Engineering

Abstract: Building orientation is a key determinant of residential energy performance; however, in consoli-dated urban areas it is largely constrained by historical urban planning decisions. This study quan-tifies the energy impact of inherited building orientations and evaluates balconies as passive so-lar-control strategies in Mediterranean residential buildings. A parametric simulation was conducted in DesignBuilder/EnergyPlus using a representative apartment building located in Alicante (Spain). Sixty models were generated by combining three dwelling positions, multiple building orientations, three balcony depths (0.0, 1.5 and 2.5 m), and two façade coverage ratios (50% and 100%). Annual heating and cooling demands were analysed under identical climatic conditions. Based on the sim-ulation results, three complementary indicators were developed: the Urban Orientation Energy Index (UOEI), which quantifies the energy effect of urban orientation; the Architectural Solar Correction Index (ASCI), which evaluates the effectiveness of balcony-based solar-control strategies; and the Urban Planning Energy Legacy (UPEL), which expresses the long-term energy implications of in-herited urban layouts. The results confirm that orientation is the primary factor governing annual energy demand, while appropriately designed balconies significantly reduce cooling demand in unfavourable orientations. The proposed framework provides a transferable methodology for as-sessing orientation-related energy performance and supporting climate-responsive architectural and urban planning decisions.

Abstract: Generative artificial intelligence (GenAI) is increasingly being explored for construction safety; however, its maturity and practical reliability remain unclear. This study systematically reviews GenAI applications in construction safety and assesses whether existing evidence supports reliable, source-grounded, human-supervised, and deployable applications. The review followed a predefined protocol and PRISMA 2020. Scopus, Web of Science, IEEE Xplore, and Google Scholar were searched in May 2026 for studies published from 2019 to August 2026. After duplicate removal and screening, sixty-five studies were included. Data on GenAI methods, safety applications, validation, deployment, and limitations were extracted and synthesized thematically. Six application areas were identified: safety knowledge retrieval and decision support; incident report analysis and risk prediction; multimodal hazard reasoning; synthetic data generation; worker-centered training, communication and behavior support; and generative optimization, monitoring, and safety control. Several studies provided proof-of-concept evidence using historical datasets, simulations, laboratory experiments, or expert assessments, while evidence of generalizability, field deployment, hazard-exposure reduction, and incident prevention remained limited. GenAI showed potential to support hazard recognition, decision support, and compliance-related safety tasks, but field validation, source grounding, and integration with preventive controls are needed for reliable deployment. A conceptual framework linking GenAI capabilities, and preventive safety controls is proposed.

Abstract: Circular architecture increasingly treats existing buildings as resources whose value can be prolonged through maintenance, adaptation and successive cycles of use. In historic public buildings, however, material retention must be reconciled with heritage significance, functional change, safety, technological upgrading and future adaptability. This study investigates selective conservation as a framework for making these decisions explicit within adaptive reuse. The framework combines four intervention modes—retain, adapt, integrate and replace—with seven whole-life criteria and a four-level evidence code. It is tested on Buildings 5 and 15 within the regeneration of the former Sant’ Osvaldo Psychiatric Hospital in Udine, Italy, using project documentation as empirical evidence. The application shows that circularity cannot be reduced to maximum material retention: selective replacement and new additions can extend the useful life of retained building resources when they are compatible, maintainable and appropriately documented. The framework therefore reads selective conservation as a decision logic linking treatment of existing resources to their consequences over time, without collapsing heterogeneous design choices into a single circularity score.

Abstract: Retractions show how a discipline deals with its own errors, yet the construction literature has not been examined from this perspective. This study analyzed the complete Retraction Watch database distributed through Crossref (72,389 records) and extracted 1385 records assigned to the architecture, engineering, and construction (AEC) sector, defined by the subject categories of the database and a manually compiled list of journals and conference series. Publication channel, reasons for retraction, time to retraction, and geography were examined. Retractions in construction do not form a single phenomenon. Conference papers account for 41.2% of the set; they are retracted quickly (median 35 days), in entire volumes, and 77.1% of their notices give no specific reason. Journal articles are retracted after a median of 529 days, and their notices state specific reasons, including paper mills (19.8%) and computer-generated content (14.5%). A title search for building information modeling (BIM) returned 52 records concentrated in a single publisher and a single country of affiliation. Systematic reviews in construction should verify the status of conference papers, the channel most affected by retractions, while notices without a justification prevent the discipline from distinguishing error from misconduct and learning from its own retractions.

Abstract: Courtyard-style campus buildings often suffer from poor outdoor wind environments affecting pedestrian comfort. This study proposes a deep learning-based rapid prediction and optimization framework to bypass the computational expense of traditional numerical simulations. A parametric dataset of 216 pedestrian-level wind fields is constructed via CFD simulations, varying building width-to-height ratio and wind incidence angle. A pix2pix conditional generative adversarial network is trained as a surrogate model to learn the direct mapping from building geometry to wind speed distribution. Unlike conventional regression networks that produce over-smoothing, the adversarial training preserves sharp velocity gradients and local extremes—features critical for comfort assessment. The model achieves a single-case inference time of 1.4 seconds. Results reveal that summer comfort coverage reaches 98.4%, while winter coverage drops sharply to 19.5%, identifying winter as the critical season for intervention. Parametric analysis shows that wind incidence angle exerts a pronounced effect: comfort area increases from 12.9% at perpendicular incidence to 24.1% at 45°. The optimal design is W/H = 1.4 and θ = 45°. The framework has been implemented as a prototype software system. This study provides a valuable reference for climate-adaptive design of campus courtyard buildings.

Abstract: This critical perspective review examines five major classes of carbon-sequestering cementitious construction materials: accelerated carbonation curing systems, geopolymers and alkali-activated materials, recycled cement derived from construction and demolition waste, biochar-modified composites, and reactive magnesia cements. The review draws principally on peer-reviewed literature published between January 2023 and early 2026. A strict distinction is made throughout between emissions avoidance, carbon capture and utilisation, and carbon dioxide removal, since failure to separate these categories continues to inflate climate claims in this field. Reported sequestration values are converted, where possible, to common functional units under explicitly stated assumptions, while the accounting level of each claim — binder, concrete, or composite — is identified in the corresponding tables and figures. Technology readiness is assessed through an evidence matrix covering laboratory validation, pilot deployment, commercial installations, standards availability, and field durability data. An assumptions-driven scenario analysis, using conservative, moderate, and optimistic variants, suggests that these material classes could contribute approximately 0.4–3.5 Gt CO₂/year by mid-century through combined emissions avoidance and carbon removal. However, strict atmospheric carbon dioxide removal is bounded at a substantially lower level, approximately 0.1–1.5 Gt CO₂/year, and remains constrained by feedstock availability, low-carbon CO₂ supply, processing energy, standardisation gaps, and unresolved durability evidence. Durability is treated not as a secondary technical issue but as a central condition for structural adoption. Particular attention is given to reinforcement corrosion risks in carbonated systems, efflorescence and alkali leaching in alkali-activated binders, strength penalties at higher biochar dosages, and moisture sensitivity in MgO-based systems. The regulatory analysis considers the revised EU Energy Performance of Buildings Directive, the EU Taxonomy, and investor disclosure frameworks, while extending the discussion to China, India, ASEAN countries, the Gulf region, and sub-Saharan Africa, where most future cement demand growth is expected to occur. The 2026 energy-price shock is discussed only as an illustrative stress test of a broader structural vulnerability: the high exposure of clinker-based materials to energy-price volatility. Overall, the review argues that carbon-sequestering cementitious materials can make a meaningful contribution to climate mitigation only if sequestration claims are accounting-transparent, durability-verified, and deployment-ready under real construction-market conditions.

Abstract: This paper reports the development and multi-dimensional experimental characterisation of post-consumer wood-based PLA (PCW-PLA) composite panels for prefabricated façade systems, presenting results across a panel design and four sequential assessment phases during early-stage development activities: design integration, manufacturing workability, acoustic characterisation, and mechanical performance. Two panel typologies were investigated: solid panels (5 mm, compression moulding) and honeycomb panels (9 mm total thickness: 2 mm face sheets with 5 mm hexagonal core, FDM additive manufacturing), benchmarked against conventional market-reference products across four prefabricated façade configurations. Phase 1 (design integration) confirmed thermal equivalence across all configurations (U = 0.35-0.158 W/m²·K per ISO 6946), within the gate criterion of ΔU ≤ 10%. Phase 2 (manufacturing workability) established that conventional on-site tools are incompatible with the material's hardness characteristics, constraining the product to off-site prefabrication, with CNC milling as the optimal route. Phase 3 revealed that the honeycomb panel achieves broadband sound absorption α = 0.40-0.55 at 1500-3500 Hz, qualitatively distinct from the solid configuration (α < 0.05 broadband). At façade system level, FS-W2W-01 achieves insertion loss 26-35 dB across 100-2000 Hz; FS-W2W-02 exhibits resonance instability at ~750 Hz and ~1200 Hz, hypothesised to originate from local mass-spring-mass coupling within the honeycomb stratigraphy, requiring acoustic design optimisation. Phase 4 characterised the honeycomb panel: bending strength fm,mean = 15.5 N/mm², modulus of elasticity E = 2640 N/mm², and internal bond ft,mean = 0.34 N/mm² (CoV ≈ 38%), with face/core interface bonding identified as the critical development parameter. The results confirm the technical feasibility of PCW-PLA composite panels for prefabricated façade applications at TRL 3–4, establishing thermal equivalence with market-reference systems, off-site CNC manufacturing as the only viable production pathway, and broadband acoustic absorption of the honeycomb geometry as the primary differentiating performance characteristic. Limitations inherent to the TRL 3–4 context – prototype material quantities, non-standard test configurations, and indicative mechanical results – are documented alongside phase outcomes and identified as development priorities for next-TRL advancement.

Abstract: University campus buildings consume significant energy through chilled water (CHW) systems, yet facility and energy managers lack tools that connect sensor data to actionable, spatially grounded decisions without requiring programming expertise. We present BearcatAI, an end-to-end analytics platform developed at the University of Cincinnati that unifies machine learning demand prediction, a large language model (LLM) conversational interface, an interactive 3D GIS-derived campus visualization, and a ROS-based robotic inspection system into a single web-accessible workflow. The platform is informed by ten years of monthly CHW records (July 2015–July 2025) and validated against 271 days of daily meter data across 70 UC buildings (June 2025–February 2026). An ensemble of XGBoost and Random Forest models achieves R2=0.987 and MAPE =21.8% on DAAP, the priority building identified by the UC utility engineering team. Beyond building-specific models, we introduce use-type models that pool buildings by functional category (Academic, Research, Administrative, Housing, Recreation, Medical) and learn a transferable CHW intensity pattern without any building-specific identity features. Leave-One-Building-Out (LOBO) cross-validation shows that the Research type model generalizes to unseen research buildings with median R2=0.911, and the Academic model achieves R2=0.918 on DAAP without having seen it during training, validating the hypothesis that a functional type model can serve as a generalizable benchmark for campus CHW consumption. A natural language interface powered by BearcatGPT, UC’s hosted instance of GPT-4.1, free for students and faculty, allows utility managers to query anomalies, request forecasts, and generate building charts through conversation. Anomalous buildings are automatically highlighted in a 3D campus model spanning UC Main Campus, Medical Campus, and Digital Futures. A ROS-based inspection robot, trained in a Digital Futures digital twin via reinforcement learning, is under ongoing development and is described as a future integration target. The platform is designed for open deployment on JetStream2 academic cloud infrastructure.

Abstract: Green facades and green roofs are increasingly promoted as core strategies for building decarbonisation, energy-efficient design, and ecological building concepts, with their environmental performance extensively documented through Life Cycle Assessment (LCA). However, the social dimension of these interventions occupant wellbeing, labour conditions embedded in the production of renewable and recycled materials, equity of access to green retrofits, and community acceptance remains largely absent from mainstream sustainability assessment. This paper conducts a systematic review, following PRISMA 2020 guidelines, of the LCA literature on green roofs and living walls published between 2015 and 2025, screening 264 records identified in Scopus and Web of Science down to 45 included studies. The review finds that environmental, economic, and technical performance evaluations dominate this literature, and that the evidence base is heavily concentrated on green roofs (91% of included studies) rather than green facades or living walls (9%). Only one included study applies a full Social Lifecycle Assessment (S-LCA) methodology to a green roof, confirming that social sustainability evaluation of vegetated building envelopes exists but remains an isolated, non-cumulative contribution rather than an established practice. Drawing on the identified findings, this paper develops a preliminary S-LCA framework for green roofs and living walls, modifying the five UNEP/SETAC stakeholder categories to better reflect the specific characteristics and social implications of vegetated envelopes. The framework provides a more comprehensive approach for researchers and policy-makers to assess the broader impacts of building greening strategies beyond conventional environmental performance indicators.

Abstract: Hospital fire safety standards in Saudi Arabia remain primarily prescriptive and do not explicitly account for patient mobility constraints, complex building geometry, and time-dependent fire dynamics. The 2015 Jazan Hospital fire, which claimed 25 lives and injured more than 100 people, demonstrated that code compliance alone cannot guarantee occupant safety. This study presents a four-phase BIM framework integrating Autodesk Revit, PyroSim/FDS fire simulation, and Pathfinder agent-based evacuation simulation, applied to a 250-bed hospital in Qatif, Saudi Arabia. The FDS computational mesh was assessed at D*/δx = 12.2 against NIST NCSTAR 1-5E benchmark data (mean errors: temperature 6.8%, smoke optical density 7.2%), and Available Safe Egress Time (ASET) was determined using the multi-criterion ISO 13571:2012 methodology. Under the examined SBC 2018 baseline (Scenario 1, four corridor-end exits), smoke penetrated the ICU corridor within 60 seconds, yielding ASET = 118 s and a Monte Carlo Required Safe Egress Time (RSET) of 412 ± 35 s (n = 50; approximate 95% CI for the mean: 402–422 s), an ASET–RSET deficit of 294 seconds. Scenario 2, incorporating fire-rated ICU/OR compartmentation and ten distributed emergency exits, extended the reported ASET beyond 185 s and reduced mean RSET to 281 ± 22 s (approximate 95% CI for the mean: 275–287 s), a descriptive reduction of 32%. On the reported ASET range of 185–225 s, however, RSET remains greater than ASET; the residual deficit is 56–96 s. Sensitivity analysis across three ignition locations indicates that the Scenario 1 deficit is not specific to one fire origin. The findings support performance-based BIM simulation as a complement to prescriptive code assessment in hospitals.

Abstract: Integrating Wire Arc Additive Manufacturing (WAAM) into the Selective Paste Intrusion (SPI) process enables the fully additive fabrication of reinforced concrete structures with complex geometries. However, the WAAM process transfers thermal energy into the reinforcement and the surrounding concrete matrix, which may affect the bond between steel and concrete. This study addresses whether the heat introduced by the welding of reinforcement segments impairs the bond between reinforcing steel and SPI-fabricated concrete. To isolate this thermal effect from the geometric process effects covered in a companion study, two groups of 16 mm reinforcing bars embedded in SPI-fabricated concrete were compared: bars whose segments were joined during printing by manual welding at seven joints per bar (process-integrated thermal exposure), and bars with adhesively bonded segments without thermal input as the reference. Eight specimens per group (n = 8) were tested in push-through. Thermography corrected for the emissivity of the steel indicates estimated bar-surface temperatures of 215 to 275 °C near the joints immediately after welding and of 125 to 162 °C after 2 to 3 min, and the bars did not cool completely to ambient temperature between cycles. Despite these temperatures and the repeated heating and cooling cycles during printing, no loss of bond capacity relative to the adhesively bonded reference group was resolvable: the welded group reaches a mean maximum bond stress of 20.5 MPa, the adhesively bonded reference 22.4 MPa, and the difference of 1.9 MPa lies within the scatter of the two groups (coefficients of variation 22% and 21%). The failure mechanism was likewise the same for welded and adhesively bonded specimens: all failed by bar push-through with the same bond-slip sequence and a similar slip at maximum bond stress. Relative to the compressive strength of 72.4 MPa determined for the same mixture on cast prisms, the bond capacities correspond to roughly 30%. The segment joints were loaded in compression only, and their tensile and fatigue capacity was not characterised. Moreover, since the procedure investigated here resembles a stud welding process rather than genuine WAAM fabrication, the results are promising but do not yet provide sufficient evidence that WAAM can be fully integrated into the SPI process. In genuine WAAM, by contrast, welding occurs far more frequently than in the present procedure because of the smaller layer heights, so that heat accumulates and overlapping thermal cycles can produce even higher sustained temperatures at the steel–concrete interface.

Abstract: Progressive Design-Build (PDB) differs from traditional lump-sum Design-Build (DB) by allowing project design, scope, and cost to be progressively developed before the final construction price is established. However, empirical comparisons of these two delivery methods across infrastructure sectors remain limited. This study compares the performance of lump-sum DB and PDB projects in the transportation, water/wastewater, and aviation sectors using data obtained from the Design-Build Institute of America database. After data screening, the final sample consisted of 95 lump-sum DB and 37 PDB projects. We evaluated cost growth, schedule growth, and construction intensity using descriptive statistics and nonparametric statistical tests. PDB projects exhibited significantly lower mean cost growth than lump-sum DB projects (1.76% vs. 3.24%, respectively; p = 0.006). Construction intensity was also significantly lower for PDB projects (\(141K/day) than for lump-sum DB projects (\)358K/day) (p < 0.001). No statistically significant difference was found in schedule growth (p = 0.131). Overall, the findings suggest that PDB may offer improved cost control while maintaining schedule performance comparable to lump-sum DB for infrastructure projects.

Abstract: High-density residential areas require environmental measures that address multiple anthropogenic pressures while remaining spatially feasible and compatible with economic and social objectives. Existing selection approaches often treat these requirements separately and rarely incorporate built-form morphology as an explicit constraint. This study develops a multi-criteria methodology based on a combined two-stage optimization model. The method classifies measures targeting chemical, physical, and biological hazards; links them to functional, technological, and evaluation attributes; calculates environmental relevance and scenario-specific sustainability scores; and excludes spatially infeasible options through deterministic geometric filtering. The methodology was tested in a residential area in Rostov-on-Don, Russia. The results demonstrate that alternative priority weights produce distinct but comparable sets of measures while preserving the same measurements and feasibility rules. The proposed approach supports the reproducible selection and spatial allocation of measures during the preparation of residential-area improvement projects. The algorithm was also implemented in an online platform that enables stepwise data entry and scenario-based recommendations.

Abstract: Artificial-intelligence infrastructure increasingly uses prefabricated modules, yet deployment boundaries need not match the different renewal clocks of computing and supporting assets. Here we introduce Modular Integrated AI Infrastructure (MiAI), a paradigm that co-designs physical assemblies, functional interfaces and renewal boundaries across computing, power, cooling, energy storage, communication and sensing-and-control. Drawing on modular integrated construction (MiC), multi-trade integrated mechanical, electrical and plumbing (MiMEP), and prefabricated modular data centres, MiAI combines factory integration and testing, transportable assemblies and standardized, reversible physical--digital connections to support staged deployment, selective renewal, relocation and multi-cycle reuse. It is physically modular, functionally integrated and temporally decoupled. We demonstrate its lifecycle-renewal mechanism through a differential lifecycle global-warming potential framework for compute transitions and exposed support infrastructure. Service-weighted effective retention, \(R_{\mathrm{eff}}=q\Delta r v\), combines transition-exposed support scope, additional physical retention relative to a deployment-oriented modular baseline that already retains support assets, and qualified remaining service. Favourable outcomes arise above the burden threshold, \(R_{\mathrm{eff}}>R_{\mathrm{eff}}^*\), so that retained-service credits exceed interface and transition burdens. In an illustrative normalized same-site configuration, the additional physical-retention threshold within the transition-exposed subset is \(\Delta r^*=0.20\). Combining this break-even condition with physical retention limits separates favourable, burden-dominated and physically infeasible configurations. MiAI makes the renewal boundary an explicit design object, extending modularity from how AI infrastructure is delivered to what it preserves across technology generations.

Abstract: The early-stage design process in architecture is critical for shaping conceptual, aesthetic, and functional ideas, and it is also the most influential phase for determining embodied and operational carbon emissions. Key variables such as structure, materials, form, and size are largely established before construction. However, in architectural education, students often prioritize form and aesthetics while underestimating the integration of sustainability concepts needed to reduce carbon emissions. This is particularly evident in relation to acoustics, which can contribute to buildings energy performance, but it is often treated separately from sustainability-oriented design education. In response to this gap, this exploratory pilot educational study examines a design studio exercise to understand how early-stage acoustic design decisions influence the building performance. The study adopts a comparative case-study approach, investigating four comparable student projects developed by 13 students in the Building Utilities 2 course at the American University of Ras Al Khaimah in the United Arab Emirates, which focuses on acoustic and lighting design. A quantitative method is applied to analyze the design studio exercise through building performance simulations. Based on building simulation results, the acoustic design strategies reveal varied outcomes across four projects. Some projects show reduction in total building carbon, and annual energy use intensity, while embodied carbon increases in some cases because of material selection decisions. These findings suggest that acoustic design can support performance-based decision making during the early-design stages within a studio environment. The study also highlights the need for architectural education to integrate sustainability strategies that encourage future architects to consider carbon, energy, material, and acoustic performance together in their design processes. In this way, simulation-based design pedagogy can strengthen students’ understanding of embodied carbon reduction strategies and low-carbon building design.

Abstract: Assessing potential copyright infringement or plagiarism in architectural design requires the systematic comparison of visual information, including building appearance, form, exterior materials, and opening configurations. However, conventional review practices often rely on qualitative expert judgment, which can limit objectivity and repeatability. This study proposes a deep learning-based information extraction and similarity review support system that automatically extracts building image information and supports the retrieval of visually similar buildings, rather than directly determining legal plagiarism. The proposed system consists of data, architecture, exterior material, structural, and frame modules. The data module refines web-crawled images using rule-based filtering, DINOv2 feature extraction, and K-Means clustering. The architecture module combines SAM, ResNet-50, and DeepLabV3+ to identify building objects and extract building regions and contour information. Experimental results showed that 10,026 images were retained as the final database from 42,172 collected images. The ResNet-50 classification model achieved more than 80% Top-1 accuracy, while the DeepLabV3+ segmentation model achieved more than 85% aACC and mACC and approximately 80% mIoU. The findings demonstrate the feasibility of image-based building information extraction as a support framework for architectural design similarity review. Future work should further refine quantitative similarity scoring and its connection to legal assessment criteria.

Abstract: Complex construction processes have many state variables, a lot of redundant monitoring indicators and after occurring process transitions, the relationships between monitoring indicators are not observable. In this study, a 36 dimensional state vector is created, which is an observable digital twin model of construction processes (CP-ODT) that can be used to model schedule deviations, resources loading, component orientations, equipment operating states, work environments in construction, etc. State transitions between different construction phases, such as structural construction, MEP installation, envelope construction, and interior work, are modeled through the use of phase transition matrices. The model takes the observability Gram matrix and assesses which 92 candidate indicators are contributing to identifying each state, and determines which is the minimal set of metrics and where to place sensors, given the cost of the measurements and node failure. A total of 268 days of experimental data were gathered from four building projects for the 7, 936 component hoists, 1, 824 work packages, 124 sensor nodes, and 19.86 million records. Finally, CP-ODT was able to keep 27 metrics and 46 sensor nodes with the minimum eigenvalue of observability Gram matrix improving from 0.011 to 0.058. The error in work package completing time reduced from 1.21 days to 0.54 days, the error in hoisting cycle time reduced from 79.6s to 32.8s, the error in component orientation reduced from 0.94° to 0.39° and the error in equipment load reduced from 8.6KN to 3.7KN compared with the fixed-measurement-point state model. The results show that CP-ODT is able to identify key states in complex construction processes with a small number of measurement metrics and continues to monitor processes in the transitions.

Abstract: Deploying macro-scale Digital Twins (DT) in the AECOO sector requires bridging micro-level engineering precision (BIM) with macro-scale geospatial dynamics (GIS). This review investigates the technological evolution of GeoBIM integration as a foundation for macro-scale DTs based on 2024–2026 developments. Evaluating the idealized paradigm of fully bidirectional DTs, the paper categorizes digital maturity levels (Digital Models, Digital Shadows, and true Digital Twins) across major infrastructure cases, including smartBRIDGE Hamburg, the M-30 Highway, the Zurich City Twin, and Virtual Singapore. Furthermore, university campuses are evaluated as multi-scale urban living labs. An empirical case study of the 72-hectare Warsaw University of Life Sciences (SGGW) campus demonstrates how integrating GIS spatial analyses with parametric BIM models enables diagnostic accessibility modeling, universal design, and urban resilience planning for vulnerable populations. To overcome cross-lifecycle data fragmentation, the DOLCE-grounded BIM-Phase ontology is highlighted for preserving physical element identity across temporal states via OWL 2 DL. Finally, the study outlines technical barriers and development trajectories toward Cognitive Digital Twins (CDT) augmented with Artificial Intelligence and Large Language Models.

Abstract: Showering represents a major component of residential water consumption and domestic hot-water energy demand. Although thermostatic mixing valves (TMVs) are widely used to improve thermal comfort and reduce the risk of scalding, their influence on showering behaviour and associated resource savings remains poorly quantified. This study proposes a conceptual model linking shower duration to temperature adjustment activities, flow interruptions during soaping, and user responses to hydraulic disturbances. Model parameters were estimated through two experimental campaigns involving 40 volunteers and 200 instrumented showers, complemented by an online survey of 1,000 French respondents. Four faucet technologies were compared under controlled laboratory conditions: a thermostatic mixing valve, a mechanical mixer tap, a two-handle mixing tap and a pressure-balance valve. The results showed that TMVs significantly reduced the initial temperature adjustment duration compared with all other faucet technologies, with reductions ranging from 45% to 60%. TMVs also led to significantly lower mixed-water temperatures, averaging 2.05 °C below those observed with conventional faucets. While no significant reduction in total shower duration was observed during experimental testing, the analysis revealed that TMVs substantially decrease adjustment-related activities during showering. The proportion of users interrupting water flow while soaping was estimated at 66%, whereas 41% of respondents reported being affected by hydraulic disturbances in their dwelling. Based on representative French showering conditions, the proposed model predicts annual savings of approximately 1.6 m³ of water and 96 kWh of energy per person. Extrapolation to the French residential building stock suggests a theoretical saving potential of 69 million m³ of water and 4.1 TWh of energy per year. These findings demonstrate that TMVs can reduce the environmental footprint of showering while maintaining user comfort and highlight their potential as a large-scale water- and energy-efficiency measure in residential buildings.

Abstract: The paper presents the results of monoaural and binaural acoustic measurements and speech intelligibility tests conducted in two classrooms representative of typical school buildings in Italy. The sample of two classrooms was selected based on a statistical analysis of reverberation time measurements taken in more than one hundred Italian schools. In the two sample classrooms, measurements were taken both when the room was empty and when it was occupied by pupils. The binaural measurements were conducted on two students who were fitted with special headphones equipped with binaural microphones, with sound sources placed in various positions within the classroom, both in the acoustically untreated classroom (with a long reverberation time) and in the acoustically treated classroom (with a short reverberation time). The results of the speech intelligibility tests conducted on 25 elementary school students confirmed that speech intelligibility improves significantly in the acoustically treated classroom.