC control. In Proceedings of the 2019 International Conference on Internet of Things (iThings) and IEEE
Green Computing and Communications (GreenCom) and IEEE Cyber, Physical and Social Computing (CPSCom) and IEEE Smart
Data (SmartData), Atlanta, GA, USA, 14–17 July 2019; pp. 1192–1198. [CrossRef]
86. Keshavarzi, M.; Afolabi, O.; Caldas, L.; Yang, A.Y.; Zakhor, A. GenScan: A Generative Method for Populating Parametric 3D Scan
Datasets. In Proceedings of the 26th International Conference of the Association for Computer-Aided Architectural Design Research
in Asia Online and Global (CAADRIA 2021), Hong Kong, China, 29 March–1 April 2021; Volume 1, pp. 91–100. [CrossRef]
87. Mei Yee, J.N.; Khean, N.; Madden, D.; Fabbri, A.; Gardner, N.; Hank Haeusler, M.; Zavoleas, Y.; Engineering Sydney, A. OPTIMISING
IMAGE CLASSIFICATION Implementation of Convolutional Neural Network Algorithms to Distinguish Between Plans and Sections
within the Architectural, Engineering and Construction (AEC) Industry. In Proceedings of the 24th International Conference of the
Association for Computer-Aided Architectural Design Research in Asia (CAADRIA), Wellington, New Zealand, 15–18 April 2019.
88. Ajayi, A.; Oyedele, L.; Owolabi, H.; Akinade, O.; Bilal, M.; Delgado, J.M.D.; Akanbi, L. Deep Learning Models for Health and
Safety Risk Prediction in Power Infrastructure Projects. Risk Anal. 2020, 40, 2019–2039. [CrossRef]
89. Yang, X.; Zhao, Q. Influence of virtual reality and 3D printing on architectural innovation evaluation based on quality of
experience evaluation using fuzzy logic. J. Intell. Fuzzy Syst. 2021, 40, 8501–8509. [CrossRef]
90. Sierra, E.A.; Argentina, D.D.I.D.C.N.; Hossian, A.A.; Labriola, C.V.; Martinez, R.G. Building automation by intelligent control of
its environment. Renew. Energy Power Qual. J. 2007, 1, 21–24. [CrossRef]
91. Muqeem, S.; Bin Idrus, A.; Khamidi, M.F.; Siah, Y.K.; Saqib, M. Application of Fuzzy expert systems for construction labor
productivity estimation. In Proceedings of the 2012 International Conference on Computer & Information Science (ICCIS), Kuala
Lampur, Malaysia, 12–14 June 2012; Volume 1, pp. 506–511. [CrossRef]
92. Amiri, M.; Ardeshir, A.; Zarandi, M.H.F. Fuzzy probabilistic expert system for occupational hazard assessment in construction.
Saf. Sci. 2017, 93, 16–28. [CrossRef]
93. Koo, T.K.; Tiong, R. An expert system for assessing the performance of RC beams and slabs. Constr. Manag. Econ. 1993, 11,
347–357. [CrossRef]
94. Ko, C.-H.; Cheng, M.-Y.; Wu, T.-K. Evaluating sub-contractors performance using EFNIM. Autom. Constr. 2007, 16, 525–530.
[CrossRef]
95. Li, H. Case-based reasoning for intelligent support of construction negotiation. Inf. Manag. 1996, 30, 231–238. [CrossRef]
96. Zhang, Y.; Yuen, K.V. Applications of Deep Learning in Intelligent Construction. Struct. Integr. 2022, 21, 227–245. [CrossRef]
97. Kruachottikul, P.; Cooharojananone, N.; Phanomchoeng, G.; Chavarnakul, T.; Kovitanggoon, K.; Trakulwaranont, D. Deep
learning-based visual defect-inspection system for reinforced concrete bridge substructure: A case of Thailand’s department of
highways. J. Civ. Struct. Health Monit. 2021, 11, 949–965. [CrossRef]
98. Wagner, H.J.; Alvarez, M.; Kyjanek, O.; Bhiri, Z.; Buck, M.; Menges, A. Flexible and transportable robotic timber construction
platform–TIM. Autom. Constr. 2020, 120, 103400. [CrossRef]
99. Krieg, O.D.; Lang, O. Adaptive automation strategies for robotic prefabrication of parametrized mass timber building components.
In Proceedings of the 36th International Symposium on Automation and Robotics in Construction (ISARC 2019), Banff, AB,
Canada, 21–24 May 2019; pp. 521–528. [CrossRef]
100. De Soto, B.G.; Agustí-Juan, I.; Hunhevicz, J.; Joss, S.; Graser, K.; Habert, G.; Adey, B.T. Productivity of digital fabrication in
construction: Cost and time analysis of a robotically built wall. Autom. Constr. 2018, 92, 297–311. [CrossRef]
Energies 2024, 17, 182 21 of 21

101. Jung, K.; Chu, B.; Hong, D. Robot-based construction automation: An application to steel beam assembly (Part II). Autom. Constr.
2013, 32, 62–79. [CrossRef]
102. Firth, C.; Dunn, K.; King, M.; Haeusler, M.H. Development of an Anthropomorphic End-effector for Collaborative use on
Construction Sites. In RE: Anthropocene, Design in the Age of Humans: Proceedings of the 25th International Conference on Computer-
Aided Architectural Design Research in Asia, CAADRIA 2020; Association for Computer Aided Architectural Design Research in
Asia: Singapore, 2020; Volume 2, pp. 363–372.
103. Li, H.; Luo, X.; Skitmore, M. Intelligent Hoisting with Car-Like Mobile Robots. J. Constr. Eng. Manag. 2020, 146, 04020136.
[CrossRef]
104. Kontovourkis, O.; Konatzii, P. Environmental and cost assessment of customized modular wall components production based on
an adaptive formwork casting mechanism: An experimental study. J. Clean. Prod. 2021, 286, 125380. [CrossRef]
105. Hong, Y.; Xie, H.; Bhumbra, G.; Brilakis, I. Comparing Natural Language Processing Methods to Cluster Construction Schedules.
J. Constr. Eng. Manag. 2021, 147, 04021136. [CrossRef]
106. Kim, T.; Chi, S. Accident Case Retrieval and Analyses: Using Natural Language Processing in the Construction Industry. J. Constr.
Eng. Manag. 2019, 145, 04019004. [CrossRef]
107. Sonetti, G.; Naboni, E.; Brown, M. Exploring the Potentials of ICT Tools for Human-Centric Regenerative Design. Sustainability
2018, 10, 1217. [CrossRef]
108. Lee, J.; Yi, J.-S.; Son, J. Development of Automatic-Extraction Model of Poisonous Clauses in International Construction Contracts
Using Rule-Based NLP. J. Comput. Civ. Eng. 2019, 33, 04019003. [CrossRef]
109. Zhang, J.; Chen, Y.; Hei, X.; Zhu, L.; Zhao, Q.; Wang, Y. A RMM based word segmentation method for Chinese design specifications
of building stairs. In Proceedings of the 2018 14th International Conference on Computational Intelligence and Security (CIS),
Hangzhou, China, 16–19 November 2018; pp. 277–280. [CrossRef]
110. Zhang, Z.; Chong, A.; Pan, Y.; Zhang, C.; Lam, K.P. Whole building energy model for HVAC optimal control: A practical
framework based on deep reinforcement learning. Energy Build. 2019, 199, 472–490. [CrossRef]
111. Vázquez-Canteli, J.R.; Ulyanin, S.; Kämpf, J.; Nagy, Z. Fusing TensorFlow with building energy simulation for intelligent energy
management in smart cities. Sustain. Cities Soc. 2019, 45, 243–257. [CrossRef]
112. Aghalari, A.; Morshedlou, N.; Marufuzzaman, M.; Carruth, D. Inverse reinforcement learning to assess safety of a workplace
under an active shooter incident. IISE Trans. 2021, 53, 1337–1350. [CrossRef]
113. Soman, R.K.; Molina-Solana, M. Automating look-ahead schedule generation for construction using linked-data based constraint
checking and reinforcement learning. Autom. Constr. 2022, 134, 104069. [CrossRef]
114. Egwim, C.N.; Alaka, H.; Egunjobi, O.O.; Gomes, A.; Mporas, I. Comparison of machine learning algorithms for evaluating
building energy efficiency using big data analytics. J. Eng. Des. Technol. 2022. [CrossRef]
115. Egwim, C.N.; Alaka, H.; Pan, Y.; Balogun, H.; Ajayi, S.; Hye, A.; Egunjobi, O.O. Ensemble of ensembles for fine particulate matter
pollution prediction using big data analytics and IoT emission sensors. J. Eng. Des. Technol. 2023. [CrossRef]
116. Debrah, C.; Chan, A.P.; Darko, A. Artificial intelligence in green building. Autom. Constr. 2022, 137, 104192. [CrossRef]
117. Rahkovsky, I.; Toney, A.; Boyack, K.W.; Klavans, R.; Murdick, D.A. AI Research Funding Portfolios and Extreme Growth. Front.
Res. Metrics Anal. 2021, 6, 630124. [CrossRef]
118. Kılkış, Ş. Better security and protection for people and ecological systems: Integrated approaches for decoupling urban growth
from emission pressures in megacities. In Sustainable Mega City Communities; Butterworth-Heinemann: Oxford, UK, 2021;
pp. 73–93. [CrossRef]
119. Egwim, C.N.; Alaka, H. A Comparative Study on Machine Learning Algorithms for Predicting Construction Projects Delay. In
Proceedings of the Edmic 2021: Environmental Design And Management International Conference: Confluence Of Theory And
Practice In The Built Environment: Beyond Theory Into Practice, Helsinki, Finland, 19–23 August 2002.
120. Egwim, C.N.; Alaka, H.; Demir, E.; Balogun, H.; Ajayi, S. Systematic review of critical drivers for delay risk prediction: Towards a
conceptual framework for BIM-based construction projects. Front. Eng. Built Environ. 2022, 3, 16–31. [CrossRef]
121. PwC. The Potential Impact of Artificial Intelligence on UK Employment and the Demand for Skills. A Report by PwC for the
Department for Business, Energy and Industrial Strategy. BEIS Res. Rep. Number 2021/042, no. August, 2021. Available online:
www.pwc.com/structure (accessed on 22 December 2023).
122. Johnson, M.; Jain, R.; Brennan-Tonetta, P.; Swartz, E.; Silver, D.; Paolini, J.; Mamonov, S.; Hill, C. Impact of Big Data and Artificial
Intelligence on Industry: Developing a Workforce Roadmap for a Data Driven Economy. Glob. J. Flex. Syst. Manag. 2021, 22,
197–217. [CrossRef]
123. Saka, A.; Taiwo, R.; Saka, N.; Salami, B.A.; Ajayi, S.; Akande, K.; Kazemi, H. GPT models in construction industry: Opportunities,
limitations, and a use case validation. Dev. Built Environ. 2024, 17, 100300. [CrossRef]
124. Saka, A.B.; Oyedele, L.O.; Akanbi, L.A.; Ganiyu, S.A.; Chan, D.W.; Bello, S.A. Conversational artificial intelligence in the AEC
industry: A review of present status, challenges and opportunities. Adv. Eng. Inform. 2023, 55, 101869. [CrossRef]

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.