A blockchain 4.0 paradigm for coordinating centralized construction digital twin and decentralized project management

Abstract

The construction industry is fundamental to global economic growth and societal progress. In the construction industry, construction project management (CPM) is vital in coordinating complex projects and ensuring successful outcomes. However, CPM is inherently complex and faces numerous challenges, including fragmented information, complex stakeholder dynamics, and uncertainty in a changing environment. While technologies such as digital twin (DT) have been applied to enhance CPM through real-time data sharing and analyses, there remains a dilemma between the centralized nature of CPM DTs and the inherently distributed and dynamic nature of CPM. To address this dilemma, the construction industry needs new technologies to adapt effectively to these changing demands.

Blockchain is an emerging distributed ledger that ensures secure, immutable, and transparent transactions through encryption. Thus, blockchain’s key benefits include enhanced data security, comparability, and trust between stakeholders with conflicting interests. These properties of blockchain are exactly what CPM needs for data integration and security, especially in dynamic and distributed environments. However, current blockchain solutions like Hyperledger Fabric and Ethereum face high latency and depend on stable networks, which CPM environments often lack. Therefore, innovative blockchain technology is needed to address these challenges and meet CPM’s unique demands.

This thesis aims to propose the Blockchain 4.0 paradigm for overcoming dilemmas related to coordinating centralized DT and decentralized CPM. Four key objectives are designated: (1) challenges in coordinating DTs with decentralized CPM are to be identified, (2) a Blockchain 4.0 paradigm that incorporates artificial intelligence (AI) for CPM is to be designed, (3) the paradigm is to be implemented with typical real-world construction scenarios, and (4) the effectiveness of the paradigm in improving CPM practices is to be evaluated.

Two case scenarios demonstrated the Blockchain 4.0 paradigm’s effectiveness in CPM. Case Scenario A focused on remote inspections of 952 MiC modules, ensuring full project data traceability and achieving a 99.8% reduction in response time for uploading and 99.2% for querying inspection records. Case Scenario B used 23,858 rows of inspection data to predict MiC module completion. Automated Machine Learning (AutoML) achieved an R² score of 0.974 and an RMSE of 8.305 days. Interpretable Machine Learning (IML) identified nine key features affecting production, validated through a qualitative questionnaire, confirming Blockchain 4.0’s potential for AI-driven CPM in MiC projects.

The contribution of this thesis is four-fold. First, the outcomes of this thesis extend the knowledge frontier of CPM with a novel Blockchain 4.0 paradigm coordinating ordinary blockchain technology and AI in CPM. Secondly, according to quantitative experimental results, the newly proposed Blockchain 4.0 paradigm addresses the fundamental challenges, such as being overly reliant on stable Internet connection, high latency, and insufficient analytics for CPM decision-makers. Furthermore, it incorporates IML to enhance the interpretability of machine learning. Third, CPM practitioners, such as Field experts and stakeholders, agreed that the proposed paradigm has great potential to improve the efficiency and reliability of the CPM process. Finally, the study adopts a robust mixed-method approach that combines case studies with empirical validation to demonstrate the system’s effectiveness.