Modular construction uses off-site modules for on-site assembly; computer vision supports progress monitoring, inspection, and assembly. These tasks may occur in poorly lit spaces, where low light degrades image quality, while prior work has not addressed low-light enhancement for this setting. We propose RetinexRefiner, a two-stage framework that first performs illumination-based enhancement and then refines the result to suppress residual artefacts such as colour cast and noise. Trained end-to-end, it achieves improved visibility on night construction images, reaching 28.07 dB PSNR and 0.83 SSIM, and is applicable to construction sites and similar dim environments.
Indian steel and aluminium megaprojects frequently
experience multi-year schedule overruns despite detailed
planning and risk registers. This paper presents a
predictive delay-risk framework that learns from 20
completed projects using contextual similarity and
interaction-aware modelling of 39 delay events. Projects
are encoded through execution context, while events are
represented using normalised delay impacts and a
structured dependency network derived from empirical
co-occurrence and expert judgement. A prototype tool
generates probabilistic completion envelopes and ranked,
interaction-aware risk lists for new projects. Initial results
align with observed completion behaviour and highlight a
small set of high-leverage risks for targeted mitigation.
Potholes are a significant cause of vehicle damage and road safety accidents. Existing inspection practices rely on manual surveys using inspection vehicles or vehicle-mounted cameras, which are time-consuming, non-scalable, and often provide incomplete visibility. This paper presents a comparative study of instance-segmentation-based deep learning models for pothole detection using UAV imagery, enabling wider and more consistent coverage. A custom dataset was developed and augmented using geometric transformations. YOLOv8 and transformer-based RF-DETR models were evaluated using precision, recall, and mAP@50. The best model achieved mAP@50 of 66.3%, highlighting challenges in detecting shallow potholes under limited training data.
Requirement verification in integrated contract models, such as Design & Construct, remains a manual, document-centric, and error-prone activity. While technical verification through Automated Code Compliance has advanced significantly, process requirements are typically verified through human inspection of textual documentation. This research addresses this gap by developing a novel artefact architecture that integrates general-purpose Large Language Models, the RASE mark-up language, and Knowledge Graphs to automate traceability between contractual obligations and contractor deliverables. Tested through iterative Design Science Research, the final prototype achieved a precision of 1.000 and accuracy of 0.968, offering a transparent, "white-box" framework for contractual auditability.
Biobased construction materials store atmospheric carbon, yet current life cycle assessment and environmental product declaration practices represent biogenic carbon as aggregated indicators with limited traceability across value chains. This limits transparency, attribution, and credibility of carbon-related claims. This paper explores how blockchain-enabled system principles can address these deficiencies by treating biogenic carbon as a persistent, traceable state linked to material flows. Through an exploratory mapping onto bamboo value chains, we identify how such ledger-based logic can formalize carbon continuity, prevent double counting, and support auditable life cycle transitions. The paper contributes a conceptual framework for carbon ledger approaches in construction.
The Architecture, Engineering, and Construction (AEC) sector faces complex regulatory frameworks like building energy performance certification under the European Energy Performance of Buildings Directive (EPBD). Traditional approaches suffer from fragmented documentation, procedural ambiguities, and poor traceability and auditability, causing outcome inconsistencies and automation challenges. This paper proposes a decoupled conceptual architecture for digitally actionable and traceable regulatory workflows, leveraging compliance modeling, semantic web technologies, and linked data. By separating semantic representation, reasoning, execution, and interaction layers, it improves compliance, interoperability, and adaptability. Demonstrated on Estonian Energy Performance Certificates (EPCs), it addresses key gaps and supports broader application beyond energy efficiency.
The Industry Foundation Classes (IFC) standard underpins BIM interoperability, but its STEP/EXPRESS stack is ill-suited to cloud-native, transactional workflows. Building on buildingSMART community-driven requirements and proposals, review of alternative solutions, and iterative prototyping, we propose the modernisation of IFC. The work includes a proposal for a data schema, JSON serialisation, and exchange mechanisms, while preserving IFC’s role as a semantically rich neutral data model. Innovation includes a compositional, layered federation model allowing incremental updates, multi-author collaboration, versioning, and fine-grained, traceable ownership. We compare the proposal to existing standard and discuss the potential, trade-offs and implications.
Requests for Information (RFIs) remain a critical yet poorly integrated component of construction workflows, typically managed as unstructured text disconnected from Building Information Modeling (BIM) data. Existing studies primarily focus on reducing RFI frequency or response time, with limited focus on linking RFI content to specific BIM elements. This study presents a data-driven framework that combines large language model (LLM)-based intent extraction with BIM metadata reasoning through a staged matching pipeline. By separating semantic cue extraction from rule-based element selection, the framework enhances interpretability and robustness. Validation on experimental and real-world datasets demonstrates reliable element and family-level retrieval.
Residential prosumer buildings with PV and high-load subsystems exhibit irregular demand, yet current residential digital-twin and baselining studies rarely address missing-data realism and causal deployment constraints. This paper presents a monitoring-oriented operation-phase digital twin analytics pipeline for an instrumented villa in Segrate, Italy, linking whole-building demand, PV output, pool-circuit power, and outdoor air temperature to a one-hour-ahead baseline task. A fixed Random Forest is evaluated across eight dataset configurations against persistence and Multiple Linear Regression. Under deployment-realistic causal conditions, the best configuration achieves RMSE 576.5 W, demonstrating the value of a traceable leakage-safe baselining workflow for residential prosumer settings.
Data-driven approaches increasingly inform urban policy, yet the integration of intangible cultural heritage remains limited due to object-centric documentation and the separation of tangible and intangible domains. Existing frameworks conceptualize data as static records, inadequately supporting the semantic articulation of cultural meaning within urban contexts. This paper applies a semantic narrative approach based on a triadic model linking physical elements, procedural knowledge, and social practices. The framework structures heterogeneous cultural data into interpretable relational sequences. Through ecclesiastical heritage in Karachi, Pakistan, the study shows improved coherence between cultural data and policy-oriented interpretation and contributes to knowledge applicable to urban transformation.
