Purpose This study presents an integrated systematic mapping of artificial intelligence (AI) applications in construction management, clarifying technique-domain pairings, data and validation practices, and deployment readiness across the construction lifecycle, addressing fragmentation in prior reviews. Design/methodology/approach Following the preferred reporting items for systematic reviews and meta-analyses (PRISMA) guidance for study identification and screening and the structured literature review approach, 148 peer-reviewed articles published between 2007 and 2025 across 14 journals were analyzed through bibliometric and thematic synthesis. An open companion web application (https://cm.electriai.com/) was developed, providing bibliometric visualizations, a searchable paper database, and a retrieval-augmented generation (RAG) AI assistant. Findings The review shows that AI research in construction management accelerated after 2022 but remains uneven in its technical and methodological development. Technique-domain mapping revealed lock-in patterns: vision-based deep learning (DL) was concentrated in safety, and traditional machine learning (ML) in cost estimation, while optimization, graph neural networks (GNNs), and generative AI remained underexplored. Data constraints persist: 40.5% of studies used datasets below 500 samples, 65.5% relied on private or unreported data, cross-project validation appeared in 1.4% of papers, and explainable AI (XAI) appeared in 3.4%. Practical implications Construction stakeholders require coordinated investment in open benchmark datasets, cross-project validation protocols, XAI integration and interoperable standards for building information modeling and Internet of Things systems. Originality/value This review integrates AI techniques, construction domains, data and validation practices, supported by an open companion web application for verifying, filtering, and querying the evidence base.
Purpose This paper proposes a domain-specific design science research (DSR) methodological framework for the built environment, integrating artefact taxonomy, evaluation strategy selection, phase-to-output guidance, and a researcher self-assessment checklist for construction management, BIM, and facility management contexts. Design/methodology/approach A PRISMA compliant systematic literature review across five databases retrieved 327 papers; 248 underwent full-text assessment and 30 met inclusion criteria. Structured data extraction and thematic synthesis identified patterns in DSR framework adoption, artefact types, knowledge contributions, evaluation methods and critical success factors across built environment sub-domains. Findings Instantiation artefacts dominate the corpus (37%, n = 11) and improvement contributions prevail (67%, n = 20), revealing a field of applied strength and limited theory-building engagement. Four primary frameworks require contextual adaptation for construction settings. Three pragmatic evaluation alternatives address timescale and access constraints: prospective validation, embedded knowledge transfer partnership (KTP) designs, and multi-site replication. The framework is grounded in systematic analysis of 30 empirical DSR studies conducted in built environment contexts. Originality/value Whilst prior DSR frameworks offer process guidance and prior built environment studies develop domain-specific artefacts, the two bodies of work have not been synthesised into a consolidated methodological resource simultaneously addressing artefact taxonomy, evaluation strategy selection, phase-to-output mapping, and researcher self-assessment. This paper addresses that gap through empirically grounded synthesis from 30 built environment DSR studies.
Purpose This study develops and evaluates an integrated framework for modular construction and Just-in-Time (JIT) logistics under narrow street and constrained site conditions in urban regeneration areas. Design/methodology/approach A simulation-based optimization approach is proposed by integrating discrete event simulation (DES) (SimPy), genetic algorithm optimization and Monte Carlo analysis. Three construction strategies, namely conventional construction, modular bulk delivery and modular JIT delivery, are comparatively analyzed under stochastic uncertainties. Findings Results show that modular construction significantly reduces project duration, cost, and urban externalities compared to conventional construction. Modular bulk delivery consistently outperforms JIT delivery on duration, cost, street closure and noise exposure; JIT’s sole measurable advantage is the elimination of on-site storage requirements. Accordingly, JIT should be regarded as a constraint-driven fallback strategy, appropriate when staging space is unavailable, rather than a performance-superior alternative to bulk delivery. Under the JIT configuration, optimal parameters are identified as a 10-hour buffer time and a 6-module (minimum feasible) storage capacity, though this configuration still yields an 18.75% late-delivery rate, reflecting an inherent reliability trade-off in space-constrained JIT operations. Originality/value Rather than proposing new simulation or optimization techniques, the study’s contribution lies in applying and calibrating established methods, namely DES, genetic algorithms and Monte Carlo analysis, to the previously under-examined problem of JIT logistics parameter tuning under severe narrow-street spatial constraints, and in providing empirical evidence that JIT delivery functions as a constraint-driven fallback strategy rather than a generally preferable alternative to bulk delivery.
Purpose Evidence remains limited on how unsafe behaviors connect and progress directionally within construction accident sequences. This study investigates these relationships using 329 written fall-from-height (FFH) accident investigation reports from Jiangsu Province, China. Design/methodology/approach Text mining was integrated with complex network analysis. Unsafe behavior nodes were identified through coding and topic modeling, while a convolutional neural network was used to extract progressive and parallel relations from accident narratives. The resulting directed network comprised 71 nodes and 217 edges and was analyzed using degree, node strength, betweenness centrality, and shortest-path measures. Findings The resulting network exhibits a heterogeneous structure, with a limited number of unsafe behaviors occupying prominent positions across multiple structural measures. The analysis further identifies recurrent behavioral pathways linking unsafe acts during FFH accident development, indicating that documented accident processes involve interconnected rather than isolated behaviors. The identified directional relations represent reported behavioral progression in accident narratives. Practical implications The findings can support safety interventions that move beyond isolated unsafe acts by targeting structurally important behaviors together with their associated behavioral pathways. Application of the identified priorities to other settings should consider differences in local construction and accident-reporting contexts. Originality/value This study develops an evidence-linked approach for constructing unsafe behavior networks directly from accident narratives. By extracting directional behavioral relations from documented accident sequences rather than defining network connections solely through co-occurrence or expert judgment, the approach provides a reproducible basis for examining the organization and progression of unsafe behaviors in construction accidents.
Purpose This study aims to develop a probabilistic cost-benefit analysis framework for mine slope reinforcement, addressing the growing challenges of slope instability in coal mining areas. The model evaluates the impact of reinforcement measures on the probability of slope failure and economic performance, providing a decision-making tool for selecting cost-effective stabilization strategies. Design/methodology/approach A two-dimensional slope model was constructed using the Limit Equilibrium Method, incorporating failure probabilities derived under varying geological and mechanical assumptions. The framework integrates slope stability analysis, uncertainty modeling, and cost estimation, including initial investment, reinforcement expenses, risk-related losses, and post-failure restoration costs. A case study from the Jianbei Coal Mine in Shaanxi Province, China, was used to compare multiple reinforcement schemes combining different numbers and lengths of upper and lower anchors, with and without slope cutting. Findings The results indicate that anchoring reinforcement generally exhibits superior cost-effectiveness compared to slope cutting in the studied case. However, an optimal reinforcement level is identified, beyond which additional anchoring leads to diminishing returns and unnecessary cost escalation without corresponding safety benefits. The model captures the interaction between geological uncertainty, reinforcement design, and economic performance. Originality/value This study presents a novel probabilistic cost-benefit analysis method that explicitly incorporates slope failure probability into reinforcement decision-making. The proposed model provides a practical tool for evaluating reinforcement strategies under uncertainty, supporting optimized and economically efficient slope stabilization in mining engineering.
