Exploratory literature review of blockchain in the construction industry

First academic publications on blockchain in construction instantiated in 2017, with three documents. Over the course of several years, new literature emerged at an average annual growth rate of 184%, surmounting to 121 documents at time of writing this article in early 2021. All 121 publications were reviewed to investigate the expansion and progression of the topic. A mixed methods approach was implemented to assess the existing environment through a literature review and scientometric analysis. Altogether, 33 application categories of blockchain in construction were identified and organised into seven subject areas, these include (1) procurement and supply chain, (2) design and construction, (3) operations and life cycle, (4) smart cities, (5) intelligent systems, (6) energy and carbon footprint, and (7) decentralised organisations. Limitations included using only one scientific database (Scopus), this was due to format inconsistencies when downloading and merging various bibliographic data sets for use in visual mapping software.


Introduction
Blockchain is the technology that enables triple entry accounting, which allows multiple parties to transact across a shared synchronous ledger. Each transaction is substantiated with a digital signature to provide proof of its authenticity [1]. Blockchain includes several key features, such as decentralised, distributed, and consensus [2]. A typical public blockchain is comprised of thousands of computer nodes connected through a decentralised network, and it does not require a central power of authority to manage the system [3]. Blockchain is a selfsustaining network that rewards users for participating in mining, which is the process of creating new blocks and distributing them across all nodes on the network [4]. Whenever transactions are sent to the network, they are placed in a pool of unverified transactions, where they are periodically collected and validated by miners before they are placed into a block [5]. Miners apply a consensus mechanism to check each other's results prior to the inclusion of new blocks, this is to ensure that there is only one version of the ledger in existence at any moment in time [6]. Bitcoin was the first blockchain which came into existence in 2009, since then, its protocol has proved immutable to hacks and has not suffered accounting errors, such as double spending [7]. Ethereum was the second blockchain to come into existence, which emerged in 2015 and introduced smart contracts, which allowed transacting parties to codify and deploy peer-to-peer agreements without the reliance of a trusted third party [8]. Smart contracts include a unique property, in that they cannot be changed once deployed, which mitigates against users unfairly withdrawing from signed agreements [9]. Smart contracts disallow external entities from interfering with peer-to-peer contracts and enables atomic transactability. The codified terms of a smart contracts are transparent and open for auditing, which allows transacting parties to verify agreements for consistency.
The timescale of this review spans from 2017 to 2021 and incorporates 121 academic documents. A bottom-up method was implemented to assess the existing environment through a literature review, which includes an exploratory investigation of the progression of the topic across a wide range of application categories. The document types used in the review are comprise of journal articles, conference papers, and book chapters. Non-academic sources such as company reports were not included into the review as they do not include the same level of scientific rigor as peer-reviewed content, furthermore, the quantity of documents attainable from academic sources were of sufficient quantity.
Two search queries were conducted on the Scopus scientific database, which was used to obtain all of the reviewed documents. The research method chapter displays the structure of these queries diagrammatically; furthermore, the search string for retrieving the results is available in the appendix, which allows users to replicate the search results. Other scientific databases that were considered include Web of Science (WoS), IEEE Xplore, Science Direct, Directory of Open Access Journals (DOAJ), and JSTOR [10]. Based on the topic of blockchain in construction, Scopus and WoS included the largest quantity of results by a substantial margin. In a comparison of these, it revealed Scopus with 53% more content, and with 85% of the WoS data already existent in Scopus. Both databases included a balanced range of top tier journals (top 25% based on Scientific Journal Ranking indicator), while Scopus included a larger number of mid to lower tier journals.
The first academic literature on blockchain in construction emerged in 2017 within the categories of Building Information Modelling (BIM) [11], smart cities [12], and peer-to-peer energy markets [13]. The quantity of new publications on topic increased at an annual growth rate of 184% each year since 2017. The quantitative aspect of this article provides data on the expansion of the topic through statistics and scientometrics. VOS-Viewer was used to present scientometric data through visual mapping. The literature review chapter was structured around application categories of blockchain in construction. Each category was substantiated by a minimum of three documents to ensure a level of academic consensus was achieved. Altogether, 33 application categories were investigated and organised into seven subject areas, which are (1) procurement and supply chain, (2) design and construction, (3) operations and life cycle, (4) smart cities, (5) intelligent systems, (6) energy and carbon footprint, and (7) decentralised organisations. An exploratory method was implemented to encapsulate a wide range of categories to investigate the existing environment through a macro-orientated approach. This method aligns with the quantitative analysis that was conducted as part of this review.

Related works
From the 121 reviewed documents in this article, six included reviews of similar nature and are displayed in Table 1. From these, four delimited their results to academic documents, while two incorporated a combination of academic and non-academic sources. The Non-academic material included company and organisation reports [14]. An expansive literature review of 121 documents on blockchain in construction from academic publications have only recently been feasible since 2021, as there is now an established body of work on the topic. Blockchain is a fast-evolving technology, and this article builds upon the work displayed in Table 1 to provide an updated review on the contemporary state of the topic.
Bhushan et al., conducted a comparative literature review of blockchain in smart cities, published in Sustainable Cities and Society journal, which outlined six subject areas and eight categories [15]. Hunhevicz & Hall, produced a literature review of blockchain in construction, published in Advanced Engineering Informatics journal, which included seven categories and 24 use-cases [16]. Kiu, et al., composed a systematic review of blockchain in construction, published in the International Journal of Construction Management, and outlined six subject areas [14]. Li et al., composed a systematic literature review published in Automation in Construction, which extrapolated seven built environment application categories; furthermore, three use-cases were substantiated through interviews with academics and industry practitioners, such as "automated project bank accounts", "regulation and compliance", and "single shared-access BIM model" [17]. Perera, et al., produced a literature review article on blockchain in construction published in the Journal of Industrial Information Integration, and identified 18 categories, extracted from academic and non-academic sources [7]. Yang et al., included a literature review in their blockchain proof of concept article published in Automation in Construction, which summarised four subject categories for managing business processes [18].

Research method
Content was collected from journal articles, book chapters, and conference proceedings. Scopus was selected as the scientific database for extracting documents, as it contained the largest bibliographic index of academic literature on the topic, and is reputably owned by publishing organisation Elsevier [19]. Reason for using only one scientific database is due to format inconsistencies when merging data sets from various databases. When conducting a parallel search on Scopus and Web of Science (WoS) (the top two largest academic indexes on the topic) [20], it revealed Scopus with 53% more content, and with 85% of WoS documents already existent in Scopus, thus Scopus was selected as the database of choice. Fig. 1 displays the two search queries. Search one incorporated inputting the ISSN and ISBN numbers of journals and books within the subject category of architecture, building and construction, and civil and structural engineering, followed by the key words shown in the search query column in Table 1. The ISSN and ISBN number is a unique identifier given to each journal and book, which can be downloaded from https://www.scimagojr.com. The SCImago web portal provides an index of academic publishers for each specific subject area [21]. The Scopus web portal allows users to search for documents according to a predefined list of subject areas, in this case SUBJAREA(engi) was implemented into query two, with key terms such as blockchain and construction. Two queries were used to increase the accuracy of results from Scopus, which returned to a combined total of 412 documents. Upon removing duplicates and filtering content for suitability, the final result surmounted to 121 publications.

Quantitative analysis
Fig. 2 displays the quantity of documents published each year, documents types, and scientific journal rankings (SJR). SJR is the impact factor of each journal, which is calculated through a network analysis of citations [22]. SJR is measured in quartiles, whereby, Q1 represents the top 25% of journals, while Q4 is the lowest 25% [22]. The statistics in Fig. 2 were obtained through conducting a search using the queries listed in Fig. 1. The results in Fig. 2 are based on full complete years, in this case 2017-2020. This article was written in 2021, thus results from that year were not included.
The subject areas and categories of the literature review are displayed in Fig. 3. Each category was substantiated by a minimum of three documents to ensure a level of academic consensus was achieved. These categories were further organised into seven subject areas for the purpose of adding structure when organising correlating categories together. Fig. 4  Note: a Includes content from non-peer reviewed sources (e.g., reports). b Includes many use-cases that were not individually itemised by its author.   with 28 documents each includes internet of things (IoT), supply chain management, and smart grids. While peer-to-peer energy markets is third place with 27 documents. The newest categories which emerged in 2020 included machine learning, water management, physical waste management, geospatial, and Integrated Project Delivery (IPD). The topical coverage of each of the 121 reviewed documents were manually recorded and transferred into visual mapping software VOSviewer, to produce the Fig. 6 visual map. VOS-viewer algorithmically maps data using natural language processing techniques [23]. Fig. 6 is broken down into three parts, which includes categories (shown as circular nodes), colour clusters (shown as the groups of nodes displayed in blue, green, yellow, or red), and links (which are the lines that connect the nodes together). Each of the reviewed documents typically covered a range of categories. Illustrating the overlap/co-occurence of these categories is the purpose of the Fig. 6 co-occurrence map. Colour clusters are assigned when a group of categories frequently co-occur in the reviewed documents. Categories with a high number of shared links naturally gravitate to the centre, as a central position has greater equidistance with its shared links. However, categories also gravitate to each other based on their link strength, whereby, if two categories appear frequently together in literature, they will be positioned close to each other on the Fig. 6 map. Blockchain was positioned most centrally as it shares links with all of the 33 categories. BIM was also positioned centrally as it shared links with 32 out of the 33 total categories. Whereas IPD, carbon accounting, fintech and off-site construction were all positioned in the outskirt, due to their low number of shared links with the overall categories. Table 2 displays the results from Fig. 6. The table is sorted from largest to smallest according to links, followed by link strength, then occurrences. The Link strength is calculated by the number of times each category co-occurs with another. While the occurrences is calculated by the number of times each category appears in literature irregardless of its link strength. The results show that 89% of the reviewed documents included multiple categories in their paper, while 11% focused their attention solely on one category. Fig. 7 displays which blockchain platforms were most utilised in the reviewed documents. 18 documents developed solutions for Ethereum [8], while 14 developed solutions for Hyperledger [24]; additionally, one publication investigated utilising both platforms [18]. Ethereum emerged in 2015 as a public blockchain platform; furthermore, it is currently the leading platform for decentralised applications and includes the largest population of blockchain developers [25]. Hyperledger, by the Linux Foundation, instantiated their own variant in the same year (2015) using a private blockchain protocol [26]. Less popular platforms in the reviewed material include Multiledger [27], Bitcoin [28], Corda [29], and IOTA [30]. Fig. 8 displays the various types of data collection implemented in the reviewed documents. A conceptual framework was incorporated in 46% of documents, which was used as a foundation to formulate highlevel ideas [31]. Case studies were also a popular method used in 27% of documents, which included joint ventures between academia and industry [32]. Literature reviews were used in 26% of the documents, which were typically implemented as a prerequisite to support the development of conceptual frameworks [33], such as with the Brooklyn micro-grid project, which used a literature review to assess the existing environment prior to the implementation of a case study [34]. Statistics was incorporated in 23% of documents, such as with measuring the performance of blockchain-based network systems [35]. The types of data collection which appeared less frequently included systematic reviews (12%), proof of concepts (12%), interviews (7%), surveys (7%), and questionnaires (1%). Fig. 9 displays a visual map showing the co-occurrences of the data collection types shown in Fig. 8. Fig. 9 displays links shown in red numerals and link strength shown in blue numerals. From analysing the diagram, the top three data collection types which co-occurred most frequently in the reviewed literature included conceptual frameworks, statistics, and case studies, demonstrated through their high link strength count shown in blue numerals. The outer position of systematic reviews revealed that it co-occurred less frequently than literature reviews, however, this particular statistic can be misleading, as both systematic and literature reviews are terms used interchangeably throughout research; however, the author ensured not to interfere with the terminologies provided in the reviewed documents. 12 publications conducted a proof of concept (PoC), which surmounts to 10% of the reviewed documents. The data collection types with the least number of co-occurrences included questionnaires, systematic reviews, and surveys. Altogether, 55% of the reviewed documents incorporated multiple data collection types in their research, while 45% included only one. Through conducting this review, the author noticed that papers which included higher numbers of data collection types were typically less technical overall, such as literature/systematic reviews. While papers which included only one data collection type were typically more in-depth, such as with a PoC. Table 3 is to be read in conjunction with Fig. 9, and is organised according to link count, total link strength, and occurrences. Link count refers to the quantity times a particular type of data collection co-occurs with another; however, it does not take into account the weight if each link. Whereas link strength factors in the weight, which refers to the cumulative total of when each link co-occurred with another. The occurrences column represents the quantity of times each data collection type occurred in literature regardless of its links or link strength.

Literature review
The literature review is broken down into seven sections, which is represent by the seven subject areas listed in Fig. 3, these are (1) procurement and supply chain; (2) design and construction; (3) operations and lifecycle; (4) smart cities; (5) intelligent transport; (6) energy and carbon footprint; and (7) decentralised organisations. Each subject area includes several application categories, these were grouped according to their correlation. The subject areas and categories were selected following a bottom-up approach. This was conducted without a predefined or systematic strategy on which topics to cover, provided that it was in conjunction with the construction industry or built environment. The process followed an organic progression through manually making note on a spreadsheet the topical coverage of each of the review documents, as shown in the shared Google spreadsheet following the link provided below.

Procurement and supply chain
This section is comprised of six application categories grouped into the procurement and supply chain subject area. Altogether, this subject area was discussed in 57 out for the 121 documents and is focused on pre-construction activities.
Procurement, bid, and tender (discussed in 12 documents). In a survey conducted by Kim, et al., based on theme of lifecycle, project management, and blockchain, and from respondents in construction industry, the top three applications for blockchain emerged as bidding, procurement, and change management [36]. Lack of trust is particularly evident in procurement, and current management practices requires innovating to improve the ability to track provenance of faults, trace contract alterations, and drawing revisions, while minimising information asymmetry during the tender process [37]. Based on a questionnaire and survey by Isikdag, of 64 industry practitioners in construction industry, consisting of architects, engineers, contractors, and subcontractors, eprocurement appeared to offer very few benefits compared to its nonelectronic counterpart, furthermore, the primary barrier to e-procurement includes a lack of trust in supply chain, unsatisfactory legal infrastructure, and inadequate cybersecurity for storing confidential data [38]. Moreover, Isikdag, stated that blockchain can potentially be used to provide the vital infrastructure required to support privacy without the risks associated with centralised storage; furthermore, he discussed how e-procurement lacks standardisation from regulatory bodies [38].
Logistics, scheduling and programme (discussed in 16 documents). Logistics management has become increasingly complex due to globalisation [39]. Kifokeris, et al., performed a case study of seven Swedish logistics consultancies, which outlined that "delivery failure, imprecise  Table 2 Presents the values of the categories displayed in Fig. 6. The colours labelled in the 'Clusters' column is representative of the colour clusters shown in Fig. 6 data, delays in time, inefficient flows and data transfers between systems" are limitations in existing logistics processes, and discussed the lack of cyber-physical systems integration and analytics in managing on-site assets [39]. Moreover, he proposed a blockchain solution for logistics, using a crypto-economic model to incentivise collaboration [39]. Lanko, et al., considered that existing centralised computer systems are  susceptible to data manipulation, and proposed a framework which incorporated blockchain with RFID for managing logistics of readymixed concrete on-site, whereby, RFID tags are used to record stages of delivery, such as pouring, transportation, handling, quality inspections, and mould forming, with all data exchanges recorded on the blockchain [40]. Blockchain in logistics provides opportunities in offering improved service to clients through automating the process of storing and authenticating data with increased trust, furthermore, decentralised applications potentially reduce the resource requirements in managing systems efficiently [41].
Cash flow and payments (discussed in 14 documents). Chong & Diamantopoulos conducted a case study on a project in Melbourne, Australia, that used blockchain to automate payments; Furthermore, works included the delivery of 5000 building façade panels tracked with Bluetooth sensors to monitor live location of each panel from factory in China to on-site, with BIM used to monitor installation of each panel, while smart contracts executed payments at delivery checkpoints [42]. Additionally, this integrated with a mobile phone application which allowed project participants to view progress of installation in real-time [42]. Ahmadisheykhsarmast developed an add-in for Microsoft Project using programming language C-sharp and Visual Studio, which allowed smart contracts to integrate with mainstream project management software; furthermore, blockchain platform Ethereum, with its native programming language Solidity, was used to link the front and back-end functions of the user application that connected blockchain to Microsoft Project [43].
Late payments is a major problem in construction, caused by contractors performing cash farming, which is the process of withholding supply chain payments to sustain positive cashflow while aggressively investing in new work [44]. Das, et al., proposed a conceptual framework that enabled smart contracts to control the release of payments to supply chain which includes integration with banking systems, furthermore, he discussed the potential to integrate with strategies such as Project Bank Account (PBA) [45].
Digital and automated contracts (discussed in 14 documents). McNamara & Sepasgozar conducted an interview of industry practitioners in the construction industry and derived that trust, risk, and dispute management were ubiquitous concerns in almost all projects, with main contractors exerting dominance through unfair contract conditions [46]. In a survey conducted by Badi et al., of 104 respondents in the UK construction industry, regarding the use of smart contracts, the main factors which determined its adoption in enterprise is competitive edge and commercial value [47]. Hunhevicz, et al., proposed a digital contracting framework which simulated the decision points of a typical design-bid-build project in Switzerland, which included the client, owner, planner, contractor, and supplier, all interacting with smart contracts to control the approvals and validations process of contract activities, such as project definition, design coordination, tendering, supplier selection, and contract signing; furthermore, this was prototyped through a web-based application connected to the Ethereum blockchain [48].
Supply chain management (discussed in 30 of documents) Qian & Papadonikolaki conducted interviews of industry practitioners in the construction industry that are knowledgeable in supply chain and blockchain, and identified that blockchain can potentially be used to mitigate the trust problem in construction, through data traceability, non-repudiation, and disintermediation; furthermore, it was projected that blockchain can save up to 70% on costs associated with data processing and management, through automating compliance checking, payments, and analytics on project performance [49]. Sheng, et al., proposed a framework which allowed project participants to assess compliance to standards and monitor information exchanges through a user application, where project participants would upload data associated with contract documents, project schedule, and cost; furthermore, the application would autonomously notify users of their responsibilities to upload or approve works, which automated the processing of payments and completion certificates [50]. Dutta, et al., conducted a systematic review of blockchain in supply chain and identified several key attributes where blockchain can improve performance, such as evidentiary trail of delivered works substantiated with immutable data, resilience from network disruption, improved data synchronicity, data trust in cyber-physical systems, business process automation through smart contracts, and improved tracking of product revisions [51].
Standards, regulation, and compliance (discussed in 10 documents). The transparent and irrefutable properties of the blockchain make it a suitable technology for trialling whether smart contracts can be used to automate the compliance checking of objects in BIM models [52]. Nawari & Ravindran proposed an automated regulation and compliance checking framework for BIM, whereby, modelling elements are scanned and cross-checked with client specifications, which autonomously notifies designers of their obligations to make design alterations [53]. Blockchain can also be used as a decentralised authority to provide BIM objects with copyright verification, through a lookup service that checks the intellectual property signature of a BIM object, and cross-checks it with data stored in a distributed database; furthermore, designers and contractors working on a BIM model can be instantaneously notified of any copyright infringement of model objects [54].

Design and construction
The design and construction subject area consists of five application categories discussed in 44 of the review documents. This section is focuses on the capital expenditure stage of construction projects.
Building Information Modelling (BIM) (discussed in 41 documents). One of the fundamental reasons for the slow adoption of BIM is a lack of traceability in model revisions, as the current systems is based on manual data entry and relies on trust from designers to keep track of changes [55]. The ability for multiple users in a project to update a BIM model simultaneously is extremely challenging using existing centralised cloud systems, furthermore, the coupling of BIM with blockchain further creates bandwidth limitations, which is due to blockchain's consensus properties, whereby, majority of the nodes on the network need to agree on changes before data can be revised [56]. Zheng, et al., proposed a mobile device application which allowed users to verify on their portable computing device (e.g., phone, tablet, laptop) whether a BIM model is the most recent version, whereby, a hash of the BIM model is stored on the blockchain which allows a lookup service to cross-check the hash of a downloaded model with the hash stored on-chain, afterwards, the application would provide users with a verification receipt stating the model's validity [57]. On another note, a case study by Mason, et al., discussed how the effective logging of geometry and volume in BIM models can transition effectively into computable code for smart contracts [58].
IFC-based interoperability (discussed in 6 documents). IFC is a data standard format registered by the International Standards for Organisation (ISO), which is used for saving BIM model files [59]. Table 3 Presents the values of the data collection types displayed in Fig. 9. The numerals highlighted in bold in the 'Total link strength' column are the same values as the blue numerals shown in Fig. 9 BuildingSmart is an organisation that promotes digital workflow through utilisation of IFC, while OpenBIM is a set of common agreed workflow standards for BIM projects, for the purpose of increasing supply chain collaboration and standardising data exchange processes [59]. Hunhevicz, et al., produced a prototype which incentivised users to produce high quality data sets following the OpenBIM standard, this incorporated the use of smart contracts to provide financial rewards based on the quality of data provided by its users [48]. Ye, et al., produced a prototype which incorporated an IFC model that interoperated with smart contracts, which executed payments autonomously based on elements quantified within the BIM model; furthermore, readable text was maintained as it transferred into smart contracts, which allowed users the ability to intuitive cross-reference IFC data in blockchain code [60]. A study was conducted by Xue & Lu which investigated whether IFC semantics can be substantially minimised to allow for potential storage of IFC code on-chain, and whether small portions of the IFC code can be partitioned away from its original syntax while still remaining readable for purpose of isolating model revisions, which resulted in a semantic reduction of 99.98% of its original size; however, the consensus properties of blockchain proved to be problematic due to its low throughout with data processing, even when tested on a private blockchain network [56].
Integrated Project Delivery (IPD) (discussed in 3 documents). IPD operates through onboarding the construction supply chain with a shared risk and reward contract for the purpose of promoting collaborative workflow [61]. Hunhevicz, et al., discussed how the characteristics of IPD integrate effectively with the ideologies of common pool resource (CPR) and the Ostrom principles for flat organisational structures, which incorporates mutual and economical benefit for project participants who work together to achieve a common goal, whereby, projects which implement blockchain in IPD contracts include potential to reward participants with tokenised and non-tokenised incentives, such as financial rewards for collaborative delivery, transparent agreements, and automated payments upon validated completion of works [62]. Elghaish, et al., conducted a simulated proof of concept which incorporated blockchain in an IPD contract for managing supply chain payments, using private the blockchain platform Hyperledger Fabric (HLF); Whereby, financial operations such as reimbursed cost, profit pool, cost saving pool, and risk pool, were programmed into smart contracts which automated the dispensation of funds according preagreed terms, such as validated completion of works from appointed authorities and project milestone dates [61].
Off-site construction (discussed in 4 documents). Off-site construction includes strong topical overlap with Internet of Things, blockchain, BIM, AI, robotics, and 3-D printing [63]. According to Turk, R. Klinc, the primary application for blockchain in off-site construction is supply chain management, with a projected average saving of 70% through reduced processing costs, which is amassed through improved systems integration, automation through smart contracts, and real-time data traceability [63]. Wang et al., proposed a framework using blockchain platform Hyperledger Fabric for the management of precast construction activities through a user interface, which allowed real-time querying of scheduling, production, and transportation [64]. Additive manufacturing, synonymously called 3-D printing, includes potential to integrate with off-site construction and blockchain for the production, cataloguing, and copyrighting of customised building components [65].
Geospatial, 3-D scanning, and point cloud (discussed in 4 documents). Geospatial technologies such as "remote sensing, LiDAR, internet mapping, GPS and GIS" have strong implications working in conjunction with autonomous vehicles due to their rapid response in scanning geographical landscapes; furthermore, it interoperates effectively with BIM models, smart infrastructure, and cyber-physical systems [66]. 3-D scanning allows assets and geographical locations to be imported into BIM models; however, there is currently a lack of technological capacity for scanned objects to be autonomously cross-referenced with registered objects in a database [63]. Copeland and Bilec proposed a conceptual framework which integrated assets with geospatial sensors and blockchain to produce what they called "buildings as material banks", which utilises sensors affixed to building components which records metadata regarding its condition for reusability, using blockchain as the trusted system for authenticating components and materials within built assets [67].

Operations and lifecycle
The operations and lifecycle subject area is comprised of four categories and consists of 24 documents. This section is focused on the operational expenditure stage of an asset's lifecycle.
Facilities management and maintenance (discussed in 6 documents). Li, et al., proposed a framework for the semi-automated procurement of replacement parts during the operations phase of a built asset, which includes the integration of Internet of Things (IoT) sensors and a computer aided facilities management system (CAFM) for the automated identification of faulty parts; furthermore, a request for replacement parts is processed through a decentralised autonomous organisation, while an e-marketplace handles the bidding and appointment of prospective contractors [68]. Blockchain includes the ability to transact on and off-chain for the purpose of increasing the performance of data exchanges in a decentralised network. Bai, et al., proposed a framework for managing the communications between IoT and blockchain for asset maintenance, which uses on-chain for immutable hash storage and smart contracts, while off-chain handles data storage, computational processing, and analytics [69]. Integrating off-chain applications with blockchain allows for greater transaction throughput, lower transaction fees, and greater control over system operations such as privacy controls.
Life cycle and circular economy (discussed in 11 documents). Shojaei discussed how metadata recorded of raw materials extracted from source can be appended onto the blockchain for end-to-end lifecycle assessment, which allows for a complete and uninterrupted data stream from each handling merchant to end-user to provide proof of provenance from source to construction [70]. Asset data such as specifications, standards, and contract agreements include potential to integrate with blockchain for post-occupancy evaluation, utilising BIM as the data repository for the built environment asset and blockchain as its corresponding data validator [71]. Copeland & Bilec proposed a framework which utilised RFID, BIM, and blockchain to provide components with an evidentiary trail of data throughout its lifecycle, through sensors periodically recording data at key stages, such as installation, decommission, provenance, and metadata regarding supplier, manufacturer, and handling checkpoints [67]. This includes potential to integrate with a crypto-economic incentive scheme for the recycling of assets, with data verified by blockchain.
Construction waste management (discussed in 3 documents). Surplus waste generated by the construction industry is a global issue; furthermore, there is a lack of systems that can accurately account for material waste, which make it an acceptable by-product despite its carbon impact and incurred costs on projects [7]. However, blockchain includes potential to increase the accountability of waste through its ability to verify its lifecycle from source to disposal [7]. Despite this, a proposed solution on who would supply the systems which allows supply chain to quantitatively account the unused material was not discussed in the reviewed papers.
Real estate and property registry (discussed in 10 documents). Dakhli, et al., conducted a case study of 56 residential properties and concluded that blockchain has potential to achieve construction cost savings of 8.3%, which is higher than a typical property developer's net margin of 6%; furthermore, the projected cost savings were attributed to the use of smart contracts and a decentralised autonomous organisations (DAO) to manage and automate business processes [72].
The management of land registries in many developing countries is an unnecessarily complicated process which is prone to fraud and manipulation [73]. Land management was identified in the World Bank's Ease of Doing Business report as a one of the main services that affects the economic growth of a country, furthermore, blockchain was discussed as having the potential to provide a single source of truth to land records, thus reducing administrative overheads in data processing and alleviating risk of fraud [73].

Smart cities
The smart cities subject area is comprised of four categories and consists of 27 documents. This section is focused on how city infrastructure networks can interoperate to provide a data-rich ecosystem of connect devices for managing built environment assets.
Smart cities (discussed in 16 documents). Ahad, et al., conducted a literature review on the topic of smart cities and suggested that they are driven by network-based technologies that integrate to support the delivery of industry 4.0 [66]. These technologies include Internet of Things (IoT), big data, cyber-physical systems, 5-G technology, artificial intelligence (including machine learning and deep learning), blockchain, cloud/edge computing, and geospatial technologies [66]. The interconnected network of devices in a smart city increases the demand for trusted data, therefore, a new business model is required that is more resilient to hacks and central point of failure [74]. This can potentially be supported through the traceable, immutable, and decentralised properties of blockchain [74]. Fu & Zhu proposed a conceptual framework which integrated technologies such as cloud platforms, blockchain, and IoT to form a trusted platform for monitoring live data from infrastructure services, such as geographic information systems (GIS), safety devices, and weather monitoring systems that relay information to city infrastructure services such as transport, communication, and utility [75].
Smart homes and buildings (discussed in 4 documents). Moretti, et al., proposed a conceptual framework that incorporated the use of ultrasonic sensors for the purpose of monitoring indoor activity of a building, which includes sensors placed in rooms to monitor usage, occupancy, and maintenance, which integrate with analytics to provide automated reporting of indoor activity; furthermore, the author discussed the potential to incorporate a blockchain-based management system, through using smart contracts to provide automated payments upon successful delivery of maintenance works [76]. Roy, et al., proposed a prototype for a smart home ecosystem, which included the aggregation of a home device network, blockchain platform, and maintenance service system; furthermore, the home network was comprised of smart meters, IoT, and actuators; the blockchain was used to store and validate results received from the home devices; while the maintenance system provided facility management through identifying when replacement parts were required and provided credentials of prospective suppliers [77].
Intelligent transport (discussed in 15 documents). López & Farooq proposed a smart mobility blockchain framework for managing transportation data, which was comprised of five layers such as (1) privacy, which gives users control of their data when using location revealing applications such as Google maps; (2) contract layer, which controls how smart contracts use user data; (3) communication layer, which appends digital identifiers to communication channels between network nodes; (4) incentive layer, which rewards users for participating in the blockchain network; and (5) consensus layer, which allows nodes to upload data verified by its users [24]. Implications of this included privacy between users and transportation system hosted on a decentralised network [24].
Supplying battery recharge to electric vehicles based on a fast-charge system is technologically challenging, as current recharge systems need to be designed for both intermittent and continual usage [78]. Zhang, et al., conducted a 15 month study at University of California, Los Angeles (UCLA), which implemented a blockchain platform that incentivised users to charge their electric vehicles at specific timescales, which mitigates energy providers having to store unused energy in batteries for extended periods of time; moreover, a user interface provided users with a ranking system based on their record of renewable energy consumption, which rewarded users with discounts and the ability to choose flexible recharge schedules [78].
Water management (discussed in 3 documents). The infrastructure for wastewater management in cities is reaching the end of its lifespan in many countries, which is caused by old treatment plants and damaged pipes which excrete sewage into environmentally sensitive areas that cause health and safety and wildlife concerns [79]. Berglund, et al, discussed how the construction of new water management systems can potentially benefit through innovations such as Internet of Things (IoT), smart meters, and blockchain, to provide live data feed on the performance of water management systems, with implications in improving lifecycle maintenance of infrastructure assets [79]. Perera, et al., discussed how WaterChain, a water utility blockchain network in the United States, allows their participants to invest in water recycling plants and allows them to benefit through the dividends supplied by its service; furthermore, the management of the plant is transparent and can be investigated by the community at any time and dividends are automated through smart contracts, this merges the boundary between consumer and producer and allows the opportunity for communities to self-sustain and self-own their utilities [7].

Intelligent systems
The intelligent systems subject area includes six categories and consists of 46 documents altogether. This section focuses on advanced computer systems, information processing, and the benefits of data-rich networks.
Big data (discussed in 6 documents). The amount of new data produced each year is increasing exponentially, furthermore, the construction industry is under additional pressure to exploit the benefits of data-driven economies whilst in a resource deficit caused by poor margins in construction projects [80]. Blockchain offers a new type of data model which reduces the resource requirements for storing data securely, through bypassing the need to use heavily centralised systems to authenticate data [66]. Network systems such as internet of things (IoT) and smart technologies include the potential to integrate with blockchain to provide increased trust in authenticating data, which is achieved without reliance on oversight from centralised technology companies [24]. Concerns regarding privacy is mitigated through private blockchain protocols such as Hyperledger Fabric, which uses an enterprise-centric model that provides platform developers with control over the privacy features on their network [81]. Alternatively, public blockchain protocols, such as Ethereum, include advanced cryptographic methods such as zero-knowledge-proofs which allow private data exchanges to occur on a public network [82]. Big data integrated with blockchain includes practical applications in off-site construction and supply chain management, through improved contract management, compliance checking, traceability of data in project reports, and reliable data for use with analytics [63].
Artificial intelligence (AI) (discussed in 8 documents). AI, alongside additive manufacturing (synonymously called 3-D printing), autonomous vehicles, blockchain, drones, and Internet of Things, are the fundamental components that form to create the emerging industry 4.0, which were points first discussed in the 2011 report by Germany's economic development agency [65]. Car manufacturers use AI powered robots that work alongside humans in production plants; furthermore, companies such as General Electric and Caterpillar are developing AI solutions to equip workers with robotic exoskeletons to assist with labour intensive jobs [65]. AI is progressively being used in industries to streamline workflow and improve decision making, such as with JP Morgan, who developed a software algorithm called COIN, that scans thousands of contract documents instantaneously to provide judgement on written agreements [83]. A practical use-case for blockchain in AI is the ability to safeguard its coding through placing it in a smart contract, D.J. Scott et al. which mitigates the risk of unauthorised manipulation of the code without permission from authorised actors, effectively, creating unbreakable codified laws which govern the functionality of AI; simultaneously, AI can also be used to debug smart contracts and improve blockchain's protocol design [84].
Cloud computing and electronic document management system (EDM) (discussed in 13 documents). EDM allows companies to manage, store, and process documents electronically [30]. EDM platforms are limited with their potential to interoperate with other technology suppliers, which is due to its centralised systems architecture; conversely, a blockchain-based EDM is built with interoperability as its core and is not financially driven by sales of its product [14].
Cloud computing is a fundamental driver of logistics 4.0 (a branch of industry 4.0), which encompasses global standards, digitisation of business processes, and cyber-physical systems that interoperate with supply chain and logistics networks [14]. Blockchain-based decentralised cloud platforms provide the ability for users and enterprises to store data with greater privacy, this is achieved without risk of hacks or data mining from service providers; however, due to its nascency, decentralised storage solutions may lack in its ability to modularise its functions to suit business workflows [85]. Singh, et al., proposed a framework for managing the data flows of cyber-physical systems in a smart city network, which integrates cloud computing, software-defined networking, and blockchain for trusted data exchanges [29].
Cybersecurity (discussed in 12 documents). The decentralised characteristics of blockchain puts the responsibility in the custody of its users to manage their digital keys competently, which requires users to keep their private-key secret and not reveal the personal identity behind their public-key [86]. Xiong, et al., proposed a "secret-sharing-based key protection" protocol which allows users with compromised or lost private-keys to retrieve access to their account, which involves a stepby-step multiparty verification process, whereby, each party anonymously and privately reveals a small portion of the key, which altogether combines to produce the entire lost private-key [28].
The immutable property of blockchain also comes at the cost of low scalability (measured in transactions per second) and limited capacity to store large amounts of data on-chain [87]. To mitigate this, Bai, et al., proposed a framework which consists of on-chain and off-chain functionalities, which included a "smart predictive maintenance" and a "sharing service of equipment status data" model, whereby, the hashes (unique identifiers) of files are stored on-chain, while off-chain handles high-volume data storage and computational processing [69]. This includes the use of a lookup service which connects the hashes stored onchain to data repositories off-chain, which amalgamates the immutable properties of blockchain with large capacity data storage [69].
Machine learning (ML) (discussed in 3 documents). The procurement and management process of road construction in India is challenged with political corruption and fraud, through lack of compliance checks, material fraud, and unsupervised labour that leads to incomplete works [88]. Shinde, et al., discussed how ML can be used to forecast material quantities, labour requirements, and delivery schedules, while blockchain can be used as the trusted system to verify the authenticity of data sets without reliance on a trusted third party; furthermore, ML coding can be stored in a smart contract or decentralised repository, which can be designed to allow authorised parties to jointly contribute to updating and verifying the code through consensus [88]. ML is used in construction for statistical decision making, irregularity detection, and deriving insight from historic records [89]. Woo, et al., identified five software applications that use ML in the construction industry, these are (1) GenMEP, by Building Systems Planning, which uses ML for the automation of mechanical, electrical, and plumbing data in a Revit model; (2) BIM 360 IQ, by Autodesk, which uses ML to forecast and calculate the impact of subcontractor risks in construction projects; (3) SmartTag, by Smartvid.io, uses ML to automate the process of labelling/ tagging of site assets from pictures and videos; (4) Smart Construction, by Komatsu and NVIDIA, uses ML to simulate the construction process for health and safety and programme analysis; followed by, (5) IBM Watson IoT, who uses ML for proposing energy efficiency and occupancy enhancing solutions in buildings [89].
Internet of things (IoT) (discussed in 31 documents). Wang, et al., discussed how IoT and blockchain can potentially integrate with building information modelling (BIM) to provide a central hub for managing and authenticating data received from built environment sensors; furthermore, the BIM model can be used to map the position of each sensor in a digital model, which provides a 3-D map for maintenance suppliers to utilise [63]. IoT can also be fitted onto the wearables of personnel on construction sites to provide quantitative insight on the environmental conditions and geographic positioning of on-site workers, with blockchain used to hash and timestamp data received from the IoT [30]. Fu & Zhu proposed a smart city framework which incorporates the use of IoT to provide a system which integrates and monitors geographic, safety, and weather, which altogether feed data to a user interface to provide live analytics for use in construction and asset management [75].

Energy and carbon footprint
The energy and carbon footprint subject area includes four categories and consists of 38 documents altogether. This section is focuses on how blockchain can integrate as part of a system to better manage energy, renewables, and carbon.
Peer-to-peer (P2P) energy markets (discussed in 30 documents). P2P energy markets are designed around homeowners buying and selling excess renewable electricity through a local network, which provides neighbourhoods with self-sufficiency and promotes decarbonisation [90]. Esmat, et al., proposed a conceptual framework for a P2P energy marketplace hosted on blockchain, which includes automated uniform pricing and real-time settlements [91]. Ableitner, et al., conducted a 4month field study of 37 households in Switzerland to assess the outcome of a micro-grid prototype, which was a joint effort between academia and industry; furthermore, each of the households were supplied with renewable energy production technologies, smart meters, and a P2P energy trading application hosted on the blockchain [92]. Afterwards, the results were analysed through questionnaires, interviews, and statistics, which displayed active involvement from the participants with the blockchain application and an eagerness from the households to continue with the study after it concluded [92]. Energy trading can also occur between machine-to-machine (M2M) for the purpose of achieving full automation without the reliance of appointing users to authorise the trade, as shown in a conceptual framework by Sikorski, et al., which included a study of two energy suppliers that operate in tandem to provide consumers with the most economically priced electricity [13]. Despite the immutable property of blockchain, P2P markets are at potential risk from producers manipulating the power measurements recorded at connection points; However, to mitigate this, Saha, et al., proposed a blockchain-based distributed verification algorithm that penalises inconsistent measurements of current [93].
Smart grids (discussed in 29 documents). 'Peer-to-peer (P2P) energy' and 'smart grids' are discussed interchangeably; however, the former relates to trading markets, while the latter relates to energy infrastructure and smart meters. The integration between decentralised microgrids and the main power grid is made possible through a demandside management (DSM) application proposed by Noor et al., whereby consumers are able to supply their own smart energy appliances and battery storage and utilise the DSM application to connect their local grid to the main grid [94]. Christidis, et al., conducted a case study of 63 solar panel fitted homes, situated in Texas, United States, which compared the efficiency of a semi-centralised versus a decentralised energy grid market, which included the former consisting of high transactions speeds with lower security, while the latter included low transaction speeds with higher security, which resulted in the blockchain approach being less efficient due to its high latency in processing transactions [81]. A similar framework was proposed by Foti & Vavalis, which investigated how a blockchain-based smart grid would perform with 1000 participants transacting on the Ethereum blockchain testnetwork, which resulted in the centralised grid being efficient at providing lower cost electricity due to the mining fees associated with blockchain, however, when factoring in the lifecycle cost of managing systems, the decentralised approach was discussed as potentially being more cost-effective and resilient to external threats such as cyber-attacks [95].
Renewable energy solutions (discussed in 3 documents). The energy industry is experimenting with new business models that transition from centralised to decentralised, which includes the integration of smart devices, micro-grids, blockchain, and energy recycling technologies [96]. A combined heat and power (CHP) system provides energy recycling through combining electricity and heat generation into one system, which integrates fittingly with renewable production technologies such as photovoltaic and wind turbines for the purpose of reducing carbon footprint [97]. Furthermore, in the event of natural disasters such as flooding, high winds, earthquakes, wild fires, snow/ice, and extreme temperature, CHP maintained performance most consistently in comparison with photovoltaic, wind turbine, standby generators, and biogas [97]. The demand for renewable energy increases with the depletion of oil and rise in global warming. Perrons et al., stated that the geothermal energy sector has received pressure from stakeholders to innovate renewable production methods and management systems, with blockchain discussed as a potential candidate to improve the software aspect of it [98]. Keivanpour investigated two off-shore wind farms in United Kingdom, called Robin Rigg, and Walney Phase 1, and concluded that the current delivery method of industrial scale renewables is unnecessarily expensive due to longstanding supply chain processes, and discussed the innovation potential with blockchain, Internet of Things, and big data [99].
Carbon accounting and decarbonisation (discussed in 7 documents). Khaqqi, et al., proposed a carbon emission trading framework, where a government organisation would issue construction companies with a limited number of carbon credits to expend on a construction project, whereby, each credit is representative of a tonne of carbon emissions; furthermore, companies are able to buy or sell excess carbon credits to each other through a decentralised online marketplace, which incentivises renewable companies, while at the same time penalises nonrenewable companies [27]. Rodrigo, et al., conducted an interview with three industry practitioners, each with over 13 years of experience in information technology, which concluded that the inherent properties of blockchain, such as auditability, security, and decentralisation, is a suitable tool for embodied carbon estimating [100]. Hua, et al., proposed an energy trading framework that rewards carbon credits to prosumers of a micro-grid network, whereby, energy producing technologies are linked to the blockchain to record the carbon footprint at time of production; furthermore, each prosumer is provided a set quantity of carbon credits which their permitted to expend during production, which incentivises prosumers to act sustainably [101].

Decentralised organisations
The decentralised organisations subject areas is comprised of four categories and consists of 19 documents altogether. This section is focused on decentralised services and autonomous organisations. Some of the topics in this section are more general purpose than the previous sections, nevertheless, they included strong overlap with the construction industry and each category was discussed several times in the reviewed documents.
Decentralised Autonomous Organisation (DAO) (discussed in 5 documents). DAO is an autonomous blockchain entity with decentralised governance at its core, which rewards users with tokenised incentives for participating in the network and operates entirely through smart contracts [102]. The construction industry is particularly known for incurring change orders and programme alterations, which is problematic for smart contracts due to their unalterable properties once deployed; furthermore, translating written agreements into codified form creates linguistical challenges between contract managers and programmers, whereby, each party may not understand the industryspecific culture differences of the other, such as terminologies and processes [68]. Dounas, et al, produced a prototype which utilised DAO and smart contracts to automate the awarding of works for architectural design submissions, which involved a simulated study where stakeholders submit a request for a built environment asset through a DAO platform, followed by submission of the designs from prospective contractors or architects, and finally, the autonomous calculation of the winning proposal through a predefined scoring system and awarding of work through a smart contract [103]. Similarly, DAO also includes the potential to integrate with the construction or operations phase of a built asset, through semi-automating the procurement process for obtaining new materials or replacement parts, whereby, DAO is used as the medium for connecting prospective suppliers to new work, managing payments, cross-checking compliance certificates, and quantitatively assessing the risk of each supplier through their track record of delivered works [104].
Identity and certificate authentication (discussed in 5 documents). The fundamental properties of blockchain (traceability, transparency, and immutability) make it a suitable technology for incorporating identity authentication services, as centralised systems are prone to hacks and data manipulation [86]. Private blockchains include privacy controls as a fundamental feature to its protocol. Whereas, public blockchains include cryptographic functions such as zero-knowledge proofs which permit private transactions to occur on a public network, however, this incurs additional transaction fees added onto the existing mining fee [82]. Nawari & Ravindran discussed how private blockchain Hyperledger is suited for identity management services in construction due to its modular architecture, which allows automated compliance checking of identities on the network [53]. Similarly, Shojaei, et al., discussed how Hyperledger's certificate authority can be used to maintain an active lists of supply chain participants in a construction project, which can be reused across multiple projects [105].
Blockchain allows the creation of non-fungible tokens (NFTs), which can be used as a digital certificate that represents the ownership of a physical asset; furthermore, this NFT can hold additional data such as title deeds, lifecycle data, building certificates, and any other associated data [106]. Implications include substantial reductions in data retrieval for insurers, estate agents, facility managers, and building inspectors [72]. Due to the immutable properties of blockchain, data stored in the NFT is append only, thus leaving an intact evidentiary trail of data throughout its lifespan.
Financial technology & banks (discussed in 7 documents). The emergence of decentralised finance in 2020 allows banks to extend their portfolio to include additional commercial products for customers [12]. Yao, et al., proposed a conceptual framework which discussed the viability for banks to provide blockchain-based supply chain finance, through using blockchain to verify the regulatory compliance of their customers, track signed agreements, and trace pending invoices [107]. Blockchain can be used to maintain an accurate and irrefutable record of transactions without risk of ledger inconsistencies, such as reconciliation errors and double spending; furthermore, banks can potentially provide escrow services through smart contracts, which allows transacting parties to formalise agreements amongst themselves while under oversight from regulatory controls, this ensures compliance to fair business terms and legal standards [15]. Smart contracts also include the potential to automate tax duties, such as with the legal movement of goods across international borders, whereby compliance certificates would be autonomously awarded upon payment of taxes [108].
Crowdsourcing (discussed in 4 documents). Blockchain-based crowdsourcing is a decentralised alternative to acquiring project funding, which includes benefits such as providing opportunities for skilled talent in economically disadvantaged nations, reduced intermediaries, and codified agreements with auditable terms for the purpose of supporting fair contract executions [109]. Public blockchains provide free protocol infrastructure that allows users to develop platforms and raise funds through initial coin offerings (ICOs), which is similarly compared to the initial public offerings (IPOs) offered in stock markets when private companies transition to PLC [110]. However, ICOs have been a target for criminal activity due to their ability to raise funds from anonymous users and lack of regulation checks, such as know your customer (KYC) and anti-money laundering (AML). Hassija, et al., discussed how the crowdfunding platform, BitFund, allows investors to propose a problem to a public community of programmers and include project-specific parameters such as budget, timescale, use-case, etc., afterwards, the awarding of works is conducted algorithmically through smart contracts to ensure a fair selection process of the development team [109].

Discussion
An exploratory approach was implemented into this review for the purpose of understanding which categories in construction are most influenced by blockchain. This review explored 33 application categories of blockchain in construction. Each category was substantiated by a minimum of three documents. These categories were further organised into seven subject areas, which include (1) procurement and supply chain, (2) design construction, (3) operations and life cycle, (4) smart cities, (5) intelligent systems, (6) energy and carbon footprint, and (7) decentralised organisations. When assessing the types of data collection used in the reviewed documents (as shown in Fig. 8), synonymous data collection terminologies were merged together for simplicity, such as conceptual frameworks, which included conceptual models and theoretical frameworks. Similarly, proof of concepts (PoC) included pilot studies and prototypes. The first three subject areas of this review are sequential stages that occur in a construction project, such as subject area one procurement and supply chain, which includes implementing blockchain in the digital tendering process [111], contract and cashflow management [43], and automated checking of compliance to standards [68]. Subject area two design and construction, incorporated using blockchain for trusted data exchanges [112] and traceability of deliverables throughout the supply chain [113]. While subject area three life cycle and circular economy, included how blockchain can be used as part of the assessment and management of a built asset during its operational expenditure stage [114]. Subject area four smart cities, and subject area five intelligent systems, included a macro-orientated approach, assessing how multiple built environment assets and services interact through a smart city network, which includes the interoperability of various systems such as utility [115], transport [116], Internet of Things (IoT) [117], and smart technologies [88]. Subject area six energy and carbon, focused attention on peer-to-peer energy trading models [118], sustainable technologies for the built environment [119], and carbon accounting strategies [120]. And finally, subject area seven decentralised organisations, incorporated decentralised autonomous organisations (DAO) and decentralised services [103]. DAO is difficult to precisely classify in the current environment, as its definition is dynamic in translation and its development is constantly evolving; however, in construction, many of its activities (for now) overlap with the responsibilities of a main contractor, therefore, for simplicity, DAO can be described as a decentralised contractor.
The aforementioned 33 categories and seven subject areas were not distinctly siloed and included substantial topical crossovers. E.g., the supply chain management category overlapped with all of the subject areas, however, based on the scientometric analysis conducted (as per Fig. 6), it was positioned most quantitatively relative in the procurement subject area, due to its high number of shared link with other categories in that area [121,122]. IoT also strongly overlapped with several subject areas, which include smart cities [7], energy and carbon [123], design and construction [113], procurement [85], and decentralised organisations [79]; however, IoT was placed in the intelligent systems subject area due to its strong correlation with the other categories in this area. The categories electronic document management systems (EDMS) and digital/automated contracts were placed in separate subject areas despite their similarities, as the former is characterised by the digital management of documents on a centralised system, while the latter utilises smart contracts on a decentralised protocol, thus dissimilar systems architecture [124].
A smart medical record system, which includes managing patient records and sharing healthcare data with hospitals, is a category supported by two authors [15,75]; however, blockchain for healthcare is an entirely different subject area and a vast topic suited for a separate literature review altogether [115]. Health and safety monitoring of site conditions and historic records of on-site accidents were discussed in two documents [79,102]; however, despite its practical applications in construction, it also lacked content for substantiation. Another topic that was excluded despite its interest in two documents is smart governance, which incorporates governmental organisations implementing blockchain to automate the compliance checking and auditing of built environment assets [17,115]. Multi-category applications of blockchain in construction that were not included due to its general-purpose nature include transaction immutability, digital notarisation, decentralised applications (dApps), smart contracts, and information sharing, as effectively, these topics are already integrated within all of the reviewed categories and do not require itemising [102].
As blockchain is a decentralised technology, appropriate incentivisation techniques must be applied to encourage platform interaction through a crypto-economic model [102]. The integration of blockchain in enterprise in the current environment is reliant on dApps harmonising with existing centralised systems, however, as blockchain matures, the transition to complete decentralisation is likely to increase. This assumption is based on assessing the growth and expansion of blockchain in construction since its emergence in academic literature, and the intensifying global interest in blockchain. In a report regarding impact of blockchain, it was identified as potentially transforming 58 industries globally, which includes the construction industry [125].
Business operations are entirely based on risk management activities, which includes economic risks through investments in new business models, social risk through job losses, legal risk through dispute resolution and corporate liability, environmental risk through sustainability and ecological sensitivity, and technical risk through increased pressure to integrate systems and provide data-driven solutions [85]. Blockchain mitigates against centralised hacks, data manipulation, accounting errors, and provides a foundation for trusted data without reliance on a trusted third party [126]. An area which lacked discussion from the review documents was the integration capabilities of blockchain with existing enterprise systems, as blockchain is considered a high-risk technology due to its decentralised design and lack of standards. Trust is a term that appeared most frequently in the reviewed literature when describing the characteristics of blockchain, such as "stakeholder trust" [122], "peer-to-peer trust" [127], "trust in collaboration" [128], "information trust" [26], "removal of trusted authority" [11], and "trusted distributed ledger" [129]. Other commonly used terms include transparency, traceability, immutability, security, automation, auditability, decentralisation, and disintermediation [9,[118][119][120]123,130,131].
Over the course of 2017-2020, the rate at which new documents were published on blockchain in construction was recorded at an average of 184%; however, the sample number of years is small, and this level of growth cannot be maintained long-term. A 10-year period would provide a more statistically comprehensive result. Fig. 4 documented the annual expansion of new categories on topic since its emergence in 2017, which displayed six new categories in 2017, nine in 2018, 13 in 2019, followed by five in 2020. It is likely that the expansion of new categories on the topic has almost reached a plateau, therefore, over the next consecutive years, it is envisaged that existing categories will undergo maturity as more attention is focused on testing and developing earlier ideations.

Conclusion
New academic documents on blockchain in construction increased at an average of 184% each year since 2017, surmounting to an accumulated total of 121 documents at time of writing this article in 2021. An exploratory approach was implemented to investigate all 121 publications to examine the contemporary environment of the topic. This review identified 33 application categories, these were organised into seven subject areas and included (1) procurement and supply chain; (2) design and construction; (3) operations and life cycle; (4) smart cities; (5) intelligent systems; (6) energy and carbon footprint; and (7) decentralised organisations. To support the literature review, statistics and scientometrics were incorporated to display the progression of the topical area. This includes visual maps that display the co-occurrences of the categories (as shown in Fig. 6) and data collection types implemented in the reviewed documents (shown in Fig. 9). A complete list of the 121 reviewed documents, along with their category coverage, document type, data collection type, and impact factor, is provided in the shared Google spreadsheet link provided below.
https://docs.google.com/spreadsheets/d/1V4UICRdoyWycaGENH9 rnuxukRNQJFIArQ-feV7NM0a4/edit?usp=sharing Limitations included using only one scientific database, Scopus, due to the inconsistencies which emerged when amalgamating information from various scientific databases for use in visual mapping software. In a comparison of the search results from seven scientific databases and based on the topic of blockchain in construction, Scopus overshadowed its competition by a large margin; furthermore, up to 85% of the documents indexed in other scientific databases were already existent in Scopus. Another limitation was the restricted capacity to conduct indepth investigation on one particular subject area within the topic, this was due to the exploratory nature of the study, which covered a wide range of application categories. Despite this, the findings provided a solid foundation for aggregating all of the research areas of blockchain in construction in the contemporary environment.
Content for this exploratory review was obtained predominantly from documents published from 2017 to 2020, as this article was written in early 2021; however, further work includes an extended review following the progression of the topic over the next consecutive years.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. 00284939 or 22113444 or 13693999 or 20541236) AND TITLE-ABS-KEY ("blockchain*" OR "block chain*" OR "distributed ledger*" OR "smart contract*") Search query two: Search query two used a simpler method, which included using one of the predefine subject areas available on Scopus, followed by specific key words. The limitation to using this search query is the high number irrelevant documents that accompany the results.
The string of text for query to consists of: SUBJAREA(ENGI) AND TITLE-ABS-KEY("Blockchain" AND "Construction").