A Blockchain 3.0 paradigm for digital twins in construction project management
Rui Zhao1, Zhe Chen2, and Fan Xue3*
This is the peer-reviewed post-print version of the paper:
Zhao, R., Chen, Z., & Xue, F. (2023). A blockchain 3.0 paradigm for digital twins in
construction project management. Automation in Construction, 145, 104645. Doi:
10.1016/j.autcon.2022.104645
The final version of this paper is available at: https://doi.org/10.1016/j.autcon.2022.104645.
The use of this file must follow the Creative Commons Attribution Non-Commercial No
Derivatives License, as required by Elsevier’s policy.

Abstract
Construction project management (CPM) is inherently complex and distributed, while digital
twin and blockchain are recognized as promising solutions for information-reliant CPM. By
learning from the lessons of Blockchain 1.0 and 2.0 paradigms in the literature, such as slow
synchronization and failed offline functions, this paper proposes ChainPM as a Blockchain
3.0 paradigm. ChainPM extends Blockchain 2.0 with innovative indexing, query, and analysis
function sets for key CPM data. Experimental results from a pilot study of a modular
construction project showed that the information synchronization latency was reduced by
99.2% to 99.8%, and query and analytical functions worked equally well without network
connections. ChainPM contributes to a novel trend of Blockchain 3.0 paradigms for CPM
digital twins, emphasizing indexing key CPM data, combinatorial query, digital authorship,
and fast response without downgrading the ‘single source of truth.’ For practitioners,
ChainPM addresses key barriers of Internet reliance and information delay to CPM digital
twins.

Rui ZHAO, MPhil Student
Department of Real Estate and Construction, The University of Hong Kong, Pokfulam, Hong Kong SAR, China
Email: jeremyrui.zhao@connect.hku.hk, ORCiD : https://orcid.org/0000-0003-1993-3129
1

Zhe CHEN, PhD Student
Department of Real Estate and Construction, The University of Hong Kong, Pokfulam, Hong Kong SAR, China
Email: cz77@connect.hku.hk ORCiD : https://orcid.org/0000-0002-8720-2638

2

Fan XUE, Assistant Professor, Ph.D.
Department of Real Estate and Construction, The University of Hong Kong, Pokfulam, Hong Kong SAR, China
Email: xuef@hku.hk, ORCiD : https://orcid.org/0000-0003-2217-3693
3

*: Corresponding author, Tel: +852 3917 4174, Fax: +852 2559 9457, Email: xuef@hku.hk

1

Keywords
Blockchain 3.0; Construction project management; Digital twins; Smart contracts; Modular
construction

Highlights
⋅

Blockchains 1.0 and 2.0 fail to meet the demands of CPM due to high latency and
Internet reliance.

⋅

A Blockchain 3.0 paradigm ChainPM is proposed for CPM digital twins.

⋅

ChainPM has unique two-step workflow: a fast response by local Cached Twin before
Blockchain.

⋅

Three decentralized CPM applications are designed for contractors, manufacturers,
and regulators.

⋅

Experiments showed ChainPM saved 99.2– 99.8% time and worked offline as a
Blockchain 3.0.

2

1 Introduction
Construction project management (CPM) is inherently complex (Bryde et al. 2013; Oraee et
al. 2017). Large quantities of construction information are often susceptible to a variety of
interrelated factors, such as multiple stakeholders with conflicts of interest (Taylan et al.
2014), complex roles of participants (Vrchota et al. 2021), uncertainties in a changeable
environment (Stanitsas et al. 2021), and out-of-date models and schedules (Xue et al. 2018).
Ballard and Howell (1994) claimed that 50% to 80% of construction site problems occur due
to a delayed receipt or lack of available information. Therefore, CPM has been questioned for
construction projects’ unsatisfactory efficiency and productivity. In comparison with the
projects in other industries, construction projects have been deemed to achieve inadequate
efficiency gains in the past 50 years (Daboun et al. 2022).
Several reasons lie behind the phenomenon of low efficiency. The first and foremost reason is
that construction is becoming increasingly complicated as technology advances, with
significant structural and building projects of unprecedented complexity (Wang et al. 2020).
Meanwhile, the intricacy of the designs has necessitated a higher level of project management
that had not been required earlier. Furthermore, project partners often have highly specialized
expertise and collaborations in modern construction projects (Liu et al. 2019). Finally, the
execution of such a project challenges prevalent practices in teamwork, trust, transparency,
and regulation (Owusu et al. 2019). Thus, the innovative CPM must learn to adapt to the
changing environment.
The recent key trend is the digital twin (DT) (PwC 2020) for information sharing, simulations
(Xue et al. 2020b), and collaborative decision-making (Wollschlaeger et al. 2017) to meet the
new management demands of the sector. Construction is an increasingly information-reliant
industry, where modern CPM depends heavily on the data shared between contracting
organizations during the life cycle of a project (Bryde et al. 2013). Secured, authentic, up-todate, and complete information in DTs enables computational simulations of the components
and systems and, thus, enhances managers’ decision-making and management with the latest
information for evidence-based decisions at any location (Kochovski & Stankovski 2018).
Successful examples include a random forest–based predictive model for the projects’ carbon
emissions at the early design stage (Fang et al. 2021), a DT of construction sites at the
disaster preparedness stage (Kamari & Ham 2022), and a fire risk assessment for the building

3

at the operation and maintenance stage (Wang et al. 2021). However, there exists a dilemma
between the centralized CPM DT and the distributed and dynamic nature of CPM.
Consequently, the industry requires systematic and fundamental changes to respond to the
dilemma.
Blockchain is an influential emerging distributed ledger technology that provides many
benefits of data security, autonomy, transparency, audit ability, privacy, immutability, and
efficiency (Scott et al. 2021). Blockchain can thus lay a foundation for distributed
administration and the creation of an irreversible ledger to facilitate collaborations. Moreover,
an identical ledger exists for all nodes participating in the blockchain network (Lu 2018). The
distributed data storage mechanism can effectively reduce the risk of data corruption, loss,
and tampering (Elli et al. 2018). The evolution of blockchains leads to four maturity levels
(i.e., Blockchain 1.0, 2.0, 3.0, and 4.0), whereas the existing blockchain applications in CPM
are confined to levels 1.0 and 2.0, as far as we are concerned (Elghaish et al. 2021). Briefly,
the focal point of Blockchain 1.0 is finance, which is often referred to as cryptocurrency
(Figueiredo et al. 2022). In Blockchain 2.0, smart contracts execute many agreement terms
automatically (Li et al. 2021b). Blockchain 3.0 emphasizes the decentralized nature of
blockchain for functions such as web user interfaces of Decentralized Applications (DApps)
(Hunhevicz & Hall 2020). Finally, in Blockchain 4.0, blockchain becomes intelligent and is
applied together with artificial intelligence, other data analytics, and Industry 4.0’s
technologies.
However, the currently adopted blockchain levels (i.e., 1.0, 2.0, and web DApps) in the
literature cannot meet the requirements for CPM DTs. The most notable drawbacks are high
latency and performance loss (Sanka & Cheung 2021). In the same blockchain network, the
nodes participating in the network have a complete ledger that records every record
completely for traceability. Therefore, as time advances, the large amount of transaction data
will make the blockchain lengthy, leading to performance loss. Furthermore, in Blockchain
2.0, the network transactions are run in a smart contract, automatically processed and
executed by the blockchain (Lu 2018). The common web DApps, though they are Blockchain
3.0, rely heavily on the Internet as well. Thus, a stable network environment is required for
every data transaction and query in Blockchain 2.0 and web DApps. However, the complex
physical environment on construction sites challenges Internet connection stability (Jin et al.
2020) and hinders the adoption of any blockchains at the levels of 1.0, 2.0, and web DApps

4

(Zhong et al. 2022).
This paper aims to present ChainPM, a Blockchain 3.0 paradigm with fast synchronization
and offline functions for CPM DTs. The key innovative mechanism in ChainPM is CPM data
transaction schema-based functions, including indexing, query, and analytics, specified for
CPM. The major contribution of this paper is twofold. First, ChainPM extends existing
Blockchain 2.0 technology to 3.0 by introducing indexing for CPM data transactions, digital
authorship, combinatorial query, and expeditious response. Second, ChainPM is an effective
management method enabling offline decision-making functions to streamline distributed
CPM collaboration rather than a ‘single source of truth’ on the Internet, addressing the
barriers of Internet reliance and information delay. The proposed ChainPM framework was
demonstrated via a case of modular construction.
The rest of the paper is organized as follows: The systematic review of CPM and blockchain
is given in Section 2, and Section 3 proposes the ChainPM framework for distributed CPM.
Section 4 presents a case study of a modular construction project, followed by experimental
results and analyses in Section 5. Discussions appear in Section 6, and the conclusion is
drawn in Section 7.

2 Literature review
2.1 Construction project management as distributed activities
A construction project, a series of long-term but closely linked organized processes, is
regarded as complex traditional activity (Walker 2015). The basic procedure of a construction
project starts with investment and bidding, then goes through the life cycle stages of design,
survey, construction, transportation, and transfer to use (Li et al. 2021a). As a result, it
inherently leads to heavy reliance on extensive data exchange (Bryde et al. 2013).
CPM is the systematic management of physically and temporally distributed activities
(Martínez-Rojas et al. 2016). In particular, many construction projects involve multiple—
sometimes global—stakeholders (e.g., architects, engineers, surveyors, and clients) with
diverse professions, in full-time or momentary employment, who are engaged in the
processes (Walker 2015). Meanwhile, stakeholders are responsible for different stages during
off-site or on-site work. As a result, various barriers impeded the recording and sharing of an

5

abundance of data, such as inaccurate and belated recording, inadequate information sharing,
ineffective communication, and inconvenient query (Daboun et al. 2022). Notably, the pains
in information sharing were magnified by the reduced face-to-face collaboration due to the
2019 coronavirus disease (Covid-19) outbreak.
The recent development of DT (Opoku et al. 2021; Lee et al. 2021; Xue et al. 2020b),
sometimes synchronized by the Internet of Things (IoT), has brought new opportunities for
real-time information gathering, timely data accessing, and efficient collaboration and
coordination among shareholders. IoT can provide accurate and timely information
collection, such as using radio frequency identification (RFID) tags in prefabricated
construction logistics (Li et al. 2022b) and real-time kinematic positioning for bridge
maintenance (Elnabwy et al. 2013). However, project data used to be stored in a centralized
database in traditional ways, where the data accessibility and security do not meet the
requirements of DT (Xue & Lu 2020a). For instance, one user may find it exhausting to
synchronize a 0.5 GB Building Information Modeling (BIM) file daily. Popular cloud
services can partly resolve the security issue but endangers the accessibility issue
simultaneously—assuming that Internet speed is slower than a project’s intranet (Kochovski
& Stankovski 2018). However, construction projects have a dynamic nature that is affected
by unstable environments and changing situations (Xue & Lu 2020a). Therefore, there exists
a dilemma between the centralized CPM DT and the distributed and dynamic nature of CPM.
2.2 Blockchain as a new distributed technology
Blockchain is an emerging technology for distributed applications (Elli et al. 2018). Three
key elements contribute to its innovative practice. First, distributed ledgers represent a shared
approach to recording transactions launched by peer nodes, through the shared approach
datasets with interacting data blocks could be recognized (Scott et al. 2021). Second,
consensus mechanisms ensure that added transaction data can join the blockchain only after
meeting a series of previously defined protocols and formulating a consensus (Anjum et al.
2020). Therefore, every new block is verified by preset conditions and its ability to meet
specific rules, such as Proof of Work (PoW) (Sanka & Cheung 2021). The third element is
cryptography, where the hash algorithm is a typical representative (Xue & Lu 2020a).
Therefore, it serves as an eligible gatekeeper for assuring data in the blockchain are
information-encrypted, modification-detected, and content-authorized, thus making the data

6

unable to be tampered with or hacked maliciously at the slightest chance.
Table 1 The evolution of blockchain (Figueiredo et al. 2022; Zhao et al. 2019)
Time

Maturity (gen.)

Example

Enabler

2008

Blockchain 1.0

Bitcoin

Distributed consensus

Transaction cost

2013

Blockchain 2.0

Ethereum

Smart contracts

Added services

2015

Blockchain 3.0

Decentralized
application (DApp)

Organizational boundaries

2018

Blockchain 4.0

Blockchain
database
Blockchain for
Industry 4.0

Decentralized AI

Value driver

Automatic AI decisionmaking

As shown in Table 1, the past few years witnessed blockchain development at an increasing
pace. Blockchain 1.0, a prominent example, is used for cryptocurrencies (e.g., Bitcoin). It is
the first symbolic component in blockchain through which financial applications realize
secure, anonymized, peer-to-peer transactions (Scott et al. 2021). Heralded by the rise of
Ethereum, blockchain embraced its 2.0 stage wherein smart contracts ‘live’ in the blockchain
in the manner of smart computer programs (Wang et al. 2020). Its decentralized structure
reduced dependency on third parties in executing and monitoring contract terms. Meanwhile,
the smart contract, such as automated protocol in construction payment
(Ahmadisheykhsarmast & Sonmez 2020), reaches the lower cost of verification, execution,
and arbitration, allowing transparent contract definition that prevents moral hazard problems.
In the Blockchain 3.0 age, decentralized storage and communication are integrated as a selfenforced blockchain platform (Maesa & Mori 2020), referred to as DApp (usually webbased). For example, secure personal health record sharing (Wang et al. 2021b), food supply
chain management (Zhao et al. 2019), and regulation compliance (Zhong et al. 2022). DApps
can digitize and customize blockchain-based solutions for business processes; the
development of Blockchain 4.0 is expected to involve automatic AI decision-making.
The conventional CPM includes paper-based or electronic management methods, which
constantly suffer from the lack of approaches supporting sufficient information integration
and security (Das et al. 2022). For instance, Sun (2020) proposed a file management system
to eliminate CPM’s ambiguity and loss of information; however, traditional file-based CPM
is neither efficient (e.g., less cost in time and space) nor effective (e.g., better accuracy) in
exchanging and validating projects’ real-time changes. Cloud-based CPM was then
developed with Autodesk BIM 360, Google Drive, and Oracle Aconex for construction data

7

storage and distribution. However, it is also unsatisfactory for data handling due to
construction projects’ temporary and fragmented nature. Meanwhile, timely accessing and
accurate data queries are still not full-blown applications in existing CPM practice, let alone
automatically grouping and searching construction information in such datasets.
Blockchain is a promising solution for the next upgrade across the CPM (Stanitsas et al.
2021). As a trending technology, blockchain has been explored and applied to several
pioneering industrial and academic studies, though the implementation in CPM practice is
still in the initial stages. For instance, Xue and Lu (2020a) proposed a semantic differential
transaction (SDT) approach to minimizing information redundancy in BIM development. Li
(2021b) developed a blockchain-enabled model for supervising spatial-temporal operations in
construction while the latency level is currently problematic. Smart contracts also have
significant value in automating construction processes, such as updating variations
(Hamledari & Fischer 2021), automatic payments (Wu et al. 2021), and documentation
approval (Das et al. 2022). It intrinsically relied on various interactions and the participation
of project shareholders in decision-making.
2.3 A dilemma identified
There exists a dilemma between the centralized CPM DT in the literature and the distributed
and dynamic nature of CPM. The recent advancement in DT has brought potential
opportunities for the digital transformation of CPM. However, three application gaps noted
can be summarized as follows: 1) The DT can be used to serve as ‘a single source of truth’ in
a centralized manner, which fails to meet the distributed applications in CPM; 2) access to
centralized DTs relies heavily on stable Internet, which cannot be guaranteed on construction
sites with complex physical environments and various certainty (Jin et al. 2020); 3)
centralized DTs are sometimes prone to security issues such as data tampering and hacking,
which undermines the information trustworthiness and accuracy.
Blockchain, as a distributed technology, has the potential to resolve the dilemma. However,
most research reported in the literature, as explorative studies, concentrated on formulating
Blockchain 1.0 or 2.0 frameworks to store CPM data, leaving the critical point of practical
utilizations for distributed CPM activities. Two application gaps of Blockchain 1.0 and 2.0
are summarized: 1) a blockchain holding a large amount of data becomes lengthy, associated

8

with performance loss and slow synchronization; 2) stable Internet is required for data
transactions and queries, leading to failed offline functions.
Thus, this research aims to develop a Blockchain 3.0 paradigm for fast synchronization and
offline functions, including efficient communication, indexing for combinatorial query,
remote analytics, and digital authorship, for localizing distributed blockchain for distributed
and dynamic CPM activities.

3 ChainPM for CPM digital twins
The research methodology of the study is shown in Figure 1. The first step aims to analyze
the existing dilemma in CPM DT by reading the literature and supplementing it with handson experiences. Then, in the next step, the study presents ChainPM as a Blockchain 3.0
paradigm that extends Blockchain 2.0 with innovative indexing, querying, and analysis
feature sets for key CPM data to address the dilemma. Then, ChainPM is implemented in a
case study of modular construction with a CPM DT of 952 volumetric modules. Finally,
experimental results gauge the performances of the proposed ChainPM in contrast with the
traditional Blockchain 2.0.

Figure 1. Research methodology in this study

9

3.1 A schema of CPM transaction data
First, a general schema of CPM transaction data is designed, as shown in Table 2. Regardless
of the various CPM activities, four common keys are defined:
⋅

DT channel: a private channel dedicated to a particular DT involving two or more
stakeholders,

⋅

component ID: the unique ID of a construction component,

⋅

transaction time, and

⋅

user ID: the user who submitted the transaction.

Besides that, other transaction data are organized as general data in a semi-structured format,
such as JSON format. Different types of data, such as numbers, texts, and images, are
compatible with the semi-structured format.
Table 2 Example of the CPM data transaction schema in this study
DT channel

Component ID

Transaction time

User ID

Channel A

ID001

2022-02-28 14:21

User A

Channel B

ID002

2022-03-02 16:37

User B

General data (semi-structured)
{“taskID”: 1, “task”: “Inspection”,
“deviceID”: 53, “photo”: [BLOB], …}
{“taskID”: 2, “task”: “Transportation”,
“vehicleID”: 994, …}

3.2 The ChainPM architecture
Figure 2 shows the ChainPM architecture. ChainPM has three layers: application, smart
contract, and Blockchain 3.0. The design of the three layers is compatible with existing
Blockchain 2.0 paradigms. However, what makes ChainPM unique is the sublayer Cached
Twin, designed to bridge the conventional Blockchain 2.0 framework and the smart contract
layer.

10

Figure 2. Blockchain 3.0 architecture of ChainPM
Application layer. This layer offers practical CPM functions for users in terms of
smartphone (or web page) graphical applications. When the Internet is unavailable, Internetreliant actions, such as saving data on the chain, will be queued and executed in the temporal
order once the Internet resumes. After a short time of enqueuing the action, users can proceed
with their other offline operations without worrying about loss of requests and data. Each
user is required to have a unique ID associated with a certificate and a private key issued
from the certification authority to access the application layer’s functions in ChainPM.
Smart contract layer. This layer realizes smart contracts, consisting of algorithms, statistics,

11

scripts, and pre-processing, for CPM DTs. The smart contract layer aims to build a bridge
between the application layer and the Blockchain 3.0 layer for data transfer and analysis.
Multiple algorithms are included in the smart contract to implement the different functions of
CPM, such as adding records and querying records. CPM statistics (e.g., volumes,
ownerships, and progress) depict the CPM DT in general based on multi-source data without
unfolding all the blockchain transactions and cryptographic hashes. Scripts can help users
more easily update and deploy smart contracts for possible security hazards. Based on the
features of smart contracts, CPM information can be pre-processed and analyzed
automatically. The algorithms, statistics, scripts, and pre-processing together also ensure data
integrity, eliminate duplication, and improve data security for CPM.
Blockchain 3.0 layer. Three sublayers are integrated into this layer:
1) Blockchain 2.0: ChainPM includes and extends the Blockchain 2.0 sublayer. For
example, Hyperledger Fabric is a matured Blockchain 2.0 with a versatile design and
many functional modules. For CPM DT, the Blockchain 2.0 sublayer can reuse the
off-the-peg consensus mechanism for data security and membership services for
multiple stakeholders.
2) Cached Twin: The Cached Twin sublayer aims to improve the indexing and analytics
functions, particularly for fast synchronization and offline functions. Indexing and
analytics are vital to Blockchain 3.0 in ChainPM. Usually, a query to a three-node
blockchain network consumes a few seconds, which is much higher than most users’
expectations for productivity and response (Nah 2004).
3) Data adaptor: The role of the data adaptor sublayer is to exchange transactions
between Blockchain 2.0 and the Cached Twin. The data adaptor regularly handles the
synchronization between the on-chain and off-chain data against different scenarios
(e.g., two synchronization requests every second). The success of the synchronization
depends on the availability and speed of the Internet connection.
ChainPM is designed for utilizing DTs for distributed CPM. Stakeholders are authorized to
join the ChainPM as nodes. Smart contracts provide stakeholders with trusted transactions
without a third party. A ChainPM node determines the authenticity of each transaction
according to predefined rules and the consensus protocol, while certain rules are designed as
well for offline mode. Integrating the transparency and immutability features of Blockchain
2.0 and the low latency and rich analytics from the new sublayer, the proposed ChainPM can

12

ensure timely access and analysis of CPM DTs. Last but not least, the information in the
Blockchain 3.0 Layer in ChainPM is as equally accurate and secured as the employed
Blockchain 2.0 backbone.
3.3 Two-step workflows in ChainPM
ChainPM has unique two-step workflows that return results twice. In existing Blockchain 2.0
frameworks, smart contracts follow a sequential execution order based on a consensus
protocol setup, where the data connection between the application layer and smart contract
closes when the response is received. The two-step workflows in ChainPM return an
immediate response from the Cached Twin before the blockchain’s final response, as shown
in Figure 3, which are different from those in Blockchain 2.0.

Figure 3. Message sequence charts of two-step workflows in ChainPM. (a) Upload records;
and (b) queries and analytics
Figure 3(a) shows the message sequence chart of the two-step workflow of uploading
records. The smart contract verifies the function and pre-processes data first according to the
predefined contract before sending pre-processed data to Cached Twin. Cached Twin
acknowledges the request promptly and then communicates with the data adaptor for

13

synchronization with the Blockchain 2.0 network. After the impacts on the blockchain, the
blockchain’s result is returned as the second response. In the meantime, the new data are
indexed asynchronously. The time-consuming executions of block packaging and key data
indexing do not block the application layer because the application layer receives the
acknowledgment of the request; therefore, the unnecessary waiting time for blockchain
packaging in conventional Blockchain 2.0 is avoided.
Figure 3(b) shows queries and analytics are handled by the smart contract in two steps. First,
the Cached Twin in ChainPM immediately returns the indexed data for the smart contract.
Then, it confirms the slow-but-trustworthy results from the blockchain network using the
indexed hash value. If no results are found in the cache, the query request is forwarded to the
blockchain for a full-chain query. In the offline situation, the data will be temporarily stored
in Cached Twin. When the network is recovered, the data adapter will synchronize with the
blockchain network at regular intervals. In CPM DT, synchronization and data querying are
particularly important (Martínez-Rojas et al. 2016).

4 A case of modular construction
4.1 Three application functions
A modular construction project was selected in Hong Kong, focusing on quality and safety
inspections to validate the proposed ChainPM. The project was a student residence at the
Wong Chuk Hang site for the University of Hong Kong with two 17-story towers, including
952 prefabricated dormitory rooms and support facilities, such as prefabricated restrooms and
common areas. First, a manufacturer, Yahgee Modular House Company Limited, produced
modular components in Zhuhai, China. Then, the components were shipped and preassembled into a volumetric module plant in Tuen Mun, Kowloon. Later, the assembled
modules would be delivered in batches to the construction site in Wong Chuk Hang District
on Hong Kong Island. Thus, the DT, in this case, involved spatially and temporally
distributed CPM activities.
Three typical CPM functions were included for validating the functionality for different
stakeholders without loss of generality. Figure 4 shows the screenshots of the DApp of
ChainPM. The first function was designed for manufacturer users to upload inspection
records of new modules to the CPM DT. The records involved a variety (consisting of facts,

14

signatures, and photos) and a big volume (> 500 MB) of inspection data. Each upload
operation, therefore, took roughly 5 to 10 minutes to store on a Blockchain 2.0 framework. It
was a painful practice in that the average number of uploads was about 40 per day. Figure
4(a) shows that ChainPM works well with various data types for CPM DT for manufacturer
users.

Figure 4. Screenshots of demanded CPM functions of ChainPM. (a) For manufacturer users;
(b) for main contractor users; and (c) for government regulator users
The second function was project statistics for main contractor users, as shown in Figure 4(b).
The main contractor demanded to check the modules’ production, delivery, pre-assembly, and
installation information on the chain for transparent and traceable project progress. Figure
4(b) shows the statistical overview of the work-in-progress modules at different stages. For
example, Figure 4(b) shows that 733 prefabricated components have been produced, 18 of
them are in shipping, and no components have been installed. The weekly statistics are
represented in a wave chart, while the project’s overall progress is shown in a gauge chart.
The third function was the approval of records for government regulator users, as shown in
Figure 4(c). The regulatory authority is expected to review and approve the documents

15

submitted by stakeholders, including the quality inspection records and assembly records.
Figure 4(c) shows daily inspection records pending approval on the DApp interface of
ChainPM. Records’ details, such as signatures of endorsers and inspection photos, are
available from the blockchain network as well.
4.2 Implementation of ChainPM
ChainPM was implemented with several software environments, including Node.js (version
16.15.1) for smart contracts, SQLite (version 3.31.1) for indexing in the Cached Twin, and
Hyperledger Fabric (version 2.2) for the Blockchain 2.0 framework, which consisted of three
organizations (i.e., a manufacturer, the main contractor, and regulatory authority), as shown
in Figure 5(a). The cryptogenic of Hyperledger Fabric is shown in Figure 5(b). At the same
time, channel configuration includes organization and profile configuration, as shown in
Figure 5(c). The default asymmetric encryption was employed for safety and effectiveness.

Figure 5. Blockchain 2.0 network configuration in ChainPM. (a) Organizations; (b) host; and
(c) channel configuration
The Cached Twin sublayer adopted a relational data structure for indexing the CPM DT, as
shown in Figure 6. Figure 6(a) shows that two more keys, i.e., taskID and taskType, were
indexed for the case study. Besides, tranID was the unique code referring to a transaction ID,
as shown in Figure 6(c). Figure 6(b) shows a set of examples of indexed data. Then, Cached
Twin utilizes a lightweight SQL service, SQLite, to index all the selected CPM on-chain data.

16

Figure 6. Example of indexing tables in ChainPM. (a) Keys to index for on-chain data
transaction; (b) data in Cached Twin; and (c) transactions in the blockchain network
With the Cached Twin sublayer and supporting data adaptor, ChainPM elevates the
conventional Blockchain 2.0 to a 3.0 paradigm, emphasizing synchronous data indexing and
advanced queries for DApps. First, the smart contract layer automatically synchronizes the
CPM transactions with the Cached Twin sublayer without user interventions for data sorting
or block packaging. Advanced queries, such as ‘selecting all the records uploaded by User 55
on 25 June 2022,’ are therefore enabled by the Cached Twin. The combinatorial conditions
can be applied to all the columns in Figure 6(b). Furthermore, the DT’s digital authorship is
cached as an enormous network of component-time-author nexus, using componentID,
tranTime, and usrID. Consequently, multiple analyses can be realized based on the
transactions and data sources, such as monitoring project schedules, worker productivity, and
digital authorship.
ChainPM safeguards data authenticity in the Cached Twin using a data adaptor module, as
shown in Figure 7. Figure 7(a) shows the indexed keys, which extended the general schema
defined in Sect. 3.1. A block in the Blockchain 2.0 framework comprises three parts,
specifically a block header, block data, and block metadata, as shown in Figure 7(b). The list
of ordered transactions of the current block is contained in block data. The data adaptor in
ChainPM synchronizes the transactions in block data to Cached Twin. Figure 7(b) shows the
attributes of an example transaction, including all the key columns in Table 2 and privacy
data like signImg, which is the user’s signature photo encoded in Base64 format. Such
privacy data were waived from indexing in Cached Twin and can only be verified for
authorized persons (e.g., government regulators) via the tranID. Moreover, the semistructured formValue is a JSON dictionary comprising the submitted transaction details.

17

Figure 7. Example data for blockchain data adaptor. (a) Target structured CPM data to index;
and (b) source blockchain data

5 Experimental results
5.1 Experimental settings
All experiments were conducted on a Docker engine (version 19.03.13) with an Intel i712770KF CPU (3.6 GHz, 8-thread), 16 GB memory, and Ubuntu 18.04.1 LTS (Linux version
5.4.0-58) operation system. For gauging the efficiency of synchronization and offline
functions, the performance metrics in this study included storage cost and response time,
which were adopted from Blockchain 2.0 applications in the literature, such as Kuzlu et al.
(2019) and Li et al. (2021b).
Storage was calculated by measuring the space changes in the operating system hard disk.
The configuration of Hyperledger Fabric was set to default, where the maximum block
interval was 2 s, the maximum number of block transactions was 500, the maximum capacity
of each block was 2 MB, and if the size of a transaction exceeded 10 MB, the transaction

18

would not be packaged (Elli et al. 2018). Experiments on storage costs were conducted with
100,000 same data transactions recorded independently for Blockchain 2.0 and ChainPM,
where 100,000 was the approximate ceiling number of all project tasks for all the 952
modules.
Response time refers to the smart contract API’s waiting execution time in milliseconds for a
transaction, which is a key performance measure of the project management system.
Response time was measured through the system Unix timestamp. The first response time of
ChainPM indicates the time the user needs to get feedback from Cached Twin, while the
second response time indicates the time the blockchain needs to finish data storage. The
metric offline response time, which is new to the existing studies, aimed to gauge the
performance of the proposed ChainPM without Internet connections. The expiration time for
the smart contract to identify the offline mode was set to 5,000 ms. The experiments on the
response time were conducted with 100 independent requests of (i) uploading new inspection
reports, (ii) listing the records of all modules, and (iii) listing specific records of a given new
module for gauging the performances.
5.2 Experimental results
5.2.1 Storage cost
The results showed that ChainPM consumed more disk space than conventional Blockchain
2.0. In the experiments, the conventional Blockchain 2.0 took 297.0 MB of extra space to
store the 100,000 data transactions. In comparison, ChainPM occupied 505.2 MB of extra
disk space, which was about 1.7 times the size of conventional Blockchain 2.0. Modern
smartphones and DApps can handle both sizes effectively and efficiently.
5.2.2 Response time
Figure 8 shows the box chart of the response time of uploading inspection records of new
modules (100 independent requests). The first response time of ChainPM was 3.53 ms on
average, while the average second response time was 2,258.4 ms. For Blockchain 2.0, the
online response time was 2,212.2 ms on average and relatively stable, whereas the offline
response time was over 5,000 ms, according to the settings. On average, ChainPM saved >
99.8% response time against conventional Blockchain 2.0, while the second response time

19

was almost the same as that of Blockchain 2.0.

Figure 8. Box chart of the response time of uploading inspection records of new modules
(100 requests per test)
Figure 9 shows the box chart of the response time of querying the records of new modules.
Figure 9(a) illustrates that Blockchain 2.0’s native query function was about 150.1 ms on
average for listing the records of all modules; Figure 9(b) shows that querying the records of
a specific new module was, on average, 157.9 ms, slower than listing all modules. In contrast,
ChainPM took merely 1.1 ms for both tasks, in which the time cost saved was > 99.2%.
Furthermore, ChainPM’s 1.10 ms for a specified module was faster than listing all data (1.13
ms) due to the efficiency of SQLite indexing for the 100-run experiments. The significant
(99.2%–99.8%) time-saving confirmed the fast synchronization of the proposed ChainPM
framework as a Blockchain 3.0 paradigm.

20

Figure 9. Experimental results for data query. (a) Listing the records of all modules and (b)
regulator’s specific record query of a new module (100 requests per test)
Regarding the offline settings, ChainPM returned the same results in the same computational
time, whereas Blockchain 2.0 failed to work. As shown in Figure 10, the average time cost of
Blockchain 2.0 for the offline cases was slightly over 5,000 ms, which indicated the user had
to wait more than 5 seconds for a ‘no Internet’ message. Figure 10(a) shows the offline mode
identified by ChainPM through a long tunnel; Figure 10(b) displays the offline mode
simulated on a smartphone emulator.

21

Figure 10. Example screenshots of offline ChainPM. (a) Offline mode identified through a
long tunnel; (b) offline mode simulated on a smartphone emulator
Overall, experimental results confirmed that ChainPM is efficient and stable for bridging the
emerging blockchain technology and DTs for distributed stakeholders and CPM activities.
Thus, ChainPM can facilitate project stakeholders in terms of communication and
collaboration, particularly where the Internet connection is unstable. The proposed ChainPM
framework and the case DApp realized a Blockchain 3.0 paradigm for CPM DTs.

6 Discussion
In contrast with the existing Blockchain 2.0 frameworks in the literature, ChainPM has three
advantages. Firstly, experimental results demonstrated that ChainPM has high-level
flexibility for variant types of DT information and meets the demands of distributed CPM.
Secondly, existing Blockchain 2.0 frameworks rely heavily on the Internet, though the
Internet connections are unstable on real construction plants and sites. In contrast, the
proposed ChainPM synchronizes the key index data with Cached Twin for online and offline
settings, while the subsequent data synchronization operation can be performed immediately

22

when the network is available. Thirdly, the DApp realizing ChainPM confirms that the
Blockchain 3.0 paradigm is feasible and adequate for CPM—for example, in building
component quality inspection, project progress monitoring, and government regulation.
Therefore, CPM researchers and practitioners can refer to the ChainPM to adopt DTs and
tackle the advent of the blockchain era.
However, ChainPM also has several limitations. Firstly, as an exploratory study, the ChainPM
system lacks large-scale experiments. Therefore, unexpected situations were not considered
thoroughly, necessitating further explorations in future research. Secondly, the indexed keys
in the Cached Twin sublayer were a subset of structured data selected from the CPM data
transactions. Due to the encryption mechanism of Blockchain 2.0, the full data transactions
were not accessible even though their briefs (as the indexed keys) were listed in offline mode.
Thus, another tractions cache can be designed for hosting the recent or highly likely 100
transactions. Thirdly, ChainPM does not fully utilize multidimensional BIM; subsequent
research thus is suggested on integrating semantically rich BIM and city information
modeling (Xue et al. 2021). Fourthly, the proposed ChainPM introduces the public smart
contract design in Hyperledger and a single-chain multi-node blockchain network. A future
direction is thus recommended on extra privacy protection and cross-chain and multi-chain
methods to ensure the blockchain’s security, scalability, and performance. Lastly, the Cached
Twin in the case study involved two additional structured data columns, i.e., taskType and
formValue, which led to manual SQL database alternation by an expert. A user-friendly
Cached Twin ‘remote controller’ will be developed in the future.

7 Conclusion
Traditional CPM is challenged by inadequate integration, laborious data accessibility, and
unsatisfactory information security. The current adoption of ICTs and DTs has brought
opportunities for upgraded information digitization and data integrity. Harnessing blockchain
technology is a potential solution to fill the gaps. The explorative applications (i.e.,
Blockchain 1.0 and 2.0) can theoretically solve the problems regarding proof of nonrepudiation and traceability for distributed CPM. However, Blockchain 2.0 also relies on a
smooth network. Yet, unstable Internet connections, especially in complex physical
environments in construction sites, can jeopardize DTs and blockchains of serviceable uses.
Therefore, this paper proposes a ChainPM framework for CPM DT. ChainPM incorporates

23

three innovative layers: (i) a smart contract layer for handling DTs with Blockchain 3.0 using
two-step workflows, (ii) a distributed Cached Twin sublayer with CPM indices and analytics,
and (iii) a blockchain data adapter layer optimized for an unstable Internet environment.
Through the unique two-step workflows in ChainPM, on-chain CPM data transactions can be
efficiently indexed and analyzed with Cached Twin services for the first response for
stakeholders’ access while not losing the authenticity and trust backed by the blockchain
network (via the second response). ChainPM explores an uncharted domain of designing and
applying Blockchain 3.0 for CPM DTs, which is vital to the practical adoption of blockchain
in the construction industry and, in turn, contributes to the scalability and extension of
blockchain theory studies in the literature.
Experimental results showed that ChainPM saved > 99.8% response time in uploading new
records and > 99.2% time in querying CPM data, respectively, where the baseline was the
well-known Blockchain 2.0 system on Hyperledger Fabric, while ChainPM used 70% more
disk space. In summary, ChainPM drastically improved the time performance of Blockchain
2.0 at a tolerable cost in disk size. Future research can be directed to (i) larger-scale
experiments with real projects, (ii) building another tool for hosting the most recent
transactions, and (iii) multidimensional data sources such as OpenBIM and city information
modeling (Li et al. 2022a), and (iv) more privacy preservation means other than data channels
in consortium blockchains.

Acknowledgment
The work presented in this paper was supported by Hong Kong Research Grants Council
(RGC) (No. 17200221) and in part by Hong Kong Innovation and Technology Fund (ITF)
(No. ITP/029/20LP). A generalized version of the Cached Twin sublayer of ChainPM is opensourced at: https://github.com/jeremyRZ/ChainPM.

References
Ahmadisheykhsarmast, S. & Sonmez, R. (2020). A smart contract system for security of
payment of construction contracts. Automation in Construction, 120, 103401.
doi:10.1016/j.autcon.2020.103401
Anjum, H. F., Rasid, S. Z., Khalid, H., Alam, M. M., Daud, S. M., Abas, H., Sam, S. M. &

24

Yusof, M. F. (2020). Mapping research trends of blockchain technology in healthcare.
IEEE Access, 8, 174244-174254. doi:10.1109/ACCESS.2020.3025011
Ballard, G. & Howell, G. (1994). Implementing lean construction: stabilizing work flow. In
Lean Construction (pp. 105-114). Rotterdam: A.A.Balkema.
Bryde, D., Broquetas, M. & Volm, J. M. (2013). The project benefits of Building Information
Modelling (BIM). International Journal of Project Management, 31(7), 971-980.
doi:10.1016/j.ijproman.2012.12.001
Daboun, O., Yusof, A. M. & Khoso, A. R. (2022). Relationship management in construction
projects: Systematic literature review. Engineering Management Journal, in press.
doi:10.1080/10429247.2022.2041962
Das, M., Tao, X., Liu, Y. & Cheng, J. C. (2022). A blockchain-based integrated document
management framework for construction applications. Automation in Construction,
133, 104001. doi:10.1016/j.autcon.2021.104001
Elghaish, F., Hosseini, M. R., Matarneh, S., Talebi, S., Wu, S., Martek, I., Poshdar, M. &
Ghodrati, N. (2021). Blockchain and the ‘Internet of Things' for the construction
industry: research trends and opportunities. Automation in Construction, 132, 103942.
doi:10.1016/j.autcon.2021.103942
Elli, A., Artem, B., Vita, B., Christian, C., Konstantinos, C., Angelo, D. C., David, E.,
Christopher, F., Gennady, L., Yacov, M., Muralidharan, S., Murthy, C., Nguyen, B.,
Sethi, M., Singh, G., Smith, K., Sorniotti, A., Stathakopoulou, C., Vukolić, M., Cocco,
S. & Yellick, J. (2018). Hyperledger fabric: a distributed operating system for
permissioned blockchains. Proceedings of the Thirteenth EuroSys Conference (p. 30).
New York, NY, USA: Association for Computing Machinery.
doi:10.1145/3190508.3190538
Elnabwy, M. T., Kaloop, M. R. & Elbeltagi, E. (2013). Talkha steel highway bridge
monitoring and movement identification using RTK-GPS technique. Measurement,
46(10), 4282-4292. doi:10.1016/j.measurement.2013.08.014
Fang, Y., Lu, X. & Li, H. (2021). A random forest-based model for the prediction of
construction-stage carbon emissions at the early design stage. Journal of Cleaner
Production, 328, 129657. doi:10.1016/j.jclepro.2021.129657
Figueiredo, K., Hammad, A. W., Haddad, A. & Tam, V. W. (2022). Assessing the usability of
blockchain for sustainability: Extending key themes to the construction industry.
Journal of Cleaner Production, 343, 131047. doi:10.1016/j.jclepro.2022.131047
Hamledari, H. & Fischer, M. (2021). Measuring the impact of blockchain and smart contracts

25

on construction supply chain visibility. Advanced Engineering Informatics, 50,
101444. doi:10.1016/j.aei.2021.101444
Hunhevicz, J. J. & Hall, D. M. (2020). Do you need a blockchain in construction? Use case
categories and decision framework for DLT design options. Advanced Engineering
Informatics, 45, 101094. doi:10.1016/j.aei.2020.101094
Jin, R., Zhang, H., Liu, D. & Yan, X. (2020). IoT-based detecting, locating and alarming of
unauthorized intrusion on construction sites. Automation in Construction, 118,
103278. doi:10.1016/j.autcon.2020.103278
Kamari, M. & Ham, Y. (2022). AI-based risk assessment for construction site disaster
preparedness through deep learning-based digital twinning. Automation in
Construction, 134, 104091. doi:10.1016/j.autcon.2021.104091
Kochovski, P. & Stankovski, V. (2018). Supporting smart construction with dependable edge
computing infrastructures and applications. Automation in Construction, 85, 182-192.
doi:10.1016/j.autcon.2017.10.008
Kochovski, P. & Stankovski, V. (2018). Supporting smart construction with dependable edge
computing infrastructures and applications. Automation in Construction, 85, 182-192.
doi:10.1016/j.autcon.2017.10.008
Kuzlu, M., Pipattanasomporn, M., Gurses, L. & Rahman, S. (2019). Performance analysis of
a hyperledger fabric blockchain framework: throughput, latency and scalability. 2019
IEEE International Conference on Blockchain (pp. 536-540). IEEE.
doi:10.1109/Blockchain.2019.00003
Lee, D., Lee, S. H., Masoud, N., Krishnan, M. S. & Li, V. C. (2021). Integrated digital twin
and blockchain framework to support accountable information sharing in construction
projects. Automation in Construction, 127, 103688.
doi:10.1016/j.autcon.2021.103688
Li, M., Xue, F. & Wu, Y. &. (2022a). A room with a view: Automatic assessment of window
views for high-rise high-density areas using City Information Models and deep
transfer learning. Landscape and Urban Planning, 226, 104505.
doi:10.1016/j.landurbplan.2022.104505
Li, X., Chi, H. L., Lu, W., Xue, F., Zeng, J. & Li, C. Z. (2021a). Federated transfer learning
enabled smart work packaging for preserving personal image information of
construction worker. Automation in Construction, 128, 103738.
doi:10.1016/j.autcon.2021.103738
Li, X., Lu, W., Xue, F., Wu, L., Zhao, R., Lou, J. & Xu, J. (2022b). Blockchain-enabled IoT-

26

BIM platform for supply chain management in modular construction. Journal of
Construction Engineering and Management, 148(2), 04021195.
doi:10.1061/(ASCE)CO.1943-7862.0002229
Li, X., Wu, L., Zhao, R., Lu, W. & Xue, F. (2021b). Two-layer Adaptive Blockchain-based
Supervision model for off-site modular housing production. Computers in Industry,
128, 103437. doi:10.1016/j.compind.2021.103437
Liu, M., Le, Y., Hu, Y., Xia, B., Skitmore, M. & Gao, X. (2019). System dynamics modeling
for construction management research: critical review and future trends. Journal of
Civil Engineering and Management, 25(8), 730-741. doi:10.3846/jcem.2019.10518
Lu, Y. (2018). Blockchain: A survey on functions, applications and open issues. Journal of
Industrial Integration and Management, 3(4), 1850015.
doi:10.1142/S242486221850015X
Maesa, D. D. & Mori, P. (2020). Blockchain 3.0 applications survey. Journal of Parallel and
Distributed Computing, 138, 99-114. doi:10.1016/j.jpdc.2019.12.019
Martínez-Rojas, M., Marín, N. & Vila, M. (2016). The role of information technologies to
address data handling in construction project management. Journal of Computing in
Civil Engineering, 4(30), 04015064. doi:10.1061/(ASCE)CP.1943-5487.0000538
Nah, F. F.-H. (2004). A study on tolerable waiting time: how long are web users willing to
wait? Behaviour & Information Technology, 23(3), 153-163.
doi:10.1080/01449290410001669914
Opoku, D.-G. J., Perera, S., Osei-Kyei, R. & Rashidi, M. (2021). Digital twin application in
the construction industry: A literature review. Journal of Building Engineering, 40,
102726. doi:10.1016/j.jobe.2021.102726
Oraee, M., Hosseini, M. R., Papadonikolaki, E., Palliyaguru, R. & Arashpour, M. (2017).
Collaboration in BIM-based construction networks: A bibliometric-qualitative
literature review. International Journal of Project Management, 35(7), 1288-1301.
doi:10.1016/j.ijproman.2017.07.001
Owusu, E. K., Chan, A. P., DeGraft, O. M., Ameyaw, E. E. & Robert, O. K. (2019).
Contemporary Review of Anti-Corruption Measures in Construction Project
Management. Project Management Journal, 50(1), 40-56.
doi:10.1177/8756972818808983
PwC. (2020). Digital twins in smart city. Hong Kong: PricewaterhouseCoopers. Retrieved
from https://www.pwccn.com/en/research-and-insights/publications/digital-twins-insmart-city.pdf

27

Sanka, A. I. & Cheung, R. C. (2021). A systematic review of blockchain scalability: Issues,
solutions, analysis and future research. Journal of Network and Computer
Applications, 195, 103232. doi:10.1016/j.jnca.2021.103232
Scott, D. J., Broyd, T. & Ma, L. (2021). Exploratory literature review of blockchain in the
construction industry. Automation in Construction, 132, 103914.
doi:10.1016/j.autcon.2021.103914
Stanitsas, M., Kirytopoulos, K. & Leopoulos, V. (2021). Integrating sustainability indicators
into project management: The case of construction industry. Journal of Cleaner
Production, 279, 123774. doi:10.1016/j.jclepro.2020.123774
Sun, J., Lei, K., Cao, L., Zhong, B., Wei, Y., Li, J. & Yang, Z. (2020). Text visualization for
construction document information management. Automation in Construction, 111,
103048. doi:10.1016/j.autcon.2019.103048
Taylan, O., Bafail, A. O., Abdulaal, R. M. & Kabli, M. R. (2014). Construction projects
selection and risk assessment by fuzzy AHP and fuzzy TOPSIS methodologies.
Applied Soft Computing, 17, 105-116. doi:10.1016/j.asoc.2014.01.003
Vrchota, J., Řehoř, P., Maříková, M. & Pech, M. (2021). Critical Success Factors of the
project management in relation to Industry 4.0 for sustainability of projects.
Sustainability, 13(1), 281. doi:10.3390/su13010281
Walker, A. (2015). Project Management in Construction (6th ed.). John Wiley & Sons.
Wang, L., Li, W., Feng, W. & Yang, R. (2021). Fire risk assessment for building operation
and maintenance based on BIM technology. Building and Environment, 205, 108188.
doi:10.1016/j.buildenv.2021.108188
Wang, Y., Zhang, A., Zhang, P., Qu, Y. & Yu, S. (2021b). Security-aware and privacypreserving personal health record sharing using consortium blockchain. IEEE Internet
of Things Journal, 9(14), 12014-12028. doi:10.1109/JIOT.2021.3132780
Wang, Z., Wang, T., Hu, H., Gong, J., Ren, X. & Xiao, Q. (2020). Blockchain-based
framework for improving supply chain traceability and information sharing in precast
construction. Automation in Construction, 111, 103063.
doi:10.1016/j.autcon.2019.103063
Wollschlaeger, M., Sauter, T. & Jasperneite, J. (2017). The future of industrial
communication: Automation networks in the era of the internet of things and industry
4.0. IEEE Industrial Electronics Magazine, 11(1), 17-27.
doi:10.1109/MIE.2017.2649104
Wu, W., He, H., Chu, Z., Liu, Y., Shi, L., Gu, M., Zhang, T., Lin, Z. & Yang, L. (2021).

28

Blockchain-based smart payment scheme of power infrastructure funds. Energy
Reports, 7(S7), 725-733. doi:10.1016/j.egyr.2021.09.199
Xue, F. & Lu, W. (2020a). A semantic differential transaction approach to minimizing
information redundancy for BIM and blockchain integration. Automation in
Construction, 118, 103270. doi:10.1016/j.autcon.2020.103270
Xue, F., Chen, K., Lu, W., Niu, Y. & Huang, G. Q. (2018). Linking radio-frequency
identification to Building Information Modeling: Status quo, development trajectory
and guidelines for practitioners. Automation in Construction, 93, 241-251.
doi:10.1016/j.autcon.2018.05.023
Xue, F., Lu, W., Chen, Z. & Webster, C. J. (2020b). From LiDAR point cloud towards digital
twin city: Clustering city objects based on Gestalt principles. ISPRS Journal of
Photogrammetry and Remote Sensing, 167, 418-431.
doi:10.1016/j.isprsjprs.2020.07.020
Xue, F., Wu, L. & Lu, W. (2021). Semantic enrichment of building and city information
models: A ten-year review. Advanced Engineering Informatics, 47, 101245.
doi:10.1016/j.aei.2020.101245
Zhao, G., Liu, S., Lopez, C., Lu, H., Elgueta, S., Chen, H. & Boshkoska, B. M. (2019).
Blockchain technology in agri-food value chain management: A synthesis of
applications, challenges and future research directions. Computers in Industry, 109,
83-99. doi:10.1016/j.compind.2019.04.002
Zhong, B., Guo, J., Zhang, L., Wu, H., Li, H. & Wang, Y. (2022). A blockchain-based
framework for on-site construction environmental monitoring: Proof of concept.
Building and Environment, 217, 109064. doi:10.1016/j.buildenv.2022.109064

29