Special Issue on Data Analytics Applications for Architecture and Construction

1. Introduction

Construction processes are lengthy, complex and involve a variety of stakeholders. Each stakeholder holds different roles and responsibilities. As a result, large amounts of data are generated in a variety of formats to serve a variety of purposes during construction projects. Some of these data have specific standards to manage them, whereas others do not; similarly, some data must be manually classified or managed, while others can be automated. Progress in the development of digital devices and software also plays an important role in how these data are generated and managed, which creates inconsistencies. The construction industry is undergoing a digital transformation, with big data emerging as a critical enabler across organizational, costing, and design practices. Despite this, the path to adoption of data analytics is not without obstacles. In Australia, construction professionals have identified over 40 barriers to Big Data implementation, spanning people, technology, and environmental factors. Mapping these challenges offers firms a strategic lens to anticipate and mitigate risks during deployment. In the realm of project costing, automation is revolutionizing how data are extracted and analyzed. New methods now enable the rapid, accurate extraction of cost and quantity data from complex documents, achieving up to 97.5% accuracy. In infrastructure, machine learning is being used to benchmark railway project costs and uncover key cost drivers. Moreover, legal frameworks in public works, such as those in Spain, reveal how regulatory environments influence cost overruns—highlighting the need for data-informed policy reform. With reference to design and construction, big data is enhancing both safety and building efficiency. From semi-automated modeling of residential buildings in Italy to real-time hazard detection using deep learning and IoT wearables, data-driven tools are reshaping how we create the built environment and protect those that use it. Natural language interfaces such as VISA4D are simplifying BIM interactions, while performance evaluation systems and bibliometric analyses are helping teams assess design-phase effectiveness and track research trends. Together, these studies underscore a pivotal shift: Big data is no longer a futuristic concept but a practical necessity. As the industry continues to digitize, embracing data-driven tools will be essential for improving outcomes, reducing costs, and fostering innovation across the construction lifecycle. As part of this Special Issue, 10 papers have been collated, categorised into three main themes: (1) big data application and practice in organizations [1,2], (2) construction project costing practice and [3,4,5], and (3) design and construction practice [6,7,8,9,10].

2. Overview of Contributions

2.1. Big Data Application and Practice in Organizations

Mapping the Barriers of Big Data Process in Construction: The Perspective of Construction Professionals [1]: In this study, the authors identify and map the barriers to big data processes in the construction industry from the perspective of construction professionals in Australia. Interviews were conducted with these professionals, noted for their varying experiences in the big data process. Through qualitative data analysis, 40 barriers were identified and mapped, including 26 newly identified, across five themes: people, knowledge, technology, data, and environment. The barriers were further mapped, with some covering more than one stage in the big data process. Many of these barriers have not been empirically identified in previous studies. By implication, mapping the barriers across the big data process enables professionals/construction firms to visualize the potential lapses before and/or during implementation. The findings of this study offer professionals/construction firms strategic insights and operational perspectives for planning and deploying big data processes.

Big Data Value Proposition in UK Facilities Management: A Structural Equation Modelling Approach [2]: Big data analytics (BDA) is transforming industries by improving efficiency and enabling client-focused services. In this article, the authors examine how UK facilities management (FM) organizations are adopting BDA to drive innovation and add value, particularly in customer service and decision-making. Using interviews and a survey, the study authors developed and validated a fifteen-variable model for BDA outcomes, identifying three key dimensions: improved client value, added value in FM operations, and increased efficiency. The findings highlight how digitization enables FM services to deliver more personalized, client-centric solutions and better resource management, positioning analytics as a key differentiator, making a significant contribution to the literature.

2.2. Construction Project Costing Practice

A Method to Enable Automatic Extraction of Cost and Quantity Data from Hierarchical Construction Information Documents to Enable Rapid Digital Comparison and Analysis [3]: Despite efforts to standardize construction cost data, many firms still use manual methods, causing inefficiencies and errors in large projects. In this study, the authors introduce a fully automated system that extracts and structures cost and quantity data into a consistent, machine-readable format. Data mining is used to detect document structure and combines data science with expert input for classification. Tested on 5770 assets across 18 files, the method achieved 97.5% accuracy, with most errors due to source data issues. Operating without human input, the system offers a scalable, reliable solution for fast cost comparison and benchmarking in construction.

An Automated Method for Extracting and Analyzing Railway Infrastructure Cost Data [4]: Extracting and standardizing project cost data is vital for accurate forecasting, benchmarking, and deeper cost insights. Manual data entry and a lack of standardization make this process inefficient and unreliable, however. To address this issue, a new method is introduced that combines data mining, statistics, and machine learning to analyze railway infrastructure cost data. Using 23 historical projects from Network Rail, the approach enables cost comparability and aids in the identification of key cost drivers. Machine learning techniques support efficient benchmarking and reveal critical relationships in cost structures. The findings demonstrate the value of automated data extraction for improving analysis, decision-making, and understanding of infrastructure project costs.

The Relationship between Cost Overruns and Modifications for Construction Projects: Spanish Public Works and Their Legal Framework [5]: Cost overruns in public works are often linked to design changes, yet the legal framework’s role in these modifications is underexplored. In this study, the authors examine Spanish public works projects, analyzing how legal limits affect cost changes and the alignment between initial bids and final costs. They also investigate strategic behavior by construction firms that adjust costs to match bid prices. Using statistical tools such as Spearman correlation and visual analysis, the study authors uncover a strong link between bid and final prices. Comparing two legal periods reveals that less restrictive laws lead to higher overruns, offering insights for policymakers and future research.

2.3. Design and Construction Practice

Semi-Automated Dataset Generation for Residential Buildings Using Graph-Based Topological Modelling [6]:

Italy’s residential buildings largely predate modern safety and energy regulations, posing challenges for policymakers deciding between renovation and reconstruction. Both choices have significant socio-economic and environmental impacts and require detailed architectural data, which are currently scarce. To address this issue, the study authors introduce a semi-automated method using graph theory to model residential floor plans as connectivity graphs enriched with functional data. Tested on 98 buildings in Bologna, the method achieved an 89.8% success rate, proving effective in data-limited contexts. The resulting dataset enables spatial configuration analysis and geometric attribute extraction, paving the way for future machine learning applications in building typology and function detection.

Leveraging Deep Learning and Internet of Things for Dynamic Construction Site Risk Management [7]:

The construction industry faces ongoing safety challenges due to complex, dynamic site conditions. The authors of this study present a deep learning-based hazard warning system using YOLOv7 image recognition and IoT-enabled smart wearables for real-time monitoring and alerts by analyzing 8663 construction site images. Achieving a mean average precision of 0.922 and an F1 score of 0.88, the system detects hazards in under one second. It uses the DIKW framework for data management and CNNs for robust feature extraction. Innovations such as perspective projection coordinate transformation and a security assessment module enhance accuracy. Validated through field tests, the system improves safety, reduces accidents, and supports scalable construction automation.

A Novel Framework for Natural Language Interaction with 4D BIM [8]:

Natural language interfaces can improve construction project efficiency by making information more accessible and reducing administrative workload. In this paper, the authors present the Voice-Integrated Scheduling Assistant for 4D BIM (VISA4D), which combines speech recognition and NLP with Building Information Modeling to streamline schedule updates. VISA4D enables voice and text inputs, integrates with Autodesk Navisworks, and visually tracks progress using color-coded BIM elements. Tested on a Canadian office project, VISA4D achieved 89% accuracy in classifying construction commands, with 92% of users finding it easy to use. The tool advances AI-driven construction management by addressing practical challenges in real-world field operations.

Performance Evaluation System for Design Phase of High-Rise Building Projects: Development and Validation Through Expert Feedback and Simulation [9]:

In this study, the authors develop a performance evaluation system for the design phase of high-rise building projects, addressing a gap whereby performance is usually only measured during construction. The process involved identifying 21 key performance indicators, creating standardized indicator sheets, and building a dashboard-based evaluation tool in Excel. Validated by experts and tested on an AI-generated simulated project, the system effectively detected performance issues across planning, cost, time, quality, and people categories. The results showed strengths in quality and stakeholder collaboration but weaknesses in cost and time. The tool is practical and adaptable for real projects, though broader international validation and real-world application are recommended for future studies.

Bibliometric Analysis of Hospital Design: Knowledge Mapping Evolution and Research Trends [10]:

Hospital design significantly influences patient outcomes, clinical efficiency, and infection control. Since the COVID-19 pandemic, research in this area has grown rapidly, with increasing interdisciplinary collaboration. In this study, the authors perform a bibliometric analysis using CiteSpace (version 6.2.R3), Excel, and Tableau (version 2024.3) to examine 877 publications from 1932 to 2025. The authors explore trends in publication, collaboration networks, co-citation structures, and keyword evolution. The findings show accelerated development since 2019, with multipolar global collaboration but low institutional connectivity. Evidence-based design remains foundational, while healing environments, biophilic design, and patient-centered strategies emerge as key themes. The study offers systematic insights into the field’s evolution, knowledge base, and future directions for researchers and practitioners.