<?xml version="1.0" encoding="utf-8"?>
    <rss version="2.0">
      <channel xmlns:content="http://purl.org/rss/1.0/modules/content/">
        <title>Frontiers in Built Environment | Construction Management section | New and Recent Articles</title>
        <link>https://www.frontiersin.org/journals/built-environment/sections/construction-management</link>
        <description>RSS Feed for Construction Management section in the Frontiers in Built Environment journal | New and Recent Articles</description>
        <language>en-us</language>
        <generator>Frontiers Feed Generator,version:1</generator>
        <pubDate>2026-08-21T11:04:06.702+00:00</pubDate>
        <ttl>60</ttl>
        <item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fbuil.2026.1886913</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fbuil.2026.1886913</link>
        <title><![CDATA[Factors influencing cloud computing adoption in Saudi Arabia’s construction industry: a hybrid PLS-SEM and ANN approach]]></title>
        <pubdate>2026-08-19T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Mosaab Alaboud</author>
        <description><![CDATA[Despite the capacity of cloud computing to improve productivity and sustainability within the construction sector, its implementation in developing nations remains inadequately understood, especially within the distinctive regulatory, infrastructural, and cultural framework of Saudi Arabia. This disparity is noteworthy considering the emphasis on digital transformation across all economic sectors under Saudi Arabia’s Vision 2030. Grounded in the Technology-Organization-Environment (TOE) framework and supported by elements of the Diffusion of Innovations theory and the Technology Acceptance Model, this study identified and statistically analyzed the factors influencing cloud computing adoption in the Saudi construction industry. A methodically designed questionnaire was distributed among 70 professionals—comprising civil engineers (n = 38), architects (n = 24), electrical engineers (n = 4), and mechanical engineers (n = 4)—affiliated with contractor organizations, consultancy firms, design entities, governmental institutions, and semi-governmental organizations throughout Saudi Arabia. The dataset underwent a comprehensive analytical procedure employing a multifaceted methodological framework that incorporated Exploratory Factor Analysis, Partial Least Squares Structural Equation Modeling (PLS-SEM), and Artificial Neural Networks (ANN), alongside a 10-fold cross-validation technique to effectively counteract the potential for overfitting. The PLS-SEM findings indicate that industry-specific (β = 0.631) alongside organizational elements (β = 0.514) exert the most significant direct influence on the adoption of cloud computing (R2 = 0.172). Concurrently, the ANN sensitivity analysis identifies competitive pressure as the paramount driver, succeeded by cost/performance and technological advancement factors. This study contributes to the literature by (1) extending the TOE framework to the Saudi construction context, (2) demonstrating the complementary predictive power of ANNs over linear PLS-SEM models for capturing nonlinear relationships, and (3) establishing a context-specific hierarchy of adoption factors. The findings yield practical implications for decision-makers and stakeholders in the industry who are striving to improve the integration of cloud computing into digital transformation strategies.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fbuil.2026.1888974</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fbuil.2026.1888974</link>
        <title><![CDATA[Emerging risk factors affecting real estate development delivery in Accra, Ghana]]></title>
        <pubdate>2026-08-12T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Kingsley Ofori</author><author>Simon Ofori Ametepey</author><author>Clinton Aigbavboa</author>
        <description><![CDATA[IntroductionThis study examined emerging risk factors affecting real estate development delivery in Accra, Ghana, focusing on economic, institutional, market, and environmental risks. The study was motivated by the increasing complexity of real estate development environments in emerging economies, where project performance is shaped by interconnected structural uncertainties beyond traditional construction risks.MethodsA quantitative research design was adopted. Data were collected from 98 real estate development and construction professionals, including architects, civil engineers, quantity surveyors, project managers, contractors, and estate surveyors, using structured questionnaires. Data were analyzed using descriptive statistics, reliability analysis, exploratory factor analysis, Spearman correlation, one-sample mean tests, and multiple regression analysis.ResultsThe findings showed institutional risks were the most critical, indicating governance and regulatory inefficiencies are the main challenges in Ghana’s real estate sector, followed by economic, market, and environmental risks which also contribute to uncertainty. Respondents rated all risk dimensions above the neutral midpoint, showing emerging risks are widely perceived as significant and persistent, while development performance was rated below the midpoint, indicating weak outcomes. One-sample tests confirmed risk levels are higher than expectations, while performance is lower. Factor analysis showed risks form a single interconnected system rather than independent factors. Correlation results indicated that higher risk levels are strongly associated with lower development performance. Regression analysis confirmed all risk dimensions significantly reduce performance, with institutional risk having the strongest effect.DiscussionThe study concludes that emerging risks operate as an integrated systemic structure that significantly constrains real estate development delivery performance in Ghana.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fbuil.2026.1864359</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fbuil.2026.1864359</link>
        <title><![CDATA[Assessing sustainability practices across construction and post-construction phases in Nigeria’s AEC industry]]></title>
        <pubdate>2026-08-10T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Enoch I. Obanor</author><author>Adefemi O. Adeodu</author><author>Jesse B. Zachiang</author><author>Promise K. Egba</author><author>Emmanuel Ajakanu</author><author>Joseph O. Dirisu</author><author>Enoabasi A. Umofia</author>
        <description><![CDATA[PurposeThis study evaluates the implementation of sustainability practices during and after the construction phases of building projects within Nigeria’s Architecture, Engineering, and Construction (AEC) sector. It aims to compare phase-based sustainability performance and identify implementation gaps that hinder lifecycle integration.Design/methodology/approachA structured cross-sectional survey was administered to 100 AEC professionals, including architects, engineers, builders, and sustainability consultants. Descriptive statistical analysis was conducted to assess twelve sustainability practices in both the construction and post-construction phases using mean scores and standard deviation measures.FindingsResults indicate moderate to high implementation levels across both phases, with post-construction practices marginally outperforming construction-stage practices (construction: M = 3.68/5, 73.6%; post-construction: M = 3.72/5, 74.4%; both p < 0.001). Despite this progress, behavioural engagement, IoT-enabled monitoring, and structured sustainability reporting remain inconsistently implemented across both phases.OriginalityThis study advances sustainability discourse in developing economies by providing a comparative, phase-based quantitative assessment of construction lifecycle practices, highlighting implementation disparities rarely examined within Nigeria’s AEC sector.Practical implicationsThe findings underscore the need for institutional reforms, structured capacity building, smart technology integration, and post-occupancy performance monitoring systems to strengthen lifecycle-based sustainability implementation.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fbuil.2026.1868534</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fbuil.2026.1868534</link>
        <title><![CDATA[Mapping psychosocial work exposures across construction: a multilevel evidence gap map with implications for offsite construction]]></title>
        <pubdate>2026-08-07T00:00:00Z</pubdate>
        <category>Review</category>
        <author>Yize Guo</author><author>Don Amila Sajeevan Samarasinghe</author><author>Wajiha Mohsin Shahzad</author><author>Ravi Reddy</author><author>Zechen Guan</author>
        <description><![CDATA[Empirical research on psychosocial work exposures in construction remains fragmented, with limited connection to offsite construction settings and supply chain related outcomes. Most of the research results are disconnected from the offsite construction (OSC) environment and supply chain related results. This is a key gap: the OSC supply chain relies on the connection between manufacturing, logistics and assembly nodes, and psychosocial factors play the role of soft risk in it, which may disrupt performance and resilience. However, due to the lack of structured evidence, it remains difficult to determine which psychosocial exposures have been actually measured, which organisational levels they come from, and whether the existing knowledge base is sufficient to guide the risk management of the OSC supply chain. The purpose of this review is to: (a) identify empirically tested psychosocial work exposures in construction research; (b) classify them using the Job Demand Resource (JD-R) × individual-group-leadership-organisation (IGLO) framework; and (c) evaluate the alignment between existing evidence and outcome domains relevant to supply chain operation. According to the PRISMA 2020 guidelines, a total of 160 eligible studies were retrieved, and 245 unique psychosocial exposures and 394 study instances were found. There is structural selectivity in the existing evidence: the demand field dominates (63 exposure factors; 129 instances); the organisational level exposure factors account for 71.4% of all unique exposure factors. Only one study has clearly examined the prefabricated or offsite construction environment. The correlation of results mainly focuses on health (37.6%) and wellbeing (27.9%), with limited attention to performance (10.4%) and resilience. These findings show that the growing evidence base does not cover the crucial multilevel coordination processes in the prefabricated or OSC supply chain, and these processes lack a conceptual basis for understanding the psychosocial work exposure factors in offsite construction delivery systems as soft risks. This review is limited to English peer reviewed empirical studies published since 2010. The evidence atlas captures the exposure of the prevalence in each study, rather than the size of the combined effect or the causal relationship.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fbuil.2026.1890441</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fbuil.2026.1890441</link>
        <title><![CDATA[Associations between safety management system implementation and accident rates in a US Army Corps of Engineers center: a conceptual replication study]]></title>
        <pubdate>2026-08-06T00:00:00Z</pubdate>
        <category>Brief Research Report</category>
        <author>Joshua Michael Moskowitz</author><author>Jeremy Bart McCranie</author><author>Richard Todd Shilling</author>
        <description><![CDATA[Safety management systems have become progressively popular as systems of choice to develop systematic safety programs and prevent accidents. The U.S. Army established the Army Safety and Occupational Health Management System (ASOHMS) to standardize safety programs and practices across subordinate commands. The U.S. Army Corps of Engineers (USACE) was an early adopter of the initiative which may affect safety outcomes for USACE employees and contractors globally. This brief report and replication study sought to replicate the first ever study on ASOHMS’ associations with employee and contractor accident rates. This research retrospectively analyzed accident data within a USACE center through descriptive statistics and correlation testing. This study found moderate negative correlations between implementation progress of ASOHMS with all employee accident rates under study and a lack of statistically significant correlations when compared with contractor accident rates. These findings suggest that ASOHMS may share relationships with employee accident rates. USACE organizational structures and the intensity of interaction with contractors may be a major factor in the potential to influence contractor safety outcomes.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fbuil.2026.1808188</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fbuil.2026.1808188</link>
        <title><![CDATA[Predicting a pavement condition index for municipal streets using a sigmoid deterioration model]]></title>
        <pubdate>2026-08-04T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Muhammad Amjad Afridi</author><author>Sigurdur Erlingsson</author>
        <description><![CDATA[This study applies sigmoidal models to predict pavement condition index (PCI) deterioration over time for residential and non-residential (main, collector and industrial) streets using pavement condition data from Skellefteå Municipality, Sweden. The dataset includes windshield-survey pavement assessments conducted in 2014, 2018 and 2022. Two modelling configurations were applied: temporal models calibrated using 2014 and 2018 observations to predict 2022 PCI ratings, and full dataset models developed using all available observations to represent overall deterioration behaviour. Separate sigmoidal curves were developed for residential and non-residential streets, as well as for the specific maintenance-treatment categories applied exclusively to non-residential streets. Sigmoidal curves for residential streets showed lower predictive performance and higher variability, potentially associated with heterogeneous pavement conditions, inconsistent pavement age and localized utility cuts, whereas sigmoidal curves for non-residential streets demonstrated comparatively better predictive capability due to more consistent deterioration patterns. Model performance also varied across maintenance-treated pavement surfaces, partly due to limited sample sizes. Sigmoid best-fit curves for street pavements treated with special treatments (ST) showed the highest predictive performance, while milling and resurfacing (MR)-treated pavements demonstrated relatively good agreement with measured PCI ratings. Surface levelling (SL) is a relatively low-cost treatment and may be suitable for earlier stages of deterioration, although traffic and pavement characteristics should also be considered when selecting treatments. Differences in unit cost between treatment types are considered only for contextual interpretation. Overall, sigmoidal curves may support more timely maintenance planning in data-limited municipal pavement management and reduce the need for major reconstruction.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fbuil.2026.1884560</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fbuil.2026.1884560</link>
        <title><![CDATA[The impact of perceived design origin attribution (Human vs AI) on trust and performance in augmented reality assembly tasks among construction professionals]]></title>
        <pubdate>2026-08-03T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Rahul Chaudhari</author><author>Paul Goodrum</author><author>Matt Jones</author><author>Siddharth Bhandari</author><author>Nolan Brady</author><author>Matthew Hallowell</author><author>Tom Yeh</author>
        <description><![CDATA[As the construction industry accelerates its digital transformation, it is essential to thoroughly investigate the value proposition of emerging technologies. This study investigated the complex interplay between the perceived origin of a design (AI vs. Human), the information format (AR vs. Paper), user trust, and work performance. The objective was to understand how user trust is affected by the perceived origin of a design and whether it influences user work performance in an AR-assisted environment. To explore this, we conducted a controlled experiment with one hundred practicing industry craft workers performing a full-scale Mechanical, Electrical, and Plumbing (MEP) assembly task. With the use of deception, the perceived origin of the design was manipulated as either AI or human-generated. Participants used traditional paper drawings or one of two AR models to complete the task. We recorded self-reported trust, both before and after the task, as well as work performance factors such as task time, rework, and errors. The findings revealed that user trust is not uniform: participants’ perceptions of design accuracy remained stable regardless of origin, even when they noticed errors, whereas trust in the design’s safety decreased significantly when it was attributed to AI. Regarding performance, the AR interface notably enhanced accuracy by reducing rework, regardless of the design’s perceived origin. Task speed, however, was mainly influenced by the user’s innate spatial cognitive ability. These outcomes suggest that successful integration of AI into construction will not depend just on the capabilities of the algorithms but on the entire sociotechnical system, with the user interface playing a crucial mediating role. Therefore, to unlock the full potential of these technologies, experts must address underlying human factors such as trust and user perception.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fbuil.2026.1812897</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fbuil.2026.1812897</link>
        <title><![CDATA[Smart cities and the infrastructure metaverse: can data markets and securitization help close the financing gap?]]></title>
        <pubdate>2026-07-30T00:00:00Z</pubdate>
        <category>Review</category>
        <author>Peter Adriaens</author><author>Juri Mattila</author>
        <description><![CDATA[The increasing integration of sensors and digital infrastructure in city assets to enable climate resilience, infrastructure health monitoring or venue flow of people and goods is fueling the growth of a wide range of data types differing in latency, scope, reliability and frequency. While data-driven business models in this new metaverse of cities have received attention in engineering design, policy, and legislation, the connection between this information supply chain and the capital markets that finance infrastructure remains largely unexplored. This paper addresses three research questions: how do smart city data value chains connect to infrastructure financing models; how can data securitization leverage market mechanisms to support new structured finance instruments; and what conditions enable transactional data markets to move beyond bilateral agreements toward liquid price discovery? By leveraging the literature on asset-backed securities (ABS) and data markets, a multi-layer securitization framework is proposed to structure contracted revenues from smart cities data and services to raise capital for next-generation infrastructure delivery. The framework maps the full digital infrastructure stack onto a structured capital raise: physical sensing and digital twin (Layers 1–2) data that are productized into contracted revenue streams (Layer 3), pooled in a special purpose vehicle and issued as rated debt tranches (Layer 4), under data stewardship and credit enhancement governance (Layer 5). A use case of the framework is illustrated using digital transportation infrastructure (smart pavements) design and financing. Although no rated data contract securitization for infrastructure has yet closed, individual elements of the framework are operational across public and private contracting contexts. Policy implications address data stewardship governance, regulatory frameworks for digital twin-based finance, and the structural preconditions. These include standardized data quality scoring and independent measurement, reporting and verification (MRV), required for capital markets to treat infrastructure data as a creditworthy asset class. Key limitations include the absence of a recognized rating agency methodology for data-backed ABS, legal ambiguity over data ownership on public rights-of-way, and government counterparty appropriations risk.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fbuil.2026.1870384</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fbuil.2026.1870384</link>
        <title><![CDATA[The role of digital technologies in engineering procurement: a systematic literature review]]></title>
        <pubdate>2026-07-15T00:00:00Z</pubdate>
        <category>Review</category>
        <author>Imoleayo A. Awodele</author><author>Molusiwa S. Ramabodu</author><author>Nathaniel Ayinde Olatunde</author><author>Emmanuel C. Eze</author><author>Iruka C. Anugwo</author><author>Yesica P. Espinosa</author><author>Temitope O. Kehinde</author><author>Bamidele T. Arijeloye</author><author>Alice T. Ogunlolu</author>
        <description><![CDATA[The increasing complexity of engineering procurement and the growing demand for efficiency, transparency, and resilience in supply chains have intensified the need for digital transformation in procurement systems. Despite the rapid advancement of Industry 4.0 technologies, existing studies on digital procurement remain fragmented, often focusing on individual technologies or general supply chain contexts rather than engineering procurement environments. This study examines the role of digital technologies in engineering procurement and provides an integrated understanding of their impact, associated challenges, and transformation pathways. A systematic literature review (SLR) approach was adopted, guided by PRISMA principles, using peer-reviewed studies published between 2015 and 2024 from major academic databases. A thematic and cross-study synthesis was employed to analyse the conceptual evolution of digital procurement, the role of key technologies, and barriers to adoption. The findings reveal that technologies such as artificial intelligence (AI), blockchain, the Internet of Things (IoT), Building Information Modelling (BIM), robotic process automation (RPA), and e-procurement systems are transforming procurement from a transactional function into a strategic, data-driven capability. These technologies enhance procurement performance by improving operational efficiency, cost optimization, transparency, and supply chain resilience. However, their adoption is constrained by key challenges, including system integration complexities, organizational resistance, skills gaps, and regulatory uncertainties. The study further demonstrates that the effectiveness of digital procurement depends on the integration and complementarity of technologies within interconnected ecosystems, supported by organizational readiness and technological capabilities. This study contributes to Procurement 4.0 literature by providing an integrated synthesis, developing a conceptual framework, and proposing a digital procurement maturity perspective to guide transformation in engineering procurement.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fbuil.2026.1656743</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fbuil.2026.1656743</link>
        <title><![CDATA[Web-based information management system for streamlining Indian road transport infrastructure projects]]></title>
        <pubdate>2026-07-09T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Anshul Gupta</author><author>Prasanna Venkatesan Ramani</author>
        <description><![CDATA[The infrastructure sector is an extremely data-intensive and event-driven industry. Implementing traditional and modern digital techniques creates huge amounts of data, from which information can be extracted for efficient information flow, especially for complex infrastructure projects like tunnels. A project management information system (PMIS) plays a vital part in managing the massive scale of operations in tunnel and road transport construction projects. This study crafted a comprehensive framework for implementing integrated web-based project management information systems, especially for road infrastructure. A prototype was developed incorporating tools providing backend, frontend, and database services for transportation infrastructure based on their project life cycle phases. Recognizing the critical need for efficient information management in construction projects throughout the project life cycle, the frameworks integrated key web-based PMIS development principles with specific requirements. The system supports information services across six life cycle phases: planning, design, pre-construction, construction, post-construction, and operation and maintenance. A prototype MERN (MongoDB, Express.js, React, and Node.js)-style implementation was developed using a MongoDB document database (NoSQL) for flexible schema, Node.js for backend services (REST APIs), and React.js for the user interface. The database schema and API endpoints were designed to support six lifecycle phases (planning to operation and maintenance). Validation was performed on five real transport-infrastructure projects with functional, load (up to 50 concurrent users), and security tests for real-time information structuring using a common platform throughout the project life cycle phases and retrieval. It contributes to the field of infrastructure informatics by offering a unified platform that improves collaboration, transparency, and decision making across the project life cycle.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fbuil.2026.1837617</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fbuil.2026.1837617</link>
        <title><![CDATA[A governance-calibrated machine learning pipeline for predicting change order ratio under extreme data scarcity in public construction]]></title>
        <pubdate>2026-07-08T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Mohammed A. KA. Al-Btoush</author><author>Hashem Khaled Almashaqbeh</author><author>Imad Al Shalout</author><author>Ja’far A. Aldiabat Al-Btoosh</author><author>Taiseer Mustafa Rawashdeh</author><author>Amani Ismail Alsmadi</author>
        <description><![CDATA[IntroductionPredicting change order magnitude in fragmented public construction systems remains methodologically unreliable, particularly under extreme data scarcity and correlated project attributes. This study develops a governance-calibrated predictive architecture for estimating the Change Order Ratio (COR) using objective contractual and design records from 21 public projects in Jordan, and is therefore framed as preliminary, proof-of-concept evidence from a small, single-context sample rather than externally generalizable inference.MethodsTo prevent overfitting and validation bias, a leakage-free nested Leave-One-Out Cross-Validation (LOOCV) framework integrates regularized regression and nonlinear Support Vector Regression (SVR).ResultsThe unregularized baseline achieves strong in-sample fit (R2 of 0.84) but deteriorates under cross-validation (R2 of 0.54), suggesting instability in small datasets. Regularization enhances robustness (LASSO LOOCV R2 of 0.70), while nonlinear SVR delivers the strongest generalization (LOOCV R2 of 0.76; RMSE of 1.25), suggesting curvature and interaction structures beyond linear shrinkage in this sample. Sensitivity analysis identifies nonlinear amplification of design completeness and positions Requests for Information as leading instability indicators.DiscussionThe findings suggest that change-order volatility is more closely associated with tender-stage information and documentation conditions than with project scale or contractual strictness in the studied context. The study contributes a defensible small-sample modeling framework and highlights upstream documentation maturity as the primary leverage point for enhancing predictability in data-scarce public-sector project delivery environments.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fbuil.2026.1851806</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fbuil.2026.1851806</link>
        <title><![CDATA[Predicting civil engineering project success with machine learning: benchmarking models on one million projects]]></title>
        <pubdate>2026-06-25T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Mirvat Abdallah</author><author>Chady El Hachem</author>
        <description><![CDATA[Civil engineering projects involve complex interdependencies among financial, technical, and organizational factors and directly influence the performance of the built environment. Accurate prediction of project success can support earlier risk identification, more effective resource allocation, and better planning decisions, particularly for large-scale infrastructure programs. This study evaluates several machine-learning classifiers using the Civil Engineering Global Project Dataset (Kaggle), which contains one million project records described by eleven input variables, including project cost, proximity, certification level, years of experience, company score, overtime hours, and salary bracket. The outcome variable is_good is used as a binary indicator of project performance (high-performing versus needs improvement), reflecting overall project delivery quality and operational effectiveness. A Logistic Regression model provided as a baseline was replicated to verify reproducibility, yielding 0.7975 accuracy and 0.7976 ROC–AUC. Model evaluation was further supported using precision, recall, and F1-score metrics to provide a more comprehensive assessment of classification performance. Random Forest and XGBoost models were then compared, along with Random Forest variants using polynomial feature expansion (degrees 2 and 3) and extensive hyperparameter tuning. Across the full dataset, accuracy and ROC–AUC values remained tightly clustered around 0.79–0.80, indicating a performance plateau and suggesting that algorithmic complexity alone provides limited incremental benefit with the current feature set. A cost-stratified analysis, motivated by the bimodal cost distribution, revealed distinct regime-dependent error profiles, highlighting the importance of threshold calibration and cost-aware decision rules for practical deployment. Overall, the findings emphasize that future gains are more likely to come from domain-driven feature engineering, richer project descriptors, and cost-sensitive evaluation, supporting more reliable decision-making for infrastructure delivery and resilience in the built environment.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fbuil.2026.1844340</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fbuil.2026.1844340</link>
        <title><![CDATA[Exploratory data analysis and interpretable machine learning for anomaly detection in Epb tunnel boring machine operations]]></title>
        <pubdate>2026-06-18T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Murat Gunduz</author><author>Khalid Kamal Naji</author><author>Amr Mohamed</author><author>Omar Khan</author>
        <description><![CDATA[IntroductionEarth-Pressure-Balance Tunnel Boring Machines (EPB TBMs) generate high-dimensional sensor streams during excavation, yet operational anomalies are difficult to identify early because the signals are noisy, coupled, and change with geology and operating mode.MethodsThis study develops an exploratory data analysis framework for EPB TBM operations using ring-based data from a 7 km tunneling project comprising more than 4,400 rings. After preprocessing and z-score normalization, Pearson and Spearman correlation analyses confirmed strong torque–thrust coupling (r ≈ 0.85). Time-series visualization with z-score outlier screening identified rings showing high thrust but low penetration, indicating potential tool inefficiency, wear, or challenging ground conditions. An XGBoost regression model was used to predict thrust force, and SHAP analysis quantified feature contributions after excluding thrust-derived and internally redundant indices, identifying cutterhead torque, alignment position, and penetration per revolution as the principal independent predictors.ResultsThe model achieved R2 = 0.872, RMSE = 381.8 kN, and an a20-index of 0.993 on the held-out test set.DiscussionThe framework captures both expected physical couplings and orthogonal behaviors, enabling interpretable anomaly identification and providing actionable insights for monitoring and decision support in EPB TBM operations.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fbuil.2026.1846314</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fbuil.2026.1846314</link>
        <title><![CDATA[BuildSafe: a generative AI framework for joint construction safety, permit, and community-impact reasoning from linked NYC DOB-OSHA-311 data]]></title>
        <pubdate>2026-06-18T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Shahid Ali</author><author>Ammar Alzarrad</author><author>Hwapyeong Song</author><author>Husnu S. Narman</author>
        <description><![CDATA[BuildSafe investigates whether domain-adapted generative models can turn routine administrative text permit filings, incident narratives, and citizen complaints into early-warning signals for construction risk in New York City. The study links Department of Buildings (DOB) records, OSHA enforcement cases, and 311 complaints into a shared schema. It uses Gemini 1.5 Flash to generate three-section annotations covering job-site hazards, permit/code requirements, and community impacts. Three open models (Gemma-3-1B, Llama-3.2-3B, and Mistral-7B-Instruct-v0.3, 4-bit) are then fine-tuned with LoRA on this corpus using a unified prompt template, and evaluated with BERTScore, ROUGE, BLEU, METEOR, and bootstrap confidence intervals on a 2,833-record held-out test set, and supplemented with a 150-record two-expert consensus validation subset used for targeted human evaluation of factual accuracy, completeness, and hallucinations. Across the short-run LoRA adaptation setting, all models showed small train-validation-test gaps, suggesting stable proof-of-concept adaptation without obvious overfitting rather than evidence of full convergence. Llama-3.2-3B is the clear leader on lexical overlap metrics, Mistral-7B-Instruct delivers the strongest semantic fidelity and most structured narrative outputs, and Gemma-3-1B offers competitive semantic performance at a fraction of the compute cost, suggesting potential value for edge or real-time triage workloads. A qualitative case study of Manhattan permit records illustrates how these models produce multi-section safety narratives and reveals trade-offs among completeness, specificity, and verbosity. The paper highlights key limitations, most notably the reliance on machine-generated reference labels, the limited scale of human expert evaluation, and the NYC-specific regulatory context, and proposes a roadmap for future operational validation rather than claiming demonstrated deployment impact. Priorities include richer and better-linked data sources, retrieval-augmented and cascaded inference architectures, impact- and fairness-oriented evaluation in agency pilots, and human-in-the-loop governance suited to high-stakes public-sector AI systems.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fbuil.2026.1827664</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fbuil.2026.1827664</link>
        <title><![CDATA[Benchmarking YOLOv8-YOLOv12 models for image-space near-miss detection in construction safety]]></title>
        <pubdate>2026-06-12T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Shahid Ali</author><author>Ammar Alzarrad</author><author>Sudipta Chowdhury</author><author>Husnu S. Narman</author>
        <description><![CDATA[Automated near-miss detection can support proactive construction safety monitoring by identifying hazardous worker-equipment interactions before injuries occur. This study presents a systematic evaluation framework for comparing 24 publicly available pretrained YOLO object detection models, spanning YOLOv8 through YOLOv12, for image-based near-miss detection in construction environments. The evaluation used 38 publicly available construction-related videos, standardized to 1920 × 1,080 resolution and sampled at 1 frame per second, resulting in approximately 903 evaluation frames across varied work-zone conditions, camera perspectives, object scales, clutter, motion blur, and partial occlusion. A rule-based pipeline identified person-equipment interactions using centroid-based image-space proximity under a primary 250-pixel threshold. Because complete camera calibration metadata were unavailable, this study treats this threshold as an operational image-space criterion rather than a universally calibrated physical distance. This study evaluates object-detection performance using precision, recall, and F1 score against a human-corrected consensus ground truth derived from YOLOv8x candidate annotations and independently reviewed by three domain experts. The analysis assesses near-miss performance using frame-level recall, spatial overlap, fuzzy centroid matching, and threshold-sensitivity analysis. Results showed that YOLOv8x, YOLOv8l, and YOLOv9c achieved the strongest near-miss performance. The findings highlight the importance of stable frame-level detection and provide an exploratory, reproducible benchmark of comparative YOLO model behavior under selected image-space construction-video conditions.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fbuil.2026.1840012</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fbuil.2026.1840012</link>
        <title><![CDATA[Institutional gatekeeping and change order prevention in public infrastructure projects: evidence from a governance reform in Jordan]]></title>
        <pubdate>2026-06-04T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Mohammed A. K. A. Al-Btoush</author><author>Hashem Khaled Almashaqbeh</author><author>Taiseer Mustafa Rawashdeh</author><author>Jafar A. Aldiabat Al-Btoosh</author><author>Imad Al Shalout</author><author>Mahmoud Ali Alsubeh</author>
        <description><![CDATA[This study examines change order behavior in public road infrastructure as a two-stage process separating occurrence from conditional severity. Using an official dataset of 157 projects in Jordan (116 pre-policy, 41 post-policy), the analysis applies exact nonparametric inference and a two-part modeling framework suited to zero-inflated, heavy-tailed outcomes. Results reveal a sharp structural shift following governance reform: nonzero-change order occurrence declines from 40.5% to 4.9%, with complete suppression in externally funded projects and near-elimination in the Ministry of Public Works and Housing (MoPWH) funded projects, while decentralized projects show no change. Qualitative evidence further indicates that strengthened approval thresholds act as the primary downstream gatekeeping mechanism, while design accountability and contractor prequalification operate as upstream controls reducing the generation of change-inducing conditions. In contrast, conditional severity remains highly dispersed once changes are approved, indicating limited governance influence at this stage. These findings establish a clear frequency severity separation and demonstrate that governance effectiveness operates primarily through suppressing initiation rather than reducing magnitude, highlighting the limitations of aggregate cost-based metrics in evaluating policy impact.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fbuil.2026.1842237</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fbuil.2026.1842237</link>
        <title><![CDATA[Development of a hierarchical fuzzy assessment model for quantifying GenAI risks in sustainable construction projects]]></title>
        <pubdate>2026-05-28T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Mohamed Abdelwahab Hassan Mohamed</author><author>M. K. S. Al-Mhdawi</author><author>Ghazal Shadlooye</author><author>Udechukwu Ojiako</author>
        <description><![CDATA[PurposeThe purpose of this study is to identify and evaluate the significance of emerging risks associated with the integration of Generative Artificial Intelligence (GenAI) into risk management (RM) within sustainable construction projects (SCPs).MethodologyThe research followed a four-stage methodology to collect and analyse data, including: (1) a systematic literature review to identify a comprehensive list of GenAI-related risks in the context of SCPs; (2) the development of a multi-criteria risk assessment model to determine key evaluation criteria; (3) the distribution of a structured survey to 80 construction experts to assess the identified risks based on these criteria; and (4) the development of a fuzzy-based model to quantify and rank the significance of the risks.FindingsA total of 30 risk factors were identified and subsequently classified into five categories: input quality, technological adaptability, ethical and governance, information integrity, and financial-related risks. Furthermore, the fuzzy analysis revealed that the most significant risk factors were human error, data unavailability, insufficient training, data breaches, and lack of awareness.ImplicationsThe study provides both theoretical and practical contributions by developing a novel risk assessment framework tailored to GenAI integration in sustainable construction. The fuzzy set theory approach enhances the accuracy of decision-making in high-uncertainty environments and aids project managers and policymakers in prioritising critical risks. The findings also offer actionable insights for developing mitigation strategies and fostering responsible GenAI implementation.Originality/valueThis study makes a unique contribution by being among the first to systematically analyse the risks of GenAI integration in construction RM using fuzzy logic. It advances the understanding of AI-driven risks in the built environment and provides a replicable framework for future risk assessments. Moreover, it encourages further exploration of regional and contextual variations in GenAI risk perceptions and supports the broader digital transformation agenda in sustainable construction.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fbuil.2026.1815172</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fbuil.2026.1815172</link>
        <title><![CDATA[Text mining and natural language processing in construction research: a scientometric analysis and qualitative review]]></title>
        <pubdate>2026-05-25T00:00:00Z</pubdate>
        <category>Review</category>
        <author>Yuxuan Yuan</author>
        <description><![CDATA[Text mining (TM) and natural language processing (NLP) have emerged as powerful analytical approaches for extracting value from the large volumes of unstructured textual data generated throughout the construction industry, leading to a growing body of research focused on addressing sector-specific challenges using these technologies. This paper presents a systematic review integrating scientometric analysis and qualitative synthesis to map the research landscape of TM/NLP applications in construction. The objectives are to summarize existing research achievements, identify prevailing research trends, evaluate academic influence patterns, and propose future research directions. A dataset of 153 peer-reviewed journal articles was compiled from 2015 to 2025 and retrieved from the Web of Science (WoS), Scopus, and Google Scholar databases following duplicate removal and screening procedures. The analysis investigates publication growth trends, collaboration networks among researchers, countries, and institutions, and thematic structures through keyword co-occurrence and co-citation analysis to identify emerging research topics. Results reveal that the field has entered a high-growth stage since 2022, with annual publications surging sharply; the most dominant research theme is construction safety risk identification and accident prevention, followed by technical applications of machine learning and NLP. The findings highlight leading contributors, influential institutions, and major publication outlets, providing guidance for academic collaboration and research dissemination. Furthermore, the study offers a reference framework for researchers and practitioners to select appropriate TM- and NLP-based solutions for specific construction challenges, while also supporting policymakers and journal editors in prioritizing future research and development directions. Overall, this review enhances understanding of the application landscape of TM and NLP in construction and contributes to advancing the digital transformation of the industry.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fbuil.2026.1805761</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fbuil.2026.1805761</link>
        <title><![CDATA[Risk management as a qualitative evaluation criterion in public procurement of construction works]]></title>
        <pubdate>2026-05-19T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Jana Korytárová</author><author>Vít Hromádka</author><author>Petr Marvan</author>
        <description><![CDATA[This paper focuses on the application of risk management as a qualitative criterion for evaluating public construction contracts using the Best Value Approach (BVA). This method is being progressively implemented in the Czech Republic as a strategic tool to enhance the quality and efficiency of public procurement. The primary objective is to present the “Risk Assessment Plan” criterion and its role in improving project preparation. The analysis is based on practical experience gained from construction contracts executed by a specific public institution. The study examines the principles, parameters, and evaluation process of the “Risk Assessment Plan” criterion, assessing its functionality and effectiveness within the real-world context of the Czech public procurement environment. The criterion proved to be an effective tool, providing the contracting authority with direct insight into bidders’ capabilities to identify, describe, and manage project risks. Key findings include verification of the professional readiness and expertise of suppliers, effective transfer of market knowledge and experience into the project preparation process, a significant increase in the project’s readiness for the implementation phase and enhanced transparency and objectivity in the bid evaluation process. The article also highlights critical pitfalls and common mistakes made by both bidders and contracting authorities –notably the confusion between project-specific risks and general supplier risks, insufficient documentation structure, and a fundamental misunderstanding of the criterion’s purpose. Based on these observations, the paper provides methodological recommendations for the correct definition of the criterion, scoring mechanisms, and its subsequent use during project execution. The results confirm that a well‐structured risk management criterion is essential for more effective project delivery and overall higher standards in public construction works.]]></description>
      </item><item>
        <guid isPermaLink="true">https://www.frontiersin.org/articles/10.3389/fbuil.2026.1768360</guid>
        <link>https://www.frontiersin.org/articles/10.3389/fbuil.2026.1768360</link>
        <title><![CDATA[Evaluating GeoAI tools for urban resilience and sustainability: a survey-based analysis of disaster preparedness, climate adaptation, and adoption barriers using the technology acceptance model (TAM)]]></title>
        <pubdate>2026-05-12T00:00:00Z</pubdate>
        <category>Original Research</category>
        <author>Abdulrazzaq J. Alkherret</author><author>Islam A. Alshafei</author><author>Mohammad Alhusban</author><author>Ayah A. Alkhawaldeh</author><author>Rawan Allouzi</author><author>Safa’ Al-Kfouf</author>
        <description><![CDATA[IntroductionThis study examines user-reported perceptions of GeoAI tools in supporting urban resilience applications, particularly in disaster preparedness, climate adaptation, and sustainable development. It explores how Perceived Usefulness (PU) and Perceived Ease of Use (PEOU), based on the Technology Acceptance Model (TAM), influence the adoption of GeoAI tools in professional practice.MethodsA survey was conducted with 149 professionals from the fields of urban planning, geomatics, and artificial intelligence. The study evaluated three widely used GeoAI tools—ArcGIS Pro, CityEngine, and QGIS with AI plug-ins—focusing on their perceived effectiveness in real-world urban resilience scenarios. Statistical analysis was used to assess the influence of PU and PEOU on tool adoption.ResultsArcGIS Pro was rated as the most effective tool, particularly in disaster preparedness and climate adaptation, with a mean score of 4.6 out of 5.0. CityEngine and QGIS with AI plug-ins also received positive evaluations but exhibited higher barriers to adoption. Key challenges identified include data quality (65%), integration issues (55%), cost (50%), and lack of training (45%). The findings indicate that both PU and PEOU significantly influence adoption, with Perceived Usefulness emerging as the strongest predictor.DiscussionThe results highlight the critical role of usability and perceived value in the adoption of GeoAI tools for urban resilience. Addressing barriers such as data integration, cost, and training is essential to enhance adoption rates. Improving these factors can support more effective implementation of GeoAI technologies, ultimately strengthening urban resilience planning and decision-making.]]></description>
      </item>
      </channel>
    </rss>