Content uploaded by Sururah Apinke Bello
Author content
All content in this area was uploaded by Sururah Apinke Bello on Apr 05, 2025
Content may be subject to copyright.
Full Length Article
Cloud computing for chatbot in the construction industry: An
implementation framework for conversational-BIM voice assistant
Sururah A. Bello
a,*
, Lukumon O. Oyedele
b,*
, Lukman A. Akanbi
c,*
, Abdul-Lateef Bello
d
a
Faculty of Computer Science and Engineering, Obafemi Awolowo University, Ile-Ife, Nigeria
b
Big Data Enterprise and Articial Intelligence Laboratory (Big-DEAL), Bristol Business School, University of the West of England Bristol, Frenchay Campus, Bristol BS16
1QY, United Kingdom
c
College of Business, Digital Transformation and Entrepreneurship, Faculty of Business, Law and Social Sciences, Birmingham City University, Birmingham United
Kingdom
d
School of Architecture and Environment, University of the West of England Bristol, Frenchay Campus, Bristol BS16 1QY, United Kingdom
ARTICLE INFO
Keywords:
Software project management
Amazon web services
Cloud computing
Building information modelling (BIM)
Conversational AI
Construction industry
Framework implementation
Chatbot
construction workers
Design thinking methodology, Focus group,
Stakeholders management
ABSTRACT
This study presents a structural framework for selecting cloud services for the Conversational AI system
implementation in the construction industry using Design Thinking Methodology. A focus group discussion
approach was used to obtain user requirements from construction workers to implement the Conversational AI
for BIM. This resulted in ve factors: nance, speed of operation, privacy, estimation, and interface. The user
specications were mapped into technical modules, which were used to select cloud services employed to
implement the virtual assistant for the construction industry. The study thus presented the comprehensive re-
quirements for the different categories of construction workers to implement the Conversational-BIM Chatbot
(Conversational-BIM) system. Furthermore, the study presented the architecture of Conversational-BIM using
Amazon Web Services. The study is useful to researchers and IT developers in implementing chatbots for the
construction industry as it presents the relevant considerations for conversational AI applications in the industry.
1. Background and introduction
Voice user interfaces employ voice-based commands [1] to exhibit
hands-free and eyes-free interaction with great intuitiveness and exi-
bility [2]. According to Ruan et al. [3], humans speak faster than typing.
Also, Lee and Nass [4] opined that hearing synthesised speeches as re-
sponses created a strong social presence to users. No doubt, voice-based
technology’s speed, efciency, and convenience is leading towards the
less screen-interaction era [5]. Thus, this is not surprising that the
adoption of voice-based technology is soaring, with about 3.25b digital
voice assistants in use in 2019 and a forecast of 8b users by 2023 [6].
Digital Voice Assistants, also called Chatbots or Conversational Articial
Intelligent systems [7], have recorded successful applications across
retailing [8], marketing [9], education [10] and healthcare ([11] &
[12]) sectors. In addition to carrying out business operations like
administration, billing, payroll, amongst others within an ofce, con-
struction industries also carry out several operations on sites. No doubt,
the construction industry is daily evolving and desiring to benet from
the latest technological innovation for maximum operational efciency.
However, the industry is slow in adapting to change, especially adopting
advances in IT [13,14]; nevertheless, this tide appears turning nowa-
days. As the construction industry has been benetting from emerging
technologies, i.e., from BIM [15,16], Digital Twins [17], Articial In-
telligence [18], Big Data [19,20,21], IoT [22], Machine Learning [23,
24], Deep Learning [25–27], Augmented Reality/Virtual Reality [28,
29], Robotics [30,31] to Cloud Computing [32], the industry is also
poised to benet from the recently developed digital voice assistant
technology. In practice, voice-based technologies enable construction
workers to perform diverse activities on sites using voice command with
hands-free and eye-free operations [7]. Furthermore, voice-based tech-
nologies can improve site worker’s productivity as interacting with
voice is more natural [33], unlike other mediums of interactions like the
keyboard and mouse.
Voice user interfaces employ voice-based commands [1] to exhibit
* Corresponding authors.
E-mail addresses: apinkebello@gmail.com (S.A. Bello), Ayolook2001@yahoo.co.uk (L.O. Oyedele), lukman.akanbi@bcu.ac.uk, laakanbi2010@gmail.com
(L.A. Akanbi).
Contents lists available at ScienceDirect
Digital Engineering
journal homepage: www.sciencedirect.com/journal/digital-engineering
https://doi.org/10.1016/j.dte.2024.100031
Received 16 July 2024; Received in revised form 20 October 2024; Accepted 19 December 2024
Digital Engineering 5 (2025) 100031
Available online 25 December 2024
2950-550X/© 2024 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC license (
http://creativecommons.org/licenses/by-
nc/4.0/ ).
hands-free and eyes-free interaction with great intuitiveness and exi-
bility [2]. According to [3], humans speak faster than typing. Also, Lee
and Nass [4] opined that hearing synthesis speeches as responses created
a strong social presence to users. No doubt, voice-based technology’s
speed, efciency, and convenience is leading towards the less
screen-interaction era [5]. Thus, this is not surprising that the adoption
of voice-based technology is soaring, with about 3.25b digital voice
assistants in use in 2019 and a forecast of 8b users by 2023 [6]. Digital
Voice Assistants, also called Chatbots or Conversational Articial
Intelligent systems, have recorded successful applications across
retailing [8], marketing [9], education [10] and healthcare ([11] &
[12]) sectors. However, the construction industry is slow in adapting to
change, especially adopting advances in IT [14]; nevertheless, this tide
appears turning nowadays. As the construction industry has been
benetting from emerging technologies, i.e., from BIM [15], Big Data
[19,20], IoT [22], Deep Learning [26], Augmented Reality/Virtual Re-
ality (Delgado et al., 2020), Robotics (Delgado et al., 2019) to Cloud
Computing [32], the industry is also poised to benet from the recently
developed digital voice assistant technology. In practice, voice-based
technologies enable construction workers to perform diverse activities
on sites using voice command with hands-free and eye-free operations.
Furthermore, voice-based technologies can improve site worker’s pro-
ductivity as interacting with voice is more natural [33], unlike other
mediums of interactions like the keyboard and mouse.
BIM is a modelling tool comprising 3D visualisations, time and cost
projections to result in 5D designs (Eastman et al., 2008). BIM is a shared
database containing specialised documentation on architecture design,
landscape design, construction and installation design, scheduling, bills
of quantities, and cost estimates [34], accessible for collaboration by
stakeholders to produce complete project documentation. The use of
BIM software on construction sites has become an established routine.
Meanwhile, interacting with BIM with the traditional keyboards and
touchscreens has slowed down the adoption of the technology [35]. This
slow adoption is due to inconvenient methods of interaction, thus
limiting the use of the technology. Construction workers are already
working with hands on site; however, using hands to hold tools and
interact with BIM simultaneously is quite difcult. Akinade et al. [15]
argued that interfacing the voice with BIM will widen the adoption of
BIM for improved productivity and fast delivery of projects, as inte-
gration of Voice Assistance with BIM would provide a more natural
interface for construction workers to interact with BIM on site.
Regrettably, BIM originally is not accessible in real time, as tradi-
tional BIM is a standalone system not readily accessible to diverse
construction stakeholders, hence, not widely adopted [36]. In contrast,
cloud computing is a key enabling technology to facilitate easy access
and collaboration for BIM [37]. Also, the government’s strategic plan for
BIM Level3 will not materialise without the adoption of cloud
computing across construction industries. Consequently, interfacing
BIM with conversational technology will improve interaction, while
cloud computing technology will enhance BIM accessibility [38].
Thus, the need to develop a conversational BIM application, which
will enable construction workers to have unhindered interaction with
BIM software using speech inputs and responses requires a cloud tech-
nology platform. Maximising the opportunities from technologies such
as cloud services is dependent on prudent decisions which requires
considerations of strategies, facilities and requirements of organisations
(Costa, 2013). There are numerous cloud computing services from
different providers due to the increasing demand for the technologies
(Bello, 2012). Consequently, there has been numerous attempts to select
appropriate cloud services; [39,40,41,42,43,44] amongst others. Most
of these existing studies on selecting cloud services are for general ap-
plications whereas selecting cloud services is better with a partic-
ular/specic use case in focus [45]. However, there is no known study
that selects cloud service to implement Conversational-BIM application
and also focusing on the need of the construction workers, hence this
study. This study’s objectives are to develop a framework that
establishes the relevance and timeliness of Conversational-BIM in con-
struction, formulate its technical requirements, and map the technical
requirements to appropriate cloud services for efcient delivery and
improved productivity.
The rest of paper is as follows; in Section 2 existing use of cloud
computing in construction is discussed. This is followed by discuss on
the case study, that is Conversational BIM (Conversational-BIM). Section
4discusses the methods employed to establish criteria for selecting
cloud services to implement Conversational-BIM. Next to this is full
discussion of the result, then motivations as well as challenges for the
use of Conversational-BIM is discussed. Section 7 discusses the signi-
cance of the study while Section 8 concludes the paper.
2. Cloud computing in construction
2.1. Cloud computing
According to National Institute of Standards and Technology (NIST)
“Cloud computing is a model for enabling convenient, on-demand
network access to a shared pool of congurable computing resources
(for example, networks, servers, storage, applications, and services) that
can be rapidly provisioned and released with minimal management
effort or service provider interaction” [46]. Generally, cloud services are
popularly offered as three services, Infrastructure-as-a-service (IaaS),
Platform-as-a-service (PaaS) and Software-as-a- service (SaaS). These
services could be deployed at the public, private, hybrid or community
level, and they are of different forms; some are available on all plat-
forms, whereas some require third-party applications to run effectively.
Also, some are simple to set up and use, and some allow for collaboration
within a group of users exhibiting a groupware feature. Some cloud
services are suitable for small business or enterprise use, while some are
for personal use. Lastly, synchronisation is another form, which allows
for fast access to cloud storage. Consequently, cloud services offer
different control rights to users, depending on the technical ability of
users. A highly tech-savvy user may desire more control of the cloud
resource, while a novice may opt for lesser control of the service. Cloud
services have different price variants. Some have free start then followed
by subscription payment, while some start with a subscription offers
only. In all, the decision makers’ interest [47] is a strong factor in
selecting a cloud service for adoption, and selecting the best cloud ser-
vice could be inuenced by decision makers’ experience and knowledge
[48].
2.2. Existing applications of cloud computing in construction
In Bello et al. [32], cloud computing has been helpful in the pre-
construction, construction, and post constructions stages, specically for
managing energy, waste, health and safety, supply chain, and project
communication for built assets. Azambuja et al. [49] solved the accu-
mulating large inventories problem resulting in material wastage on
construction sites using cloud technologies. Also, Redmond et al. [50]
employed cloud technology to alleviate the limited access problem to
existing construction information that usually results in resource
wastage. Cloud technology was employed to solve the inaccurate com-
ponents delivery problem due to the lack of coordination among parties
in precast construction [51].
Getuli et al. [52] employed cloud technology to effectively monitor
construction activities with location information for improved safety on
site. Park et al. [53] also employed cloud technology to detect unsafe
conditions on construction sites and prevent potential hazards to site
workers. Tang et al. [54] used cloud technology to solve the problem of
irregular and untimely site inspection. Furthermore, Guo et al. [55]
evolved a safety system using cloud technology to observe workers
during metro construction, while Li et al. [56] used cloud technology in
underground construction for timely and accurate recognition of safety
risks during preconstruction.
S.A. Bello et al.
Digital Engineering 5 (2025) 100031
2
Cloud computing has been used to manage energy in different stages
of construction [57,58]. For instance, Khajenasiri et al. [59] employed
cloud technology to intelligently control building energy in smart cities,
while Rawai et al. [60] used cloud computing to reduce both energy
consumption and CO
2
emissions during construction, Cho et al.
[61]
employed the
technology to manage energy systems for a sustainable decision support
system. Wang et al. [62] reported the cloud technology application to
realise the Building Management operations of green buildings. Balaras
et al. [63] utilised cloud technology to develop a Virtual Energy Labo-
ratory for simulation designs of energy-efcient buildings. Furthermore,
Curry et al. [64] employed cloud technology to provide a unied
interface to manage building data from diverse integrated sources, and
Naboni et al. [65] employed cloud technologies for parametric simula-
tion of a building’s energy performance.
Cloud computing has also been employed to solve the problem of a
gap in the supply chain that causes delays in project delivery as a result
of an uncoordinated traditional material supply chain [66]. Fathi et al.
[67] employed cloud technology for accurate information transfer to
parties in the construction supply chain process, while Azambuja and
Gong [68] used cloud technologies to evolve cost-efcient management
of supply chain data. Grilo and Jardim- Gonclaves [69], used cloud
technology to improve interoperability among stakeholders in the pro-
curement process. Ko et al. [70] and Sahin et al. [71] employed cloud
technologies to provide an affordable tracking system for material
movement on construction sites. Hemanth et al. [72] also employed
cloud technology to solve the misperception of information in the pre-
cast industry.
Cloud technology has been employed to solve the low construction
quality problem as a result of poor communication and coordination
among stakeholders [73]. Ferrada et al. [74] used cloud technologies to
formalise knowledge transfer among local construction companies. Petri
et al. [45] used cloud technologies to coordinate multi-site construction
activities involving varied organizations and individuals. Alaka et al.
[19] employed cloud technologies to analyse data for predicting the
failure of construction businesses. Ahn et al. [75] employed cloud
technology to integrate information from construction sites together
with ofce work to aid decision making. Petri et al. [76] employed cloud
architecture to improve data access during construction, and Beach et al.
[77] employed cloud technologies to store and manage building data for
improved security. Polter and Sherer [78] used cloud technologies for
SMEs to provide affordable data transfer systems, while Nú˜
nez et al.
[79] employed the technologies for SMEs to manage lessons from pre-
vious projects. Jiao et al. [80] employed cloud technologies manage-
ment to provide a cost-effective life-cycle data management system for
the AEC/FM sector.
These various applications of cloud technologies revealed that the
technology had been used to solve a number of problems/issues in the
construction industry (Ajayi et al., 2016; [81]). No doubt, cloud
computing adoption has grossly impacted the efciency of construction
companies as the industry is beginning to feel the relevance of cloud
computing, which holds considerable benets for the industry, including
the use of voice-based devices on construction sites. The technology
reduces the use of papers for documentation in project delivery and
reduces energy even to discard waste papers [82]. Cloud adoption has
minimised the energy consumed by individual in-house servers for op-
erations and cooling. Also, cloud computing reduces the commuting of
construction workers and reduces carbon emission and carbon footprint
[83]. Thus, cloud computing adoption reduces operational and main-
tenance costs resulting in increased ROI for the construction industry.
2.3. Generative articial intelligence in the construction industry
No doubt the current wave of generative AI adoption is hitting the
construction industry, as its use is being found in the various stages of
construction. In 2022, Hayman [84] proposed the ChatGPT to simplify
the art of gathering information varieties of topics in architecture and
design not limited to building materials, construction methods, or design
trends. Rane et al. [85] also opined that the construction industry like
manufacturing, nance, retail and transportation could benet from the
use of generative AI for project planning, design enhancement and risk
mitigation. Thus, [86] demonstrated the use of GPT in construction
optimize material selection in the design phase of construction.
Furthermore, Priesto [87] employed ChatGPT to generate a construction
schedule that resulted in positive interaction experience thus indicating
the potential of generative AI tool to automate repetitive tasks in con-
struction industry. Additionally, You et al. [88] also proposed a
sequence planning system for construction tasks leveraging on the
advanced reasoning capabilities of the ChatGPT. This was corroborated
by Parm [89] who demonstrated the use of Generative AI to ease the
process of designing project planning.
Meanwhile, Beach et al. [90] had earlier advanced GPT models
leveraging in their NLP capabilities to discern textual relevant infor-
mation for digitized reasoning can be valuable in automated regulatory
compliance during construction activities. Despite the diverse features
requires for efcient resource allocation in project management [91,18]
suggested that Generative AI can provide necessary parameters for
project impact analysis in project quality management. Furthermore,
Zheng and Fischer [92] agreed that GPT models is capable of scruti-
nizing historic data and integrating it with new information on the
current state of the project for a comprehensive project risk assessment.
Meanwhile, Xia et al. [93] had proposed the use of AI to overcome the
limitation in analysing extensive datasets that could miss concealed
defects resulting in safety issues in assessing structures.
Akinosho et al. in 2020 gave an insight into how deep learning
techniques could benet the AEC industry in image processing, com-
puter vision and natural language processing. [94] gave insight into how
Computer Vision being a branch of AI can benet the construction in-
dustry for better prediction accuracy for onsite health and safety ana-
lytics. [95] further demonstrated the CGPT to provide safety education
and training for unrecognized hazards on construction sites that can
result in unexpected safety incidents construction professionals to
improve hazard recognition levels.
To improve building usability, [92] propound the GPT analysing
historical data, weather, occupancy and sensor data on equipment to
detect waste, predict future usage to optimize energy management and
customize resource recommendation for building users. In addition,
Saka et al. [7] demonstrated the ChatGPT prompt as a Facility Chatbot to
collect and sort occupant requests to improve facility management. No
doubt, demolition process is characterised with structural dismantling,
debris removal and hazardous material management [96] thus requiring
a comprehensive risk assessment as against the traditional risk assess-
ment which may be subjective and time consuming [97]. Thus, the GPT
leveraging on its advanced NLP and ML capabilities could provide a
more accurate risk assessment to identify and mitigate potential dangers
in construction demolition.
3. Integrating BIM with conversational AI (Conversational-BIM)
Conversational AI enables humans to interact freely with computers
in the most natural forms, this was earlier demonstrated by ELIZA in
1966 (Gentsch 2019). This process involves the use of voice commands
to interface with computer systems as advances in natural language
processing has enabled computer systems to identify valuable informa-
tion from human speeches [98]. Conversational-BIM (Con-
versational-BIM) enables interaction with the building systems
represented in BIM les with voice commands. It leverages articial
intelligence, natural language processing, and deep learning technolo-
gies to evolve an automated interaction system using voice and text
between humans and BIM les in a human-like conversational ow [99].
This technology allows construction workers the opportunity to interact
with the construction database with voice commands and text options;
and support workers to simultaneously manoeuvre between getting
S.A. Bello et al.
Digital Engineering 5 (2025) 100031
3
design information, procedure implementation, and equipment
handling. Thus, a worker holding tools with his hands can query the BIM
database with voice commands and also receive responses as voice and
text outputs. This technology thus has the potential to save time and
improve productivity in construction. Voice assistance technology has
become widespread with the diffusions from technological giants like
Amazon Alexa, Google Assistant, Microsoft Cortana, and Apple Siri. The
evolution of construction workers interaction from traditional to the
BIM era then to the Conversational- BIM era is illustrated in Fig 1. In the
traditional era, construction workers gather around to discuss con-
struction documents produced from computers starting from around
1957 (Cherkaoui, 2017).
BIM crept into the construction industries in the late 1990s [100] and
attained about 75 % increase in usage between 2007 and 2009 (Kihong
and Mojtaba, 2011). Meanwhile, with BIM evolution, various stake-
holders could interact with construction data (BIM platform) using a
keyboard and mouse. Conversational interfaces came in around 2014
(Klopfenstein et al., 2017), now culminating into the
Conversational-BIM to enable voice interaction with BIM on construc-
tion sites while possibly holding tools with hands.
3.1. Selecting cloud service for conversational-BIM
There are several cloud computing providers due to the increase in
demand for cloud technology services [101]. There are numerous
criteria for evaluating cloud services providers. Bello and Reich [102]
discussed some criteria that could be used in evaluating a cloud pro-
vider. However, these criteria could be incompatible, i.e., keeping down
operational cost while attaining high performance and service security
[103]. Though a cloud user can evaluate a cloud provider based on user
requirements ([104,105,106,107]& [108]). However, this selection may
not be made intuitively [109] as a number of decision frameworks
([110] & [111]) and selection criteria [112,113,114,115,116,117] are
available for consideration.
Attempts in selecting cloud services have been classied as decision
problems, which contain conicting criteria. A number of approaches
exist to select appropriate cloud services. Menzel et al. [43] employed an
ANP approach, while Dastjerdi et al. [39] used a logic-based method to
select an IaaS service. Godse and Mulik [109] employed an AHP-based
method to select a SaaS. Zheng et al. [118] proposed a cloud service
architecture and core algorithms to select a cloud service. Limam and
Boutaba [119] employed trust worthiness-based approach to select a
cloud service. Also, Saripalli and Pingalli [110] ranked cloud service
alternatives and adoption with simple additive weights. Martens et al.
[120] suggested community platforms approach, and similarly, Jung
et al. [121] proposed a recommendation platform to select a cloud
service. Sundareswaran et al. [44] employed a greedy algorithm-based
approach, while Yang et al. [122] described the dynamical adjustment
methods to select a cloud service. Also, Quinton et al. [123] presented an
automatic selection approach while Lee and Seo [124] employed a
hybrid MCDM model, using a balanced scorecard (BSC), fuzzy Delphi
method (FDM), and fuzzy analytical Hierarchy Process (FAHP) to select
IaaS for IT managers. Also, Ji et al. [40] considered the multi-criteria
decision-making analysis (MCDM/MCDA), and Liu and Wang [42]
employed a fuzzy algorithm to resolve uncertainties around information
for MCDM/MCDA. Fuzzy numbers are used to represent linguistic values
depicting the weights of the criteria.
Finally, intuitionistic fuzzy set [125] was employed in Liu et al.,
[41]; Liu, [126] and Liu and Li, [127]. Nevertheless, all the above
studies are geared towards selecting cloud services for general applica-
tions. However, Patil et al. [128] have opined that selecting a reliable
cloud service is dependent on the use case, thus, provided a comparative
study of cloud platforms to develop a chatbot system. Abdel-Basset et al.
[129] further in 2018 proposed an improved framework to select cloud
services for an e-learning platform. These studies though addressed
specic use cases, focused on the requirements of system developers,
whereas our study employs decisions of Conversational-BIM software
end-users to select appropriate cloud services for Conversational-BIM
implementation. Thus, this study attempts to select cloud services for
the Conversational-BIM employing/using the needs/requirements of the
construction workers who are Conversational-BIM target users. This
study establishes a structural framework for selecting cloud services for
implementing Conversational-BIM in the construction industry using
Design Thinking Methodology.
Fig. 1. Evolution of Interactions on Construction Sites.
S.A. Bello et al.
Digital Engineering 5 (2025) 100031
4
4. Research methodology
The study involved focus group discussion among practitioners in the
construction industry. The output of the discussion was analysed
thematically to obtain the user requirements for conversational BIM
implementation in the construction industries. The users’ requirements
were translated into the necessary technical specications which was
matched with appropriate cloud services. The discussion also gave the
motivations and challenges for the adoption of BIM Voice Assistant in
the Built Industry. The roadmap for the study is as depicted in Fig 2.
4.1. Establishing criteria for the selection of cloud services and providers
In gaining an in-depth understanding of industry experts’ experience
[130], a descriptive research approach was employed in this study to
obtain rst-hand information relating to real-life experiences of practi-
tioners [131]. This study employs design thinking methodology which
involves conducting focus group discussion with domain experts within
the construction industry. This is to facilitate the opportunity to inter-
relate with competent participants to gain their common understanding
of the subject matter. This approach contrasts the biased understanding
of a single individual or researcher, or marketer [132]. The approach,
therefore, assists in getting comprehensive information from industry
practitioners [133] on the criteria to select cloud services for Conver-
sational- BIM. Creswell in 2013 opined that an in-depth interview with
individual participants or interview with multiple participants (focus
group discussions) could be used to carry out data collection. Hence, this
study employed focus group discussions to support inter-subjective
opinions among participants to arrive at a common understanding, as
it avails participants to build on one another’s opinion in the course of
the discussion [134]. According to [135] purposeful sampling was
employed to determine relevant participants whose understanding is
key for the study. Selection criteria for participants were based on job
role, years of experience, interest in Conversational-BIM, eagerness, and
convenience to participate in the study. The established network of re-
searchers within the industry was used in reaching out to the partici-
pants. A similar sampling technique was found in [20,136–140]. As
such, participants were selected based on critical sampling to ensure all
professions involved in using Conversational-BIM are considered. The
professionals selected for the focus group discussion are Engineers,
Builders, MEP Professionals, and Project Managers.
In qualitative research, Polkinghorne [141] had recommended 5 to
25 participants; a total of 24 participants were involved in this study.
The participants (depicted in Table 1) were selected from the UK con-
struction rms and had 6 to 20 years of experience. The participants
have been involved in several projects within the last ve years and are
committed to using Conversational-BIM software. Two members of the
research team moderated each of the focus group discussions. Table 1
shows the number of participants in each discussion group.
The discussion started with a demonstration of a Conversational
software. The participants were then asked to state the requirements for
implementing a conversational software in the construction industries.
To spun the discussion, some set of questions were provided for the
participants. The discussion basically is to state the users’ requirements
or expectations from Conversational-BIM to be implemented in the
construction industry. The discussion lasted between 50 and 70 min. The
discussion was recorded for ease of transcription and analysis with the
consent of the discussants
4.2. Thematic analysis
Qualitative data analysis involves a systematic procedure that en-
ables a researcher to move from a narrow unit of analysis to broader
units [135]. The analytical process emanates from the identication of
signicant statements to broader units of units. In order to accomplish
this, the voice data were transcribed into written statements and read
over many times to bring out remarkable statements and crucial themes
that explain how participants would be able to present their specica-
tions which will be transformed to technical requirements to determine
appropriate cloud services to implement Conversational-BIM. To ach-
ieve this, a content-driven thematic analysis [142] was employed to
explore and identify both implicit and explicit recommendations coming
from the data. This process, otherwise called “horizonalisation” [130],
was followed by expanding clusters of meaning to highlight the basic
criteria for selecting cloud services for Conversational-BIM. Table 2
Fig. 2. Mind map of the study.
Table 1
Participants in the Focus Group Discussion.
Focus
group
Categories of participants No of
experts
Years of
experience
1. Civil and Structural
Engineers
○ Site based Engineers
5 7 −10
2. Builders 6 6–13
3. MEP Professionals
○ 3 Mechanical Engineers
○ 4 Electrical Engineers
○ 2 Plumbing Engineers
9 7–12
4. Construction Project
Managers
4 9–20
Total 24
S.A. Bello et al.
Digital Engineering 5 (2025) 100031
5
summarises the user requirements identied through the focus group
discussion.
4.3. Grouping users specications for conversational-BIM
A total of ve clusters of user’s requirements for Conversational-BIM
were identied from the discussion. These are (i) Interface (ii) Privacy
(iii) Finance (iv) and Estimation (v) Speed. The outcome from the focus
group discussion is expected to provide a guide into the user re-
quirements to be translated into technical requirements to implement
Conversational-BIM. The user’s specications that are concerned with
the interface and interaction with the software are grouped as interface.
This category involves the views, handling, and navigating of the soft-
ware to ensure ease of use of the Conversational-BIM system. The users
are also concerned about the safety and integrity of the system, this is
categorised under the privacy requirements. This module is also con-
cerned about protecting the data, as users are concerned about the
location of the data centre housing the data. The discussion also revealed
that the nancial implication of the Conversational-BIM system as a
requirement, thus bringing in nance as user’s specication. Since the
burden of the nancial implication of Conversational-BIM will be passed
to the project. Thus, the users emphasized a cheaper option like Pay-As-
You-Go that will not involve a signicant take-off cost. Also, users are
interested in a trial period that will enable a smooth take off to allow for
smooth experimentation and try-out of the Conversational-BIM system.
The focus group discussion revealed that users expect some auto com-
putations for recurring operations, for example, components selection,
list of components, Bill of Quantities Export components, Quantity Take-
off and so on. This is categorised as estimation requirement for the
Conversational-BIM system. The user’s specication also discussed the
responsiveness of the system, which refers to the speed of operation
(latency). That the latency of the Conversational-BIM system is required
to be minimal so as to reduce delay in the system’s operation. In all, ve
areas are identied in the discussion as the germane factors that inu-
enced the choice of tools to implement Conversational-BIM.
5. Framework for selecting cloud services for conversational-
BIM
This section discusses the actionable insights from the thematic
analysis of the discussion. Here, the study translates the user’s speci-
cations obtained in Section 4.1 into technical requirements and further
identied the cloud services required to implement Conversational-BIM
based on the user’s specication as depicted in Fig 3. The section also
presented the selection of a cloud service provider for Conversational-
BIM and nally the architecture of Conversational-BIM using the
selected Amazon Web Services.
5.1. Technical requirements identied from users’ specications
The requirements from the users are categorised as the target di-
mensions. The various specications obtained from the users are hereby
translated into technical requirements for the CovBIM system. The
technical requirements are further implemented with cloud services as
shown in Fig 3. Adapting Repschlaeger et al. [143], interface need is
mapped into interaction dimension, the nancial consideration is
regarded as the cost of operation, the estimation need is classied as the
functionality of the system, the need for a fast and responsive system is
categorised technically as the latency, and the privacy requirement is
translated as security dimension.
5.1.1. Interaction
The discussants want the Conversational-BIM system to be able to
accept both the speech [1] and queries as this is the growing trend [144]
and the traditional text. This implies that the system is able to accept
speech or text as input and also give out responses as text or speech. As
the user talks to the system, the recognised spoken words are displayed
on the screen. The voice response from the system is also displayed on
the screen. The interaction needs to have a clear and easy to operate
interface as the aesthetics of the interface is equally important as the
content. The interface must not be clumsy, unappealing or difcult to
use. Conversational-BIM should be intuitive [2], presenting users with
clear choices and not faced with guesses that may require spending more
time to gure out the exact message being passed. Also, the character of
the personality in Conversational-BIM should be easy to perceive. Such
that the personality should be interesting and memorable to aid the
interaction. The conversation needs to reect some sort of empathy to
create a natural scenario. “It could warn a plumber that the scaffolding is
not properly coupled, hence standing on it may be dangerous”.
Conversational-BIM interface should reect the evolving use of lan-
guages by digital technologies like emojis, GIFs in an appropriate
manner. Since users interact with construction models, these models are
of different sizes and formats. Hence, Conversational-BIM is required to
recognise construction models and convert models to usable formats for
easy manipulation, consequently justifying the need for a model con-
version module. The technical specications for the interface re-
quirements are visualisation, speech processing and model conversion
modules.
5.1.2. Functionality
The construction workers desire some functionality from the system,
expects that the system is able to automate some routine tasks, i.e.,
components selections. Conversational-BIM is expected to prove its
worth by displaying the true value of the technology and making con-
struction work easier in some ways. Conversational-BIM is desired to
meet the expectation of construction workers by computing the
Table 2
User s specications for conversational BIM.
Concerns Raised/User’s Specications Focus Group
1 2 3 4
Quick Response Time
Low latency
Fast Response
✓ ✓ ✓ ✓
Bill of Quantities Export
Analysis on Bill of quantities
Quantity Take off
✓✓ ✓
Easy Navigation
Easy Manoeuvre
Handle
Pilot
✓✓
Ease of Use
Easy to operate.
Easy to manipulate.
Not too technical
✓ ✓ ✓
Intuitive User Interface
built-in interface
Spontaneous interface
Graphical interface
Friendly interface
✓✓ ✓
Easy Model Manipulation ✓ ✓ ✓ ✓
Zoom and Pan ✓ ✓ ✓ ✓
Consistency with current branding ✓✓
Data Locality
Data in vicinity
Data within surrounding area
Data Centre must reside in UK
User must log in with email and password
✓ ✓ ✓
Data Privacy Data Isolation
Data Secrecy Data Separation
✓ ✓ ✓
Data Security Data safety Data Reliability
Data protection
Price ✓ ✓ ✓ ✓
Budget ✓ ✓ ✓
Try-out Trial period
Cooling off period Experimentation Testing
✓✓ ✓
Pay as You Go ✓
S.A. Bello et al.
Digital Engineering 5 (2025) 100031
6
regularly and repeatedly engaging operations. The focus group discus-
sion revealed that construction workers expect some sort of automation
for recurring tasks on construction site. These can be components se-
lection, list of components, suggesting the next task from the sequence of
operations, quantity take-off, bill of quantities export, and analysis on
the bill of quantities/Quantities. Conversational-BIM should be able to
display some level of intelligence even outside the essential queries.
Even for queries outside the construction, the Conversational-BIM sys-
tem should have a fall-back response that will not frustrate the user but
rather engage them in a meaningful manner. Conversational-BIM should
be able to provide/offer a way forward, especially when the request
from the user is not understood and could not nd an appropriate
answer.
5.1.3. Latency
Response time is a recognized component to measure the usability of
IT systems [145]. This is in consonance with Mishra et al. [146], where
latency was considered in the choice of cloud congurations for smart
home applications. A lengthy response time may cause lower satisfac-
tion and poor productivity among system users, which may eventually
lead to discontinuing the use of the system [147]. This requirement,
however, reects the need for the site workers that will use the system
on the construction site, as a highly responsive system will improve the
efciency of site workers. However, very rapid responses could also lead
to higher error rates [145]. Recent advances in hardware speed and
communication bandwidth have really improved response time and
system performance for IT systems. However, the speed of conversa-
tional cloud service may require more than these. For example, the
location of the server, the type of storage mechanism and efciency of
the back-end processing service will determine the overall response of
the system. To overcome the latency problem may require the need for
temporary storage to hold the model in use instead of contacting the
server for every operation. Also, this might necessitate choosing a
nearby data centre. Speed of operation is critical for site workers in order
to achieve the schedule for a given period.
5.1.4. Security
Security is concerned with authorised access and privacy of the
system. Privacy concerns were raised as important in the use of the
Conversational-BIM system during the discussion, and this is in tandem
with [148,149] and [150]. As personal voice being recorded also re-
quires protection since personal data as voice is no different to other
types of data. Although, voice-enabled technologies are expected to wait
for an activation word to prompt the system to listen and respond. Since
microphones are always on, if an activation error occurs, person-
al/private conversation could be recorded and exposed to the cloud.
Moreover, as voice data is also prone to a number of cyberattack [151].
Studies have shown that exposing an individual’s voiceprint is posed to
security risks such as spoong attacks [152] and reputation attacks
[153]. Also, the use of voice technology is improving/increasing; the
ability to mimic voice for nefarious purposes or fraud like “replay
attack” is on the increase. However, effort is on to alleviate some of the
privacy and security concern over speech data processing as demon-
strated in (Han et al., 2020).
Authentication is concerned with the process of gaining access to the
system. There are a number of ways to identify/verify users of the
platform. It could be with PIN, OTP, Password or biometric features.
Biometric authentication involves using body parts (ngerprints, voice
or iris) to authenticate a user. These body parts are not identical, even
for identical twins. However, the construction workers chose to use
organisation emails and passwords to authenticate users. This is un-
derstandable as an onsite worker may not nd it easy to swipe a nger or
adjust the iris in order to be authenticated to use a system. Rather an
organisational email may require no extra effort to be entered into the
Conversational-BIM system for authentication even while working on
site. Though being a password authentication system, it requires a
password to be strong and, if possible, not repeated for other platforms.
Storing construction data in the cloud implies that the regulation of
Fig. 3. Translating User Specications into Conversational-BIM Technical Tools.
S.A. Bello et al.
Digital Engineering 5 (2025) 100031
7
the physical location housing the server prevails over the stored data.
Since personal speech reects biological and behavioural characteris-
tics, thus speech can be classied as sensitive data as reected in the
GDPR policy [154]. Nautsch et al. [155] has argued that there is
increasing privacy concern over processing speech data. The physical
location of the server housing the speech data dictate the regulation that
will apply to the data. Thus, the discussion opined the need to choose a
provider that can store the Conversational-BIM data in a location only
approved by the construction company.
5.2. Required implementation tools for conversational-BIM
5.2.1. Authentication module
Authentication is to establish, create, manage and use identities
through a secured delivery model. It is an access and authentication
platform. It enables the construction company to control access to
Conversational-BIM resources, which are on a centralised access
authentication policies and single sign on. Authentication service in-
cludes transparent and authentication methods for identity assurance.
Authentication service in Conversational-BIM uses a username-
password matching identication system. Authentication policies
describe the level of access, which part of the Conversational-BIM can be
accessed by the different category of users. A secure and scalable
authentication platform for user identity is required. The authentication
service satises the requirements of the users for data security, as data is
only available for authorised users after due verication.
5.2.2. Storage/Database module
The need to store outside the user’s device and still achieve low la-
tency necessitates a storage medium. Storage service could be for block,
le or object storage. The service is available for various use cases
ranging from data lakes, websites, backup and restore, archive, enter-
prise applications, IoT devices, big data analytics to mobile applications.
The storage service enables Conversational-BIM data to be stored on a
storage platform outside the device of the user. The agility of
Conversational-BIM demands scalable storage as its performance can be
affected by static storage facilities. The storage provides adequate data
durability as it creates and stores copies across multiple systems. This
satises the user’s requirement for a low latency service, as this storage
enables data to be retrieved quickly thus achieving fast response.
5.2.3. Backend computing module
Back-end computing is required to process the entire application in
order to give the desired output. This module is the application layer at
the back that applies business logic to the data, culminating in imple-
menting the entire Conversational-BIM. Back-end service consumes
other necessary services in order to actualise the entire system. The
Backend computing service allows for codes to be run virtually on any
application without provisioning or managing servers. Codes can be
uploaded into the service or written in the service code editor. Codes can
be automatically triggered from other cloud services or can be called
directly from any web or mobile app using only the required resources.
Execution time can be optimised by choosing the right memory size of
functions as functions can be kept initialised in the computing service.
The Backend computing service will enable users to perform regular and
on the y computations, i.e., list of components, components selection,
exporting Bill of Quantities, estimating Quantity Take off, and deter-
mining the next task to perform.
5.2.4. Bot building platform module
Bot building platforms are employed to build and deploy interactive
conversational robots, otherwise called chatbots. Bot platforms allow
users to determine behaviours and program reaction. Bot platforms also
enable intelligent maintenance and updates. Bot service embeds voice
and text to build a conversational interface for applications. It leverages
the functionalities of automatic speech recognition (ASR) and the
natural language functionalities to create a Speech-Language Under-
standing System for conversational interactions. This cloud service takes
in the input, understand the intent and realize/accomplish/satisfy the
intent by invoking appropriate response. Bot service can easily deploy a
chatbot on mobile devices, web apps and chat services. The service or-
chestrates the dialogue by prompting for the appropriate slot to build
multi-turn conversations. The service simplies complex conversations
by dynamically transferring control from one intent to another based on
the user input. The service accepts queries and delivers response in voice
and text as required by the users.
5.2.5. Speech-to-Text module
Users of Conversational-BIM interact with the system using the
conversation which produces audio les. The Speech-to-Text service is
to convert the audio inputs from Conversational- BIM users to text les
to enable the system analyse and process them further. The service is
required as it is not possible for the Conversational-BIM system to pro-
cess audio les. This service employs automatic speech recognition
(ASR), a deep learning process, to convert speech to text accurately. The
service is suitable for audio input from microphones, audio les, video
les either as live audio streams or batch audio content. It afxes
punctuation and formatting to enable the output to closely match the
quality of the manual transcription. The service could return a time-
stamp for each word to improve the search and analysis process. It also
accepts a corpus of data to enable users to build and train customised
language models.
5.2.6. Text-to-Speech module
This Text-to-Speech service converts strings, words and sentences in
text les to audio les to enable a human user to comprehend the sound.
This service is required by Conversational-BIM to convert the text les
into synthetic human speech to be played as audio to the
Conversational-BIM user. This cloud service converts text into human
speech to create speech- enabled applications. The service uses deep
learning technologies to produce natural sounding human voices in
several languages. The cloud service supports both a Newscaster reading
style for narration use cases as well as a Conversational speaking style,
for two-way communication. The cloud service can build and train a
customised voice model.
5.2.7. Interaction module
The user specications include an intuitive user interface for easy
navigation and also easy manipulation of the construction model. This is
into two parts, interaction with the construction model and interaction
with the system interface. These requirements are met with a separate
implementation. A UI/UX application was developed to achieve the
intuitive user interface, while the Autodesk Forge API (Autodesk Forge,
[156]) service was used to manipulate the construction model, as
demonstrated in Zhang et al. [157].
5.2.8. Monitoring module
This is an analytics platform to monitor used services usage, price,
troubleshooting)-no of times called, price implications. This platform is
a monitoring and observability service that provides data and actionable
insights for applications running in the cloud, thus giving a unied view
of its operational health to respond to resource utilisation and perfor-
mance optimisation changes. The cloud service detects anomalous
behaviour, set alarms, display logs, troubleshoot issues in applications
running in the cloud environment.
5.3. Selecting cloud services provider for conversational-BIM
This section describes four big players in the cloud service world for
conversational service; that include Amazon Web Services (AWS),
Microsoft Azure, Google Cloud Services and IBM Cloud Services. The
four players were compared using the ve identied users’ specication
S.A. Bello et al.
Digital Engineering 5 (2025) 100031
8
in the study. Table 3 presented some of the features of these conversa-
tional services according to the providers while Table 4 summarises
experiences of some users of the services. On a very high-level com-
parison, all the four providers are all fast with good accuracy and
acceptable prices, AWS offers an unlimited free one-year trial period
that allows more time for experimentation with the Conversational-BIM
software, while others offered limited free usage on a monthly basis. The
interaction describes the easiness of setting up the service, the type of
input acceptable to the service, and the platform is the conversational
service can work on. The four providers under consideration accepts
both voice and text input. The AWS works with mobile service texting,
while Google Cloud and IBM Cloud uses website interface, Facebook
Messenger in addition to Google Assistant and SMS, and Nance uses the
Microsoft Azure User account.
All the providers provide a visual interface for easy navigation,
though AWS interfaces are a bit less intuitive (Matteo, 2021). Going by
the recommendations of Repschlaeger et al. [143], that price trans-
parency, price granularity and location of data are high-level priorities
for consideration when selecting a cloud provider. Consequently, AWS
with more data centre, presented an added advantage over the coverage
of the service, as this gives exibility to users in terms of choosing
location to store the Conversational-BIM data for improved latency. This
also, allows for easy compliance with the government regulations and
policies of storing data within local territory. AWS has been chosen for
Conversational- BIM software implementation as it provides a mix of the
required functionalities.
5.4. AWS cloud services for conversational-BIM
The identied cloud services necessary for Conversational-BIM were
implemented on AWS. Hence Amazon Cognito, Amazon Polly, Amazon
S3, Amazon Transcribe, Amazon Lex, Amazon Lamda and Amazon
Cloud Watch were employed to implement the Conversational-BIM
system.
Conversational BIM uses Amazon Cognito for users’ login manage-
ment, including authentication and authorization. It allows
Conversational-BIM to provision user identity management and easily
integrates with BIM 360. Furthermore, once a user gets authenticated, it
releases a token that allows access management to the various resources
available on conversational BIM. Different roles are created for users on
the Conversational-BIM app and Cognito is used to map users to
different resources based on their dened roles.
Amazon Polly provides Text-to-Speech service by using advanced
deep learning to synthesize natural sounding human speech. The
Conversational BIM uses Polly to convert user’s query fullment text to
lifelike speech and employ Polly’s Neural Text-to-Speech (NTTS) voices
to deliver conversational style readout to the user. This is also called the
AWS speech synthesizer/TTS service.
Amazon Transcribe provides quick, accurate and automated speech
to text functionality using a deep learning process called automatic
speech recognition (ASR). The Conversational BIM app. Employs Tran-
scribe to convert users’ query to text that can be passed to Amazon Lex
for further processing. Hence, the audio input captured from users are
rst sent to Transcribe to obtain the corresponding text transcript.
Amazon S3 is an extremely reliable and persistent cloud storage
system that allows object storage with scalability and security.
Conversational BIM uses S3 to store application assets and resources. In
addition, data extracted by Amazon Lambda during query fullment is
temporarily stored on Amazon S3 before serving it to the user. Thus,
bringing the data nearer to reduce latency and improve the speed of the
operations.
Amazon Lex is a service for building conversational interfaces. It
provides the advanced deep learning functionalities of automatic speech
recognition (ASR) for converting speech to text and natural language
understanding (NLU) to recognize the intent of the text. The Conversa-
tional BIM application uses Amazon Lex to extract intent and slots from a
user query. It parses the user query to understand the intent behind the
query and then extract actionable intent to enable query fullment. Lex
can be referred to as the Amazon Bot/conversational interfaces building
platform.
Amazon Lambda enables the execution of code without provisioning
or managing servers. Expert functions and business logic implementa-
tions for query fullment in conversational BIM all reside and run on
Lambda. It allows modular development of functions that are then
attached to each Lex intent. Lambda is also called the AWS Serverless
computing platform.
AWS object storage platform. Route S3 is a highly available and
scalable Domain Name System (DNS) web service. It enables reliable
routing of users’ request to valid and appropriate endpoints resources on
the conversational BIM application.
Conversational BIM application uses AWS CloudWatch for moni-
toring and building operational health overview of the various AWS
services that Conversational BIM uses. It allows detection of anomaly
behaviour and detailed troubleshooting of operational issues. Overall,
CloudWatch enables Conversational BIM to easily track the performance
of all the AWS services and allows step-tracing when issues develop.
5.5. Implementation of conversational-BIM on AWS
The Amazon Cognito is implemented in the Conversational-BIM
frontend module for users’ identity management, the implementation
architecture is illustrated in Fig 4. An authenticated user issues a verbal
query into the Conversational-BIM system through the Dashboard
module. The received verbal query (speech) is sent to the Amazon
Transcribe service for conversion to texts. The text le is passed onto the
Amazon Lex service to analyse and interpret the query’s intent and then
draws out an actionable intent to full the query. The actionable intent
is passed onto the Amazon Lambda service to effectuate the intent. The
BIM model is stored in readily accessible storage (Amazon S3) to the
Lambda service for its computations, and the fullled intent is returned
to the Amazon Polly service through Amazon Lex. The Amazon Polly
service then converts the texts from Lex to speech and returns an audio
output to the query issuer through the frontend dashboard. This process
is illustrated in Fig 5. For example, a query to compute the GFA of a
Table 3
Some features of selected cloud service providers.
Cloud
provider
Languages
supported
Mode of
operation
Custom vocabulary
support
Multi speaker
identication
Openness Data center
location
Google Over 120 Real time and
batch
Yes Yes Closed-
source
US, Asia,
Europe
Australia
Amazon 31 Real-time Yes Yes Closed-
source
US, UK, Europe, Ireland, Japan, China, Singapore,
Australia, Brazil
IBM 7 Real-Time Yes Yes Closed
-source
US, Europe
Azure 43 Real-Time Yes Yes Closed
-source
USA
(Uses Microsoft Azure data center)
S.A. Bello et al.
Digital Engineering 5 (2025) 100031
9
Table 4
Comparing the selected conversational service providers.
Google cloud Amazon web service Microsoft azure IBM cloud
Cost •Limited free usage per month
(chatbotbusinessframework.com)
•Pay per use
Free one year
Pay per use
•Limited free usage per month
(chatbotbusinessframework.com)
•Pay per use
Limited free usage/month
(chatbotbusinessframework.com)
Pay per use
Speed Good speed Good speed Good speed Good speed
Security/Privacy US, Asia,
Europe
Australia
US, UK, Europe, Ireland,
Japan, China, Singapore,
Australia, Brazil
US US, Europe
Computation strength Very good Machine applications and AI
capabilities (Scott, 2020)
Well spread infrastructure
(Scott, 2020)
Vast
Enterprise application (Scott, 2020)
Strong for Social Business (Scott,
2020)
Interaction Ease of use Easy to set up (g2.com) A bit less intuitive user
interface (Matteo, 2021)
Easy to set up [45] Not easy to set up. High quality
interaction [45]
How can
users
engage
voice and text (Ahern, 2021) voice and text (Ahern,
2021)
voice and text voice and text (Ahern, 2021)
Platform Websites, Facebook Messenger
Account, SMS, google Assistant (Ahern,
2021)
SMS (Ahern, 2021) Microsoft Azure Account Websites, Facebook Messenger
Account, SMS (Ahern, 2021)
Fig. 4. Architecture of Conversational-BIM Using Amazon Web Services.
Fig. 5. Backend of the Conversational-BIM System showing some BIM Projects.
S.A. Bello et al.
Digital Engineering 5 (2025) 100031
10
model issued through the frontend dashboard is converted to text by the
Amazon Transcribe service. Amazon Lex receives the text, extracts and
passes the intent to recognise the user’s needs (GFA model to Lambda).
The Lambda service then pulls the BIM model from Amazon S3 and
computes the GFA. The computed GFA is returned as a text via Amazon
Lex to Amazon Polly. Amazon Polly converts the text to an audio
equivalent for onward transmission to the query issuer via the dash-
board module. Figs. 5, 6 and 7, show the screenshots of the backend and
frontend of the Conversational-BIM system. The backend (Fig 6) shows a
project page with three BIM models uploaded. The backend is where the
project information and related BIM les are setup. Fig 6is platform that
allows users to query the BIM model through voice and text conversa-
tion. Fig 7 shows an example of conversation between a user and the
system.
6. Motivations and challenges for conversational-BIM in the
construction industry
The focus group discussion came up with issues that can induce the
adoption of Conversational-BIM as well as the challenges that may arise
with the adoption of the technology in AEC industries. This section
discusses the motivations and challenges of adopting Conversational-
BIM in the construction industry as elucidated from the study.
6.1. Motivations for conversational-BIM in the construction industry
According to the stakeholders that participated in the study, the
following are the incentives that can accelerate the adoption of BIM
ChatBot in the Construction Industries.
6.1.1. Improved compliance with health and safety guidelines
Since Conversational-BIM eliminates the need to memorise com-
mands or navigate through several menus for interaction, prompt safety
guidelines can be obtained for real-time usage of site workers. Con-
struction workers could get updates or regulations guiding the use of
some equipment in a timely and concise manner during use.
Conversational-BIM could provide handy information that can be com-
prehended easily instead of having to read pages of documents. Thus,
making safety guidelines accessible to workers for immediate
compliance.
6.1.2. Improved productivity
Construction workers spend a huge time doing repetitive tasks and
answering queries about these tasks [158,159]. Meanwhile, it is not easy
for construction workers to get data while on construction sites.
Therefore, using project information stored in the Conversational-BIM
database, repetitive queries about equipment, design and implementa-
tion can be easily answered by Conversational- BIM. Thus, with
Conversational-BIM providing real-time responses to these queries,
construction workers can do more meaningful activities and improve
productivity
6.1.3. Tracking real-time activity progress
Conversational-BIM is capable of accessing the efciency of a worker
to give the performance on the assigned schedule. Feedback from in-
dividuals can be used to monitor the progress of construction worker
[160]. Conversational-BIM can leverage its analytics ability to give the
progress of a project. Conversational-BIM can monitor available re-
sources and give notications about the current status. This feature will
help to replenish depleted resources without having to wait till
exhaustion which can delay project. A construction worker can ask
Conversational-BIM about the status of any equipment to verify that
everything is up and running.
6.1.4. Improved training
Conversational-BIM has the potential to aid on-site training of con-
struction workers. Conversational-BIM, as an expert knowledge base in
construction can take, for instance, site workers through the training on
the use of equipment. Conversational training is easy and accessible on
construction sites. Site workers may not nd it easy to make time for
extra learning [161]. Conversational-BIM on hands-on training can
assist a team to create BIM models where a human instructor may not be
Fig. 6. Conversational-BIM System’s Frontend showing Conversation between a User and the System.
S.A. Bello et al.
Digital Engineering 5 (2025) 100031
11
readily available for clarications and assistance
6.1.5. Improved collaboration for easy tracing of problems
Since Conversational-BIM could automate routine tasks in con-
struction process, also to improve collaboration, it can be used to
schedule meetings for stakeholders by scanning through team members
calendars and proposing a convenient time that suites everyone’s
schedule. Conversational-BIM can also track equipment inventory,
equipment status updates, delivery timing schedules, and RFI (Request
for Information – open, updated, close) status. Conversational-BIM can
browse through databases for patterns to help in automated reports
generation. These reports can give a clue to some persistent problems.
Conversational-BIM can give updates both on the project and on the
equipment in an on-the-y manner.
6.2. Challenges of conversational-BIM in construction
The discussion identied some of the exceptions that must be taken
into consideration for effective use of Conversional BIM in the AEC
sector as presented in this section.
6.2.1. Language
The ability of the conversational service to accept voice input,
recognise a language, process the intent and produce an audio output is
very crucial/fundamental to maximise/optimise the Conversational-
BIM system. Since Conversational-BIM accepts natural language input,
the choice of language accepted by the conversational service is of
importance. Conversational-BIM accepts the English language for now.
Though most of the available conversational service accepts varieties of
languages, some are still in the developmental stages. Nevertheless, the
choice of language to interact with the system is increasing. Under-
standing the semantics of a sentence is paramount to the success of
voice-based technologies. Since conversation service allows developers
to dene custom vocabularies [162], the received audio input is
matched with sample utterances from congured intents. This ability
has made intent identication relatively easy and, thus, reducing
misinterpretation of the voice commands.
6.2.2. Accents
A wide range of accents occurs in every language, as same words can
be pronounced differently, with the syllables and phonetics varying,
thus making it harder for the software to recognise. Languages are
spoken differently by different tribes, such that variation in accents can
pose a challenge, for example, the ability to understand American or
Scottish English. Conversational systems require a clear and discernible
voice as a simple mispronunciation could trick the software. Modern
conversational systems are able to accommodate a varied number of
accents, though could also be challenged when voice inputs deect too
much from the average. Construction workers using the software will
need to speak clearly and enunciate each word. A simple cold can affect
the output of a human voice. Kambeyanda et al. (1997) once argued that
using conversational software may require the user to maintain constant
pitch, volume and inexion, keeping the vocal tract musculature in a
xed position and avoid throat-clearing before starting a conversation.
Fig. 7. Screenshot of Voice Conversation Querying BIM Model.
S.A. Bello et al.
Digital Engineering 5 (2025) 100031
12
6.2.3. Background noise
Background noise is a common feature at construction sites as varied
works are going on simultaneously. Background noise blurs the sound
and could create confusion and inhibit the efcient processing of the
input sound. The conversational software should be able to lter noise
from the actual user’s input. Nevertheless, most available conversational
services employ sophisticated algorithms to lter out unwanted sound
from the user’s input.
6.2.4. Convincing the managerial team
It is true that there are many different highlighted benets of
Conversational-BIM to the construction industry, getting to convince the
managerial team for its implementation may not be an easy task. Efforts
to integrate new technologies or digital change are usually met with
internal resistance, especially from superiors Hess et al. (2017). Though,
observed that rms still face notable challenges in digital transformation
even if internal leadership are well disposed to the transformation
processes (Warner and Wager, 2019).
6.2.5. Failover management
A proper contingency plan is required to be in place as
Conversational-BIM, like any digital technology, can sometimes mal-
function as a result of bugs. However, this may be minimised through
feedback from construction workers on their experience and utilizing
such for improving the system.
7. Implications of the study
This study has implication for both academic and practice. The study
developed robust selection criteria for the implementation of cloud
services in the development of Conversational-BIM solutions based on
contributions from the stakeholder in the construction industry. This is
no doubt an addition to the academic literature in the eld of Conver-
sational AI application development. The study has also provided IT
developers with guidelines for implementing chatbots for construction
industries as it presents the relevant considerations for the development
of conversational AI applications in the industry.
This result of the study reects the comprehensive nature of the
study, as it encompasses the different categories of workers in the con-
struction industry thus reecting the different requirements of each
category of construction workers. The cost is very important to the
managerial category as this will affect the price of the project. Whereas
the speed of operation is quite germane to site workers as this will
improve their efciency. The functionality requirement analysis is
essential to site workers as well. The security requirement is important
to all categories of construction workers. Also, the visualisation re-
quirements reect the needs of all categories of construction workers.
Thus, the study presented a comprehensive requirement for the different
categories of construction workers to implement the Conversational-
BIM system.
8. Conclusion
The construction industry is embracing the voice-based Conversa-
tional-BIM to improve its productivity as seen in other profession;
healthcare, aviation, tourism, etc. The success of the Conversational-
BIM applications in the construction industry is hinged on selecting
appropriate tools for its implementation by carefully considering the
peculiarities of the industry. There are numerous incompatible criteria
to evaluate cloud service adoption to implement Conversational-BIM.
This study has interacted with construction stakeholders to get their
expectations for cloud services to implement Conversational-BIM, as
against the existing selection criteria based on the views of system de-
velopers. The study found out that speed, security, functionality, visu-
alisation and cost are important specications for construction
stakeholders and are thus employed for cloud service selection to
implement Conversational-BIM. The comprehensive study brought to
bear, the different requirements of diverse categories of construction
workers. The four leading cloud service providers for conversational
service were compared, a careful selection results in the choice of AWS.
The study went further to develop an architecture for Conversational-
BIM using AWS cloud services. Further study will evaluate the ef-
ciency of the Conversational-BIM application to different categories of
workers in the construction industry.
CRediT authorship contribution statement
Sururah A. Bello: Writing – original draft, Validation, Investigation,
Formal analysis, Data curation, Conceptualization. Lukumon O. Oye-
dele: Writing – review & editing, Validation, Supervision, Software,
Resources, Project administration, Investigation, Funding acquisition,
Conceptualization. Lukman A. Akanbi: Writing – review & editing,
Resources, Project administration, Methodology, Formal analysis.
Abdul-Lateef Bello: Writing – review & editing, Visualization, Valida-
tion, Resources, Methodology, Formal analysis, Data curation.
Declaration of competing interest
The authors declare that they have no known competing nancial
interests or personal relationships that could have appeared to inuence
the work reported in this paper.
Acknowledgement
The authors would like to express their appreciation to the Engi-
neering and Physical Science Research Council (EPSRC) United
Kingdom (Grant Ref: EP/S031480/1) for providing the nancial support
for this study.
Data availability
Data will be made available on request.
References
[1] Interaction Design Foundation, 2018. Voice user interfaces [WWW Document].
[2] C. Pearl, Designing Voice User Interfaces, O’Reilly Media, 2016.
[3] S. Ruan, J.O. Wobbrock, K. Liou, A. Ng, J.A. Landay, Comparing speech and
keyboard text entry for short messages in two languages on touchscreen phones,
in: Proc. theACM Interactive 1, Wearable Ubiquitous Technol, Mobile, 2018,
pp. 1–23.
[4] K. Lee, C. Nass, Social-psychological origins of feelings of presence: creating
social presence with machine generated voices, MediaPsychology 7 (2005)
31–45, https://doi.org/10.1207/S1532785XMEP0701_2.
[5] H. Feng, K. Fawaz, K. Shin, Continuous authentication for voice assistants, in:
ACM 23rd Annual International Conference on Mobile Computing and
Networking, 2017, pp. 343–355.
[6] Department Statista Research, 2019. Number of digital voice assistants in use
worldwide from 2019 to 2023. [WWW Document].
[7] A.B. Saka, L.O. Oyedele, L.A. Akanbi, S.A. Ganiyu, D.V.M. Chan, S.A. Bello,
Conversational articial intelligence in the AEC industry: a review of present
status, challenges and opportunities, Adv. Eng. Inf. 55 (2023) 101869.
[8] M.H. Wen, A conversational user interface for supporting individual and group
decision-making in stock investment activities, in: 4th IEEE International
Conference on Applied System Innovation 2018, ICASI, 2018, pp. 216–219.
[9] A. Pradana, G.O. Sing, Y.J. Kumar, SamBot - Intelligent conversational bot for
interactive marketing with consumer-centric approach, Int. J. Comput. Inf. Syst.
Ind. Manag. Appl 9 (2017) 265–275.
[10] F. Ali Amer Jid Almahri, D. Bell, M. Arzoky, Personas design for conversational
systems in education, Informatics 6 (2019) 46.
[11] E.V. Comendador, B.M.B. Francisco, J.S. Medenilla, S.M.T. Nacion, T.B.E. Serac,
Pharmabot: a pediatric generic medicine consultant Chatbot, J. Autom. Control
Eng. 3 (2015) 137–140.
[12] L. Laranjo, A.G. Dunn, H.L. Tong, A.B. Kocaballi, J. Chen, R. Bashir, D. Surian,
B. Gallego, F. Magrabi, A.Y.S. Lau, E. Coiera, Conversational agents in healthcare:
a systematic review, J. Am. Med. Inf. Assoc. 25 (2018) 1248–1258.
[13] S.A. Ganiyu, L.O. Oyedele, O. Akinade, H. Owolabi, L. Akanbi, A. Gbadamosi,
BIM competencies for delivering waste-efcient building projects in a circular
economy, Dev. Built Environ. 4 (2020) 100036.
S.A. Bello et al.
Digital Engineering 5 (2025) 100031
13
[14] B. Kumar, J. Cheng, L. McGibbney, Cloud computing and its implications for
construction IT, in: W. Tizani (Ed.), 13th International Conference on Computing
and Building Engineering/17th International European Group Workshop on
Intelligent Computing in Engineering, Nottingham University Press, 2010,
pp. 315–320.
[15] O.O. Akinade, L.O. Oyedele, K. Omoteso, S.O. Ajayi, M. Bilal, H.A. Owolabi,
Alaka, L. Ayris, J.H. Looney, BIM-based deconstruction tool: towards essential
functionalities, Int. J. Sustain. Built Environ. (2017).
[16] J.M. Davila Delgado, L.O. Oyedele, BIM data model requirements for asset
monitoring and the circular economy, J. Eng. Design and Technology 18 (5)
(2020) 1269–1285.
[17] J.M. Davila Delgado, L.O. Oyedele, Digital Twins for the built environment:
learning from conceptual and process models in manufacturing, Adv. Eng. Inf. 49
(2021) 101332.
[18] S.O. Abioye, L.O. Oyedele, L. Akanbi, A. Ajayi, J.M. Davila Delgado, M. Bilal, O.
O. Akinade, A. Ahmed, Articial intelligence in the construction industry: a
review of present status, opportunities and future challenges, J. Build. Eng.
(2021) 44.
[19] H. Alaka, L. Oyedele, H. Owolabi, M. Bilal, S.O. Ajayi, O. Akinade, A framework
for big data analytics approach to failure prediction of construction rms, IEEE
Trans. Eng. Manag. 66 (4) (2019) 689–698, https://doi.org/10.1109/
TEM.2018.2856376.
[20] M. Bilal, L.O. Oyedele, J. Qadir, K. Munir, S.O. Ajayi, O.O. Akinade, H.
A. Owolabi, H.A. Alaka, M. Pasha, Big Data in the construction industry: a review
of present status, opportunities, and future trends, Adv. Eng. Inf. 30 (2016)
500–521, https://doi.org/10.1016/j.aei.2016.07.001.
[21] M. Bilal, L.O. Oyedele, H. Kusimo, H.A. Owolabi, L.A. Akanbi, A.O. Ajayi, O.
O. Akinade, J.M. Davila Delgado, Investigating protability performance of
construction projects using big data: a project analytics approach, J. Build. Eng.
26 (1) (2019) 100850.
[22] A. Gbadamosi, L. Oyedele, A. Mahammadu, H. Kusimo, O. Oladimeji, The Role of
the Internet of Things in Delivering Smart Construction. Paper presented at: ; Jun
17–21; International council for Rese, CIB World Building Congress, Hong Kong
SAR, 2019.
[23] M. Bilal, L.O. Oyedele, Guidelines for applied machine learning in the
construction industry: a case of forecasting prot margins, Adv. Eng. Inf. 43
(2020) 101013.
[24] X.J. Luo, L.O. Oyedele, A.O. Ajayi, O.O. Akinade, Comparative Study of Machine
Learning-Based Multi-Objective Prediction Framework for Multiple Building
Energy Loads, 61, Sustainable Cities and Society, 2020 102283.
[25] L.A. Akanbi, A.O. Oyedele, L.O. Oyedele, R.O. Salami, Deep learning model for
demolition waste prediction in a circular economy, J. Clean. Prod. 274 (2020)
122843.
[26] T.D. Akinosho, L.O. Oyedele, B. Muhammad, A.O. Ajayi, M.D. Davila Delgado, O.
O. Akinade, Deep learning in the construction industry: a review of present status
and future innovations, J. Build. Eng. 32 (2020) 101827.
[27] J.M. Davila Delgado, L.O. Oyedele, Deep learning with small datasets: using
autoencoders to address limited datasets in construction management, Appl. Soft
Comput. 110 (2021) 107587.
[28] J.M. Davila Delgado, L. Oyedele, T. Beach, P. Demian, Augmented and virtual
reality in construction: drivers and limitations for industry adoption, J. Constr.
Eng. Manag. 146 (2020) 04020079.
[29] J.M. Davila Delgado, L.O. Oyedele, P. Demian, T. Beach, A research agenda for
augmented and virtual reality in architecture, engineering and construction, Adv.
Eng. Inf. 45 (2020) 101122.
[30] J.M. Davila Delgado, L.O. Oyedele, A. Ajayi, L. Akanbi, O. Akinade,
B. Muhammad, H…A. Owolabi, Robotics and automated systems in construction:
understanding industryspecic challenges for adoption, J. Build. Eng. 26 (2019)
100868.
[31] J.M. Davila Delgado, L.O. Oyedele, Robotics in construction: a critical review of
the reinforcement learning and imitation learning paradigms, Adv. Eng. Inf. 54
(2022) 101787.
[32] S.A. Bello, L.O. Oyedele, O.O. Akinade, M. Bilal, J.M.D. Delgado, L.A. Akanbi, A.
O. Ajayi, H.A. Owolabi, Cloud computing in construction industry: use cases,
benets and challenges, Autom. Constr. (2021) 122.
[33] C. Kamm, User interfaces for voice applications, Colloq. Pap. 92 (1995)
10031–10037, https://doi.org/10.1073/pnas.92.22.10031.
[34] Z. Adamu, K.O. Ayinla, Bridging the digital divide gap in BIM technology
adoption, Eng. Constr. Archit. Manag. 25 (2018) 1398–1416.
[35] N. Azhar, R. Farooqui, S. Ahmed, Cost overrun factors in construction industry of
Pakistan. First International Conference on Construction In Developing Countries
: Advancing and Integrating Construction Education, Research & Practice, 2008,
pp. 499–508. Karachi, Pakistan.
[36] S. Azhar, Building information modeling (BIM): trends, benets, risks, and
challenges for the AEC industry, Leadersh. Manag. Eng. 11 (2011) 241–252.
[37] A. Ghaffarianhoseini, Building Information Modelling (BIM) uptake: clear
benets, understanding its implementation, risks and challenges, Renew. Sustain.
Energy Rev. 75 (2017) 1046–1053.
[38] A. Konanahalli, M. Marinelli, L.O. Oyedele, Big data value proposition in UK
facilities management: a structural equation modelling approach, Buildings 14
(7) (2024) 2083.
[39] A.V. Dastjerdi, S.G.H. Tabatabaei, R. Buyya, An effective architecture for
automated appliance management system applying ontology-based cloud
discovery, in: 10th IEEE/ACM International Conference on Cluster, Cloud and
Grid Computing, 2010, pp. 104–112.
[40] P. Ji, H.-y. Zhang, J.-q. Wang, Fuzzy decision-making framework for treatment
selection based on the combined QUALIFLEX–TODIM method, Int. J. Syst. Sci. 48
(2017) 3072–3086.
[41] P. Liu, J. Liu, S.-M. Chen, Some intuitionistic fuzzy Dombi Bonferroni mean
operators and their application to multi-attribute group decision making, J. Oper.
Res. Soc. 69 (2018) 1–24.
[42] P. Liu, P. Wang, Some q-rung orthopair fuzzy aggregation operators and their
applications to multiple attribute decision making, Int. J. Intell. Syst. 33 (2018)
259–280.
[43] M. Menzel, M. Sch¨
onherr, S. Tai, (MC2)2: criteria, requirements and a software
prototype for cloud infrastructure decisions, Softw. Pract. Exp. 43 (2013) 1283.
–1.
[44] S. Sundareswaran, A. Squicciarini, D. Lin, A brokerage-based approach for cloud
service selection, in: 5th International Conference on Cloud Computing, 2012,
pp. 558–565.
[45] I. Petri, T. Beach, O. Rana, Y. Rezgui, Coordinating multi-site construction
projects using federated clouds, Autom. Constr. 83 (2017) 273–284, https://doi.
org/10.1016/j.autcon.2017.08.011.
[46] P. Mell, T. Grance, The NIST Denition of Cloud Computing, National Institute of
Standards and Technology, 2011.
[47] S. Wibowo, H. Deng, Consensus-based decision support for multicriteria group
decision making, Comput. Ind. Eng. 66 (2013) 625–633.
[48] C.-H. Yeh, H. Deng, S. Wibowo, Y. Xu, Multi criteria group decision support for
information systems project selection, Next-Generat. Appl. Intell. (2009)
152–161.
[49] M. Azambuja, T. Schnitzer, M. Sahin, F. Lee, Enabling lean supply with a could
computing platform - an exploratory case study, in: Annual Conference of the
International Group for Lean Construction, 2013, pp. 815–824. Fortaleza, Brazil.
[50] A. Redmond, A. Hore, R. West, Developing a Cloud integrated life cycle costing
analysis model through BIM, in: Proceedings of the International Council for
Research and Innovation in Building and Construction (CIB) CIB W078 – W102
Conference: Computer, Knowledge, Building, Sophia Antipolis, France, 2011.
[51] M. Abedi, N.M. Rawai, M.S. Fathi, M.S. Fathi, A.K. Mirasa, Cloud computing as a
construction collaboration tool for precast supply chain management, J. Teknol.
70 (2014) 1–7, https://doi.org/10.11113/jt.v70.3569.
[52] V. Getuli, S. Ventura, P. Capone, A. Ciribini, A BIM-based construction supply
chain framework for monitoring progress and coordination of site activities,
Procedia Eng. 164 (2016) 542–549, https://doi.org/10.1016/j.
proeng.2016.11.656.
[53] J. Park, K. Kim, Y.K. Cho, Framework of automated construction-safety
monitoring using cloud-enabled BIM and BLE mobile tracking sensors, J. Constr.
Eng. Manag. 143 (2016), https://doi.org/10.1061/(ASCE)CO.1943-
7862.0001223. Article 05016019.
[54] N. Tang, H. Hu, F. Xu, F. Zhu, Personalized safety instruction system for
construction site based on internet technology, Saf. Sci. 116 (2019) 161–169,
https://doi.org/10.1016/j.ssci.2019.03.001.
[55] S. Guo, H. Luo, L. Yong, A big data-based workers behavior observation in China
metro construction, Procedia Eng. 123 (2015) 190–197, https://doi.org/
10.1016/j.proeng.2015.10.077.
[56] M. Li, H. Yu, P. Liu, An automated safety risk recognition mechanism for
underground construction at the pre-construction stage based on BIM, Autom.
Constr. 91 (2018) 284–292, https://doi.org/10.1016/j.autcon.2018.03.013.
[57] X.J. Luo, L.O. Oyedele, A.O. Ajayi, O.O. Akinade, J.M. Davila Delgado, H.
A. Owolabi, A. Ahmed, Genetic algorithm-determined deep feedforward neural
network architecture for predicting electricity consumption in real buildings,
Energy AI 2 (2020) 102283.
[58] B.A. Salami, S.I. Abba, A.A. Adewunmi, U.A. Dodo, G.A. Otukogbe, L.O. Oyedele,
Building energy loads prediction using bayesian-based metaheuristic optimized-
explainable tree-based model, Case Stud. Construct. Mater. 19 (2023) e02676.
[59] I. Khajenasiri, A. Estebsari, M. Verhelst, G. Gielen, A review on internet of things
solutions for intelligent energy control in buildings for smart city applications,
Energy Procedia 111 (2017) 770–779, https://doi.org/10.1016/j.
egypro.2017.03.239.
[60] N. Rawai, M. Fathi, M. Abedi, S. Rambat, Cloud computing for green construction
management, in: International Conference on Intelligent System Design and
Engineering Applications, Hong Kong, 2013, pp. 432–435, https://doi.org/
10.1109/ISDEA.2012.107.
[61] Y.K. Cho, H. Li, J. Park, K. Zheng, A framework for cloud-based energy evaluation
and management for sustainable decision support in the built environments,
Procedia Eng. 118 (2015) 442–448, https://doi.org/10.1016/j.
proeng.2015.08.445.
[62] M. Wang, S. Qiu, H. Dong, Y. Wang, Design an IoT-based building Management
Cloud Platform for Green Buildings, Chinese Automation Congress, Juhan, China,
2017, pp. 5663–5667, https://doi.org/10.1109/CAC.2017.8243793.
[63] C. Balaras, S. Kontoyiannidis, E. Dascalaki, K. Droutsa, Intelligent services for
building information modelling- assesing variable input weather for building
simulations, Open Constr. Build. Technol. J. 7 (2013) 138–145.
[64] E. Curry, J. O’Donnell, E. Corry, S. Hasan, M. Keane, S. O’Riain, Linking building
data in the cloud: integrating cross-domain building data using linked data Adv,
Eng. Inform 27 (2) (2013) 206–219, https://doi.org/10.1016/j.aei.2012.10.003.
[65] E. Naboni, Y. Zhang, A. Maccarini, E. Hirsh, D. Lezzi, Extending the use of
parametric simulation in practice through a cloud based online service, in: IBPSA-
Italy Conference BSA, 2013, pp. 105–112. Bozen, Italy.
[66] X.J. Luo, L.O. Oyedele, Integrated life-cycle optimisation and supply-side
management for building retrotting, Renew. Sustain. Energy Rev. 154 (2022)
111827.
S.A. Bello et al.
Digital Engineering 5 (2025) 100031
14
[67] M. Fathi, M. Abedi, N. Rawai, The potential of cloud computing technology for
construction collaboration, Appl. Mech. Mater. 174 (2012) 1931–1934.
[68] M. Azambuja, J. Gong, Visualizing construction supply chains with Google cloud
computing tools. Developing the Frontier of Sustainable Design, Engineering, and
Construction, 2013, pp. 671–678. Texas, USA.
[69] A. Grilo, R. Jardim-Goncalves, Cloud-marketplaces: distributed e-procurement for
the AEC sector, Adv. Eng. Inf. 27 (2013) 160–172, https://doi.org/10.1016/j.
aei.2012.10.004.
[70] H. Ko, M. Azambuja, H. Lee, Cloud-based materials tracking system prototype
integrated with radio frequency identication tagging technology, Autom.
Constr. 63 (2016) 144–154, https://doi.org/10.1016/j.autcon.2015.12.011.
[71] M. Sahin, H. Ko, H.F. Lee, M. Azambuja, A simulation case study on supply chain
management of a construction rm adopting cloud computing and RFID, Int. J.
Ind. Syst. Eng. 27 (2017) 233–254, https://doi.org/10.1504/IJISE.2017.086269.
[72] G. Hemanth, C. Sidhartha, S. Jain, P. Saihanish, V. Rohit, AHP analysis for using
cloud computing in supply chain management in the construction industry, in:
International Conference for Convergence in Technology (I2CT), Mumbai India,
2017, pp. 1228–1233.
[73] S.A. Bello, L. Oyedele, O.K. Olaitan, K.A. Olonade, A.M. Olajumoke, A. Ajayi,
L. Akanbi, O. Akinade, M.L. Sanni, A. Bello, A deep learning approach to concrete
water-cement ratio prediction, Result. Mater. 15 (2022) 100300.
[74] X. Ferrada, D. Nú˜
nez, A. Neyem, A. Sepúlveda, M. Serpell, A lessons-learned
system for construction project management: a preliminary application, Procedia-
Social Behav. Sci. 226 (2016) 302–309, https://doi.org/10.1016/j.
sbspro.2016.06.192.
[75] J.-S. Ahn, C.-H. Park, D.-M. Lee, Y.-N. Cha, S.-Y. Chin, Development plan of
individual unit PMIS using smartphone, in: International Symposium on
Automation and Robotics in Construction, 2013, pp. 1110–1118. Waterloo.
[76] I. Petri, O. Rana, T. Beach, Y. Rezgui, A. Sutton, Clouds4Coordination: managing
project collaboration in federated clouds, in: IEEE/ACM 8th International
Conference on Utility and Cloud Computing, 2015, pp. 494–499, https://doi.org/
10.1109/UCC.2015.88. Limassol, Cyprus.
[77] T.H. Beach, O.F. Rana, Y. Rezgui, M. Parashar, Cloud computing for the
architecture, engineering & construction sector: requirements, prototype &
experience, J. Cloud Comput. Adv. Syst. Appl. 2 (2013), https://doi.org/
10.1186/2192-113X-2-8. Article 8.
[78] M. Polter, R. Scherer, Towards an adaptive civil engineering computation
framework, Procedia Eng. 196 (2017) 45–51, https://doi.org/10.1016/j.
proeng.2017.07.171.
[79] D. Nú˜
nez, X. Ferrada, A. Neyem, A. Serpell, M. Sepúlveda, A user-centered mobile
cloud computing platform for improving knowledge management in small-to-
medium enterprises in the chilean construction industry, Appl. Sci. 8 (2018) 516,
https://doi.org/10.3390/app8040516. -.
[80] Y. Jiao, Y. Wang, S. Zhang, Y. Li, L. Yang, B. Yuan, A cloud approach to unied
lifecycle data management in architecture, engineering, construction and
facilities management: integrating BIMs and SNS, Adv. Eng. Inf. 27 (2013)
173–188, https://doi.org/10.1016/j.aei.2012.11.006.
[81] H. Alaka, L. Oyedele, M. Bilal, O. Akinade, H. Owolabi, S.O. Ajayi, Bankruptcy
prediction of construction businesses: towards a big data analytics approach, in:
IEEE First International Conference on Big Data Computing Service and
Applications, 2015, pp. 347–352, https://doi.org/10.1109/
BigDataService.2015.30. Redwood City, CA, USA.
[82] A. Ajayi, L. Oyedele, J.M. Davila Delgado, L. Akanbi, M. Bilal, O. Akinade,
O. Olawale, Big data platform for health and safety accident prediction, World J.
Sci. Technol. Sustain. Dev. 16 (1) (2019) 2–21, https://doi.org/10.1108/
WJSTSD-05-2018-0042.
[83] L. Akanbi, L. Oyedele, J.M. Delgado, M. Bilal, O. Akinade, A. Ajayi,
N. Mohammed-Yakub, Reusability analytics tool for end-of-life assessment of
building materials in a circular economy World Journal of Science, Technol.
Sustain. Dev. 16 (1) (2019) 40–55, https://doi.org/10.1108/WJSTSD-05-2018-
0041.
[84] Hayman, P. (2022) Chat GPT is revolutionizing the world of architecture and
design wooduchoose https://www.wooduchoose.com/blog/ai-and-architecture/.
[85] Rane, N.L. (2023). Multidisciplinary collaboration: key players in successful
implementation of ChatGPT and similar generative articial intelligence in
manufacturing, nance, retail, transportation, and construction industry. https://
doi.org/10.31219/osf.io/npm3d.
[86] A. Saka, R. Taiwo, N. Saka, B.A. Salami, S. Ajayi, K. Akande, H. Kazemi, GPT
models in construction industry: opportunities, limitations, and a use case
validation, Dev. Built Environ. 17 (2024) 100300, https://doi.org/10.1016/j.
dibe.2023.100300. ISSN 2666-1659.
[87] S.A. Prieto, E.T. Mengiste, B. García de Soto, Investigating the use of ChatGPT for
the scheduling of construction projects, Buildings 13 (4) (2023), https://doi.org/
10.3390/buildings13040857.
[88] You, H., Ye, Y., Zhou, T., Zhu,Q. and Du, J. (2023) Robot-enabled construction
assembly with Automated sequence planning based on ChatGPT: roboGPT arXiv
preprint 10.48550/arXiv.2304.11018.
[89] Parm A.G. (2023) How ChatGPT can help in project management https://parm.
com/en/chatgpt-in-project-management/.
[90] T.H. Beach, J.L. Hippolyte, Y. Rezgui, Towards the adoption of automated
regulatory compliance checking in the built environment, Autom. Constr. 118
(2020) 103285.
[91] H. Kusimo, L. Oyedele, O. Akinade, A. Oyedele, S. Abioye, A. Agboola,
N. Mohammed-Yakub, Optimisation of resource management in construction
projects: a big data approach, World J. Sci. Technol. Sustain. Dev. 16 (2) (2019)
82–93.
[92] J. Zheng, M. Fischer, Dynamic prompt-based virtual assistant framework for BIM
information search, Autom. Constr. 155 (2023) 105067.
[93] Y. Xia, X. Lei, P. Wang, L. Sun, Articial intelligence based structural assessment
for regional short-and medium-span concrete beam bridges with inspection
information, Remote Sens. (Basel) 13 (18) (2021) 3687.
[94] S. Arshad, O. Akinade, S. Bello, M. Bilal, Computer vision and IoT research
landscape for health and safety management on construction sites, J. Build. Eng.
76 (2023) 107049, https://doi.org/10.1016/j.jobe.2023.107049. ISSN 2352-
7102.
[95] S.J. Uddin, A. Albert, A. Ovid, A. Alsharef, Leveraging ChatGPT to aid
construction hazard recognition and support safety education and training,
Sustainability. 15 (9) (2023) 7121.
[96] L.A. Akanbi, L.O. Oyedele, O.O. Akinade, A.O. Ajayi, M.D. Delgado, M. Bilal, S.
A. Bello, Salvaging building materials in a circular economy: a BIM-based whole-
life performance estimator, Resour. Conservat. Recycl., 129 (2018) 175–186.
[97] M. Alipour-Bashary, M. Ravanshadnia, H. Abbasianjahromi, E. Asnaashari,
Building demolition risk assessment by applying a hybrid fuzzy FTA and fuzzy
CRITIC-TOPSIS framework, Int. J. Build. Pathol. Adaptat. 40 (1) (2022) 134–159.
[98] Dona Kambeyanda, M., Lois Singer, D.S.P.A., Cronk, S., 1997. Potential problems
associated with use of speech recognition products. Assist. Technol. 9, 95–101.
https://doi.org/10.1080/10400435.1997.10132301.
[99] M. Bilal, L.O. Oyedele, Big data with deep learning for benchmarking protability
performance in project tendering, Expert Syst. Appl. 147 (2020) 113194.
[100] A. Konanahalli, M. Marinelli, L.O. Oyedele, Drivers and challenges associated
with the implementation of big data within UK facilities management sector: an
exploratory factor analysis approach, IEEE Trans. Eng. Manag. 69 (4) (2022)
916–929.
[101] M.F. Tutunea, SMEs’ perception on cloud computing solutions, Procedia Econ.
Financ. 15 (2014) 514–521, https://doi.org/10.1016/S2212-5671(14)00498-5.
[102] S.A. Bello, C. Reich, Cloud utility price models, in: F. Desprez, D. Ferguson,
E. Hadar, F. Leymann, M. Jarke, M. Helfert E.S. (Eds.), CLOSER 2013 -
Proceedings of the 3rd International Conference on Cloud Computing and
Services Science, SciTePress, Aachen Germany, 2013, pp. 317–320, https://doi.
org/10.5220/0004350503170320.
[103] M.-G. Avram, Advantages and challenges of adopting cloud computing from an
enterprise perspective, Procedia Technol 12 (2014) 529–534.
[104] A. Benlian, T. Hess, P. Buxmann, Drivers of SaaS-adoption – an empirical study of
different application types, Bus. Inf. Syst. Eng. 1 (2009) 357–369.
[105] M. Janssen, A. Joha, Challenges for adopting cloud-based software as a service
(saas) in the public sector, in: 19th European Conference on Information Systems
(ECIS). Helsinki, Finland, 2011.
[106] C. Low, Y. Chen, M. Wu, Understanding the determinants of cloud computing
adoption, Ind. Manag. Data Syst. 111 (2011) 1006–1023.
[107] E. Luoma, T. Nyberg, Four scenarios for adoption of cloud computing in China, in:
19th European Conference on Information Systems (ECIS), 2011. Helsinki,
Finland.
[108] M. Xin, N. Levina, Software-as-a service model: elaborating client-side adoption
factors, in: International Conference on Information Systems (ICIS), 2008, p. 8.
Paris.
[109] M. Godse, S. Mulik, An approach for selecting software-as-a-service (SaaS)
product, in: IEEE International Conference on Cloud Computing, 2009,
pp. 155–158.
[110] P. Saripalli, G. Pingali, MADMAC: multiple attribute decision methodology for
adoption of clouds, in: IEEE International Conference on Cloud Computing, 2011,
pp. 316–323.
[111] Y.-C. Lee, H. Tang, V. Sugumaran, A deployment model for cloud computing using
the analytic hierarchy process and BCOR analysis, in: 18th Americas Conference
on Information Systems (AMCIS), 2012. Seattle, WA.
[112] S. Bello, A proposed criteria for evaluating cloud-based systems, Int. J. Comput.
Innov. Res. 13 (2014) 4282–4290, https://doi.org/10.24297/ijct.v13i3.2759.
[113] P. Geczy, N. Izumi, K. Hasida, Cloudsourcing: managing cloud adoption, Glob. J.
Bus. Res. 6 (2012) 57–70.
[114] J. Hetzenecker, S. Kammerer, M. Amberg, V. Zeiler, Anforderungen Cloud
Computing Anbieter, Multikonferenz Wirtschaftsinformatik (MKWI),
Braunschweig, 2012.
[115] S. Kaisler, W.H. Money, S.J. Cohen, A decision framework for cloud computing,
in: 45th Hawaii International Conference on System Science (HICSS), 2012,
pp. 1553–1562. Maui.
[116] P. Koehler, A. Anandasivam, D. Ma, C. Weinhardt, Customer heterogeneity and
tariff biases in cloud computing, in: International Conference on Information
Systems (ICIS), 2010. St. Louis.
[117] B. Martens, F. Teuteberg, M. Gr¨
auler, Design and implementation of a community
platform for the evaluation and selection of cloud computing services: a market
analysis, in: 19th European Conference on Information Systems, ECIS 2011, 2011.
[118] Z. Zheng, X. Wu, Y. Zhang, M.R. Lyu, J. Wang, QoS ranking prediction for cloud
services, Inst. Electr. Electron. Eng. Trans. PARALLEL Distrib. Syst. 24 (2013)
1213–1222, https://doi.org/10.1109/TPDS.2012.285.
[119] N. Limam, R. Boutaba, Assessing software service quality and trustworthiness at
selection time, IEEE Trans. Softw. Eng. 36 (2010) 559–574.
[120] B. Martens, F. Teuteberg, M. Gr¨
auler, Design and implementation of a community
platform for the evaluation and selection of cloud computing services: a market
analysis, in: 19th European Conference on Information Systems, ECIS 2011, 2011.
[121] G. Jung, T. Mukherjee, S. Kunde, H. Kim, N. Sharma, F. Goetz, CloudAdvisor: a
recommendation-as-a-service platform for cloud conguration and pricing, IEEE
Ninth World Congr. Serv. (2013) 456–463.
S.A. Bello et al.
Digital Engineering 5 (2025) 100031
15
[122] J. Yang, W. Lin, W. Dou, An adaptive service selection method for cross-cloud
service composition, Concurr. Comput. Pract. Exp. 25 (2013) 2435–2454.
[123] C. Quinton, D. Romero, L. Duchien, Automated selection and conguration of
cloud environments using software product lines principles, in: 7th IEEE
International Conference on Cloud Computing, 2014, pp. 144–151.
[124] S. Lee, K. Seo, A hybrid multi-criteria decision-making model for a cloud service
selection problem using BSC, fuzzy delphi method and fuzzy AHP, Wirel. Pers.
Commun. 86 (2016) 57–75, https://doi.org/10.1007/s11277-015-2976-z.
[125] K.T. Atanassov, Intuitionistic fuzzy sets, Fuzzy. Sets. Syst. 20 (1986) 87–96.
[126] P. Liu, Multiple attribute group decision making method based on interval-valued
intuitionistic fuzzy power heronian aggregation operators, Comput. Ind. Eng. 108
(2017) 199–212.
[127] P. Liu, H. Li, Interval-valued intuitionistic fuzzy power Bonferroni aggregation
operators and their application to group decision making, Cogn. Comput. 9
(2017) 494–512.
[128] A. Patil, K. Marimuthu, R.A. Nagaraja, R. Niranchana, Comparative study of cloud
platforms to develop a Chatbot, Int. J. Eng. Technol. 6 (2017) 57–61, https://doi.
org/10.14419/ijet.v6i3.7628.
[129] M. Abdel-Basset, M. Mohamed, V. Chang, NMCDA: a framework for evaluating
cloud computing services, Futur. Gener. Comput. Syst. 86 (2018).
[130] C. Moustakas, Phenomenological Research Methods, Sage Publications, Thousand
Oak, 1994.
[131] M. Van Manen, Researching Lived Experience: Human Science for an Action
Sensitive Pedagogy, Althouse, London, Ontario, 1990.
[132] M. Crotty, The Foundations of Social Research: Meaning and Perspective in the
Research Process, Sage Publications, London, 1998.
[133] M.A. Jasper, Issues in phenomenology for researchers of nursing, J. Adv. Nurs.
(1994) 309–314, 191994.
[134] S. Kvale, InterViews: An Introduction to Qualitative Research Interviewing, Sage,
CA, USA, 1996.
[135] J.W. Creswell, Qualitative Inquiry and Research Design: Choosing Among Five
Approaches, 3rd ed., Sage, Thousand Oaks, 2013.
[136] S.O. Ajayi, L.O. Oyedele, O.O. Akinade, M. Bilal, H.A. Alaka, H.A. Owolabi,
Optimising material procurement for construction waste minimization: an
exploration of success factors, Sustain. Mater. Technol. 11 (2017) 38–46.
[137] S.O. Ajayi, L.O. Oyedele, O.O. Akinade, M. Bilal, H.A. Owolabi, H.A. Alaka,
Ineffectiveness of construction waste management strategies: knowledge gap
analysis, in: M. Okeil (Ed.), Smart, Sustainable and Healthy City. First
International Conference of the CIB Middle East and North Africa Research
Network (CIB-MENA 2014), 2014, pp. 261–280 (2014)pp. 261–280.
[138] S.O. Ajayi, L.O. Oyedele, K.O. Kadiri, O.O. Akinade, M. Bilal, H. Owolabi, H.
A. Alaka, Competency-based measures for designing out construction waste: task
and contextual attributes, Eng. Construct. Architect. Manag. 23 (4) (2016)
464–490.
[139] A. Akintoye, C. Taylor, E. Fitzgerald, Risk analysis and management of private
nance initiative projects, Eng. Constr. Arch. Manag. 5 (1998) 9–21.
[140] L.O. Oyedele, M. Regan, J.V. Meding, A. Ahmed, O.J. Ebohon, A. Elnokaly,
Reducing waste to landll in the UK: identifying impediments and critical
solutions World, J. Sci. Technol. Sustain. Dev. 10 (2013) 131–142.
[141] D.E. Polkinghorne, Phenomenological research methods, in: R.S. Valle, S. Hallin
(Eds.), Existential-Phenomenological Perspectives in Psychology, Springer,
Boston, MA, 1989.
[142] V. Braun, V. Clarke, Using thematic analysis in psychology, Qual. Res. Psychol. 3
(2006) 77–101.
[143] J. Repschlaeger, S. Wind, Z. R, K. Turowski, Decision model for selecting a cloud
provider: a study of service model decision priorities, in: Americas Conference on
Information Systems, AMCIS, Chicago, Illinois, USA, 2013.
[144] E. Corbett, A. Weber, What can I say?: addressing user experience challenges of a
mobile voice user interface for accessibility, in: 18th International Conference on
Human-Computer Interaction with Mobile Devices and Services, 2016, pp. 72–82,
https://doi.org/10.1145/2935334.2935386.
[145] B. Shneiderman, Designing the User Interface: Strategies for Effective Human-
Computer Interaction, 3rd ed., Addison-Wesley, Reading, MA, 1998.
[146] A. Mishra, T. Reichherzer, E. Kalaimannan, N. Wilde, R. Ramirez, Trade-offs
involved in the choice of cloud service congurations when building secure,
scalable, and efcient Internet-of-Things networks, Int. J. Distrib. Sens. Netw.
(2020).
[147] J.A. Hoxmeier, C. DiCesare, System response time and user satisfaction: an
experimental study of browser-based applications, in: AMCIS 2000 Proceedings,
2000, p. 347.
[148] Cho, G., Choi, J., Kim, H., Hyun, S., Ryoo, J., 2019. Threat Modeling and Analysis
of Voice Assistant Applications., in: Kang B., J.J. (Ed.), Information Security
Applications.
[149] Y. Wang, W. Cai, T. Gu, W. Shao, Y. Li, Y. Yu, Secure your voice: an oral airow-
based continuous Liveness detection for voice assistants, Proc. ACM Interact.
Mob. Wearable Ubiquitous Technol. (2019), https://doi.org/10.1145/3369811,
3, 4, Artic. 157 28 pages.
[150] S. Liao, C. Wilson, L. Cheng, H. Hu, H. Deng, Measuring the effectiveness of
privacy policies for voice assistant applications, in: Annual Computer Security
Applications Conference (ACSAC ’20)., in: Association for Computing Machinery
Annual Computer Security Applications Conference (ACSAC ’20), 2020,
pp. 856–869, https://doi.org/10.1145/3427228.3427250. New York, NY, USA.
[151] Q. Yan, K. Liu, Q. Zhou, H. Guo, N. Zhang, Surngattack: interactive hidden
attack on voice assistants using ultrasonic guided wave, in: Network and
Distributed Systems Security (NDSS) Symposium, 2020.
[152] Z. Wu, N. Evans, T. Kinnunen, J. Yamagishi, F. Alegre, H. Li, Spoong and
countermeasures for speaker verication: a survey, Speech. Commun. 66 (2015)
130–153.
[153] S. Suwajanakorn, S.M. Seitz, I. Kemelmacher-Shlizerman, Synthesizing obama:
learning lip sync from audio, ACM Trans. Graph. 36 (2017) 1–13.
[154] European Parliament and Council, 2016. Regulation (EU) 2016/679 of the
European Parliament and the Council of 27 April 2016 on the protection of
natural persons with regard to the processing of personal data and on the free
movement of such data and repealing directives 85/46/EC (General Data.
[155] Nautsch, A., Jasserand, C., Kindt, E., Todisco, M., Trancoso, I., Evans, N., 2019.
The GDPR & speech data: reections of legal and technology communities, rst
steps towards a common understanding. arXiv preprint arXiv:1907.03458.
[156] Autodesk Forge API https://forge.autodesk.com. Accessed March 2021.
[157] D. Zhang, J. Zhang, J. Guo, H. Xiong, A semantic and social approach for real-
time green building rating in BIM-based design, Sustainability. 11 (14) (2019)
3973, https://doi.org/10.3390/su11143973.
[158] L.O. Oyedele, Sustaining architects’ and engineers’ motivation in design rms: an
investigation of critical success factors, Eng. Construct. Architect. Manag. 17 (2)
(2010) 180–196.
[159] L.O. Oyedele, Analysis of architects’ demotivating factors in design rms, Int. J.
Project Manag. 31 (3) (2013) 342–354 (2013).
[160] L.O. Oyedele, K.W. Tham, B.E. Jaiyeoba, M.O. Fadeyi, Model for predicting
architect’s performance in building delivery process, J. Eng. Des. Technol. 1 (2)
(2003) 168–186.
[161] L.O. Oyedele, K.W. Tham, Clients’ assessment of architects’ performance in
building delivery process: evidence from Nigeria, Build. Environ. 42 (5) (2007)
2090–2099.
[162] Developer.amazon.com., 2018. Alexa skills kit - build for voice with Amazon.
S.A. Bello et al.
Digital Engineering 5 (2025) 100031
16
