See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/384457602

Integrating BIM and AI for Smart Construction Management: Current Status

and Future Directions

Article in Archives of Computational Methods in Engineering · January 2023

CITATIONS READS

11 469

3 authors, including:

Chelse Mario
Hofstra University
70 PUBLICATIONS 13 CITATIONS

SEE PROFILE

All content following this page was uploaded by Chelse Mario on 30 September 2024.

The user has requested enhancement of the downloaded file.

Archives of Computational Methods in Engineering (2023) 30:1081–1110
https://doi.org/10.1007/s11831-022-09830-8

REVIEW ARTICLE

Integrating BIM and AI for Smart Construction Management: Current

Status and Future Directions
Yue Pan1 · Limao Zhang2

Received: 5 March 2022 / Accepted: 8 October 2022 / Published online: 3 November 2022
© The Author(s) under exclusive licence to International Center for Numerical Methods in Engineering (CIMNE) 2022

Abstract
At present, building information modeling (BIM) has been developed into a digital backbone of the architecture, engineer-
ing, and construction industry. Also, recent decades have witnessed the fast development of various AI techniques in reliably
tackling a huge amount of data under complex and uncertain environments. Since both BIM and artificial intelligence (AI)
have attracted sustainable attention, the integration of BIM and AI can demonstrate newly added value in handling construc-
tion projects with inherent complexity and uncertainty. For a clear understanding of BIM-AI integration in boosting smart
construction management, the goal of this paper is to make a comprehensive investigation and summary of its potential
value and practical utility to drive the construction industry to catch up with the fast pace of automation and digitalization.
Through both the bibliometric analysis and information analysis, this paper provides a deep insight into the status quo and
future trends for leveraging AI during the entire lifecycle of a BIM-enabled project. It is worth noting that keywords that
are highly cited in the latest two years contain deep learning, internet of things, digital twin, and others, which means AI is
evolving as the next frontier to accelerate the revolution in the traditional civil engineering. According to keyword clusters
derived from the log-likelihood ratio, we determine six advanced research interests and discuss the state-of-the-art research,
including automated design and rule checking, 3D as-built reconstruction, event log mining, building performance analysis,
virtual and augmented reality, and digital twin. Besides, a growing force can be put on three potential directions to more
broadly adopt the BIM-AI integration, including synthesis of human–machine intelligence, civil-level digital twin, and
blockchain, aiming to make BIM and AI live up to expectations.

1 Introduction that BIM is more than a high-fidelity 3D model, which per-

mits multi-disciplinary information to be superimposed into
Building information modeling (BIM), known as a shared virtual prototyping for more efficient collaboration and com-
database and knowledge center, has been proven useful in munication [2]. Through exploration of the available project
visualizing, integrating, and analyzing the physical and func- data, BIM has opportunities to inform valuable decisions
tional characteristics of built objects across the whole build- concerning different management aspects, such as sustain-
ing lifecycle, including the phase of design, construction, able design, schedule planning, cost estimation, construc-
operation and maintenance (O&M) [1]. It is worth noting tion monitoring, building performance analysis, and others.
Since the BIM-centric project delivery has shown its out-
standing performance in digital management, it has been
* Limao Zhang widely adopted to drive a high degree of automation, intel-
zlm@hust.edu.cn ligence, and reliability of the architecture, engineering, and
Yue Pan construction (AEC) industry in the digital era [3].
panyue001@sjtu.edu.cn According to a survey from the McGraw-Hill Con-
1
Shanghai Key Laboratory for Digital Maintenance struction, the adoption of BIM is accelerating in many
of Buildings and Infrastructure, Department of Civil countries, which has gradually evolved as a central part
Engineering, Shanghai Jiao Tong University, 800 of governmental policies in infrastructure development in
Dongchuan Road, Shanghai, China the last decade [4]. For example, the United States and
2
School of Civil and Hydraulic Engineering, Huazhong the United Kingdom have the highest awareness of BIM,
University of Science and Technology, 1037 Luoyu Road, where BIM is forced to use. Germany and France also
Hongshan District, Wuhan 430074, Hubei, China

13
Vol.:(0123456789)
1082 Y. Pan, L. Zhang

witness a surge in BIM implementation, and thus BIM intelligence (AI) for further guiding the smart develop-
will be mandatory in the near future. Particularly, 87% of ment of civil engineering. This has not been summed up
BIM users in these four leading countries have confirmed yet in the existing literature.
that they experience great business value from BIM, who Notably, AI has been regarded as the next potential fron-
highlight three top benefits of BIM, including the reduced tier in the AEC industry. Generally, the core idea behind
conflicts and errors, the better field coordination, and the most AI algorithms, such as machine learning, deep learn-
flexible adjustment in construction [5]. Therefore, it is evi- ing, cognitive computing, and others, is to simulate human
dent that continuous efforts will be made to promote the thinking, which is able to deeply learn and analyze a large
growth of BIM implementation around the world, which amount of data for tasks of clustering, prediction, and opti-
can bring significant technological innovation to facilitate mization. Compared to conventional analytical approaches,
a digitalization process in the construction domain. AI is superior to tackle complex problems in the real-world
Motivated by the popularity of BIM, previous studies in a more automated and reliable way. It is helpful to facili-
have made great contributions to reviewing BIM usage in tate a smarter decision-making process under less depend-
tackling AEC problems. For example, Zhao [6] conducted ency on human experience and knowledge, aiming to maxi-
a scientometric review to gain a snapshot of the global mize the chance of success in realizing the project goal [14].
BIM research during the year 2005–2016. Li et al. [7] Hence, the nature of AI makes it to be a suitable tool to
provided an objective summary of BIM knowledge and well cope with challenges in construction projects that are
highlighted the evolution of BIM-related research. Wen highly complex and unnecessarily tedious. As a result, the
et al. [8] adopted the visualization analysis method to draw adoption of various AI technologies is expected to bring
the mapping knowledge to present the research progress sustainable benefits in terms of automation, risk mitigation,
and development trend of BIM. However, these generic improved efficiency and safety to the construction industry
studies mainly relied on the bibliometric approach to per- [15]. Although the high applicability of AI has led to an
form co-author, co-word, and co-citation analysis, which upward trend of research on the AI-powered AEC industry
were insufficient in qualitatively discussing the practical since the start of the twenty-first century, the research topic
applications and future development. To achieve a more concerning AI in the construction sector is still at an early
thorough analysis, some papers have attempted to exam- stage, which is expected to bolster its rapid growth in the
ine the contribution of BIM on a certain topic under the coming years. The reasons for the low-level AI adoption in
interaction of objective investigation and subjective inter- construction probably lie in high cost, trust, security, talent
pretation, which can make a more detailed analysis of the shortage, internet connectivity, project uniqueness, and oth-
resulting articles and the corresponding research direc- ers [16]. Meanwhile, the fragmented nature of the industry
tions to be resolved in the future. For example, Lu et al. [9] is another critical reason to impede AI incorporation into
studied BIM in supporting green building development in construction, which will lead to issues of data acquisition,
terms of applications, functions, and assessments. Wang retention, and interpretation [17]. That is to say, AI has not
et al. [10] reviewed the combination of BIM and geograph- reached its full potential in the construction domain. The
ical information systems (GIS) in a sustainable built envi- foregoing limitations inspire this study to focus on the inte-
ronment and introduced four relevant applications in data gration of BIM and AI techniques. Since the unique prop-
integration, lifecycle project management, energy manage- erty of BIM lies in holding rich sources of historical data to
ment, and urban governance. Tang et al. [11] presented deliver information throughout the project lifecycle, it can
a comprehensive review of the current situation of BIM impose a new demand on making sense of these increas-
and internet of things (IoT) device integration along with ing BIM data by AI-enabled solutions for accelerating the
the examination of limitations and prediction of future digitalized process. The engineering informatics under the
trends. Gao et al. [12] examined the current BIM abili- BIM-AI integration is believed to bridge the gap of digital
ties in supporting O&M activities of facilities and offering transformation in construction management over the full
new functionalities for facility managers. Yin et al. [13] lifecycle chain.
carried out a bibliometric-qualitative review approach to Currently, the technical premise of BIM entails digital
reveal the research needs of BIM for off-site construc- information management to produce data-rich models for
tion. However, these reviews have mainly uncovered the use in a project's life cycle [18], while AI is paving an intelli-
practicality of BIM for specific purposes or processes. It gent way to shift the vision model to realization for decision-
clearly lacks emphasis on promoting BIM development making purposes [19]. Since BIM and AI can show their
throughout the project to a higher level of intelligence. own advantages in their individual fields, a digital practice
Since the construction industry needs to embrace greater can be reasonably centered on the enormous potential of
digitalization, an essential task is to reveal the meaning- seamlessly integrating a variety of fast-growing AI meth-
ful relationship between BIM and the popular artificial ods into BIM-related projects. That is to say, the motivation

13
Integrating BIM and AI for Smart Construction Management: Current Status and Future Directions 1083

behind BIM-AI integration is to maximize great value from BIM-based projects, ultimately achieving significant con-
both BIM and AI, which creates the possibility of easily tributions to the revolution and improvement of the AEC
getting great access to available information and automati- industry in terms of efficiency, profitability, safety, intel-
cally bringing deep insights into a complex project with less ligence, and others. As for the scientific value, it is help-
manual work. Therefore, the combination of these two pow- ful to keep abreast of the current development status and
erful techniques is bound to grow in the future, which can ascertain the future research directions about BIM-AI inte-
act as a digital assistant to improve the productivity, quality, gration. As for the reference value, our review can enhance
and safety of a construction project. As reviewed. the con- the knowledge of researchers and decision-makers in the
nection between BIM and various AI methods has turned construction domain and then guide them to perform AI-
out to be an emerging area of research for the purpose of driven applications far more than machine learning algo-
revolutionizing the AEC industry. Although Zabin et al. [20] rithms in wide application ranges of BIM projects. There are
summarized the use of BIM data through machine learning, four open questions to be answered: (1) What is the current
its main weakness is that it only focuses on machine learning situation of BIM-AI integration and the development over
applications to BIM-generated project data. It is restricted time. (2) What is the potential area of BIM-AI integration
to the data analysis aspect. In other words, machine learning for improving project management. (3) What is the advanced
can be treated as a specific subcategory of AI to maximize research concerning BIM-AI integration that is worth dis-
the value of big data [21], which is responsible to retrieve cussing. (4) What are the prominent research trends in the
hidden patterns and unknown relationships from large quan- future study. For seeking answers to these questions, we
tities of data [22, 23]. In fact, the AI techniques that refer to intend to perform a comprehensive review based on a mix
a border idea to imitate human behavior are far more than of bibliometric analysis and information analysis to study
machine learning algorithms. Accordingly, the review on the the existing research works on the BIM-AI integration. For
topic of BIM-AI integration is previously unstudied. There- one thing, the bibliometric analysis as a statistical evaluation
fore, the interests need to fall into more kinds of AI methods is responsible to convert text into quantitative data, which
beyond machine learning for decision making. Additionally, can not only measure the influence of articles, authors, and
the adoption of AI into BIM is still in a nascent phase, which others within the scientific community but also capture the
has not yet reached its full potential. In light of the above- development process of the targeted topic. For another, the
mentioned issues, it is necessary to carry out a systematic information analysis focuses on interpreting the content of
and thorough review to narrow the gap in providing a full representative papers in detail, enabling to make the sum-
picture and understanding of the state-of-art development of mary and inference for assurance of in-depth investigation
BIM and AI along with their linkages. in certain key domains.
To sum up, the goal of the present research is to per- Towards the objective of our literature review, Fig. 1 visu-
form a more comprehensive and scientific review of papers alizes a general framework to critically review the synthesis
about the practical synthesis between BIM and AI in the last of BIM and AI, which can make the best use of both the
decades. More specifically, the BIM-AI integration can be quantitative and qualitative approaches to eliminate biased
understood to flexibly leverage various AI technologies in conclusions. Observably, the transparent process begins

Step 1 Step 2 Step 3

Paper retrieval Bibliometric analysis Information analysis

Knowledge discovery

Keyword search Discover the AI

potential in BIM
Selection criteria
definition Introduce the
VOSviewer CiteSpace advance research
interests
Prepared
literature Identify the future
database Country/ research directions
Journal Region Author Keyword

Fig. 1  Framework of the proposed literature review

13
1084 Y. Pan, L. Zhang

with the web search of relevant scientific articles from the mainly refers to [15, 16]. To be more specific, the two com-
Scopus database for paper retrieval. Some up-to-date pieces plementary terms BIM and AI are research objectives in this
of literature that are satisfied with the pre-defined selection review, and the topic concerning the construction industry
criteria are selected and gathered in a database serving as aims to limit the scope of papers in the engineering field.
the knowledge basis for further exploration. Subsequently, (3) There are four key points in the process of publication
two software packages named VOSviewer and CiteSpace search, aiming to prepare a high-quality dataset for ensur-
are deployed to produce informative conceptualizations, ing the reliability of analytical results. At first, the scope of
which can systematically map and discover the structural keyword search is within the full article, which can reduce
and dynamical aspects of the prepared literature dataset. the likelihood of omitting related papers. Second, a query
Finally, information analysis is carried out for knowledge statement is formulated based on the Boolean operator to
discovery from three perspectives, resulting in conclusions sufficiently cover the target topic. Third, result filtering is
to offer a deeper insight into the potential of AI in develop- performed to limit a large number of papers. For example,
ing new BIM-related topics. As a result, a roadmap as an the document type should be refined as “Article”, and thus
overall understanding of the reviewed topic about BIM-AI review papers are excluded from the prepared dataset. The
integration can be objectively created to help other research- subject area is limited to “Engineering”, “Computer Sci-
ers easily capture the status quo, identify popular research ence”, and “Environmental Science”. Fourth, the manual
interests, and trace future frontiers. screening process is necessary to further refine papers within
the research area. We quickly check the journal name, arti-
cle title, and highlight to remove irrelevant papers due to
2 Bibliometric Analysis query misinterpretation and ambiguity. As a result, a total of
2,766 scientific publications coupling three topics together
2.1 Data Acquisition are relevant to our review topics. which are extracted as our
targeted database for the bibliometric analysis. This pub-
In the stage of data acquisition, the following three criteria lication number is still limited accounting for only 0.21%
are adopted to search for the most relevant papers: (1) We of AI-related studies, 0.43% of construction-related studies,
select Scopus containing comprehensive scientific data and and 12.75% of BIM-related studies. Accordingly, substantial
literature as the practical database. It has been proven that research has concentrated on three separate topics about AI,
Scopus combines the advantages of both PubMed and Web construction, and BIM. Nevertheless, it can be concluded
of Science, which can provide 20% more publication cover- from the limited number of publications that efforts in apply-
age [24]. Moreover, Scopus acts as an ideal tool in searching ing AI to BIM for handling complex construction projects
inter-disciplinary research topics [25], and thus it is more are still at an initial stage. Since BIM-AI integration is envi-
proper for our review work on BIM and AI integration. sioned to play key roles in speeding up the digital develop-
(2) A set of keywords are determined to retrieve the most ment in the construction field, it appears to be an emerging
relevant previous literature. The search period is set from topic deserving more attention.
Jan 2000–Apr 2021, aiming to consider the development of Regarding the annual number of public studies, it
the targeted topic in the recent two decades. The keyword can be seen from Fig. 2 that interests in the individual
search strategy is tabulated in Table 1, where three major topic and the integrated topic all experience significant
concepts (including BIM, AI, and construction industry) are growth during the past 20 years. Specifically, the papers
taken into account and each concept is described by several for BIM exhibit an overall upward trend, while there is the
synonyms or similar phrases. The determination of keywords main burst of publications about AI and the construction

Table 1  Keyword search strategy and result for pieces of literature in the period of Jan 2000–Apr 2021
Concept Keyword and phrase Number of documents

AI AI OR “artificial intelligent” OR “artificial intelligence” OR “computational intelligent” OR 1,297,116

“computational intelligence” OR “data mining” OR “machine learning” OR “deep learning”
OR “neural network”
Construction industry “construction industry” OR “construction engineering” OR “construction management” OR 647,804
“architecture engineering” OR “structural engineering”
BIM BIM OR “building information modeling” OR “building information modelling” OR IFC OR 21,695
“industry foundation classes”
BIM and AI in con- Combine the keywords from the abovementioned three concepts by “AND” operator 2,766
struction industry

13
Integrating BIM and AI for Smart Construction Management: Current Status and Future Directions 1085

Fig. 2  Variation of publication 1.6x105

number per year during Jan 2000
1.4x105 AI

Number of publications per year

Number of publications per year

for AI and construction industry
2000–Apr 2021: a individual Construction industry
topic about AI, construction 1.2x105 BIM
industry, and BIM; b our review 1500
topic about the integration of 1.0x105

for BIM
BIM and AI in the construction 8.0x104 1000
industry
6.0x104
4.0x104 500

2.0x104
0
0.0

2000 2010 2020

Year
(a)
700
624
Integration of BIM and AI in construction industry
Number of publications per year

600 Trend line by SGompertz fit

95% confidence band of the trend line
500 454

400
333
286
300
223
200 175160
129
102
100 56 75
30 34
6 1 1 10 5 16 16 14 16
0

2000 2005 2010 2015 2020

Year
(b)

industry in 2016. Accordingly, the rising tendency of pub- in our reviewed topic about progressively integrating AI
lications concerning our reviewed topic roughly matches and BIM for pursuing intelligent civil engineering, and
the development of advanced AI and BIM techniques and thus a promising rise of related published works will keep
the ever-increasing attention in the construction indus- going.
try. The publication trend can be generally divided into
three stages. In the first 10 years, the annual growth was
relatively slow under an average of 12 publications every
year. From the year 2010 to 2015, this discipline field Table 2  The top 10 majority journals with the most publications
began to receive increased attention, and thus the aver- about BIM-AI integration in the construction industry
age publication number per year had risen to 95. Since
Journal Number of
2016, a pretty steep increase occurs, leading to a jump publications
in the publication number from 160 (the year 2016) to
624 (the year 2020). It is observed that the shape of the Automation in construction 340
curve in Fig. 2b generally follows the Gompertz function, Journal of Computing in Civil Engineering 96
where the growth is slowest at the beginning. Therefore, Advanced Engineering Informatics 91
such a type of function is fitted along with a 95% confi- Journal of Construction Engineering and Management 90
dence band to quantitatively capture the growing trend Engineering Construction and Architectural Manage- 79
ment
of relevant publications, resulting in a considerably high
Journal of Information Technology in Construction 47
­R 2 value of 0.849. It should be noted that the publica-
Sustainability 46
tion number has reached 286 in the first four months of
Applied Sciences 44
the year 2021 taking up around 45.8% of the previous
Journal of Cleaner Production 39
whole year, which even exceeds the annual number before
Buildings 38
2017. That is to say, there has been a surge of interest

13
1086 Y. Pan, L. Zhang

2.2 Journal, Country/Region, and Author Analysis That is to say, journals owning more related publications
have more opportunities to be cited. In short, the scope of
Most published papers about the reviewed topic come from all these top journals in Tables 2 and 3 can cover issues
the top 10 majority journals, as listed in Table 2. Among related to the integrated application of BIM and AI within
these journals, Automation in construction accounting for civil engineering, and thus they can serve as the best sources
12.3% of publications makes the greatest contribution, for researchers.
which is followed by Journal of Computing in Civil Engi- From the view of the country/region, Figs. 3 and 4 dis-
neering, Advanced Engineering Informatics. Besides, the close the distribution of publications. Evidently, the United
top 10 journals with the most direct citations are summa- States, China, United Kingdom, and Australia stand out to
rized in Table 3, which tend to exert more impacts on the be the four greatest contributors to this specific research
targeted topic. It is evident that 5 of the top 10 most cited field, which are all prominent in the number of publications
journals are listed in the top 10 majority journals providing and citations. Especially for the United States, its published
the greatest number of publications, which are Automation papers (622) and citations (11319) are far more than oth-
in Construction, Journal of Construction Engineering and ers. Hence, it can be assumed that the United States takes
Management, Advanced Engineering Informatics, Journal the leading position in exploring the BIM-AI integration for
of Computing in Civil Engineering, Journal of Cleaner smart construction. Besides, all the 9 top-ranked countries/
Production. Particularly, the first four journals under a total regions in Fig. 3 are the same, which are all more active in
citation of 3922 are also ranked at the top four in Table 2. conducting relevant research and play a greater influence.
The difference lies in the last country. Spain is ranked at the
10th place to provide the most publications, while Belgium
Table 3  The top 10 most cited journals about BIM-AI integration in owns more citations than Spain. To study cross-country col-
the construction industry laboration, an optimum co-authorship network is established
Journal Number of under the criterion that the minimum number of publica-
citations tions of a country/region is 20 and the minimum number of
citations of a country/region is 20. Therefore, 35 of the 99
Automation in Construction 1570
countries/regions meeting the threshold are incorporated in
Journal of Construction Engineering and Management 947
Fig. 4, where the bigger node donates the more influential
Advanced Engineering Informatics 752
the country/region, the thicker link represents the stronger
Journal of Computing in Civil Engineering 653
relationship between the two countries/regions, and the
Building and Environment 421
color stands for the average citations. There is a high level of
Procedia Engineering 352
cooperation among all these 35 selected countries/regions.
Energy and Buildings 308
In addition, these selected published papers can be ana-
Construction Management and Economics 290
lyzed from the author’s side. The top 8 most productive
Expert System with Applications 287
authors from Australia, Netherland, Hong Kong (China),
Journal of Cleaner Production 217
United States, Italy, and China are listed in Table 4, who

Fig. 3  Top 10 countries/regions 700

with the largest number of pub- 1 12000
lications and citations. (Note: 1 Publications
600 622 Citations
the number above each bar
stands for the top N ranking.) 10000
Number of publications

500
Number of citations

2
448 8000
400
3 6000
300
4
322
3 4 2
250
200 5 6 4000
5
9 6
197 194 7 8 8 7
9 145 142 141
100 10 2000
87 10
54
11319 4764 4797 5222 2125 3080 2555 2638 2935 903 959
0 0
a)
tes dom a Chin rea y
nited Sta Chinaited KingAustrali Italy CanadgaKong (outh KoGerman Spain Belgium
U Un Hon S
Country/Region

13
Integrating BIM and AI for Smart Construction Management: Current Status and Future Directions 1087

Fig. 4  Co-authorship network

based on the country/region

south korea canada

china
(china)

saudi ariabia united arab emirates

Average citaons

Table 4  The top 8 more productive authors

Author Country/ Institute Number of publi- Number of
Region cations citations

Xiangyu Wang Australia Curtin University 52 1513

Pieter Pauwels Netherland Eindhoven University of Technology 40 825
Heng Li Hong Kong (China) The Hong Kong Polytechnic University 38 655
Jun Wang Australia Deakin University 36 491
Martin Fischer United States Stanford University 31 530
Jack C.P. Cheng Hong Kong (China) The Hong Kong University of Science and Technology 26 659
Luigi Barazzetti Italy Politecnico di Milano 24 517
Hanbin Luo China Huazhong University of Science and Technology 22 489

have contributed over 10% of related publications in the of digitalization and intelligence in infrastructure manage-
latest two decades. Also, their papers are very influential ment during its entire lifecycle.
reaching a number of citations above 489, and thus they
can be regarded as the highly cited authors in the targeted 2.3 Keyword Analysis
research area. To be more specific, Pieter Pauwels puts a
special emphasis on smart architectural design based on A popular way to map the research hotspots and trends is
semantic web techniques. Luigi Barazzetti concentrates on drawing a network of co-occurrence keywords by feeding
computer vision and laser scanning for mapping and 3D the bibliographic data from Scopus into the VOS viewer
modeling. The rest of the authors have made great efforts tool. Under the full counting, a total of 14,748 keywords
in implementing BIM along with AI towards a high degree are retrieved from a body of the selected literature, which

13
1088 Y. Pan, L. Zhang

could unnecessarily increase the complexity of the network. correspond to the leading research interest. For example,
To attain a more readable and representative graph, we set the cluster 1 is centered around the advanced AI techniques.
the minimum frequency of keyword occurrence to be 50, The cluster 2 is associated with the purpose of BIM. The
and thus only 77 out of 14,748 keywords remain. The cre- cluster 3 is related to decision making. Keywords grouped
ated network for knowledge mapping is shown in Fig. 5, into the same cluster are more likely to occur at the same
which consists of 77 nodes and 2534 links under a total link time and belong to a similar research field. Additionally, the
strength of 23,913. As an explanation, such a network is identified high-frequency keywords in Fig. 5a are listed in
proven useful to intuitively uncover the knowledge structure Table 5 under descending order of frequency of occurrences.
behind the target research topic. A node stands for a specific For an easier understanding, Table 5 merges some keywords
keyword and its size is in direct proportion to the frequency with similar meanings into one record, where the total link
of the keyword. A pair of nodes stand for the co-occurrence strength measures the summation of a node’s link strength
of two keywords. The thickness of a line indicates the close- over other nodes connected to it. It can be seen that the key-
ness of the relationship between two nodes in terms of co- words that appear most frequently are basic concepts about
occurrence. That is to say, the two connected keywords will BIM and construction. Apart from these two keywords,
appear at the same time and their link strength indicates the the top keyword “architecture design” appears frequently
co-occurrence frequency. Herein, two informative ways of under the highest frequency of 975 and the largest total link
visualizing the keyword co-occurrence network are provided strength of 4928, revealing that an in-depth investigation has
in Fig. 5, which are discussed below. been conducted on the research focus of architecture design.
One is a static representation in Fig. 5a with a clustering Meanwhile, architecture design demonstrates a close inter-
view, where the entire network is automatically divided into relatedness with BIM, which can therefore be regarded as a
three clusters in the color of red (cluster 1), green (clus- major application domain for BIM-AI integration.
ter 2), and blue (cluster 3). The VOSviwer tool supports The other network is from the viewpoint of keyword
a network clustering method named the smart local mov- evolution over time, as given in Fig. 5b. The keywords are
ing algorithm to provide insights into the structure of the colored by the average publication year in which they had
established network, which has shown the advantages of been studied in papers. That is to say, nodes in the bluer
automatically discovering meaningful clusters with higher color represent keywords to be noted and discussed ear-
modularity values [26]. As a result, a set of closely related lier, while keywords marked in yellow color have recently
keywords can be reasonably gathered into the same clusters. captured the great attention of researchers in recent years.
Totally, 37, 23, and 17 keywords are assigned to the clus- Hence, Fig. 5b can visually demonstrate the development
ters 1–3, respectively. In particular, these extracted clusters trend of research hotspots. Based on cluster results and

Fig. 5  Keyword co-occurrence network about BIM-AI integration in the construction industry during Jan 2000–Apr 2021: a By clusters; b By
time

13
Integrating BIM and AI for Smart Construction Management: Current Status and Future Directions 1089

Table 5  Top keywords with the most frequent occurrence

Top Keyword Occurrence Total Cluster Top Keyword Occurrence Total Cluster
link link
strength strength

1 Building information modeling/bim 1281 6501 2 14 Buildings 192 991 3

2 Construction industry/construction/ 1117 5038 1 15 Ontology 190 853 2
construction projects
3 Architecture design 975 4928 2 16 Structural design 170 851 3
4 Information theory/information 624 3583 2 17 Sustainable development 167 769 3
management
5 Decision making/decision support 395 1907 3 18 Office buildings 139 849 2
systems
6 Project management 310 1437 1 19 Automation 139 644 1
7 Artificial intelligence 219 1048 1 20 Visualization 122 592 1
8 Life cycle 196 1102 3 21 Interoperability 113 567 2
9 Semantics 196 1012 2 22 Industry foundation 102 650 2
classes-ifcsystems

The cluster in this table is derived from Fig. 5a

temporal information in Fig. 5, it is interesting to find that provide references for the deeper investigation of advanced
the development of BIM-AI integration in construction can research of BIM-AI integration later. The value of Silhou-
be basically broken down into three main phases. It can be ette for each cluster is larger than 0.66, indicating the high
observed that the early research interests before 2017 are degree of homogeneity and meaningfulness embedded in
related to keywords in cluster 2. Researchers mainly regard these extracted clusters.
BIM as an information system for the basic task of archi-
tecture/structure design. During 2017–2018, more research
interests are placed on keywords in cluster 3, which high- 3 AI Potential in the Full Lifecycle of the BIM
light the major benefits of BIM-AI integration in applica- Project
tion value, such as cost analysis, decision making, energy
efficiency, optimization, scheduling, and sustainable devel- Currently, BIM is playing a leading role in revolutionizing
opment. From 2018 till now, the latest research focus has the construction industry, which incorporates technological,
shifted to keywords mainly in cluster 1, meaning that sus- agential, and managerial components to continuously digi-
tainable efforts continue to be dedicated to implementing talize the building representation process [27]. That is to say,
various AI methods for speeding up the digitalization of BIM can serve as a multifunctional set of instrumentalities
construction management in Industry 4.0. for specific purposes that are increasingly integrated [28].
To further uncover the cutting-edge development, a cita- Beyond the 3D model, BIM can offer a pool of information
tion burst analysis is performed to identify keywords that to support project management and exert substantial impacts
are most frequently cited in papers. As shown in Fig. 6, on aspects of economic, social, and environment. According
the keywords in the top three longest bursts are computer- to a survey [29], BIM brings project delivering benefits in
aided design, structural design, and information technol- the phase of planning and design, construction, and O&M
ogy, whose popularity has lasted for the first 15 years. Since accounting for approximately 55%, 35%, and 10% of BIM
2010, research content incorporating the concept of AI along adoption, respectively. In other words, BIM has gone far
with digital technology, like the algorithm, radio frequency beyond the design phase. Since BIM helps digitalize every
identification, data handling, natural language processing aspect of the building’s lifecycle, a lot of data will be accu-
system, point cloud, neural network, IoT, have obtained surg- mulated in a single phase. It is believed that the integration
ing citations. That is to say, the latest publications attempt of BIM and AI will be more and more common.
to seek digital solutions in construction engineering and Due to the nature of BIM in providing data-rich 3D
management largely based upon BIM-AI integration. To visualizations and solid data about projects, AI is an ideal
further retrieve the knowledge domain, we implement a sta- solution to be easily integrated with BIM to facilitate fast
tistic method named the log-likelihood ratio (LLR) to clus- information retrieval and analysis across the full lifecycle of
ter keywords in a more accurate manner. As summarized construction projects. It is worth noting that the great poten-
in Table 6, 8 co-citation clusters can be identified, which tial of AI is to offer smarter and more informative insights

13
1090 Y. Pan, L. Zhang

Fig. 6  Results of the top 25

keywords with the strongest
citation bursts

Table 6  Summary of the identified clusters by LLR

Cluster Silhouette Three representative keywords in each cluster Description

1 0.663 Machine learning, deep learning, convolutional neural network Various machine learning algorithms
2 0.724 Visualization, point cloud, laser scanner Computer vision
3 0.744 Computer-aided design, structure design, intelligent design Design phase management
4 0.693 Information technology, collaboration, decision Information technology and its purpose
5 0.680 Automated compliance checking, semantic system, natural language text Text data mining
6 0.667 Construction safety, risk management, safety Safety management
7 0.782 Radio frequency identification, internet of thing, smart technology Advanced technology
8 0.944 As-built documentation, facility management, energy utilization Operation & maintenance management

p value for clusters 1–7 is 1.0 × ­10–4, and p-value for cluster 8 is 1.0 × ­10–3

into the digitally built environment created by BIM, allow- the summary of AI potential in Fig. 7 is determined accord-
ing for intelligent exploration and optimization in different ing to two pieces of evidence. One is three major phases of
stages of the BIM-based project to reach higher project effi- BIM utilization, including planning/design, construction,
ciency, quality, collaboration, and sustainability. Moreover, and operation/maintenance. The other is three extracted

13
Integrating BIM and AI for Smart Construction Management: Current Status and Future Directions 1091

Fig. 7  Potential of BIM-AI BIM functionality:

integration in three phases BIM functionality:
 Fabrication
throughout a project’s lifecycle  Design
 Construction 5D
 Analysis
 Construction logistics
 Documentation
AI potential:
AI potential:
 Inter-discipline decision making
 Automated plan/design
 Risk prediction and responding
 Automated checking
measure development
 Close collaboration
 Better understanding of logistics
information and construction progress
 Smart robotics

BIM functionality: AI potential:

 Operation  Green and sustainable buildings
 Maintenance  Intelligent control of day-to-day services
 Renovation  Time-based preventive and corrective
maintenance

clusters in the keyword network of Fig. 5. Cluster 1 in red few attempts to leverage BIM and AI to support the
color indicates that AI is outstanding in its self-learning abil- automated design and drafting at different levels of
ity for construction safety management. Cluster 2 in green detail (LoD) [31]. LoD standing for the complexity of
color shows that AI is beneficial for intelligent design work. a 3D model has developed from LoD 100 to LoD500,
Cluster 3 in blue color presents that AI contributes to sus- and thus more building information in terms of orien-
tainable development. To sum up, the AI potentials in the tation, location, shape, size, quantity, and some non-
three project phases that are worthy of special attention are graphic information will be enriched in BIM [32]. An
briefly introduced as follows: issue in basic 3D modeling is that it is a little far away
from the actual project due to the lack of accurate pro-
(1) Planning and design phase: Before the start of physical ject plans and estimates. Many efforts have been made
construction, it is necessary to create detailed plans for to turn the concept of 3D BIM into 4D/5D BIM by
the project development concerning resources, sched- incorporating the additional dimension of schedule and
ule, budget, dependencies, and others. BIM can be cost, enabling better-planned and more cost-effective
introduced as a design tool to more efficiently formu- construction [33]. Meanwhile, it is also necessary to
late well-prepared plans and design schemas fitted to refer to rule inferencing and machine learning algo-
the desired client demand, time scale, and workflow, rithms for code compliance checking [34]. Therefore,
which is expected to reduce errors, cost, duration, and the problem about whether the established model meets
irrational processes in the practical project. BIM rely- the design standards and requirements can be automati-
ing on commercial software (i.e., Revit, Synchro, etc.) cally checked, which helps to lead to a high-quality
plays a vital role in transforming simple drawings to and efficient BIM-enabled design phase. Moreover, the
be digital models under the functionalities of visu- BIM-based collaborative design is another advantage
alization, navigation, and parametric modeling [30]. to be noted. To maximize the value of improved project
When the AI techniques are integrated into the BIM delivery and efficiency from collaboration, implemen-
software, the AI potential lies in automating both the tation of AI has potential to further boost the co-design
design/review process and then reducing dependency practice through exchanging design information in the
on expert knowledge and judgment. On one hand, AI standard data format among a group of participants.
provides opportunities to realize the automatic plan and (2) Construction phase: This is a phase of executing
design through generative design, which is expected physical construction. BIM helps to create a solid link
to investigate possible permutations of solutions and between the design and construction, aiming to carry
then produce design alternatives. On the other hand, out the plan made in the previous phase with fewer
AI is capable of automatically checking and analyzing unnecessary reworks, conflicts, and errors on site [35].
the detailed 3D model/animation with semantic infor- The great potential of AI is highlighted in the following
mation, which can eventually offer a comprehensive three points. Firstly, with the help of machine learning
overview of the project for easier understanding and algorithms, rich data about the ongoing project from
modification. To be more specific, there have been a BIM can be continuously explored by various machine

13
1092 Y. Pan, L. Zhang

learning algorithms to predict and prioritize high-risk tion pertaining to the O&M phase into as-built models
issues, and then prevention measures can be taken at [41]. A generic BIM-based framework encompassing
an early stage to ensure construction safety. Various the BIM model, relational database, and cloud com-
techniques of IoT, such as unmanned aerial vehicles puting can be established to maximize the value of the
(UAVs), augmented reality (AR), location tracking, collected time-series data by the full exploration and
and others, can also be combined with BIM to improve BIM model updating towards the goal of green and sus-
identification and analysis of potential field safety risks tainable buildings. In other words, the AI application
and enhance the real-time communication in managers in the O&M phase holds the opportunities to visual-
and workers. That is to say, such integration of BIM ize various aspects of the facility and comprehensively
and AI contributes to automated construction monitor- analyze the building's operational performance, which
ing and control, which is experiencing fast growth in helps to automatically recognize and predict the risk
site safety management to proactively address poten- of error or failure before its occurrence. Accordingly,
tial issues and prevent casualties. It has been proven to a wide range of O&M activities, like maintenance and
overcome limitations of manual safety checking, such repair, emergency management, energy management,
as low accuracy, discontinuity, inefficiency, and labor- and others, can be conducted in a data-driven manner
intensive [36]. Secondly, the 4D BIM simulations of to further embrace the benefits of BIM [12]. As a result,
construction schedules and activities are applicable the day-to-day services can be controlled in an efficient,
to handle construction logistics for the improvement economical, and reliable manner. Time-based preven-
of construction productivity. More specifically, the tive maintenance detects the potential risks and adjusts
adoption of BIM and AI can provide a better under- the ongoing operation prior to unexpected events. Cor-
standing of logistics information, detection of con- rective maintenance implemented after the occurrence
flicts, supervision of construction progress and supply of problems strives to repair the problematic parts and
chains, and coordination of different activities [37, 38]. get them back to the normal status as quickly as pos-
As expected, the inefficiencies and mistakes that slow sible.
things down can be eliminated, and thus the construc-
tion process will run more smoothly [39]. Thirdly,
smart robotics equipped with different knowledge can 4 Advanced Research Interests of BIM‑AI
be developed to replace humans for taking semi-auto- Integration
matic or full-automatic construction activities. They are
beneficial in decreasing worker exposure to hazardous According to LLR-based keyword clustering in Table 6, we
environments, reducing labor costs, and streamlining will focus on six advanced research interests to demonstrate
the construction process, which can further guarantee the effectiveness of BIM-AI integration in various aspects
the safety, efficiency, and sustainability of construction. of the civil engineering realm. For a clear understanding,
(3) O&M phase: When the construction is completed, the Fig. 9 displays the logic for the determination of targeted
project will enter a new phase called O&M to operate research interests, where the goal of each research inter-
and maintain a constructed facility to not only meet the est along with the most relevant clusters of keywords are
anticipated functions over its lifecycle but also ensure highlighted. In other words, a total of six emerging research
the safety and comfort of users. It is known that O&M topics regarding the BIM-AI integration are selected on a
takes the most of the time within the lifecycle, lead- basis of the captured 8 clusters (see Table 6) from Sect. 2.
ing to a large amount of cost accounting for around Bibliometric analysis. Those research topics include BIM
60% of the total project budget [40]. However, BIM automated design and rule checking, 3D as-built reconstruc-
applications for effectively operating and maintain- tion, event log mining, building performance analysis, VR/
ing facilities are still insufficient. It is well known that AR, and digital twin. Also, key points of the six determined
traditional decision-making in facility management is research interests have been summarized in Fig. 8. It should
largely based on facility managers, but it is error-prone be noted that the application range of all six topics covers
and costly. To tackle this issue, AI has great poten- the full lifecycle of a BIM-based project, including the phase
tial to provide an efficient solution for informing more of design, construction, and O&M. All the representative
reliable decisions. For example, the joint deployment pieces of literature discussed in this section and listed in
of BIM, smart IoT devices, and data mining methods the tables have high citations over 100 or they are published
under the existing BIM Application Programming in the last five years. Therefore, these selected papers can
Interface (API) has been proven useful in integrating be highly representative in BIM-AI integration, which can
the standardized information inherited from the design offer a valuable reference for digital transformation in the
and construction phase along with additional informa- construction industry.

13
Integrating BIM and AI for Smart Construction Management: Current Status and Future Directions 1093

Design phase Construction/O&M phase Design phase

1. Automated design and 2. 3D reconstruction 3. Event log mining

rule checking • To be a subfield of computer
• To ensure the practicability of • To fully explore a type of BIM
vison
BIM in intelligent and high- data called event logs
• To automatically convert 3D
quality design • To extract sequential patterns
point clouds into semantically
• To automatically create and of design command and
rich building models
analyze design and planning make predictions
• To directly learn useful
• To realize the automated rule • To study BIM collaborative
features from point clouds for
checking and compliance design from a point of social
tasks of segmentation and
reasoning network analysis
classification

Machine Computer Information Compliance Construction Radio frequency As built

learning Visualization aided design checking safety identification documentation
technology
Deep learning Point cloud Structural Semantic Risk Facility
design Collaboration system management Internet of thing management
Convolutional Laser scanner Intelligence Natural Smart Energy
neural network design Decision language text Safety technology utilization

4. Building energy 5. VR/AR 6. Digital twin

performance analysis • To link VR/AR into BIM • To form a cyber-physical
• To facilitate the green BIM environment for better project system with two ways of
practice understanding dynamic mapping
• To perform dynamic energy • To help review design • To transfer inspection data to
simulation under uncertainty content at an early stage and the virtual part in real time
• To guide the design and improve communication • To perform simulation,
operation energy-efficient among project participants prediction, and optimization
buildings by clustering, • To reduce risk during in the virtual part for data-
prediction, and optimization construction process driven decision making

Design/O&M phase Design/Construction phase Design/Construction/O&M phase

Fig. 8  Logic diagram about the selection of six advanced research interests

4.1 BIM Automated Design and Rule Checking on intuitive knowledge could be time-consuming, labori-
ous, and error-prone. Meanwhile, the modeling process
It is known that one of the most significant advancements exactly follows the plan, which is unable to automatically
provided by BIM is to extend and enhance the computer- adjust to the ever-changing conditions. Thus, with the help
aided design. Far more than the simple use of creating a of data-driven approaches, it is desired to develop an intel-
building model, BIM acting as a shared knowledge center ligent modeling framework with a higher degree of automa-
provides mass information at different levels of detail to tion and flexibility. For another, design models under a high
enrich the model. Therefore, it is meaningful to properly degree of inherent complexity need to comply with sets of
leverage the flow of design-related information incorporat- well-defined building regulations, codes, and standards as
ing geometric and semantic aspects, which can potentially well as project requirements. The current model checking
facilitate the automated creation and checking of drawings. process largely depends on manual inspection in the nature
It is worth noting that the necessity of developing BIM- of unreliability and subjectivity, which has become a costly
AI integration in the design phase comes from two main bottleneck causing undesired delays and errors during the
parts. For one thing, the manual modeling process relying project execution. To ameliorate the problem, it is essential

13
1094 Y. Pan, L. Zhang

to develop the automated rule-based compliance checking models under the accurate estimation of required facility size
approaches to perform the design assessment of a building and space availability, and then an optimization algorithm
model, which is expected to add additional value to the BIM was implemented to lower the total cost and travel distance.
practical applications. The research question can be raised It has been studied that BIM automated design is beneficial
as: how to carry out automated design and rule checking for both the architecture design and structure design tasks.
with less manual effort and high efficiency. The advantage of There are great opportunities to speed up the design proce-
this topic is to ensure the practicability of BIM level 3 with dure and eliminate errors from manual planning.
intelligent and high-quality alternative design generation and After model generation, how to automatically detect
error elimination. the potential design flaws hidden in the model with fewer
Some existing studies have attempted to answer the ques- manual interventions is another topic of interest. To check
tion of how to automate and optimize the design task. As whether a model fulfills a wide spectrum of requirements
listed in Table 7, the object of BIM-based automated design stipulated by design standards and codes, a critical step lies
is not limited to the architecture components, which can be in the rule interpretation. However, evaluation of a model
extended to the structure of a building and the construction against design codes and guidelines is not a straightforward
site. For example, from the architectural aspect, Liu et al. task due to the challenges from the inherent complexity of
[31] integrated the rule-based algorithm in consideration of building design rules and the incapability of computers in
trades' know-how with an optimization algorithm to realize directly understanding the natural language-format rule and
the automated design and planning for boarding layout of efficiently capturing key information from disorganized doc-
light-frame buildings, and thus various design alternatives uments. Currently, most checking tasks are still performed
could be assessed continuously to generate an optimal board- manually, leading to costly, cumbersome, and error-prone
ing design under the minimized material waste. Deng et al. procedures [53]. To fill the practical and technical gap, some
[42] developed a BIM-enabled framework to automatically studies concerning the automation of rule-checking proce-
integrate graphic and non-graphic drawing information and dures have been performed, as outlined in Table 7. For exam-
generate readability fabrication drawings for façade panels, ple, Zhang et al. [49] combined the NLP techniques with the
which was proven useful in better representing the curved semantic analysis and logic-based reasoning to facilitate the
panels using a new flattening algorithm. From the structural automated information processing and compliance reason-
aspect, Hamidavi et al. [43] raised a workable BIM-based ing, contributing to checking the compliance of BIM-based
solution to realize the automated creation and analysis of building designs with building codes in an automated man-
optimized structural design meeting a set of criteria, allow- ner. Ghannad et al. [50] deployed an emerging open-standard
ing for building stronger ties with disciplines of architects schema named LegalRuleOn one hand (LROn one hand) to
and structural engineers and greatly reducing design time. formalize the design rule sets with rich semantics, ontolo-
From the aspect of the construction site layout planning. gies, and parameters and then applied the Visual Program-
Kumar et al. [44] presented an automated framework to ming Language (VPL) under a graphical notation to capture
exploit information from BIM for building dynamic layout required information for the rule implementation, aiming

Table 7  Some previous studies about BIM automated design and rule checking
Purpose Method Achievement Related reference

Automatic architecture design Rule-based design algorithm; It can provide automated design support [42, 45, 46]
Evolutionary algorithm for boarding, façade, floor tile, and
others, and make optimization accord-
ing to the customized and changeable
requirements.
Automatic structure design Mathematical functions It can automatically generate optimized [43]
structural design based on the architec-
tural model and enhance collaboration
among team members.
Automatic site layout planning Metaheuristic algorithms It can facilitate more rational arrange- [44, 47, 48]
ments for a higher degree of efficiency
and safety in on-site operations.
Automatic code compliance checking Natural Language Processing (NLP); It can reduce manual efforts in extracting [49, 50, –52]
Semantic enrichment; Unsupervised regulatory information from documents
learning; Visual programming language to form logic rules and then automati-
cally provide reliable checking results at
the early stage of design.

13
Integrating BIM and AI for Smart Construction Management: Current Status and Future Directions 1095

to execute the rule-checking process against a BIM model or multi-view imagery. These acquired point clouds serve as
automatically and assess design quality iteratively. What’s evidence to locate and identify various objects in complex
more, Sydora et al. [51] proposed an integrated framework and irregular shapes. Based upon the point clouds, there
to bridge the task of automated BIM model checking and are three typical tasks to be implemented for the creation
automated interior design, where a more user-friendly rule of as-built models, including (i) to model the component
language based upon the BIM standard Industry Foundation geometry, (ii) to assign categories and materials, and (iii)
Classes (IFC) was developed to express rules under logic to build interrelationships among components. However,
operations for examining the model compliance and generat- these collected point clouds show the characteristics of ran-
ing several alternative interior designs. Accordingly, Natural domness, sparsity, and non-structure, which will inevitably
Language Processing (NLP) and semantic web technology suffer from a lot of noise and occlusion [55]. The current
are two new solutions to improve the flexibility of informa- solution for noise removal is largely dependent on manual
tion extraction and the reusability of regulatory represen- procedures. Meanwhile, expert modelers often subjectively
tations. Both of them can effectively assist in decreasing explain the point clouds and then manually build the model
the cost and time in encoding rules for the automated rule and enrich its semantics. Such manual methods could be
checking process. More specifically, NLP is responsible inefficient and unreliable. Moreover, the commonly used
for translating human-readable rules into the computer- BIM software, like Autodesk Revit, Synchro, Tekla, etc.,
processable rule format, while semantic web language, like is incapable of managing point clouds and linking them to
resource description framework (RDF), web ontology lan- the BIM model. To fill the research gap, a critical problem
guage (OWL), and others, is used to describe the complex to be addressed is how to make full use of rich data sources
knowledge of schema, instance, and relevant rules through from the raw point clouds to reproduce the 3D parametric
creating a graph, and thus extracted information from regula- as-built model automatically and accurately. Since computer
tory documents can be expressed in a logic-based manner. vision falls into an interdisciplinary scientific field to gain
For an intuitive understanding, Fig. 9 provides a technical a detailed understanding of visual data by computers, the
framework based on semantic web to perform knowledge ongoing research about BIM and 3D reconstruction can be
extraction and discovery from a BIM-based data platform. treated as a subfield of computer vision to drive the Scan-to-
BIM process to be more automatic and accurate [56].
4.2 BIM and 3D As‑Built Reconstruction Semantic segmentation is an important step in under-
standing 3D scenes, aiming to classify similar point clouds
Apart from the as-designed BIM, another type called the into homogeneous regions [57]. But point cloud segmenta-
as-built BIM acts as a semantically enriched digital model tion requires prior rules about the semantic labels, and in the
to truly reflect the as-built conditions of a facility, where the meantime, it is insufficient for irregular-shaped objects and
details cannot match the available designed models. It should complicated scenes. To address these weaknesses, recent
be noted that laser scanning and photogrammetry have been researchers have endeavored to develop numerically robust
extensively adopted to make a survey of existing facilities. and fast algorithms including mathematical models, machine
These popular yet inexpensive measurement technologies learning, and deep learning approaches, as listed in Table 8.
have given easy access to a bunch of point-cloud data con- To be more specific, the mathematical model is the early
taining metric information through distance measurability stage for reconstructing semantically rich BIM from 3D

BIM-based semantic web

contains domain models,
Model users are organized
taxonomies, classifications,
linking across domain
and dictionaries
Domain resource Domain standard ontology.
Data integration Domain ontology
(Infra/GIS/Energy) (bsDD/CityGML)

BIM-based BIM ontology

Knowledge sets Knowledge blocks Knowledge graphs Model uses
semantic web (ifcOWL)

Compliance
Logic information Ontology rules Reasoner engine
checking
BIM-based semantic web To automatically generate
relates to logical reasoning. the proofs for what is
inferred from model uses

Fig. 9  Framework of knowledge extraction and discovery in BIM [54]

13
1096 Y. Pan, L. Zhang

Table 8  Some previous studies about BIM and 3D as-built reconstruction

Purpose Method Achievement Related reference

Automatic interpretation and reconstruction of Mathematical model It can convert point clouds into BIM models with [58–60]
BIM models from point cloud data accurate geometries and
rich semantics, and thus it can handle complex
scenes with irregular-shaped and repetitive
objects.
Machine learning It can cluster and classify segments to reconstruct [61]
BIM model objects and their topology.
Deep learning It can find a direct mapping from point cloud to [62, 63, –65]
object labels under less feature engineering,
which helps to achieve better performance in
point cloud segmentation and classification in
complex 3D scenes.

point clouds. For example, Barazzetti et al. [58] presented work to accelerate the development of point cloud segmen-
a semi-automated method based on a mathematical func- tation, which can directly operate on the raw point cloud to
tion named NURBS (Non-Uniform Rational B-Splines), derive high-quality features. After that, a lot of studies for
which could generate a set of NURBS surfaces as a recon- automatic as-built BIM creation are developed based upon
struction of external object surfaces along with their geo- the improvement of PointNet. For instance, Chen et al. [62]
metric anomalies from point clouds. Then, Jung et al. [59] performed a modified PointNet architecture to understand
turned the as-built modeling into a fully automated process the indoor point cloud scene following the combined tasks
through room segmentation from point-cloud inputs, floor- of edge-based segmentation, classification with context,
wall boundary extraction under the offset-space noise fil- and component merging for building element recognition.
tering, floor-wall boundary refinement by linear regulation Ma et al. [63] verified the viability of two state-of-the-art
under the constrained least-squares method, the inside- and algorithms named PointNet and Dynamic Graph Convolu-
outside-wall component modeling. They focused on mod- tional Neural Network (DGCNN) in semantic segmentation
eling the building interiors containing multiple rooms rather of point clouds of building interiors, which could augment
than just the surface representation, and thus the 3D output the pre-existing 3D BIM for as-built modeling. Xu et al. [64]
model contained volumetric wall components with detail- developed a new 3D object detection network named Cor-
rich objects, such as windows and doors. Instead of semantic Det to concurrently extract the object-level and corner-level
segmentation, Xue et al. [60] proposed a novel semantic features using deformable convolutions and represented
registration method for semantically rich BIM reconstruc- BIM objects of indoor building primitives by class-specific,
tion, bringing two opportunities in enhancing the effective- oriented, and symmetric 3D bounding boxes, whose over-
ness and efficiency of modeling performance. One is that all performance was proven to surpass the segmentation
multimodal optimization (MMO) algorithms for finding all approaches. For an intuitive understanding, Fig. 10 pro-
optimal solutions were carried out to handle the complex vides an example of the deep learning-based reconstruction
scenes incorporating irregulated and repetitive objects; the results. Since deep learning is superior in learning high-level
other is that the architectural design knowledge hidden in features from point data itself, it has paved a more appropri-
point clouds, like symmetry, parallelism, repetition, etc., was ate way to semantically segment point clouds for reaching
fully utilized to eliminate noise and limit search space. In great efficiency and satisfactory results. That is to say, deep
general, the data-driven algorithms mentioned above can learning-enabled approaches are more powerful and robust
well process the inherently noisy, sparse, and unstructured for the automatic process of creating detail-rich models from
point cloud data, which benefit in driving the automated the point cloud data, which can meet the demand for efficient
process of converting dense point clouds into 3D volumet- and accurate as-built BIM. Overall, it has been seen that the
ric models. However, these methods too heavily rely on the increasing automated 3D reconstruction has gradually taken
combinations of hand-crafted features, and they may lack the the place of laborious operations in the scan-to-BIM process.
capability of extracting neighboring information of points.
Currently, more special attention has centered on the use 4.3 BIM and Event Log Mining
of deep learning to directly learn predictive features from
point clouds, which is particularly useful for large-scale During the BIM-enabled design process, an important type
datasets with no need for feature preprocessing. In particu- of data called BIM event logs can be automatically generated
lar, the PointNet proposed in 2017 [66] is the pioneering by BIM modeling software, such as Autodesk Revit, Tekla,

13
Integrating BIM and AI for Smart Construction Management: Current Status and Future Directions 1097

Original point
cloud data

UVAs with LIDAR

Segmentation
result

Object
recognition PointNet
result

Mesh
reconstruction
result

PointNet++

Fig. 10  Example of automated scan and 3D reconstruction [62]

and others, which will be heavily accumulated into a large crucial purposes. Its practical value is to inform data-driven
volume to bring some characteristics of “big data”. Notably, decision-making towards better management of the design
event logs in the format of pain text are similar to the web phase. The basic framework of BIM event log mining has
server logs. It is known that web logs can be interpreted as been displayed in Fig. 11. There are mainly three types of
the detailed and objective recording of all operations and functions in the event log mining, including design perfor-
events in an information system in a chronological order, mance evaluation, design activity prediction, and social net-
which has been developed to maturity for mining web usage, work analysis, as listed in Table 9.
content, and structure. That is to say, valuable information The first main task is to extract sequential patterns of
derived from web log mining has been adopted to develop design command, which is helpful in assessing the design
web content personalization and recommendation systems performance. Yarmohammadi et al. [67] firstly conducted
for web intelligence and greater surfing experiences. In the sequential pattern mining to discover the presence of com-
same way, the BIM event log is made up of process-specific mon command sequences by Generalized Suffix Trees (GST)
sequences related to the modeling activities, including cases, and then estimated the average time required to finish these
persons, time stamps, and others, which is the value-added patterns for characterizing the designer’s performance and
data to track the executed procedure that has occurred in determining the difference among designers. Zhang et al.
the entire design session. However, BIM itself is incapa- [68] retrieved the most frequent patterns of command
ble of manipulating and exploring these rich event logs for sequences using a process mining algorithm named the cus-
knowledge discovery. In this regard, the research question is tomized PATRICIA, which could serve as the baseline to
how to implement proper data mining algorithms to reveal measure design productivity. Apart from the statistical meth-
meaningful insights into the model evolution embedded in ods, Pan et al. [69, 70] referred to the novel clustering algo-
these logs. To narrow the gap between BIM event logs and rithm under the integration of Kohonen clustering network
data science, existing studies have leveraged different data (KCN) and fuzzy C-means (FCM) to obtain design behavio-
mining approaches to investigate the ever-increasing avail- ral patterns, which could easily distinguish design productiv-
ability of BIM event logs. It has been found that BIM event ity at different time periods into the high, medium, and low
log mining is a comparatively new research topic, which has level and group designers exhibiting similar efficiency into
the potential to return promising results to better understand the same cluster. In short, these high-quality extracted pat-
and improve the overall design process for the following terns can bring two great benefits in practical applications.

13
1098 Y. Pan, L. Zhang

Tom 0022 2013-03-05 16:19:30 60.28 aboujaoudei.rvt 108 Level 02 north Create A line
\\Projects\185118.000_KSU_ENG_Phase4\DESIGN\BIM\REVIT\MODELS\KSU ENGG_ARCH_INTERIOR.rvt

Tom 0022 2013-03-05 16:20:30 18.97 aboujaoudei.rvt 108 Level 02 north Other Command "AccelKey"
\\Projects\185118.000_KSU_ENG_Phase4\DESIGN\BIM\REVIT\MODELS\KSU ENGG_ARCH_INTERIOR.rvt

Tom 0022 2013-03-05 16:20:55 12.477 aboujaoudei.rvt 108 Building section corridor Delete Basic wall
\\Projects\185118.000_KSU_ENG_Phase4\DESIGN\BIM\REVIT\MODELS\KSU ENGG_ARCH_INTERIOR.rvt

Example

Data
Revit BIM event
Parsing
Event log data
Journal File logs
(CSV) preparation

Evaluation of design Prediction of design Social network

performance commands analysis
Predicted Value True Value

Cluster 1 Cluster 2 11

Cluster 3 Center 10

Log mining
9
Command Sequence

8
0
7

0.0 0.2 0.4 0.6 0.8 1.0

Probability

Fig. 11  Framework of BIM event log mining

Table 9  Some previous studies about BIM and event log mining
Purpose Method Achievement Related reference

Design performance evaluation Process mining; It can extract patterns of sequential commands to objectively evaluate [67, 68, –70]
Machine learning the design behavior of different designers and rationally arrange the
work.
Design activity prediction Deep learning It can accurately predict the next possible design command, which [71, 72]
generates data-driven command recommendations according to
designers’ preferences.
Social network analysis Statistical method; It can investigate characteristics of complex collaboration during the [40, 73]
Deep learning design process from the view of social networks for strategically
arranging work.

On one hand, the patterns reveal and highlight the important predictions about the next possible design command that
characteristics of the design behavior of different designers, could be adaptable to the changing conditions. As a result,
contributing to measuring and comparing designers’ produc- such a data-driven command recommendation can be pro-
tivity in an objective manner. On the other hand, they enable vided to potentially guide the design phase and reduce the
project managers to strategically arrange personalized and uncertainty and subjectivity in the modeling process. By
rational work, allowing for satisfying the demands of the simply following the recommended commands, it is pos-
project and enhancing the modeling efficiency. sible for designers to speed up their work and avoid some
Another point to address the concern about maximizing unnecessary mistakes.
the strength of the growing amount of BIM event log data is Moreover, since a group of designers will participate to
the deep learning-based prediction of sequential design com- ensure the success of the BIM collaborative design process,
mands. Pan et al. [71, 72] carried out the recurrent neural some efforts have emphasized social network analysis (SNA)
network (RNN) and long short-term memory neural network to monitor and evaluate the complex cooperation. Zhang
(LSTM NN) to learn sequences of the designer’s behavior et al. [40] built a social network using implicit process and
from the abundant BIM event log, resulting in reliable inter-organizational information extracted from BIM event

13
Integrating BIM and AI for Smart Construction Management: Current Status and Future Directions 1099

logs to graphically describe the information exchange and the entire project lifecycle. As expected, this topic can bring
behavior interaction among designers, aiming to achieve a benefits in various aspects of sustainability, such as (i) to
better understanding of designers’ performance and roles improve the building performance in the early design stage,
in collaborative organizations through analysis of network (ii) to reduce waste at the construction stage, and (iii) to
structural characteristics. Except for the commonly-used monitor the sustainability performance of building at the
centrality metrics, Pan et al. [73] developed a novel com- operational stage [9]. Notably, embodied phase and opera-
munity detection algorithm combining a graph embedding tional phase will take around 10–20% and 80–90% of energy
method named node2vec and a clustering method named demand in the entire project cycle, which require more focus
Gaussian mixture model to extract the potential clusters of for better energy control [75]. It has been found that the
designers who work more closely. For one thing, the SNA BIM-enabled energy analysis process is easy to follow for
results provide objective insights into the collaboration pat- facilitating sustainable development, and thus a participant
terns, individual roles, and the possible work transmission with no professional knowledge and simulation skills can
for a better understanding the complex execution of BIM perform it to formulate some environmental-friendly deci-
projects. For another, SNA offers a valuable opportunity to sions [76]. With the advent of various AI techniques, BIM
guide project managers in strategically arranging work, aim- for building energy performance analysis has played critical
ing to realize the goal of maximizing cooperation chances roles in two perspectives as shown in Table 10, resulting
and ensuring the project success. in data-driven energy-efficient design described in Fig. 12.
To sum up, BIM event log mining based on varying The first important task is the building energy simula-
data analysis approaches, including the statistical methods, tion, which is not straightforward due to a lot of uncertainty
machine learning algorithms, and deep learning algorithms, in the energy modeling process. A promising and popular
paves a new way to understand the complicated design phase solution is to integrate different kinds of probabilistic and
from a data layer, contributing to boosting the BIM-enabled statistical approaches into energy simulation, aiming to
design for higher design efficiency and quality. The practi- capture the dynamic behavior under uncertainty and under-
cal value to be highlighted is that BIM event log mining can stand the building performance objectively [85]. Chong et al.
serve as a data-driven tool, which helps project managers [77] developed the continuous-time Bayesian calibration of
define staffing strategy more objectively and helps designers energy models using data from BIM and building energy
to do the modeling process more smoothly. management systems, which could dynamically update the
established model and maintain high accuracy in predicting
4.4 Green BIM and Building Performance Analysis the monthly electrical energy consumption. However, this
kind of method could be computationally expensive in some
Since sustainability turns out to be a key factor in the con- cases. As an alternative approach, some studies have referred
struction project, green BIM practice under the integra- to machine learning models instead of physical simulation
tion of BIM and building energy performance analysis has to effectively predict energy demand in a data-driven man-
become a hot topic for green building development. The ner. For example, Singh et al. [78] implemented the proba-
core concept of green BIM is to conduct a model-based bilistic prediction under BIM-integrated machine learning
process of generating and managing coordinated and con- algorithms to estimate heating and cooling loads, and then
sistent building data towards the goal of sustainability [2]. statistical analysis was performed to determine the optimal
According to a report created by McGraw-Hill Construction design options. Geyer et al. [79] proposed a component-
Company [74], although green BIM practice is becoming based machine learning approach for predicting the thermal
an emerging trend, there remains a research gap between energy performance of buildings under high accuracy, which
the full potential of BIM and the sustainable outcome. The provided design-supporting insights to interpret and adjust
research question that has aroused is how to use BIM to sup- parameterized components of building for reaching desired
port the development of energy-efficient buildings during performance. Hammad [80] trained a deep neural network

Table 10  Some previous studies about green BIM and building performance analysis
Purpose Method Achievement Related reference

Building energy simulation Statistical It can simulate and predict the building energy performance, which can [77, 78, –80]
method; provide instant feedback to managers.
Machine learn-
ing
Sustainable design development Machin learning It can generate optimal design strategies for energy-efficient and envi- [81, 82, –84]
ronmentally conscious design, achieving more sustainable outcomes.

13
1100 Y. Pan, L. Zhang

BIM model Evaluation system BIM-based energy

development construction simulation

(ć)
2500

2000

CO2 (kg)
1500

1000

500

Optimal strategy Data-driven energy Machine learning-based

determination and control optimization energy prediction

Fig. 12  Framework of data-driven green building design

to precisely predict the expected energy consumption of a maximize indoor thermal comfort. As a result, sustainable
given building in BIM during the operation phase, which decisions about the passive design strategies and necessary
could provide evidence to make operational measures. It design adjustments can be reasonably informed to improve
has been found that the machine learning-based method can design accuracy and efficiency, contributing to promoting
show the advantage in promising accurate energy prediction the reduction of energy consumption and environmental con-
under decreased computational complexity. Also, it is more tamination at an early stage. In addition, due to the avail-
applicable to large-scale construction projects. Besides, in ability of accumulated BIM data and the powerful analytical
order to make these efficient and precise estimation tools for ability of AI techniques, it is suggested that more advanced
building energy more practical, it is expected to build them and flexible data mining approaches need to further be incor-
as a plugin in BIM software. The developed plugin helps to porated into BIM to learn from energy-related data to reveal
directly demonstrate the predicted results and give instant more comprehensive views of building energy consumption.
feedback to managers. These discovered pieces of knowledge are expected to offer
Another key task is to develop an intelligent framework more opportunities to establish a feedback loop for improv-
for desirable energy design from the BIM data layer, which ing design and facility management in a BIM environment.
can be realized by various data mining approaches, like clus- Therefore, optimal strategies leading to less energy utili-
tering, sensitivity analysis, and optimization methods. For zation in the building lifecycle can be quickly generated,
example, Peng et al. [81] developed a hybrid BIM-based which will reasonably guide the process to design and oper-
decision-making approach for extracting latent patterns in ate energy-efficient buildings.
BIM records, resulting in useful management advice for
the improvement of resource usage and maintenance effi- 4.5 BIM and VR/AR
ciency in the O&M stage of buildings. Tushar et al. [82]
quantitatively evaluated the impact of design variables on Virtual and augmented reality (VR/AR) are two emerging
the building thermal performance to find out the optimal digital tech developments to bridge the physical world and
solutions of suitable input design parameters as the guide- the cyber world. Acting as a computer-generated simulation,
line of BIM-based design, which was proven useful to not VR builds a synthetic environment to completely take the
only assure energy comfort to residents but also decrease place of users’ field vision, aiming to trigger interactions
energy costs and emissions during operation. Chen et al. [83] between the simulated world and users. Differently, AR is a
connected the NSGA-II algorithm with the fitness function mixture of computer-generated elements and the real world
from the LSSVM-based energy prediction, which can obtain relying on common equipment, like smartphones, tablets,
the optimal combination of building envelope parameters to headsets, and others, which shows the information on the
simultaneously minimize building energy consumption and real-world scene to be less immersive than VR. Recent years

13
Integrating BIM and AI for Smart Construction Management: Current Status and Future Directions 1101

have witnessed the surging trend of VR/AR applications in representation of the BIM model while trying to reserve the
various domains, including entertainment, retail, health- high consistency of the overall shape. In short, the real-time
care, education, etc., which benefit from their strengths of and automated data exchange between BIM model and VR
3D visualization, immersive and interactive environments, platform is typically composed of three parts, including BIM
personalized content, cognitive enhancement, and others. data, metadata transfer, and game engine. This is the work
Since there is a lot of geometric and semantic information basis for BIM and VR/AR integration, which needs to be
about the 3D model hidden in BIM to support virtual simu- written into commercial vendors for practicability.
lation and information interoperability, it seems reasonable Due to the nature of BIM and VR/AR, their joint usage
to link VR/AR techniques with the BIM environment [86]. is becoming a promising development to exert beneficial
Therefore, the research question focuses on the integration impacts in both the phases of design and construction, as
of BIM and VR/AR for enhancing project understanding and outlined in Table 11. For one thing, it paves a novel way of
promoting better decision making during the entire lifecycle reviewing the design schema in a realistic way from an early
of the building facility, which can eventually make break- stage, enabling engineers and managers to conduct intuitive
throughs in the construction industry. The general workflow manipulation and interaction with large data. Based on the
of BIM and AR/VR technology is summarized in Fig. 13. immersive experience in 3D architecture design, participants
Remarkably, since the heavy workload of the manual data even in lack of specialized knowledge can gain a clear under-
processing and the delay of data transmission will, unfortu- standing of the model and easily detect design errors and
nately, impede the adoption of VR/AR, a principal issue to weaknesses [90]. For example, Shiratuddin and Thabe [91]
be solved is to formulate a reliable mechanism for transfer- developed a virtual design review system for early and accu-
ring the established BIM model along with relevant data into rate 3D visualization of design within a federated BIM. Gar-
VR/AR engines in a fast and automated manner. That is to bett et al. [92] created a multi-user BIM-AR system based
say, there is a pressing need to facilitate the automatic and on an Agile Scrum software development approach to reuse
simultaneous data exchange in BIM and VR/AR. The ideal BIM data for achieving synchronous collaboration in an AR
conversion process is expected to perform real-time data environment, which could provide a unique visualization
synchronization and keep high data integrity between BIM experience for design support. It has been found that the AR
design data and VR/AR environments. To actualize this goal, implementation can improve the participants’ responses in
Du et al. [88] designed a BIM-VR real-time synchronization design by up to 20% compared to the BIM 3D visualization
system (BVRS) incorporating BIM for gathering metadata, [93]. Khalek et al. [94] proposed a hybrid BIM-AR method
cloud servers for online calculating, game engine for offer- to identify the maintainability concerns and make better
ing immersive experiences, and VR headsets for improving maintainability decisions during the design stage. Lin et al.
interpersonal communication. Chen et al. [89] adopted an [95] combined BIM, game engine, and VR technologies for
ontology-based approach to efficiently transfer BIM into a design project, which was proven useful in improving the
VR/AR under the consideration of rich semantic informa- comprehension of the design content and minimizing com-
tion, which was outstanding in simplifying the geometric munication gaps among clients, stakeholders, and designers.

Fig. 13  Framework of BIM and BIM model Real-time Specific AR/VR

use
AR/VR integration [87] Users Devices
in cloud server 3D engine application
3D/data see
BIM data transfer • AR quality control
+
Specific coded • VR 4D management
features

Table 11  Some previous studies about BIM and VR/AR

Purpose Method Achievement Related reference

Data transfer between Cloud computing; It can efficiently facilitate data transfer and synchronization for semantic [88, 89]
BIM and VR/AR Ontology method and geometric information.
Design review VR/AR techniques It can improve design understanding and collaboration based on intui- [92, 93, –95]
tive analysis and manipulation of 3D models.
Construction management VR/AR techniques It can pave a new way for project presentation and construction train- [19, 96, 97, 98, –100]
ing, contributing to enhancing construction safety and ensuring the
smoother progress of construction projects.

13
1102 Y. Pan, L. Zhang

Reviewing the current studies, it has been found that the and VR-based fire training. Accordingly, seamlessly inte-
adoption of VR/AR and BIM outperforms the traditional 2D grating the best aspects of BIM, IoT, and VR/AR is envis-
drawings by three major points. Firstly, the biggest advan- aged to offer interactive renderings, spatial coordination,
tage is to enhance the understanding of the design at its full and virtual mockups. It is gaining more and more attention
scale with the help of advanced visualization techniques. with the expectation of revolutionizing construction project
Secondly, it can achieve the interaction between the virtual management. What’s more, this topic can extend to a higher
buildings and surrounding environments. Finally, it can pro- level named BIM-enabled Extended Reality (XR), enabling
vide an immersive and interactive experience to facilitate to simulate construction projects in multidimensional digital
more efficient communication and decision making among models under the great immersion of visual, audio, touch,
designers and clients. olfactory, and others.
For another, since BIM-VR/AR contains first-hand knowl-
edge about the lifecycle activities of the project, it is advis- 4.6 BIM and Digital Twin
able to apply it in tackling issues in terms of scheduling,
site planning, safety controlling, and others, contributing to The concept of the digital twin originally arose from the
ensuring the smoothness of the project execution under rela- National Aeronautics and Space Administration (NASA) in
tively low safety risk and high quality. For example, Boton 2003 to better simulate the full life of the space capsule for
[96] proposed a neutral comprehensive framework of the mirroring and diagnosing problems that would occur in orbit
immersive VR-based collaborative BIM 4D simulation for [101]. This has been regarded as a precursor of the current
carrying out the constructability analysis meetings. Getul digital twin under the ultimate vision of modeling and test-
et al. [97] performed BIM-based immersive VR under the ing the physical entity in a virtual environment. After that,
steps of creating BIM-based modeling and planning, defin- the digital twin is rapidly developing to facilitate Industry
ing activity simulation environment, and analyzing data 4.0, and thus it is believed to be the top strategic technol-
from VR simulations by two novel algorithms, which in turn ogy trend. It can be explained as a virtual counterpart of the
provided suggestions for allocating workspaces on construc- physical assets incorporating identical fidelity with the help
tion sites meeting the safety criteria. The topic of BIM-AR of IoT devices and various AI methods. Based on its pow-
in the construction phase has been explored earlier than VR, erful capability of monitoring and synchronizing physical
and thus its practical applications are more mature and enjoy objects in real time, the digital twin provides deep insights
better usability. Jiao et al. [98] made AR work together with into the system's performance and potential problems. As
many advanced techniques, like cloud computing, web3D, shown in Fig. 14, the digital twin is typically made up of a
business social networking service (BSNS), and others, to physical model, a digital model, and an essential database
monitor the construction progress. Park et al. [99] connected connecting the cyber-physical system, which can make the
BIM with AR and an ontology-based data collection tem- utmost of bi-directional data flows. To be more specific, a
plate for defect inspection and analysis, aiming to give back digital replica of a physical counterpart that is enriched with
a proactive reduction of defect occurrence during construc- large volumes of data can dynamically imitate, model, and
tion. What’s more, Chen et al. [100] combined BIM, IoT, and analyze real-world behavior. A lot of hidden knowledge is
VR/AR technologies to build a situation awareness system bound to embed in the virtual part, which can be deeply
under an innovative immersive environment, which could explored through various data mining approaches in terms
realize three different purposes, including dynamic fire mon- of process simulation, pattern extraction, problem diagnosis,
itoring and alerting, AR-based real-time route navigation, trend prediction, and others. In return, the analytical data

Four examples of data analysis

IoT 1. Process simulation 2.Problem diagnosis
devices Real-time Data
data collection processing

Data
mining 3. Time series prediction 4. Optimization

Instruction for Real-time

improvement data analysis

Physical Model Database Virtual Model

Fig. 14  Architecture of the digital twin

13
Integrating BIM and AI for Smart Construction Management: Current Status and Future Directions 1103

results can serve as the value-added evidence to guide the an emerging integration of BIM and IoT has proven to be
formulation of response and precaution for problem solv- a significant prototype of the data-driven system for deci-
ing and process optimization, which can trigger the feed- sion making, allowing for automatic and real-time condition
back decision loop to adjust the physical part constantly and monitoring, evaluation, and improvement. Therefore, these
optimally. In short, there are three important features in the existing works can serve as a foundation to further create
digital twin that can distinguish the digital twin from BIM or a closed-loop paradigm for a complete set of digital twins.
VR/AR [102]. The first one is the emphasis on the existence As expected, a developed digital twin-based model can be
of the physical model. The second one is the twin relation- equipped with powerful abilities of constantly updating and
ship between the physical world and the virtual world. The learning real-time data for decision making.
digital object can be regarded as a corresponding represen- Due to the great potential of the digital twin towards a
tation and controlling instance of the physical object, and high level of digitization and intelligence, some existing
thus any changes in one object’s status can be concurrently papers have attempted to build the BIM-IoT-based digital
reflected in another object’s status. The third one is the full twin with intelligent functions, such as machine learn-
integration of rich and accurate data flow between the physi- ing algorithms, data analysis methods, and others. As a
cal and digital objects. result, information can be synced in real time between the
To date, the application of the digital twin in civil engi- as-designed and as-built models, aiming to support smart
neering is still in its infancy, which falls behind the industry management throughout the whole lifecycle of the BIM-
of manufacturing and aerospace. Since BIM is known as an enabled projects. For example, in the design stage, Lu et al.
information system, it owns the great potential to efficiently [105] rapidly and reliably created a digital twin for an exist-
synchronize and store mass data that is continuously col- ing reinforced concrete bridge by extracting and adopting
lected from IoT devices. In this regard, the research ques- geometric features directly from the 3D point cloud in the
tion can be formed as how to build a functional digital twin IFC format, which could address the question of manual
relying on BIM, which can facilitate a synchronized interac- digital twinning. Lydon et al. [106] carried out the coupled
tion between the physical and virtual entity for simulation, simulation of thermally active building systems to support
learning, and decision making. Some early efforts have been a digital twin, which could address building physics issues
made to integrate BIM with IoT and advanced data analysis about energy performance at the design phase of a building
methods for developing more objective decision support in project. In the construction stage, Pan et al. [108] developed
building facility management and maintenance from a data a BIM-IoT-data mining-based digital twin framework for
layer, as listed in Table 12. For example, Ma et al. [103] better understanding and optimization of the complicated
combined BIM, geographic information system (GIS), and construction work, where process mining and time series
reliability-centered maintenance (RCM) to support auto- analysis were conducted in the virtual part to discover the
matic data acquisition, updating, and analysis, which could bottleneck and predict the construction progress and then
reduce labor costs and increase objectivity in the decision- provide guidance to reasonably arrange work and staffing.
making of equipment maintenance in a business park. Cheng Matthews et al. [109] designed and implemented a real-time
et al. [104] systematically utilized the functionality of BIM object-oriented bi-directional system using cloud-based
and IoT techniques along with machine learning algorithms BIM, allowing for synchronizing information captured on-
to perceive the possible failure in mechanical, electrical, and site and improving decision-making during the construction
plumbing components of facilities, which helped managers of a reinforced concrete structure. In the O&M stage, Lu
to schedule the maintenance plan and extend the lifespan et al. [110] proposed a digital twin architecture at build-
of components in an evident-based manner. Clearly, such ing and city levels, which can successfully inform prompt

Table 12  Some previous studies aabout BIM and digital twin

Purpose Method Achievement Related reference

Design management IFC object fitting method; High-resolution It can support preliminary design with fewer [105–107]
model; Design structure matrix design defects and higher design efficiency,
which can also well predict and manage the
design changes.
Construction management Process mining; Time series analysis; Re- It can perform real-time monitoring and control [108, 109]
engineered method of construction quality and progress.
O&M management Machine learning; Deep learning It can contribute sustainability to comprehen- [104, 110–112]
sive asset management and structural health
monitoring for risk mitigation and prevention.

13
1104 Y. Pan, L. Zhang

decisions for five services, including anomaly detection in performance improvement. It is therefore urgent to drive
pumps, ambient environment monitoring, maintenance opti- digitalization and increase innovation on an urban scale,
mization, maintenance task prioritization, and urban energy which also meets the unprecedented tendency of urbaniza-
planning. Kang et al. [111] developed the twin-based ser- tion. In other words, another open question to be solved is to
vice for bridge health monitoring, which could dynamically promote smart city development, contributing to formulating
predict the defect of a deteriorating bridge and then create optimized strategies and operations along with long-term
preventive maintenance plans. To sum up, the digital twin resilience and ecological environment. The third challenge
under BIM can create a high-fidelity and dynamic digital lies in the security and transparency of data shared across all
replica of a construction project, since it is able to bring participants with different roles. Since a key feature of BIM
together and provide insights into design, construction, and to be noted is collaboration and information management,
O&M data. Hence, such a digital twin model is actually an essential task to ensure the successful BIM implementa-
responsive and continuous to evolve under the workflow of tion is to improve the transparency of data exchanges and
accurate monitoring, intelligent analysis, real-time response, contracts and keep trust among project participants. Hence,
which can potentially serve as a feasible solution to improve how to develop a more safe and collaborative way for infor-
the entire lifecycle of a construction project. However, a mation delivery under high efficiency and trust remains an
noticeable problem is that the current research about digital open question, aiming to perform a transparent, secure, and
twin development under the integration of BIM, IoT, and traceable BIM process.
advanced data analysis for the AEC industry is still limited, Notably, these three identified challenges can pro-
and its practical application is at the primary stage. Since vide clues to determine future research directions. To
digital twin shows great potential in various service per- tackle these challenges, we highlight the following three
spectives in promoting efficient design, smart construction potential research directions, including the synthesis of
comprehensive asset O&M, more attempts need to be put human–machine intelligence, city-level digital twin, and
on the creation and implementation of the reliable digital blockchain, all of which will make great contributions to
twin for boosting digitalization, automation, and safety in embracing the revolution in AEC industries. That is to say,
civil engineering. it is desired to add more bodies of knowledge and practica-
bility in these potential directions, aiming to maximize the
intelligence and automation of project implementation and
5 Future Research Directions increase the probability of project success throughout its
entire lifecycle. Although these potential directions currently
To date, construction project management is on the verge meet some challenges and stay in their nascent stage, their
of digitalization, which has been greatly influenced by tech- practical and strategic value will be increasingly revealed
nological breakthroughs from the widespread adoption of and validated in the future.
AI approaches. Therefore, it is believed that the integration
of BIM and AI will continuously be a hot area of research 5.1 Synthesis of Human–Machine Intelligence
in the following years. According to the existing papers, it
is worth noting that some challenges remain in the future This future research direction is the extension of smart pro-
development of BIM-AI integration. The first challenge is ject management. Although BIM and AI attempt to boost
the excessive dependence on data-driven algorithms. Unfor- the high degree of automation and digitalization in construc-
tunately, such a data-driven model cannot always ensure its tion projects, less consideration has been put on the human
robustness and application, which could draw conclusions intervention and communication that is an indispensable
that are contradicted to engineering mechanisms [113]. In part within the lifecycle of a project. Meanwhile, due to the
fact, human also plays an integral role in decision making. In complexity and uncertainty in practical engineering [114],
this regard, an open question is how to consider the impor- the current BIM-AI approaches are still short of sufficient
tance of the human in cooperating with various algorithms, empirical knowledge and expert interpretation, leading to
aiming to generate more reasonable results for project man- a lack of trust. In this regard, how to drive human-centered
agement. The second challenge is the application scope automation in construction management remains a great
of the current research that is mainly limited to buildings. challenge. To deal with it, the reliable development of
Technically, the leverage of BIM and AI can bring benefits human–machine intelligence is in high demand. It is nec-
in the real-world built environment and governance perspec- essary to explore the human influence and take combined
tive from the city level. In particular, this is far more than action with AI to facilitate more reliable and efficient con-
the buildings, infrastructures, and traffic systems, which is struction. Its practical value lies in the increase of situation
outstanding to offer a holistic visualization of city infrastruc- awareness and improvement of human performance, which
ture assets and gather actional insights into infrastructure will ultimately better instruct the participants in different

13
Integrating BIM and AI for Smart Construction Management: Current Status and Future Directions 1105

roles to promote the project smoothly. The goal of this deter- others, which cause great difficulties in developing an effec-
mined direction is to fully incorporate human factors, such tive framework of the city digital twin. Hence, a challenge
as behavior and psychology, into the BIM-enabled project, to be resolved is how to maximize the utilization of digital
resulting in a complicated socio-technical system for achiev- twins to improve not only the operation and sustainability of
ing human-automation interactive decision making. cities but also the quality of life. The goal of this direction
As the potential solutions, it is suggested to adopt more is to properly utilize the concept of digital twins to create a
advanced sensing technologies, such as NLP, computer- smarter city through digital simulation, analysis, and opti-
vision-based human tracking, wearable devices, and others, mization in real time. As expected, the city-level digital twin
to monitor human activities from both the physical and cog- can well reflect various functions of a city for different pur-
nitive aspects [115]. These collected data in large volumes poses, such as scenario planning, energy and waste manage-
offer a basis for understanding the uncertainty in human ment, mobility improvement, citizen satisfaction feedback,
factors, which can be tightly integrated with BIM and data and others, whose practical value comes from enhancing the
mining methods toward a human-in-the-loop cyber-physi- city’s visibility and operability in support of sustainable city
cal system [116]. Such a close loop containing cyber parts development [120].
and physical parts along with humans’ cognitive thinking As possible solutions, various IoT tools and AI methods
and reasoning capability can be reasonably regarded as the can be integrated into BIM to map the physical truth and
knowledge fusion from civil engineering, computer science, corresponding evolution of natural and built environments
and psychology. Notably, the developed human-in-the-loop on a city scale. In other words, the real-time data collec-
framework benefits in fully considering complex interactions tion from IoT devices for monitoring energy consumption,
in humans, tasks, and environments for more reliable simu- air pollution, traffic flow, and others provides valuable and
lation, analysis, and decision making. As reviewed, social- rich foundations for the creation of a digital twin city [121].
technical-based project management under human–machine Meanwhile, it brings new opportunities to capture spatiotem-
intelligence is in the initial stage. Some studies have shown poral fluctuations about human activities and natural disas-
promise in topics of resilient nuclear power plant facility ters across time and space as a complementary in situation
management [115], construction safety control [117], and evaluation, decision making, and coordination among differ-
infrastructure visual inspection [118]. It has been proved ent stakeholders. In other words, this proposed version of the
that the central role of human and human-related factors can digital twin can present a more comprehensive view of a city
potentially make the process of data acquisition and analysis through real-time monitoring, simulation, and assessment
more resilient and convincing. As a result, more promising to better understand the as-is condition of the city, which
decisions that feasibly adapt and respond to the participants, is practical in timely putting forward data-driven decisions
local conditions, and dynamic changing processes in real for disaster prevention, response, and recovery to make the
time can be informed. In conclusion, more future efforts can city smarter and resilient to extreme events [122]. Besides,
reasonably extend applications of the novel human-in-the- the city digital twin has shown its superiority in performing
loop system into all phases of construction project manage- large-scale analyses for near-real-time urban energy assess-
ment, contributing to better adapt to the changeable envi- ment and management over the traditional building energy
ronment for various purposes of diagnosis, prediction, and methods, which can effectively determine and prioritize spe-
optimization. It is expected to achieve proactive improve- cific retrofit strategies for energy efficiency improvement
ment for quality, safety, and efficiency assurance. [123]. It has been found that the USA, UK, Germany, and
Switzerland have made some efforts to reveal the potential
5.2 City‑Level Digital Twin of the city digital twin in terms of data management, visuali-
zation, situational awareness, prediction, and collaboration
This future research direction is the extension of the digital [120, 124]. There are reasons to believe that more countries
twin. The envisioned concept of the digital twin will not will focus on the establishment of digital twins at the city
be restricted to an individual building, which has prospec- level to help governments form both short-term and long-
tive potential to be extended to a higher level for augment- term resilience strategies for better city control. In conclu-
ing smart city development and management based upon sion, the development of smart cities with digital twins is
cyber-physical intelligence. Therefore, it is meaningful to currently in its infancy. It is envisioned that more attempts
combine the smart city concept and digital twin technique in the future can make the utmost of a variety of IoT data
into a single platform, and thus the current data from the along with socio-economic components and citizen feedback
smart city can drive the data analysis in the digital twin for to generate a more completely mirrored city digital twin.
intelligent decision making [119]. Compared to a building, Such a digital environment with AI-enhanced advanced
the city is a more complex living system with a wide range analytics enables a more efficient and sustainable way for
of aspects, including ecology, economic, social, culture, and resilient city management across all phases of the lifecycle.

13
1106 Y. Pan, L. Zhang

Moreover, it can potentially evolve into the urban brain with a post-disaster recovery system based on the principle of
the powerful ability of self-perceiving, self-determining, and decentralization, self-governance, and transparency, aim-
self-execution, which can more intelligently and adaptively ing to streamline a number of operations and enact prompt
support city operation and maintenance. response for a better rebuilding process [130]. Thirdly, the
blockchain-aided BIM can support sustainable building
5.3 Blockchain design collaboration and inform strategies towards sustain-
ability goals. In particular, the blockchain is responsible for
This future research direction is inspired by an emerging driving the smart contract to negotiate editing permissions
digital technology called blockchain, which is known as a and document immutable modifications to the BIM models
distributed database to store data in blocks and then chain [131]. In conclusion, the value of blockchain technology
them together in chronological order. It is revolutionizing in a BIM-enable environment is highlighted to make con-
the business practice around the world, which is promised struction management more transparent, efficient, automatic,
to secure a consensus on transactions under the significant and accountable among all participants. Since there are few
advantages of high security, automation, and decentraliza- existing studies about the blockchain potential applications
tion. Due to these inherent characteristics of blockchain, it is in the construction domain, the future interest in combina-
an ideal solution to manage the collaborative BIM platform tion of BIM and blockchain is bound to keep growing to
in the nature of interdisciplinary, multiple data contribu- boost mutual confidence and reach meaningful information
tors, and decentralized project organizational structure. In exchange especially in large and long-term engineering
particular, blockchain has developed to be an alternative projects.
approach for building trustworthy collaboration in the con-
struction industry [125]. Therefore, it can play important
roles in ensuring BIM information security through the 6 Conclusions
deployment of three critical technologies, including smart
contractors, distributed ledgers, and distributes-crypto- This paper provides a comprehensive review of the current
graphic indexing structure [126]. However, the integration development and future directions regarding the BIM-AI
of blockchain and BIM is not a straightforward task, which integration in the construction field. For one thing, BIM is
has remained a great challenge. In this regard, the goal envisioned as a novel digital paradigm to bring a remarkable
of this direction is to develop an operational and efficient shift in AEC industries for higher productivity and quality.
BIM-blockchain system. Its practical value is to reveal the It drives the construction industry into a data-intensive field.
potential value-added applications of blockchain with BIM, It provides a platform for not only collecting large volumes
enabling a built environment under a high degree of trust for of data about all aspects of the project, but also sharing,
sustainable design, smart contracts and quality assurance, exchanging, and analyzing data in real-time to achieve in-
supply chain management, and others [127]. time communication and collaboration among various par-
Some efforts have been made to extend the utilization ticipants. For another, AI is a growing trend to drive digital
of blockchain in BIM into the phases of pre-construction, transformation in many industries over the last few decades.
construction, and post-construction for well saving the The nature of AI is to invent computer programs to automati-
immutable changes, which can offer immediate benefits cally learn and think on its own, enabling complex problem
in enhancing the authenticity of records and confidence of solving and smart decision making with fewer errors and
stakeholders during the project execution. That is to say, higher efficiency. Due to the distinct advantages of BIM and
the blockchain holds the data generated by IoT in a trans- AI, it is believed that the topic of BIM-AI integration is well
parent, secure, and convenient environment, while BIM as worth exploring, where BIM can be reasonably considered
a baseline tool promises to digitize the construction pro- as a digital backbone to work with AI to further boost intel-
ject and increase collaboration [128]. The potential role of ligence in construction project management throughout the
blockchain in BIM can be performed from the following full project lifecycle, including planning and design, con-
aspects. Firstly, the marriage of BIM and blockchain has struction, and O&M phases.
shown its outstanding performance in securely tracing the According to the mixed review of bibliometric analysis
revision of BIM data for the improvement of BIM data audit, and information analysis, it has been found that the United
provenance, and accountability, and thus each participant in States, China, and the United Kingdom are the top three
a construction project is not allowed to change information countries focusing more on the targeted topic for transform-
retrospectively [129]. Secondly, since blockchain is proven ing the traditional project delivery and management towards
effective in addressing problems of data security and pri- a higher degree of automation, intelligence, and reliability.
vacy, transaction speed, modification tracking, and others, an The timezone map of keywords shows that current research-
potential application of blockchain with BIM is to develop ers turn their interests to deep learning and digital twin in the

13
Integrating BIM and AI for Smart Construction Management: Current Status and Future Directions 1107

context of Industry 4.0, indicating that digitalization in con- Research and Development Plan (No. 2022YFC3802205), and Shang-
struction is bound to be increasingly important in the future. hai Sailing Program, China (No. 22YF1419100).
Remarkably, the advanced research interests on BIM-AI
Declarations
integration for construction projects are different from one
another, including automated design and rule checking, 3D Conflict of interest The authors declare that they have no known com-
reconstruction, event log mining, VR/AR, and digital twin, peting financial interests or personal relationships that could have ap-
aiming to deliver efficient information sharing and explor- peared to influence the work reported in this paper.
ing to keep continuous updating and improvement of the
ongoing project. Based on the in-depth analysis in a range
of ways (i.e., simulation, prediction, and optimization), stra-
tegic decisions that are suitable for a certain project will References
be informed without human intervention under complicated
and uncertain environments, which is expected to generate
immediate reactions to streamline the complicated workflow,
shorten operation time, cut costs, reduce risk, optimize staff
arrangement, and others. As reviewed, the current status of
the BIM and AI integration is still at an early stage. Herein,
three potential research directions along with their practi-
cal applications, including the synthesis of human–machine
intelligence, city-level digital twin, and blockchain, have
been summarized, which can address the remaining chal-
lenges in civil engineering.
To further facilitate information digitalization in intelli-
gent construction project management, BIM can be reason-
ably considered as a digital backbone to work with AI. In the
immediate future, the integration of BIM and AI can move
the paper-based work toward online management, which
assists the traditional construction industry to catch up with
the fast pace of automation and digitalization. As expected,
it can deliver the most efficient and effective information to
keep continuous updating of the ongoing project. The solu-
tions for construction projects are different from one another.
Based on the in-depth analysis in a range of ways (i.e., simu-
lation, prediction, and optimization), strategic decisions that
are suitable for a certain project will be informed without
human intervention under complicated and uncertain envi-
ronments, which is expected to generate immediate reactions
to streamline the complicated workflow, shorten operation
time, cut costs, reduce risk, optimize staff arrangement, and
others. Meanwhile. this kind of tactical decision-making
can possibly be adapted to the changeable conditions to
optimize the project operation continuously for delivering
smarter construction management throughout the full project
lifecycle. Hence, it can be reasonably considered that the
practical value of the hybrid framework based on BIM and
AI lies in addressing challenges arising from characteristics
of construction project management, including uniqueness,
labor-intensive, dynamics, complexity, and uncertainty. This
topic of BIM and AI integration deserves more attention.

Funding The work was supported in part by the National Natural Sci-
ence Foundation of China (Nos. 72201171, 72271101), National Key

13
References
[1] Alavipanah, S. K., Minaei, B., Tavakkoli-Moghaddam, R., & Hashemi, H. (2019). A hybrid SVM-GA approach for
optimizing construction project scheduling under uncertainty. Automation in Construction, 99, 221–231.
[2] Aravantinos, V., Chatzi, E. N., Papadimitriou, C., & Ntotsios, E. (2020). A novel AI-based methodology for vibration-
based structural health monitoring. Mechanical Systems and Signal Processing, 135, 2.
[3] Asprone, D., Di Ludovico, M., Fumo, G., Prota, A., & Manfredi, G. (2018). Discovering knowledge in structural
design processes: A data mining approach. Journal of Computing in Civil Engineering, 32, 3.
[4] Asprone, D., Prota, A., Manfredi, G., & Cosenza, E. (2019). Building clustering for optimal urban resilience
enhancement. Journal of Computing in Civil Engineering, 33, 3.
[5] Azadbakht, M., Seyedpoor, S. M., & Ostadan, F. (2019). Application of a hybrid intelligent system in calibration of
finite element models of building structures for earthquake-induced response prediction. Engineering
Structures, 180, 69–83.
[6] Bakhshpoori, T., Kheyroddin, A., & Sobhani, J. (2016). Optimization of steel -concrete composite beams using a
genetic algorithm. Journal of Constructional Steel Research, 119, 1–10.
[7] Bhattacharya, S., Chowdhury, A., & Basu, B. (2019). Structural damage detection using machine learning
with application to a benchmark problem. Journal of Sound and Vibration, 455, 318–336.
[8] Bhoopalam, A., Patel, D., & Bhaskar, A. (2021). Generative design approach for low-cost pedestrian bridge:
An artificial intelligencebased structural optimization. Structures, 33, 1900–1913.
[9] Bilal, M., Farooq, S., Younas, M., & Ali, M. (2020). Role of artificial intelligence in civil engineering: A systematic
literature review. Advances in Civil Engineering, 2020, 1–14.
[10] Bishop, C. M. (2006). Pattern Recognition and Machine Learning. Springer.
[11] Parekh, Ruchit. "Constructing Wellness: Harnessing AI for a Sustainable and Healthy Future." (2024).
[12] Breiman, L. (2001). Random Forests. Machine Learning, 45(1), 5–32.
[13] Breiman, L., Friedman, J. H., Olshen, R. A., & Stone, C. J. (1984).
[14] Classification and Regression Trees. Wadsworth International Group.Cano, M., Ghorbanizadeh, O., & Guédon, S.
(2019). Soft soil consolidation settlement prediction with machine learning methods. Geotechnical and Geological
Engineering, 37(6), 5281–5296.
[15] Chen, W., Lin, S., Wang, D., Zhang, Y., & Cao, S. (2021). An optimization algorithm for parameterization of
design space in bridge engineering. Journal of Computing in Civil Engineering, 35, 1.
[16] Parekh, Ruchit. "Blueprint for Sustainability: LEED Implementation in Commercial Projects." (2024).
[17] Chen, Y., & Xie, Y. (2020). Traffic prediction using deep learning: A generative adversarial network
approach. IEEE Transactions on Intelligent Transportation Systems, 21(8), 3528–3539.
[18] Chen, Z., Zhang, Y., Li, J., & Zhu, W. (2019). An artificial neural network approach for predicting the
strength parameters of sandcement mixtures. Neural Computing and Applications, 31(4), 1293–1303.
[19] Chu, X., Sun, Y., Jin, Z., & Sun, X. (2019). Optimal design of tall buildings under wind-induced vibrations based
on hybrid optimization algorithm. Engineering Structures, 198, 2.
[20] Deb, K., Karthik, S., & Subrahmanya, S. (2016). Design optimization of truss structures using finite
element analysis and genetic algorithm. Journal of Computing in Civil Engineering, 30, 6.
[21] Dong, P., Hou, D., Chen, X., & Yu, H. (2016). Application of decision tree in slope stability analysis. Advances in Civil
Engineering, 20, 16.
[22] Dorigo, M., Poretti, S., Roa-Sepúlveda, C. A., & Tormos, P. (2020). Multiobjective optimization in civil engineeri ng:
A systematic review. Archives of Computational Methods in Engineering, 27(4), 1399–1447.
[23] Elnimeiri, M., Al-Khaleefi, K., & Elnimeiri, M. (2015). Data mining of construction delay causes. Journal of
Construction Engineering and Management, 141, 1.
[24] Fasihi, M., Malek, S., & Zolfaghari, A. (2017). Concrete compressive strength prediction using machine learning
models. Journal of Computing in Civil Engineering, 31, 6.
[25] Fayyad, U., Piatetsky-Shapiro, G., & Smyth, P. (1996). From data mining to knowledge discovery in databases.
AI Magazine, 17(3), 37–54.
[26] Feng, Y., Yuan, X., & Wang, Z. (2020). Multi-objective optimization design of bridge piers based on
geneti c algorithm. Journal of Bridge Engineering, 25, 1.
[27] Gandomi, A. H., Alavi, A. H., & Mirhosseini, R. (2015). Multi -objective self-compacting concrete mix design using
ANN and NSGA-II. Neural Computing and Applications, 26(4), 811–822.
[28] Garg, A., Rana, P., & Aggarwal, S. (2021). Urban transportation systems: Big data analytics and
intelligent transport systems. Sustainable Intelligent Transportation Systems (pp. 47–66). Springer.
[29] Goodfellow, I., Bengio, Y., & Courville, A. (2016). Deep Learning. MIT Press.
[30] Graves, A., Mohamed, A. R., & Hinton, G. (2013). Speech recognition with deep recurrent neural networks.
In Acoustics, Speech and Signal Processing (ICASSP), 2013 IEEE International Conference (pp. 6645–6649).
[31] Parekh, Ruchit, and M. Olivia. "Utilization of artificial intelligence in project management." *International
Journal of Science and Research Archive* 13.1 (2024): 1093-1102.
[32] Hassanpour, H., Saffar, B., & Asgarian, B. (2018). Multi-objective particle swarm optimization for
resource -constrained project scheduling problem. Journal of Construction Engineering and Management, 144, 9.
[33] He, X., Liao, L., Zhang, H., Nie, L., Hu, X., & Chua, T. S. (2017). Neural collaborative filtering. In Proceedings of
the 26th International Conference on World Wide Web (pp. 173–182).
[34] Hoerl, A. E., & Kennard, R. W. (1970). Ridge regression: Biased estimation for nonorthogonal problems.
Technometrics, 12(1), 55–67.
[35] Hu, Z., Li, J., & Zhang, Y. (2019). A machine learning-based approach for steel structure damage
classification. Structural Control and Health Monitoring, 26, 11.
[36] Huang, L., Han, D., & Wang, D. (2019). Landslide susceptibility mapping based on extreme learning machine with
particle swarm optimization. Engineering Geology, 261, 2.
[37] Huang, W., Xu, Z., Liu, Z., & Xu, X. (2021). An intelligent transportation system for smart cities: A systematic review
and metaanalysis. Sustainable Cities and Society, 68, 2.
[38] Huang, Y., Qi, X., Wang, W., & Xie, X. (2020). An AI-based generative design approach for energy-efficient
building layout. Automation in Construction, 117, 2.
[39] Jangid, R. S., Chaudhari, H. B., & Trikande, N. J. (2015). Artificial neural network based design response
prediction of buildings under earthquake excitation. Journal of Civil Structural Health Monitoring, 5(2), 155–166.
[40] Jin, Y., Guo, Y., He, J., & Cheng, G. (2019). Optimization of truss structures using generative design and
geneti c algorithm. Advances in Civil Engineering, 20, 19.
[41] Anderson, Lisa. "AI-Driven Innovations in Sustainable Construction." Journal of Construction Technology 20.1
(2024): 10-25.
[42] Khan, Mohammad. "Health and Safety Protocols in Green Building Projects." International Journal of
Constructi on Health 15.2 (2023): 100-115.
[43] Parekh, Ruchit. Blueprint for Sustainability: LEED Implementation in Commercial Projects. Elsevier, 2024.
[44] Garcia, Elena. "The Impact of AI on Construction Project Management." Automation in Construction 45
(2024): 150-165.
[45] Müller, Stefan. "Sustainable Materials for Urban Infrastructure." Sustainable Infrastructure Journal 12.3
(2023): 80-95.
[46] Johnson, Emily. "Enhancing Construction Safety with AI Technologies." Journal of Safety Science 33.4 (2024):
200-215.
[47] Chen, Wei. AI Applications in Construction Engineering. Springer, 2023.
[48] Smith, Olivia. "Occupational Health in Sustainable Construction Sites." Journal of Occupational Health and Safety
18.1 (2024): 50-65.
[49] Singh, Rajiv. "Machine Learning for Predictive Maintenance in Construction Equipment." Engineeri ng
Management Review 29.2 (2023): 120-135.
[50] Evans, Hannah. Sustainable Design Principles in Modern Architecture. Routledge, 2023.

2
[51] Kim, Soo-jin. "Integrating AI for Energy Efficiency in Buildings." Energy and Buildings 75 (2024): 220-235.
[52] O'Connor, Liam. "AI Tools for Construction Waste Reduction." Waste Management in Construction 14.2
(2023): 90-105.
[53] Martinez, Carlos. "Health Risks and Mitigation in Construction Projects." Journal of Construction Health
21.3 (2024): 130-145.
[54] Zhang, Li. "Sustainable Construction Practices in Urban Development." Urban Planning and Development
40.1 (2024): 70-85.
[55] Williams, Sarah. Renewable Energy Solutions for Construction Projects. Oxford University Press, 2023.
[56] Parekh, Ruchit, and Charles Smith. "Innovative AI-Driven Software for Fire Safety Design: Implementation in
Vast Open Structures." World Journal of Advanced Engineering Technology and Sciences 12.2 (2024): 741-750.
[57] Ahmed, Fatima. "AI and Robotics in Construction Site Automation." Journal of Robotics in Construction
11.3 (2024): 160-175.
[58] Parekh, Ruchit. "Comparison Analysis of Construction Costs according to LEED and non-LEED Certified
Educational Buildings." Available at SSRN 4924703 (2024).
[59] Ivanov, Alex. "Health Monitoring of Structures Using AI Techniques." Structural Health Monitoring 22.4
(2024): 210-225.
[60] Garcia, Maria. Eco-Friendly Building Materials and Their Applications. Taylor & Francis, 2023.
[61] Lee, Min-Jae. "AI-Enhanced Safety Protocols in Construction." Journal of Construction Safety 16.1 (2024):
45-60.
[62] Thompson, Grace. "The Role of AI in Sustainable Urban Planning." Urban Technology Journal 28.2 (2023):
75-90.
[63] Parekh, Ruchit. "Trends and Challenges in LEED v4.1 Healthcare Certification: A Comprehensive Analysis of
US Hospital Scores in 2024." World Journal of Advanced Engineering Technology and Sciences 12.2 (2024):
726 -740.
[64] Rodriguez, Juan. "AI Solutions for Resource Optimization in Construction." Resource Management
in Construction 19.2 (2024): 110-125.
[65] Stevens, Laura. Health and Wellness in Sustainable Building Design. McGraw-Hill, 2023.
[66] Kumar, Anil. "Predictive Analytics for Construction Safety Management." Safety Science 70 (2024): 80-95.
[67] Nakamura, Yuki. "AI Techniques in Construction Scheduling." Automation in Construction 50 (2023): 130-145.
[68] Petrov, Natalia. "Sustainable Construction Practices in Cold Climates." Journal of Environmental Engineering
38.3 (2024): 100-115.
[69] Chen, Li. "AI for Structural Design Optimization." Structural Engineering International 33.1 (2024): 60-75.
[70] Hernandez, Sofia. "Health Impacts of Indoor Air Quality in Green Buildings." Environmental Health Perspectives
27.2 (2023): 90-105.
[71] Smith, Daniel. AI and Machine Learning in Construction Management. Elsevier, 2023.
[72] Alvarez, Diego. "Implementing Sustainable Water Management in Construction." Hydrology and
Constructi on 15.4 (2024): 240-255.
[73] Taylor, Emily. "AI Applications in Construction Cost Estimation." Journal of Construction Economics 22.3 (2023):
150-165.
[74] Mustafa, Leila. "Enhancing Worker Safety with AI Wearable Devices." Journal of Occupational Safety 19.1
(2024): 70-85.
[75] Nguyen, Hoa. "Renewable Energy Integration in Building Design." Renewable Energy Journal 45.2 (2023): 180
-195.
[76] Parekh, Ruchit. Constructing Wellness: Harnessing AI for a Sustainable and Healthy Future. Elsevier, 2024.
[77] Kowalski, Piotr. "AI-Driven Risk Management in Construction Projects." Journal of Risk Management 17.2
(2024): 90-105.
[78] Evans, Zoe. "Sustainable Housing Design and Public Health." Housing and Society 30.1 (2023): 60-75.
[79] Garcia, Miguel. "AI-Based Energy Management Systems in Construction." Energy Management Journal
25.3 (2024): 140-155. 3
[80] Singh, Priya. Sustainable Infrastructure Development. CRC Press, 2023.
[81] Robinson, Mark. "AI Applications in Modular Construction Techniques." Journal of Modern Construction 9.4
(2023): 200-215.
[82] Liu, Mei. "Environmental Impact Assessments Using AI Tools." Environmental Assessment Journal 12.2 (2024):
50-65.
[83] Chen, Hao. "Advancements in AI for Construction Site Monitoring." Journal of Construction Technology 18.1
(2024): 100-115.
[84] O'Neill, Patrick. "Health and Safety in High-Rise Construction Projects." International Journal of Constructi on
Health 20.2 (2023): 80-95.
[85] Zhang, Wei. Energy Efficiency and AI in Building Design. MIT Press, 2023.
[86] Ahmed, Aisha. "AI Strategies for Waste Reduction in Construction." Waste Management Journal 16.3 (2024): 120 -
135.
[87] Brown, Jessica. "Sustainability Metrics and Their Impact on Construction." Journal of Sustainable Constructi on
11.2 (2023): 90-105.
[88] Ivanov, Sergey. "AI in Construction Equipment Automation." Automation in Engineering 27.4 (2024): 190-205.
[89] Lee, Sun-Young. "Sustainable Urban Development and AI Technologies." Urban Development Journal 35.1
(2023): 100-115.

View publication stats