ABSTRACT. At the EU level, the Digital Product Passport has been defined across several industries, which seems to overlap with material passports. The industry is divided, and confusion persists on what a material passport is, what it contains, and how it should be issued. In this article, we try to address some of these concerns and identify the data sources, tools, and methodologies that could influence the building circularity passport. To this end, we adopt two key perspectives: (1) product/component level and (2) building level. For each perspective, we analyse existing frameworks, literature, and case studies and rank the importance of the identified tools and methodologies which have become prevalent. The ranking of data sources is done considered several data domains (e.g. composition, circularity, etc.). A discussion of the two levels is provided based on the several approaches, with limitations highlighted.

ABSTRACT. The European Commission will implement Digital Product Passports (DPP) under Ecodesign and Construction Products Regulations to support sustainable goals. Products play a crucial role in sustainable growth, and built entities, such as residential or service buildings, are an assembly of products. Level(s) is a methodology focused on improving built environment sustainability by performing several assessments where product data is relevant. This research article explores how Level(s) estimation can be streamlined using DPPs, focusing on the indicators associated with circularity and cost. The findings suggest that DPPs can provide important Level(s) data and streamline the estimation processes. From a barriers and opportunities perspective, it is concluded that adjusting to existing processes and deliverables is found to be an essential strategy for success. Data availability is a common challenge for Level(s) and DPPs, where manufacturers are key stakeholders. It is upon them that many actions will enable the goal's accomplishment.

ABSTRACT. The current software landscape is not fit for the evolving requirements around sustainability reporting and the management of built assets over their life cycle. This paper builds on prior stakeholder requirements-gathering and ontology development work, detailing the implementation stages to arrive at an end-to-end Asset Life Cycle Knowledge Graph and sample queries to support several analytical use cases. The proof-of-concept graph and queries are evaluated qualitatively through broad stakeholder focus groups and the study receives resounding positivity that the technical contribution push the IT landscape in a suitable direction, but that a significant shift is required in the commercial paradigm around information management and sustainability reporting incentives. The domain insight, generated via the expert focus groups, provides useful future research directions, pointing to feasible alignment with bSI initiatives such as bsDD for intuitively visualising asset data and ensuring long-term information sustainability.

ABSTRACT. Digital building documentation plays a decisive role in sustainable and resource-saving construction and the preparation and dismantling of buildings. For the sustainable and efficient use of buildings, the construction and real estate industry needs the relevant data for the operation of the installed building products and technical systems. In the course of digitalization in the construction industry, Building Information Modeling (BIM) is a promising solution for this, which can be used to link product information with a 3D data model, among other things. In Germany, the construction industry plays a central role in the creation of building documentation. Efficient and sustainable building documentation contributes to achieving the UN's Sustainable Development Goals, making documentation even more important. As a result, the topic of sustainability in the construction industry is becoming increasingly important, as specific templates also need to be created for the documentation and evaluation of construction products from a sustainability perspective. As part of the research projects "BIM Use Cases in the construction industry" and "Building contractors as the central instance of digital building documentation - DigiBauDok", templates for the creation of building documentation were developed on the one hand. On the other hand, technical solutions for automatically linking the documents and digital product data (digital product twins) with the BIM models were developed and validated in practice.

ABSTRACT. In response to the activity-based productivity concerns in construction environments, we developed a multifaceted computer vision approach merged with BIM models. High-level process information is derived from continuously acquired site images by the following computational processing chain: (1) Worker activity is classified using the proposed vision transformer network ViTPoseActivity, leveraging human pose features to detect worker activities. (2) On-site labor activities are analyzed according to their on-site impact and fused with the corresponding BIM geometry. Our model, ViTPoseActivity, achieved 92.31 % accuracy while surpassing previous prediction speeds, demonstrating an effective trade-off between computational cost and precision in activity analysis. Unlike previous studies, our approach was deployed on a large real-world dataset, carefully investigating subtasks and affording productivity insights on reinforcement activities. Integrating as-performed and geometry information supports construction management by facilitating better decision-making regarding worker group definition and task allocation. Our research fills a crucial gap by providing a robust and efficient method to assess on-site labor productivity.

ABSTRACT. Although there is a significant increase in the use of advanced technologies in the construction industry, there remains a gap in automating the monitoring activities for construction and demolition waste. This research investigates an approach for automating the monitoring of dumping and removal events in construction and demolition waste management. Focusing on the initial phase of a complete monitoring framework, the study proposes a system that employs computer vision algorithms to detect the event of adding or removing construction waste from the static dumpsters on construction sites, or waste sorting facilities. The developed system eliminates the needs for manual monitoring, thereby reducing cost and increasing efficiency. To implement the system, data collection, cleaning, and annotation efforts were made for curating a high-quality dataset using real world data. This foundational work sets the stage for future advancements, such as integrating the proposed system with material detection models and volume estimation techniques to enable the precise identification and quantification of waste types, resulting a complete monitoring framework for construction and demolition waste management.

ABSTRACT. Building deficiencies such as heat loss, façade defects, and erosion inevitably result in inefficient space heating, contributing to high energy consumption and associated greenhouse gas emissions. Moreover, façade deficiencies may generate public safety concerns, including the risk of falling façade elements due to erosion and faulty craftsmanship. Manual inspection of such infrastructure is challenging due to accessibility constraints dictated by structural characteristics and the reliability of assessments. Consequently, manual inspection consumes significant time and financial resources while endangering the safety of façade inspectors. Utilizing unmanned aerial vehicles (UAVs) to capture visual and thermal images presents a promising solution enabled by advancements in drone technology and AI data processing capabilities. Relevantly, this study aims to categorize and summarize the standard and regulatory landscape, map requirements, describe implementation processes and outcomes, and develop conceptual abstractions to generalize insights. For this purpose, a case study was conducted in Singapore following the Periodic Facade Inspection (PFI) regime of the Building and Construction Authority to propose a solution facilitating the adoption of AI and UAVs for building inspection. The study's unique contribution is generating generalizable insights into inspection processes to commercialize AI for critical applications such as façade inspection. Stakeholders, including regulators, building owners, asset, and facility management companies responsible for ensuring building safety and efficiency, and academia seeking to inspire R&D activities, could benefit significantly from the study's outcomes.

ABSTRACT. This paper explores carpentry students' perceptions of integrating Mixed Reality (MR) and construction component tracking technologies within carpentry and assembly processes using the Twinbuild MR tool. Employing a case study approach, the research focused on the real-time assembly of a complex timber pavilion to demonstrate these technologies' practical applications and benefits. For the first time, carpentry students used this technology to assemble a section of a larger structure while the research team gathered empirical data. Various data collection methods were used to evaluate the students' mental and cognitive loads during task completion, complemented by semi-structured interviews. Findings yielded insights into the key advantages and challenges of using MR technology in construction assembly, along with recommendations for improvements. The findings highlighted the technology's potential to enhance assembly processes, reduce errors, and improve construction education and practice quality. However, areas for refinement were identified, including user interface design, graphic accuracy, and hardware. The study underscores MR and tracking technologies' transformative impact on construction training and practice. By integrating these technologies, construction education can move beyond traditional methods to offer immersive and interactive learning experiences, enhancing understanding and skill acquisition by effectively linking theory with practical application. The research emphasizes the need for ongoing development and integration of such technologies in educational curricula and industry practices to keep pace with technological progress and evolving sector needs.

ABSTRACT. Increasing use of BIM requires solutions for sharing Product Data (PD), generated and governed by manufacturers, across applications and actors of the AECO industry. PD must be available in a machine-interpretable way to facilitate digital processing, but is usually provided in proprietary data formats. Aiming to analyse manufacturers’ current approach to PD provision, a scoping literature review was conducted, supported by a case study investigating a database providing IFC BIM objects, and the possibility to represent PD using Linked Data (LD). Findings suggest that PD provision is limited for reasons spanning from lacking expertise to data management issues. While the representation of PD in IFC BIM objects as well as with LD was concluded to be possible with currently available means, it has yet to be adopted by the industry.

ABSTRACT. The absence of a standardised, neutral product catalogue format capable of representing parametric and variable product data poses a challenge in the construction industry. Searching and selecting technical building equipment products remains a laborious and manual task. Using a product example, this paper explores and evaluates the constraints of parametric and variable product representation methods in the catalogue format structure influencing configuration processes. A literature review has identified crucial information for product selection required in configuration. We propose an example configuration process that includes a building model and a product data catalogue as data resources. The result shows that selection hierarchies can only be partially mapped due to a lack of flexibility in addressing individual multi-column table values and missing data control structures. The research intends to support the development of a product catalogue file format to facilitate product decisions with the help of efficient configuration processes.

ABSTRACT. The introduction of takt production and the application of single-piece flow in the interior phase of construction projects has highlighted the problems in the frame erection phase of concrete element construction. Despite implementing flow production, the frame erection phase has become a bottleneck in shortening the lead time in residential building construction. The study investigated the implementation of the structural phase in three consecutive projects and utilised previous studies' process performance measurements and deviation analysis. The study revealed that the design and implementation of the structural phase suffer from a lack of process and product information flows. Similarly, it was found that a contributing factor to the poor level of digitalisation is the engagement of the national data architecture, which has been adopted in the industry using the Manufacturing to Stock (MTS) business model and strategy. As concrete element supply chains are based on a different production logic, Engineer to Order (ETO), the study defined significant differences in data structures. It required data architecture regarding process and product information from MTS architecture. The study proposes a national or EU-level solution model for the design and product information exchange for concrete element supply chains. Based on the case study, the research formed a national reference model for implementing decentralised data architecture and digitising inter-company data transfer. The study was conducted as part of a project by a national advocacy organisation in the construction industry aimed at digitalising the concrete element supply chain. The research results are significant in the studied market area, as they contribute to the implementation of digitalisation by demonstrating the implementation method of the data architecture for ETO supply chains and enable the same development in digitalisation for the element industry that the stock product manufacturing sector has already achieved in the digitalisation of the supply chain, wholesale operations, and the implementation of machine reading and automated data processing.

ABSTRACT. In light of the escalating emphasis on environmental stewardship, carbon management has emerged as a pivotal element in the construction industry. However, carbon emission information is often accompanied by cross-department information exchange needs, and the secure and accurate management of its data is still difficult. Therefore, this paper proposes an IFC-based embodied carbon assessment and tracking framework using blockchain to provide carbon management. This framework encompasses the IFC-based carbon emission information supplement and checking, coupled with developing a smart contract. It facilitates accurate collection of embodied carbon, grounded in BIM, across the Life Cycle Assessment (LCA) stages A1-A5, while simultaneously safeguarding model correctness and data traceability. The practicality of this framework has been substantiated through a designed derived scenario from OPARK2 in Hong Kong, with its utility corroborated by illustrating both the projected and actual embodied carbon emissions of the construction project in question.

ABSTRACT. Industries like construction have recently been making significant investments in information-driven management and control of physical systems. Such models are commonly referred to "Digital Twins". However, in the construction safety domain, a digital twin (DT) remains vastly undefined. No consensus exists on two essential aspects: (a) the connection between the physical reality of a construction site (the "physical" twin) and the corresponding computer model (the "digital" twin) and (b) the most effective selection and exploitation of real-life data for supporting the safe design, planning, and execution of construction. This paper outlines the concept for a Digital Twin for Construction Safety (DTCS), defining four essential steps in a DT workflow: (1) safe workplace design and planning for hazard prevention, (2) conformance checking for ensuring compliance, (3) risk monitoring and control for proactive prediction and alerting, and (4) continuous performance improvement for personalized- or project-based learning. DTCS should be viewed as a system-based approach enhancing the overall performance rather than exclusively integrating sensing information or generating knowledge in Building Information Modeling (BIM) for safety purposes. Our result is a DTCS including the description of its modules.

ABSTRACT. Digital twin (DT) has the potential to facilitate the reduction of carbon emissions over asset lifecycle, which is a critical target that should be instructed by real-time data. However, the trustworthy carbon data to feed into DT to support environmental impact assessment, identification of hotspots, and carbon reduction progress monitoring is largely missing in literature and practice due to the lack of a clear asset carbon information requirements (ACIR). This paper sets out to develop the ACIR for the collection of fine-granular carbon data for highway DT by reviewing relevant standards, industry guidelines, and tools and engagement with industry experts through a design-thinking workshop. A total of 74 ACIR are added to the ISO 19650 asset information requirements, among which 38 are sustainability or carbon directed and 12 are in the technical aspects. These ACIR provide a foundation for developing a carbon DT for highway assets that can facilitate real-time sustainability-orientated decision making. Future research directions to facilitate the completeness and the implementation of the ACIR are recommended.

ABSTRACT. Precast construction faces challenges that arise from differing priorities in terms of production optimization between the production and the erection functions. Factories prefer large production batches, which does not align with the needs of erection crews for sets of varied pieces at each step in erecting a building. We propose a Digital Twin Construction framework for closed-loop monitoring and control of precast concrete production and construction. A core module in the framework performs adaptive global optimization of the system. The optimization module compiles sets of production plan parameters using planning heuristics, evaluates the plans using agent-based simulations, and then optimizes for plan parameters using a genetic algorithm. A novel utility function works to minimize production waste throughout the system, rather than seeking shortest project duration. The result has minimum cost with greatest value. Once implemented, the framework as a whole may enable automated monitoring and control in a closed loop that optimizes production plans to enhance efficiency, reduce waste, and improve coordination, ultimately streamlining the entire precast construction process.

ABSTRACT. Virtual reality (VR) technology has been increasingly used in investigating pedestrian behaviors in normal and emergency scenarios. However, traditional VR experiments are confined to lab settings, limiting the number of participants coexisting and interacting in immersive virtual environments (IVEs) due to space and device constraints. We propose a novel approach - crowdsourced virtual reality (CVR)-based experimental approach, which expands traditional lab-based VR experiments to an online environment and allows to access participants in a crowdsourced manner. The proposed approach contains a web-based platform and a process model. A classic experiment on unidirectional pedestrian flow was conducted to validate the CVR experimental approach. The widely-recognized characteristics of pedestrian dynamics were successfully reproduced in the experiment. The first of its kind, the proposed approach enables researchers to access participants across barriers of geography, culture, age, and social identity, and yield representative and generalizable behavior data for pedestrian and evacuation dynamics (PED) research.

ABSTRACT. The spatial alignment of digital resources is necessary for each process in the life cycle of an asset in the built environment. However, there is currently no way to explicitly describe the spatial relationships inherent in the resources in a machine-readable and vendor-neutral way. Therefore, the paper presents a schema to superimpose heterogeneous data in a digital space. The schema defines four main classes inherent in the Architecture, Engineering and Construction (AEC) industry: Global, Asset, Document and Entity Space. Global Space represents the Asset's physical location, while the Asset Space is the digital replica of the built object superimposing all related resources of the building. Each resource defines its own Document Space, which in itself contains one to many Entity Spaces. An Entity Space is a coherent unit of information, such as a section or a model. Moreover, the schema provides a means to express the relationships between the different space types. To verify the approach, we demonstrate the application with an example bridge data set. This work should serve as a preliminary step towards automating the spatial linking of heterogeneous resources in the built environment.

ABSTRACT. Traditional 3D modeling in the AECO industry often emphasizes geometric data over semantic details, leading to potential inconsistencies. This paper introduces an innovative approach to enhancing Building Information Models by utilizing geometric intelligence to improve semantic information quality. Given that an object’s geometric features and semantic properties are intrinsically linked, our method focuses on extracting implicit semantic insights from high-quality geometric data through feature extraction followed by an AI-based semantic classification. By creating a framework incorporating this ’Geometric Intelligence,’ we reduce the dependency on manually entered data and identify inconsistencies between geometric and semantic information. This research underscores the potential of geometric data to enrich semantic content, promoting more efficient, safe, and sustainable building practices.

ABSTRACT. In Building Information Models (BIM), spaces and zones are defined to bundle similar sub-space uses or to cluster assets for efficient retrieval and control with building automation. While spaces and zones are often explicitly represented in a BIM, their boundaries are not always defined by physical elements such as walls in the physical world. This is especially true for open space layouts such as in open office or retail spaces. To enable digital building operations, Scan-to-BIM methods can be utilized to create a geometric-semantic representation with topological relationships. However, reconstructing functional space/zone geometry in open spaces remains a challenge. In this paper, we propose a spatial semantic aggregation method to reconstruct functional clusters based on the semantic similarity and spatial proximity of objects in the scene. In detail, this study proposes a computational method comprising three steps. First, a geometric graph is generated from RGB-D images using open-set object detection and 3D reconstruction to define the nodes, and the spatial proximity to define the relationships. Second, a multi-step graph clustering approach is applied to aggregate objects into functional clusters using distance and semantic coherence. Third, axis-aligned bounding boxes are fitted to the clusters to reconstruct semantic BIM zones. The method is tested on data from a medium-sized souvenir shop. The method is evaluated using standard metrics against ground truth spaces from the IFC model. The results will be analyzed and discussed with respect to their suitability for the use case of asset management and retrieval for retail building operations. A recommendation for enriching the geometric graph with further object or relationship semantics will be provided.

ABSTRACT. We propose a deep learning framework to segment and classify a voxel-based representation of IFC models.

While BIM-based building element classification have set promising avenues of research, most do not take into account the context of elements and they only learn on normalized shaped examples. We believe that element representations should be learned by providing neighboring information to the network. Some classes such as walls, columns and slabs could be very similar in a normalized input as they all can be boxes, but they are different classes that can be discriminated against from their surroundings, as a wall bounds spaces while a column only carries loads. 

However, there is no idiomatic, standardized, nor universal way to decompose the physical building into a set of elements. It depends on preferences of the modeler, functionality of the tool, construction method and phase of the development. The later the phase, the more detailed and the more akin to how the building will be constructed as opposed to how it conceptually functions. Therefore, we present an approach to classify building elements based on a classification of individual voxels (cubic elements of mass on a regular grid obtained from the building element geometries) within the full building context. That means that the method is independent of individual modeler choices on ways of subdivision and can learn even features that are not an individual element, such as e.g typically a door knob.

In this study we put a particular emphasis on the usability of the method in inference for use in applications. We argue that the voxelization step completely eliminates modeler choices as it is able to utilize the full building context and that a patch-based training approach benefits from advances in regularized 2D grid convolutional approaches.

ABSTRACT. Building Information Modeling (BIM) is providing significant opportunity to improve the construction industry by enabling efficient collaboration and information exchange among stakeholders. Developed by buildingSMART, the Industry Foundation Classes (IFC) has long been established as an open standard, facilitating data exchange and interoperability. The adoption of IFC as an open standard offers several advantages for BIM data exchange. It promotes interoperability among different software platforms, allowing stakeholders to seamlessly exchange BIMs without format conversion issues. However, ensuring the authenticity and integrity of BIM data remains a critical concern. Various types of solutions and related standards for exchanging and authenticating BIMs have been developed, but they suffer from certain flaws including limited support for object-level authentication, complex implementation processes, and maintenance issue when it comes to longevity constraints for verifying the authentication mechanism. This research explores the potential of adding digital signatures to build asset information models at the object level by investigating the IFC schema and finding the appropriate container for digital signatures. This research highlights the existing challenges regarding IFC structure to implement a fully functional solution.

ABSTRACT. The focus of this study is a standardized exchange of road quantities in the architecture, engineering and construction (AEC) domain. On the one hand, International Organization for Standardization (ISO) announced a new version of Industry Foundation Classes (IFC) earlier this year which introduces many concepts from the infrastructure domain in general, and road domain in particular. As such, it is now possible to define specifications for IFC exchanges in the infrastructure sector with clear semantic assignment of objects and their attributes. On the other hand, Regulations for Electronic Billing in Construction (REB) (Regeln für die elektronische Bauabrechnung, in German) include the procedural descriptions for the billing of roadway structures, e.g. road construction sites. REB 22.001 specifically standardises the minimum set of quantities applicable to road objects for the German market. Thus, the study at hand explores and reports on the exchange of road quantities following REB 22.001 using the IFC data model. We base our study on the Information Delivery Manual (IDM)/model view definition (MVD) development process following ISO 29481. We take the REB 22.001 guideline as an IDM defining the quantity take-off requirements for road objects. A specialized MVD dubbed 'MVD_REB_22001' has been developed for the newly published IFC version mentioned above. We limit ourselves on the set of concept templates and usages from the Reference View 1.2 MVD. As a consequence, the developed MVD can be used with slight adjustment of applicable objects also in the previous major versions IFC2x3 and IFC4. As such, we do not impose additional burden on the existing implementations of IFC export and import interfaces within commercial software solutions which enables quick adoption by the industry. The resulting MVD is light weight and can be used for exchange and quality control of REB 22.001 compliant road quantities. A prototype has been developed by the German software house AKG in their products VESTRA and KOSTRA. The plausibility of our results has been checked with the implementation of the same exchange, based on the Objekt Katalog Straße (OKSTRA®) and IFC exchange formats. The results showcase the great flexibility of IFC specification to support national use cases without the need for additional international standardisation efforts. Thus, the interoperability is secured and checking routines can be engaged for higher quality as well as a better transparency of digital data during the design processes. This study serves as a template for future endeavours showcasing how national guidelines can employ IFC data model to ensure semantically crisp and seamless information exchanges.

ABSTRACT. Artificial intelligence (AI) for Computer-Aided Design has evolved significantly since its first suggestion in the 1970s, especially with the advent of CADGPT and computational design through generative algorithms. Similarly, several emergent technologies that characterize Industry 4.0 have been adopted to facilitate the efficient delivery of construction projects to offer maximum value to clients. In recent years, the dichotomy between the design and the construction phases has become more unified for optimal financial performance, but it is not without hindrance. In such models, every project team member has a stake in every stage of the project irrespective of the extent of their roles and the occurrence of errors and omissions are mitigated effectively with minimal contractual disruptions. The study introduces a novel conceptual framework within Building Information Modeling (BIM) by integrating AI and Smart Contracts (SC). This framework utilizes AI to refine the BIM model using data from completed projects, thus enhancing the accuracy and utility of the BIM models, which underpins the construction contract. Concurrently, AI activates smart contracts to automate essential contractual processes like signoffs and payments, aligning with the construction progress tracked via the BIM model. This dual approach aims to minimize contractual errors and streamline project execution within a singular, efficient process. The paper validates the proposed framework through qualitative analysis of procurement workflows in the Architecture, Engineering, and Construction (AEC) industry, demonstrating its potential to revolutionize construction project management.

ABSTRACT. Nowadays in the Architecture, Engineering, and Construction (AEC) Industry, the construction cost is quantified using 5D BIM software that facilitates and accelerates the cost estimation. However, the cost estimation process, despite using 5D BIM tools, is still heavily reliant on subjective evaluations that are influenced by the judgment of each estimator. Another problem related to the accuracy of the cost estimation is the lack of interoperability and manual input of information, such as rules for extracting quantities. In addition, by using unstructured cost items, the association between these and their geometrical objects is based on the object-attribute relationship. This activity can be time-consuming and prone to errors, leading to ambiguities in evaluations; often, these issues impact the increase in the cost of the subsequent construction stages and legal disputes. The research aims to compare the traditional cost estimation method using Primus, one of the 5D BIM tools present in the construction market for cost estimation and developed by ACCA software, and an innovative one, carried out by the same research group through a code developed with IfcOpenShell. This new method uses structured cost items in IFC, based on a more complex architecture than a simple description in natural language, and defines a cost estimate through the relationship between cost and geometric IFC entities. The results show that this new method can reduce the limitations in the current cost estimation process and that the approach based on relations between entities (new method) results more agile and effective than the approach based on cost assignment via attributes (traditional method). The new method ensures the possibility of automating the association of the cost items to geometric entities querying the cost item based on the information contained in the geometric entities and verifying the plausibility of the link between them and the cost estimate.

ABSTRACT. In recent years, Digital Twins (DTs) have emerged as an influential and transformational digital technology in the Architecture, Engineering, Construction, and Operations (AECO) industry. This concept has gained traction among researchers and industry practitioners, reflecting its growing importance and relevance in digital transformation in this domain. Use cases play a pivotal role in implementing DTs. Recognizing the crucial role that use cases play, their definition becomes a key point in ensuring the implementation and effectiveness of Digital Twins in the AECO industry. However, there is no formal schema for DT use case definition in the AECO industry. This research aims to bridge this gap by proposing a schema for defining DT use cases through the analysis of BIM use case schemas and DT use case definitions from other industries. Moreover, attributes were added from DT definitions and frameworks. The schema was validated through a focus group and sample use cases.

ABSTRACT. Cognitive Digital Twins (CDTs) are well-established for manufacturing but have had very little implementation in the buildings sector. This paper presents one such implementation: a CDT for a mixed-use (academic/ residential) building developed from the construction BIM. Lessons learned regarding five key elements of CDT development are presented: the integrated data model and supporting ontology; building automation system data acquisition and streaming; data lake; event detection algorithms; and integration and visualization. Insights regarding approach selection, implementation considerations, limitations, and alternatives are presented for each to guide future CDT development.

ABSTRACT. Operating buildings in an optimized manner can be supported by digital twins of a building with the possibility to digitally model and predict the energy use of the building. Generating such digital twins, however, requires a large effort in terms of capturing data from the existing building. Moreover, even if owners or operators are willing to undertake such an investment, capturing the required building data is often not possible as capturing technologies might interfere with the occupants of the building. In this paper, we propose a survey based process to quickly and non-intrusively collect information about a building to allow for an initial energy assessment. We introduce the survey to collect data and show how the survey data can be converted to input for a building performance simulation model. To validate the survey, we benchmark simulated results based on models of eight different buildings across Europe. For each of the buildings we compared the results of building performance simulations of simple models generated from survey data with detailed models generated from BIM models. The results show that simulated deviations are between 0.75% to 40.57% between the two types of models. We conclude that the survey based approach can be a reasonable starting point for first quick energy assessments for most type of buildings.

ABSTRACT. The accelerating pace of digitalization of the built asset industry is predicated upon an ever-growing generation, exchange, transformation, and consumption of built asset information to support decision-making across their lifecycles. Building Information Modeling (BIM) has been hailed as a revolutionary, yet foundational concept supporting this digitalization as it acts as a central platform that enables multiple processes and services supporting the planning, delivery, and use of built assets. The advent of the Internet of Things (IoT) and connected objects as well as an increase in the presence of sensors and actuators allowing real-time coupling of digital and physical worlds, has prompted the concept of BIM to evolve towards that of Digital Twin in the built asset industry, concept which appeared in the manufacturing domain in the 2010’s. Being a nascent concept, questions still abound around digital twins in the built asset industry, their remit, application, operationalization, and most importantly their benefits. This presentation will discuss the initial findings of a research program that is aimed at better defining the application of digital twins in the built asset industry, their adoption and implementation and the capacities needed to support this process. The results include a taxonomy characterizing the different elements of the built asset digital twinning domain. These elements serve as building blocks for the development of a service-oriented digital twinning framework. The framework is validated through an action-research project with Ivanhoe Cambridge, a large real estate company, on the implementation of a digital-twin of Place Ville Marie, a large commercial complex in downtown Montreal, Quebec, Canada.

ABSTRACT. Despite extensive research and development in automating compliance checking (ACC) of building designs, a generic, scalable, and automated system remains elusive. Current ACC systems are limited to specific domains and rule types. The main challenges lie in the vast number of design requirements and the wide diversity of object concepts, attributes, constraints, and relations. This complexity demands an intelligent, autonomous system capable of independent planning, decision-making, and task execution to adapt to new situations, rather than traditional pre-programmed automation paradigms. Inspired by the agency capabilities of Large Language Models (LLMs) in solving complex tasks, this study explores their application in building compliance checking. The goal is to develop an LLM agent that can understand design requirements, plan the checking process, retrieve relevant BIM data, and execute checks autonomously. A proof-of-concept prototype was developed and an autonomous rule-checking experiment conducted with it has successfully demonstrated the potential of the LLM-driven approach.

ABSTRACT. Although Natural Language Processing (NLP), a branch of Artificial Intelligence (AI), is not a new concept, it experienced major advances long after its introduction and became one of the most active trends in the last decade. This branch of AI has numerous applications, such as machine translation, email spam detection, information extraction, text summarization, chatbots, and question answering, applied in various fields.

On the other hand, the construction industry is a fragmented and complex sector with countless pressing challenges. It has been slow in terms of digitization compared to other industries, which hinders applying NLP for various purposes. However, there is an immense potential for applying NLP to facilitate various tasks in construction.

The main purpose of this research is to investigate the potential applications of NLP in the construction industry. A mixed method approach is adopted in which both quantitative and qualitative data are gathered and analyzed. An online survey was designed and distributed to a large group of industry practitioners. In the second stage, a group of professionals were handpicked for semi-structured interviews. The selection strategy considered interviewing professionals in various roles, with varying range of work experience, and working in diverse types of construction companies. The results of the study identified and ranked potential applications, benefits, challenges, and implementation requirements of NLP in each sub-domain.

ABSTRACT. The construction industry is actively pursuing digitalization of its workflow processes. One of the routinely utilized elements in the workflow is email communication. There are plenty of emails the construction companies receive daily, which contain documents, photos, technical drawings, etc. Therefore, employees invest a significant amount of time reviewing these emails, and the absence of categorization results in a situation where some emails lack urgency, and receive high attention, while others requiring immediate responses may be overlooked. Email classification is a resource-intensive process, and dependence on manual processes makes the entire procedure ineffective and error-prone. In this study, we propose an approach that combines multimodal techniques, which can not only interpret the textual information of emails but also process most of the information in attachments, thus improving the classification accuracy. Bidirectional Encoder Representations from Transformers are employed to generate contextual embeddings that effectively capture the semantics of words within their respective contexts. Furthermore, a Bidirectional Long Short Term Memory layer is utilized to further process these embeddings, enabling the extraction of both forward and backward dependencies across the entire text. This combined approach enhances the overall comprehension of the email content by incorporating a thorough analysis of relationships within the text. The classification of emails involves the utilization of an attention mechanism to attain varied levels of feature focus. Distinct attention layers are employed for major and minor classification tasks, to capture relevant features at different hierarchical levels within the email text. This study’s innovative use of a combination of multimodal and natural language processing makes it possible to deal with complex unstructured data in real-world scenarios.

ABSTRACT. A rapid understanding of a subject within a disciplined body of knowledge is critical within today’s abundant scientific literature in research-oriented activities. This paper introduces a cutting-edge system that leverages the capabilities of Generative AI and Retrieval-Augmented Generation (RAG) frameworks to redefine construction engineering and management literature analysis. The method focuses on establishing a comprehensive augmented knowledge base to segment scholarly articles into precise context-aware chunks to build embeddings for inclusion in a vector database by conversion. The approach uses LangChain agents to unearth essential concepts, generate responsive question-answer pairs, and aggregate critical statistics, thereby enhancing the efficiency of each individual step that constitutes the scientific literature within research activities. The system has a user interface with intuitive navigation to maximize the focus and attention span. The aim is to empower researchers to rapidly observe trends and clusters of topical areas within parameters of breadth and depth. The benefits include time reduction, boosting productivity, and ample comprehension of topics that might be missed within the subject matter. The innovation of the approach resides in its provision of immediate, structured access to knowledge and integration of external and internal information sources through advanced search capabilities and AI-driven summarizations. The presented approach improves access to construction engineering and management-related research and pioneers a new benchmark for the academic literature review process, initiating a new form of scientific exploration in the construction engineering discipline.

ABSTRACT. This research paper explores the development of a Computational Design (CD) methodology for constructing adaptable homes in the permafrost landscapes of Northern Canada, particularly in regions like Nunavut. Leveraging the capabilities of Ameba and Karamba 3D, the study emphasizes optimizing structural configurations and material usage, integrating innovative approaches such as Cross-Laminated Timber (CLT) and in situ 3D printing. This strategy addresses the logistical and environmental challenges characteristic of off-site construction, where transportation costs are significant, and conditions are harsh. Additionally, the research underscores the necessity of incorporating Indigenous cultural considerations into the architectural design, ensuring that the final outcomes respect and align with local traditions and values. The paper showcases the transformative potential of CD in extreme and isolated environments. This proposal, while currently theoretical and intended for future research and potential real-world application, presents a significant step forward in environmentally and culturally sensitive architectural design.

ABSTRACT. The construction industry struggles with low productivity and limited usage of prefabrication. One obstacle is the lack of tolerance management and workable tools in BIM systems for implementing it. As most construction is primarily outsourced to subcontractors without implementation design, tolerance management cannot be effectively developed and applied in the fragmented business environment. The study proposes a tolerance management model based on two cases in which prefabricated HVAC elements were used in residential construction and renovation projects. The tolerance management model is based on research conducted in the manufacturing industry's CAD/CAM and ERP systems and manufacturing theory. The model was validated by utilising the construction industry's BIM and CAD systems in the design phase and measuring the installation of prefabricated elements on-site. The analysis of deviations in the context of reactive PDCA indicated the success of tolerance management. The study is significant, as it presents the principles of tolerance management in the manufacturing industry and applies them in a High Volume/Low Product Mix (HVLM) production system not previously seen in the construction industry. The model is suitable for takt production employing a single-piece flow in low product mix construction projects at the project and portfolio level, such as residential, office, and hotel construction, as well as their renovations. Also, from an information management and digitalisation perspective, its validation highlights the challenges and deficiencies in current BIM systems when applied to manufacturing-driven construction (Manufacturing to the Stock, MTS). The research demonstrates how the existing systems and information models can solve this problem, significantly enhancing prefabrication in construction by altering the design approach and implementing tolerance management.

ABSTRACT. This paper describes the development of a modular timber evaporative cooling unit (ECU) possessing a complex dendritic formal quality, presented as a case study, and uses the findings to propose a conceptual prototyping model for the built environment. The case study project leverages computational design methods for design detailing and material optimisation, and digital fabrication for part production. The resultant structure was successfully constructed, however challenges relating to material properties were required to be overcome, these challenges could have been mitigated through a structured approach to prototyping. Reflecting on this project, the paper proposes a conceptual model for built environment prototyping, that combines the insights from established theories to enhance project outcomes.

ABSTRACT. The adoption of semantic web technologies in the digital construction ecosystem has been advocated for facilitating modularity and extensibility of data schemas, or the use of languages and standards such as the SPARQL query language or SHACL rule verifications that are generic and developed beyond the sole AEC environment. A less known functionality of triple store and the web semantic is the use of reasoning engine that enable a dyamic enrichment of a data model through logic inference. In recent years, more and more research on using reasoning engines applied to AEC practices were developed, often through a reduced set of inference rules. In this paper, we present an experiment on using semantic web technologies, and in particular reasoning engines, to deploy an automatic generation of energy renovation scenarios to compare traditional renovation of enveloppes and renewable installation with an integrated solution developed in the ENSNARE H2020 project, at a feasability study stage. The semantic data model and the inference rules are presented and the overall strategy is discussed and compared with alternative implementations.

ABSTRACT. In recent years, there has been a growing trend towards Human-Robot Collaborative (HRC) assembly in off-site construction environment. This approach maximizes the capabilities of robots in the assembly process while leveraging the cognitive abilities, adaptability, and flexibility of humans. However, the realm of HRC assembly in off-site construction is currently in its preliminary stages of development. A complex assembly environment with various obstructing components may hinder the visibility of operators, which causes human-robot interaction challenging and unsafe. Hence, Diminished Reality (DR), a technique that visually removes obstructive elements from the operator's field of view, emerges as a promising solution. This paper investigates an innovative use of DR technology, integrated with a real-time point cloud inpainting method using a deep generative model. The method can recover occluded parts of a shared workspace from view of the operator to ensure safety and efficiency. The proposed method improves visibility, reduces distractions, and consequently, minimizes errors caused by operators. The proposed method consists of three key components:(i) real-time object detection, which detects obstacles within the operator's field of view and generates image of obstructive objects; (ii) deep generative model algorithm, which incorporates a convolutional neural network (CNN) structure with advanced generative adversarial loss, to generate missing depth data of the obstructive objects; and (iii) spatial mapping of the depth image, which reconstructs the occluded scene in point cloud format through coordinate system conversion. Experimental validation of the proposed method demonstrates its effectiveness in reconstructing occluded scene in real time within industrial HRC environments in off-site construction settings. Users experience less distractions and better visibility, which leads to increased safety and efficiency in construction assembly tasks.

ABSTRACT. Much research integrating Building Information Modelling (BIM) and prefabricated modular construction (PFMC) only examines BIM implementation in one isolated process, and not its effect on the project’s entirety. This hinders the formation of an overview of benefits and disadvantages when implementing BIM on multiple phases of PFMC projects. This paper identifies the benefits, disadvantages, and challenges of implementing BIM workflows in the design, manufacturing, and construction phases of PFMC projects, using an openBIM process with Revit as the main modelling tool and IFC files for interdisciplinary collaboration. Results show that BIM implementation in PFMC increases the productivity and quality of design and manufacturing processes, however costing more time to develop projects and revise geometry in BIM compared to 2D-based workflows. This information can help housing fabricators expand openBIM implementation beyond a single phase, inspiring future research on optimizing the application of openBIM workflows to all phases in PFMC project lifecycles.

ABSTRACT. Traditional data collection methods for progress and productivity monitoring on construction sites have mostly remained manual, involving verbal communication, routine job site walking rounds, and techniques such as work sampling (WS) studies. Such methods are costly and time-consuming for construction companies. During the past decades, various construction 4.0-derived physical and virtual technologies have emerged to support progress monitoring in construction projects, including Virtual and Augmented reality, the Internet of Things (IoT), Wearable Sensors, and Unmanned Aerial Systems (UAS). One of the technologies that stand out over others is wearable sensors. This is because they can provide real-time data on workers' movements, activities, and locations in a timely and efficient manner. This study aims to evaluate the feasibility of using smartwatches for workers’ allocation in jobsites. More specifically, the study looks to determine the benefits and barriers of adopting smartwatches to comprehend the causes of low productivity in construction sites. To achieve this goal, a case study methodological approach was adopted. Data collection from interior and exterior work was conducted on a renovation project in Fredericia - Denmark, facilitated by smartwatches. The study found that using smartwatches for tracking workers’ activities and allocating time into distinct workspaces on construction sites presents a promising path for enhancing data collection and understanding the causes of low productivity during the construction stage. Besides, the research highlighted the importance of meticulous data collection, cleaning, and treatment processes to ensure the robustness and reliability of the collected data sets.

ABSTRACT. Lean construction is increasingly adopted due to low productivity, industry fragmentation, and adversarial relationships, enhancing stability and transparency. This paper outlines future scenarios for Germany, aiding construction companies in proactive preparation. Using the scenario technique, three Lean construction scenarios were developed and compared with AI-generated results. A literature review examined Lean construction's current state, followed by identifying crucial influencing factors through megatrend analysis and a PESTEL analysis. Sustainability, digitization, and collaboration emerged as key factors. Future projections for each were developed and linked to form scenarios: A sustainability scenario which is shaped by environmental regulations and green technology for sustainability, advancements in Building Information Modeling (BIM) are shaping the digitization scenario and integrated project delivery methods form the third scenario “collaboration”. The scenarios developed in this paper support those initially suggested by an AI tool. This analysis enables construction companies to adapt oneself to future scenarios and derive actions.

ABSTRACT. Effective water asset management is crucial for ensuring the reliable and sustainable delivery of water services. However, the diverse and complex nature of water infrastructure often poses challenges in managing, exchanging, and retrieving asset information. To address these challenges, this paper proposes the development of an ontology-based approach for standardised water asset management, operations and maintenance (O&M), leveraging the Water Environment Research Foundation (WERF) and the International Infrastructure Management Manual (IIMM) as foundational references. The ontology aims to provide a structured representation of the domain knowledge and relationships between different entities of a water treatment plant (WTP) to enhance information requirements for water asset O&M, promote seamless information exchange between disparate systems, facilitate efficient information retrieval through structured queries, and sets the foundation for future integration with Large Language Models (LLMs) and BIM models to optimize decision-making processes and improve water asset management practices. The paper presents the ontology's design, development methodology, and validation method to assess its effectiveness.

ABSTRACT. Effective cost estimation for tendering plays a critical role in the building construction process, enabling efficient investment management and ensuring successful execution of the construction phase. The current practice involves the classification of building items, extracting all the quantities of the latter, collecting pricing information from construction priced list documents and manually relate these data to the building items. The objective of this paper is to support cost estimation activity by developing a tool that automates the process of assigning a cost domain description to a IFC-based BIM building objects, in such a way as to minimize the human error when manually performing this activity and speed up the process. To handle the textual data involved, the authors introduce a prompt-based framework, testing Mistral-7b language model for querying cost domain description with data in IFC format, which represent two domains characterized by different semantic.

ABSTRACT. Performing building information modeling (BIM) tasks is a complex process that imposes a steep learning curve and a heavy cognitive load due to the necessity of remembering sequences of numerous commands. With the rapid advancement of large language models (LLMs), it is foreseeable that BIM tasks—including querying and managing BIM data, 4D and 5D BIM, design compliance checking, or authoring a design, using written or spoken natural language (i.e., text-to-BIM or speech-to-BIM)—will soon supplant traditional graphical user interfaces. This paper proposes a generalized LLM-augmented BIM framework to expedite the development of LLM-enhanced BIM applications by providing a step-by-step development process. The proposed framework consists of six steps: interpret-fill-match-structure-execute-check. The paper demonstrates the applicability of the proposed framework through implementing a speech-to-BIM application—NADIA-S (Natural-language-based Architectural Detailing through Interaction with Artificial Intelligence via Speech)—using exterior wall detailing as an example.

ABSTRACT. Successful applications of Artificial intelligence (AI) highlight the importance of representing domain-specific data in formats suitable for learning. Researchers in the BIM domain often adopt an inverse approach by selecting AI techniques first and then tailoring BIM information into specific formats, which results in incomplete information and limited application scenarios. We argue that BIM-specific data representations and learning techniques are crucial to leveraging the full richness and scope of the information in BIM models for AI applications. Consequently, this article poses two fundamental questions: 1) What formats are most suitable for BIM data representation? 2) What are the corresponding learning techniques needed for BIM? To begin the exploration of the first question, we propose a graph-based approach to represent design data for storage and computation. Through an object classification experiment, we demonstrate that two AI algorithms achieve an accuracy of 95% and an F1 score of 0.95 after incorporating graph-related features.

ABSTRACT. This study reviews BIM standards across several Asian countries, focusing on those with strong government and professional support, particularly in Singapore, Hong Kong, South Korea, and the United Arab Emirates. These countries have made significant progress in implementing BIM Level 3, aligning with ISO 19650 to enhance industry-wide data sharing and collaboration. The study examines the development of BIM standards, the factors driving BIM adoption, and the importance of aligning regional approaches with global standards. The findings highlight variations in BIM standards due to differing cultural and regulatory contexts. The study concludes that effective BIM adoption is driven by government initiatives, professional support, and strategic industry integration, with alignment to international standards being crucial for seamless cross-border BIM integration and improved efficiency in the construction sector.

ABSTRACT. The construction industry is experiencing a notable growth in innovation and technology adoption, with major corporate entities increasingly prioritizing technology investment through several ways like corporate venturing or the acquisition of products from technology companies. However, despite this growing trend, the overall scale of technology investment remains relatively modest when compared to the vast scope of the industry. Additionally, there is a noticeable lack of comprehensive research on the motivations and challenges associated with this technological shift. Especially notable is the lack of studies that specifically investigate corporate ventures within the construction sector. As a response to this gap, this study aims to delve into the drivers and challenges shaping technology investment in the construction industry. The research seeks to provide a nuanced understanding by examining various investment approaches, shedding light on the dynamics within this evolving landscape. To achieve this aim, the study conducts interviews with construction companies invested in novel technologies to reveal their decision-making behavior. The research has both theoretical and practical contributions. In terms of theoretical contributions, the study fills the gap in the literature on construction companies’ decision-making approach when selecting a novel technology and innovation investment. Practically, the result of the study can be used by construction companies when collaborating with a technology startup and selecting the best service/product to improve their performance. Moreover, construction technology companies may benefit from the findings when targeting their customers and follow a solid strategy to meet customers’ expectations and thereby improve their business.

ABSTRACT. The adoption of Building Information Modeling (BIM) in architectural Brazilian offices has been rapidly increasing due to the numerous benefits it provides architects in their design decision-making process. However, when using BIM, information related to the physical and functional aspects of the building can be read and interpreted by computers. As a result, this information becomes vulnerable to intentional or accidental cyber-attacks. Considering this, it is essential for architects to employ strategies to safeguard project information. This paper proposes a maturity model for the adoption of information management and security, specifically tailored to architectural firms that utilize BIM in their project development. The model is structured based on guidelines found in the ISO 19650 series, as well as the General Data Protection Law and other relevant directives. Additionally, a risk matrix was developed in conjunction with an analysis of documents from companies that have incorporated BIM into their projects. The maturity model outlines pathways for architectural firms to implement measures that ensure information security throughout project development. These measures contribute to advancing BIM adoption while considering the risks associated with innovative processes.

ABSTRACT. Temporary works are essential parts of every construction project. This paper focuses on Scaffolding systems as elements of temporary works due to its frequent use in most construction projects and its high impact on safety measures in construction sites. This paper aims to review the current state of the scaffolding industry and identify areas within the design of scaffolding systems that require further consideration by BIM facilities. From the review it was apparent that the need for automation in the industry was high. This was due to the reactive nature of temporary works resulting in frequent geometry changes causing many extra hours of rework to both the structure design and related safety registers. It was concluded that the commercially available BIM applications are insufficient to provide fully tested temporary works. Although attempts have been made, they need to move from the current simple geometry modelling to a further functionality modelling.

ABSTRACT. This study investigates the impact of digital technology and explores a holistic framework to guide the development, integration and implementation of digital transformation strategies, products, and services in the Architecture, Engineering, Construction and Operations (AECO) industry. The study adopts an abductive and qualitative approach to data collection and analysis through a review of 78 journal articles and four semi-structured interviews with AECO experts. The findings suggest that transitioning from siloed to systems-based thinking, push to pull service models, and product-focused to integrated stakeholder-driven approaches would facilitate the strategic and efficient utilisation of digital technologies. The findings present a preliminary conceptual framework that advocates for a platform perspective in implementing digital initiatives to integrate digitally-enabled deliverables, services, and products. The study improves the understanding of factors that facilitate the successful implementation of digital transformation initiatives in the AECO industry, paving the way for better utilisation of machine-readable and interpretable building information.